AuroraMiddleware/v8
1
0
Fork 0
Code Issues 2 Pull Requests Releases Wiki Activity
11f1e12b25
v8/test/cctest/interpreter/bytecode_expectations/IIFEWithOneshotOpt.golden

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

791 lines
21 KiB
Plaintext
Raw Normal View History

Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
#
# Autogenerated by generate-bytecode-expectations.
#
---
wrap: no
top level: yes
print callee: yes
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
oneshot opt: yes
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
---
snippet: "
(function() {
l = {};
l.aa = 2;
l.bb = l.aa;
return arguments.callee;
})();
"
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
frame size: 3
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 39
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 16 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 29 S> */ B(CreateEmptyObjectLiteral),
/* 31 E> */ B(StaGlobal), U8(0), U8(0),
/* 45 S> */ B(LdaGlobal), U8(0), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(2),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 50 E> */ B(StaNamedPropertyNoFeedback), R(1), U8(1), U8(0),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 63 S> */ B(LdaGlobal), U8(0), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 70 E> */ B(LdaGlobal), U8(0), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 72 E> */ B(LdaNamedPropertyNoFeedback), R(2), U8(1),
/* 68 E> */ B(StaNamedPropertyNoFeedback), R(1), U8(2), U8(0),
/* 101 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(3),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 108 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["l"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["aa"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["bb"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
(function() {
l = {};
for (i = 0; i < 5; ++i) {
l.aa = 2;
l.bb = l.aa;
}
return arguments.callee;
})();
"
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
frame size: 3
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 67
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 16 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 29 S> */ B(CreateEmptyObjectLiteral),
/* 31 E> */ B(StaGlobal), U8(0), U8(0),
/* 50 S> */ B(LdaZero),
/* 52 E> */ B(StaGlobal), U8(1), U8(2),
/* 59 S> */ B(LdaGlobal), U8(1), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(5),
/* 59 E> */ B(TestLessThan), R(1), U8(6),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(JumpIfFalse), U8(42),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 81 S> */ B(LdaGlobal), U8(0), U8(7),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(2),
/* 86 E> */ B(StaNamedProperty), R(1), U8(2), U8(9),
/* 101 S> */ B(LdaGlobal), U8(0), U8(7),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 108 E> */ B(LdaGlobal), U8(0), U8(7),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 110 E> */ B(LdaNamedProperty), R(2), U8(2), U8(11),
/* 106 E> */ B(StaNamedProperty), R(1), U8(3), U8(13),
/* 66 S> */ B(LdaGlobal), U8(1), U8(4),
B(Inc), U8(15),
/* 66 E> */ B(StaGlobal), U8(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
/* 45 E> */ B(JumpLoop), U8(49), I8(0),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 149 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(4),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 156 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["l"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["i"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["aa"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["bb"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
(function() {
l = {};
c = 4;
while(c > 4) {
l.aa = 2;
l.bb = l.aa;
c--;
}
return arguments.callee;
})();
"
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
frame size: 3
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 68
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 16 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 29 S> */ B(CreateEmptyObjectLiteral),
/* 31 E> */ B(StaGlobal), U8(0), U8(0),
/* 45 S> */ B(LdaSmi), I8(4),
/* 47 E> */ B(StaGlobal), U8(1), U8(2),
/* 68 S> */ B(LdaGlobal), U8(1), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(4),
/* 68 E> */ B(TestGreaterThan), R(1), U8(6),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(JumpIfFalse), U8(42),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 85 S> */ B(LdaGlobal), U8(0), U8(7),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(2),
/* 90 E> */ B(StaNamedProperty), R(1), U8(2), U8(9),
/* 105 S> */ B(LdaGlobal), U8(0), U8(7),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 112 E> */ B(LdaGlobal), U8(0), U8(7),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 114 E> */ B(LdaNamedProperty), R(2), U8(2), U8(11),
/* 110 E> */ B(StaNamedProperty), R(1), U8(3), U8(13),
/* 128 S> */ B(LdaGlobal), U8(1), U8(4),
B(Dec), U8(15),
/* 129 E> */ B(StaGlobal), U8(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
/* 60 E> */ B(JumpLoop), U8(49), I8(0),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 168 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(4),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 175 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["l"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["c"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["aa"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["bb"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
(function() {
l = {};
c = 4;
do {
l.aa = 2;
l.bb = l.aa;
c--;
} while(c > 4)
return arguments.callee;
})();
"
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
frame size: 3
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 68
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 16 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 29 S> */ B(CreateEmptyObjectLiteral),
/* 31 E> */ B(StaGlobal), U8(0), U8(0),
/* 45 S> */ B(LdaSmi), I8(4),
/* 47 E> */ B(StaGlobal), U8(1), U8(2),
/* 75 S> */ B(LdaGlobal), U8(0), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(2),
/* 80 E> */ B(StaNamedProperty), R(1), U8(2), U8(6),
/* 95 S> */ B(LdaGlobal), U8(0), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 102 E> */ B(LdaGlobal), U8(0), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 104 E> */ B(LdaNamedProperty), R(2), U8(2), U8(8),
/* 100 E> */ B(StaNamedProperty), R(1), U8(3), U8(10),
/* 118 S> */ B(LdaGlobal), U8(1), U8(12),
B(Dec), U8(14),
/* 119 E> */ B(StaGlobal), U8(1), U8(2),
/* 141 S> */ B(LdaGlobal), U8(1), U8(12),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(4),
/* 141 E> */ B(TestGreaterThan), R(1), U8(15),
[interpreter] Make IterationBody StackChecks implicit within JumpLoop Since now the IterationBody StackChecks are implicit within JumpLoops, we are able to eagerly deopt in them. If we do that, whenever we advance to the next bytecode we don't have to advance to the next literal bytecode, but instead "advance" in the sense of doing the JumpLoop. Adding tests that test this advancing for wide and extra wide JumpLoops. Also, marking JumpLoop as needing source positions since now it has the ability of causing an interrupt. Bug: v8:10149, v8:9960 Fixes: v8:10149 Change-Id: Ib0d9efdfb379e0dfbba7a7f67cba9262668813b0 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2064226 Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jakob Gruber <jgruber@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66809}
2020-03-20 11:13:04 +00:00
B(JumpIfFalse), U8(5),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
/* 60 E> */ B(JumpLoop), U8(49), I8(0),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 171 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(4),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 178 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["l"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["c"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["aa"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["bb"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
(function() {
l = {
'aa': 3.3,
'bb': 4.4
};
if (l.aa < 3) {
l.aa = 3;
} else {
l.aa = l.bb;
}
return arguments.callee;
})();
"
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
frame size: 4
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 70
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 16 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
/* 29 S> */ B(LdaConstant), U8(0),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
B(LdaSmi), I8(41),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
B(CallRuntime), U16(Runtime::kCreateObjectLiteralWithoutAllocationSite), R(2), U8(2),
/* 31 E> */ B(StaGlobal), U8(1), U8(0),
/* 95 S> */ B(LdaGlobal), U8(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 101 E> */ B(LdaNamedPropertyNoFeedback), R(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(3),
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
/* 104 E> */ B(TestLessThan), R(1), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(JumpIfFalse), U8(15),
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
/* 121 S> */ B(LdaGlobal), U8(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
B(LdaSmi), I8(3),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 126 E> */ B(StaNamedPropertyNoFeedback), R(1), U8(2), U8(0),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Jump), U8(19),
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
/* 158 S> */ B(LdaGlobal), U8(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[interpreter][runtime] Avoid AllocationSites for oneshot code No need to create allocation site for literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateObjectLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I224b3a30f10361cfe9ff63129b36da8230c5e403 Reviewed-on: https://chromium-review.googlesource.com/1163615 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#55050}
2018-08-10 12:41:06 +00:00
/* 165 E> */ B(LdaGlobal), U8(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 167 E> */ B(LdaNamedPropertyNoFeedback), R(2), U8(3),
/* 163 E> */ B(StaNamedPropertyNoFeedback), R(1), U8(2), U8(0),
/* 206 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(4),
Reland "[Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code" This is a reland of 690bda84eb023c7afa63c51131c474771fbf7bda Original change's description: > [Interpreter] Do not use IC slots for property load/stores in an IIFE and top-level code > > An IIFE or top-level code is executed only once hence, there is no need to collect > type feedback. We can save some memory by not using IC slots for property Loads/Stores > within a IIFE/top-level code. This CL emits Runtime Get/Set property calls instead of LdaNamedProperty > /StaNamedProperty for the property loads within a IIFE and top-level code. > > Change-Id: I3e0ce26d05d82bb3648cb9262c4e112a2c4556c9 > Reviewed-on: https://chromium-review.googlesource.com/1146579 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Marja Hölttä <marja@chromium.org> > Reviewed-by: Camillo Bruni <cbruni@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#54949} Change-Id: I7b07ce86f7236d82191caaceafd31b86e5863ff5 Reviewed-on: https://chromium-review.googlesource.com/1167802 Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Reviewed-by: Toon Verwaest <verwaest@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55017}
2018-08-07 13:14:23 +00:00
/* 213 S> */ B(Return),
]
constant pool: [
OBJECT_BOILERPLATE_DESCRIPTION_TYPE,
ONE_BYTE_INTERNALIZED_STRING_TYPE ["l"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["aa"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["bb"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
---
snippet: "
(function() {
a = [0, [1, 1,2,], 3];
return arguments.callee;
})();
"
Reland "[interpreter] Add bytecode for leading array spreads." This is a reland of 1c48d52bb1ee9bb28e146c60eda08cd4afaa5745. It turned out that IterableToList doesn't always behave according to the ES operation with the same name. Specifically, it allows holey arrays to take its fast path, which produces an output array with holes where actually "undefined" elements should appear. This CL changes the version of IterableToList that is used for spreads (IterableToListWithSymbolLookup) such that holey arrays take the slow path. It also includes tests for such situations. Original change's description: > [interpreter] Add bytecode for leading array spreads. > > This CL improves the performance of creating [...a, b] or [...a]. > If the array literal has a leading spread, this CL emits the bytecode > [CreateArrayFromIterable] to create the literal. CreateArrayFromIterable > is implemented by [IterableToListDefault] builtin to create the initial > array for the leading spread. IterableToListDefault has a fast path to > clone efficiently if the spread is an actual array. > > The bytecode generated is now shorter. Bytecode generation is refactored > into to BuildCreateArrayLiteral, which allows VisitCallSuper to benefit > from this optimization also. > For now, turbofan also lowers the bytecode to the builtin. > > The idiomatic use of [...a] to clone the array a now performs better > than a simple for-loop, but still does not match the performance of slice. > > Bug: v8:7980 > > Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng > Change-Id: Ibde659c82d3c7aa1b1777a3d2f6426ac8cc15e35 > Reviewed-on: https://chromium-review.googlesource.com/1181024 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Sigurd Schneider <sigurds@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Georg Neis <neis@chromium.org> > Commit-Queue: Hai Dang <dhai@google.com> > Cr-Commit-Position: refs/heads/master@{#55520} Bug: v8:7980 Change-Id: I0b5603a12d2b588327658bf0a9b214bd0f22e237 Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1201882 Commit-Queue: Hai Dang <dhai@google.com> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#55639}
2018-09-05 07:50:48 +00:00
frame size: 5
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 23
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 16 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
/* 29 S> */ B(LdaConstant), U8(0),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
B(LdaSmi), I8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
Reland "[interpreter] Add bytecode for leading array spreads." This is a reland of 1c48d52bb1ee9bb28e146c60eda08cd4afaa5745. It turned out that IterableToList doesn't always behave according to the ES operation with the same name. Specifically, it allows holey arrays to take its fast path, which produces an output array with holes where actually "undefined" elements should appear. This CL changes the version of IterableToList that is used for spreads (IterableToListWithSymbolLookup) such that holey arrays take the slow path. It also includes tests for such situations. Original change's description: > [interpreter] Add bytecode for leading array spreads. > > This CL improves the performance of creating [...a, b] or [...a]. > If the array literal has a leading spread, this CL emits the bytecode > [CreateArrayFromIterable] to create the literal. CreateArrayFromIterable > is implemented by [IterableToListDefault] builtin to create the initial > array for the leading spread. IterableToListDefault has a fast path to > clone efficiently if the spread is an actual array. > > The bytecode generated is now shorter. Bytecode generation is refactored > into to BuildCreateArrayLiteral, which allows VisitCallSuper to benefit > from this optimization also. > For now, turbofan also lowers the bytecode to the builtin. > > The idiomatic use of [...a] to clone the array a now performs better > than a simple for-loop, but still does not match the performance of slice. > > Bug: v8:7980 > > Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng > Change-Id: Ibde659c82d3c7aa1b1777a3d2f6426ac8cc15e35 > Reviewed-on: https://chromium-review.googlesource.com/1181024 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Sigurd Schneider <sigurds@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Georg Neis <neis@chromium.org> > Commit-Queue: Hai Dang <dhai@google.com> > Cr-Commit-Position: refs/heads/master@{#55520} Bug: v8:7980 Change-Id: I0b5603a12d2b588327658bf0a9b214bd0f22e237 Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1201882 Commit-Queue: Hai Dang <dhai@google.com> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#55639}
2018-09-05 07:50:48 +00:00
B(CallRuntime), U16(Runtime::kCreateArrayLiteralWithoutAllocationSite), R(3), U8(2),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
/* 31 E> */ B(StaGlobal), U8(1), U8(0),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 77 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(2),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
/* 84 S> */ B(Return),
]
constant pool: [
ARRAY_BOILERPLATE_DESCRIPTION_TYPE,
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
(function() {
a = [];
return arguments.callee;
})();
"
Reland "[interpreter] Add bytecode for leading array spreads." This is a reland of 1c48d52bb1ee9bb28e146c60eda08cd4afaa5745. It turned out that IterableToList doesn't always behave according to the ES operation with the same name. Specifically, it allows holey arrays to take its fast path, which produces an output array with holes where actually "undefined" elements should appear. This CL changes the version of IterableToList that is used for spreads (IterableToListWithSymbolLookup) such that holey arrays take the slow path. It also includes tests for such situations. Original change's description: > [interpreter] Add bytecode for leading array spreads. > > This CL improves the performance of creating [...a, b] or [...a]. > If the array literal has a leading spread, this CL emits the bytecode > [CreateArrayFromIterable] to create the literal. CreateArrayFromIterable > is implemented by [IterableToListDefault] builtin to create the initial > array for the leading spread. IterableToListDefault has a fast path to > clone efficiently if the spread is an actual array. > > The bytecode generated is now shorter. Bytecode generation is refactored > into to BuildCreateArrayLiteral, which allows VisitCallSuper to benefit > from this optimization also. > For now, turbofan also lowers the bytecode to the builtin. > > The idiomatic use of [...a] to clone the array a now performs better > than a simple for-loop, but still does not match the performance of slice. > > Bug: v8:7980 > > Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng > Change-Id: Ibde659c82d3c7aa1b1777a3d2f6426ac8cc15e35 > Reviewed-on: https://chromium-review.googlesource.com/1181024 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Sigurd Schneider <sigurds@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Georg Neis <neis@chromium.org> > Commit-Queue: Hai Dang <dhai@google.com> > Cr-Commit-Position: refs/heads/master@{#55520} Bug: v8:7980 Change-Id: I0b5603a12d2b588327658bf0a9b214bd0f22e237 Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1201882 Commit-Queue: Hai Dang <dhai@google.com> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#55639}
2018-09-05 07:50:48 +00:00
frame size: 5
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 23
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 16 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
/* 29 S> */ B(LdaConstant), U8(0),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
B(LdaSmi), I8(37),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
Reland "[interpreter] Add bytecode for leading array spreads." This is a reland of 1c48d52bb1ee9bb28e146c60eda08cd4afaa5745. It turned out that IterableToList doesn't always behave according to the ES operation with the same name. Specifically, it allows holey arrays to take its fast path, which produces an output array with holes where actually "undefined" elements should appear. This CL changes the version of IterableToList that is used for spreads (IterableToListWithSymbolLookup) such that holey arrays take the slow path. It also includes tests for such situations. Original change's description: > [interpreter] Add bytecode for leading array spreads. > > This CL improves the performance of creating [...a, b] or [...a]. > If the array literal has a leading spread, this CL emits the bytecode > [CreateArrayFromIterable] to create the literal. CreateArrayFromIterable > is implemented by [IterableToListDefault] builtin to create the initial > array for the leading spread. IterableToListDefault has a fast path to > clone efficiently if the spread is an actual array. > > The bytecode generated is now shorter. Bytecode generation is refactored > into to BuildCreateArrayLiteral, which allows VisitCallSuper to benefit > from this optimization also. > For now, turbofan also lowers the bytecode to the builtin. > > The idiomatic use of [...a] to clone the array a now performs better > than a simple for-loop, but still does not match the performance of slice. > > Bug: v8:7980 > > Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng > Change-Id: Ibde659c82d3c7aa1b1777a3d2f6426ac8cc15e35 > Reviewed-on: https://chromium-review.googlesource.com/1181024 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Sigurd Schneider <sigurds@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Georg Neis <neis@chromium.org> > Commit-Queue: Hai Dang <dhai@google.com> > Cr-Commit-Position: refs/heads/master@{#55520} Bug: v8:7980 Change-Id: I0b5603a12d2b588327658bf0a9b214bd0f22e237 Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1201882 Commit-Queue: Hai Dang <dhai@google.com> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#55639}
2018-09-05 07:50:48 +00:00
B(CallRuntime), U16(Runtime::kCreateArrayLiteralWithoutAllocationSite), R(3), U8(2),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
/* 31 E> */ B(StaGlobal), U8(1), U8(0),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 62 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(2),
[interpreter][runtime] Avoid AllocationSites for Array literals in oneshot code No need to create allocation site for array literals in oneshot code since they are executed only once. The interpreter emits a runtime call to CreateArrayLiteralWithoutAllocationSite for creating literals in oneshot code instead. Change-Id: I285879c84759ff9e2ce281e9548112f52ce5e7d1 Reviewed-on: https://chromium-review.googlesource.com/1167843 Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Commit-Queue: Chandan Reddy <chandanreddy@google.com> Cr-Commit-Position: refs/heads/master@{#55349}
2018-08-23 09:50:54 +00:00
/* 69 S> */ B(Return),
]
constant pool: [
ARRAY_BOILERPLATE_DESCRIPTION_TYPE,
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
---
snippet: "
this.f0 = function() {};
this.f1 = function(a) {};
this.f2 = function(a, b) {};
this.f3 = function(a, b, c) {};
this.f4 = function(a, b, c, d) {};
this.f5 = function(a, b, c, d, e) {};
(function() {
this.f0();
this.f1(1);
this.f2(1, 2);
this.f3(1, 2, 3);
this.f4(1, 2, 3, 4);
this.f5(1, 2, 3, 4, 5);
return arguments.callee;
})();
"
frame size: 8
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 136
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 237 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 255 S> */ B(LdaNamedPropertyNoFeedback), R(this), U8(0),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 255 E> */ B(CallNoFeedback), R(1), R(this), U8(1),
/* 274 S> */ B(LdaNamedPropertyNoFeedback), R(this), U8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(Mov), R(this), R(2),
/* 274 E> */ B(CallNoFeedback), R(1), R(2), U8(2),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 294 S> */ B(LdaNamedPropertyNoFeedback), R(this), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(Mov), R(this), R(2),
/* 294 E> */ B(CallNoFeedback), R(1), R(2), U8(3),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 317 S> */ B(LdaNamedPropertyNoFeedback), R(this), U8(3),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(3),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(5),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(Mov), R(this), R(2),
/* 317 E> */ B(CallNoFeedback), R(1), R(2), U8(4),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 343 S> */ B(LdaNamedPropertyNoFeedback), R(this), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(3),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(5),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(6),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(Mov), R(this), R(2),
/* 343 E> */ B(CallNoFeedback), R(1), R(2), U8(5),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 372 S> */ B(LdaNamedPropertyNoFeedback), R(this), U8(5),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(3),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(5),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(6),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(5),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(7),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(Mov), R(this), R(2),
/* 372 E> */ B(CallNoFeedback), R(1), R(2), U8(6),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 416 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(6),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 423 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f0"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f1"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f2"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f3"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f4"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f5"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
function f0() {}
function f1(a) {}
function f2(a, b) {}
function f3(a, b, c) {}
function f4(a, b, c, d) {}
function f5(a, b, c, d, e) {}
(function() {
f0();
f1(1);
f2(1, 2);
f3(1, 2, 3);
f4(1, 2, 3, 4);
f5(1, 2, 3, 4, 5);
return arguments.callee;
})();
"
frame size: 8
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 139
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 189 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 202 S> */ B(LdaUndefined),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaGlobal), U8(0), U8(0),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 202 E> */ B(CallNoFeedback), R(1), R(2), U8(1),
/* 216 S> */ B(LdaUndefined),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaGlobal), U8(1), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 216 E> */ B(CallNoFeedback), R(1), R(2), U8(2),
/* 231 S> */ B(LdaUndefined),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaGlobal), U8(2), U8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 231 E> */ B(CallNoFeedback), R(1), R(2), U8(3),
/* 249 S> */ B(LdaUndefined),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaGlobal), U8(3), U8(6),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(3),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(5),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 249 E> */ B(CallNoFeedback), R(1), R(2), U8(4),
/* 270 S> */ B(LdaUndefined),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaGlobal), U8(4), U8(8),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(3),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(5),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(6),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 270 E> */ B(CallNoFeedback), R(1), R(2), U8(5),
/* 294 S> */ B(LdaUndefined),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaGlobal), U8(5), U8(10),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(1),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(3),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(4),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(3),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(5),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(4),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(6),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
B(LdaSmi), I8(5),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(7),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 294 E> */ B(CallNoFeedback), R(1), R(2), U8(6),
Reland "[interpreter] Separate bytecodes for one-shot property loads and stores" This is a reland of eccf18674911a3d8a84213589e4bda70bf81aeb1 Original change's description: > [interpreter] Separate bytecodes for one-shot property loads and stores > > Create LdaNamedPropertyNoFeedback and StaNamedPropertyNoFeedback > for one-shot property loads and stores. This CL replaces the runtime > calls with new bytecodes for named property load stores in one-shot code. > the runtime calls needed extra set of consecutive registers and > additional move instructions. This increased the size of > bytecode-array and possibly extended the life time of objects. > By replacing them with NoFeedback bytecodes we avoid these issues. > > Bug: v8:8072 > Change-Id: I20a38a5ce9940026171d870d354787fe0b7c5a6f > Reviewed-on: https://chromium-review.googlesource.com/1196725 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Reviewed-by: Yang Guo <yangguo@chromium.org> > Reviewed-by: Georg Neis <neis@chromium.org> > Commit-Queue: Chandan Reddy <chandanreddy@google.com> > Cr-Commit-Position: refs/heads/master@{#56211} Bug: v8:8072 Change-Id: Ie8e52b37daf35c7bc08bb910d7b15a9b783354e4 Reviewed-on: https://chromium-review.googlesource.com/1245742 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Georg Neis <neis@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#56266}
2018-09-27 13:10:08 +00:00
/* 338 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(6),
[Interpreter] Create and use CallNoFeedback bytecode for one-shot code We do not have to collect feedback for function calls in one-shot code. This CL avoids allocating CallICslots for each function call by emitting CallNoFeedback bytecodes. We save one CallICSlot (two entries in feedback vector) per function call in One-shot. Bug: v8:8072 Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic2580e5972acd5124c2e71d540985736ce797fe8 Reviewed-on: https://chromium-review.googlesource.com/1178051 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#55951}
2018-09-17 08:44:19 +00:00
/* 345 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f0"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f1"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f2"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f3"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f4"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f5"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
---
snippet: "
var t = 0;
function f2() {};
if (t == 0) {
(function(){
l = {};
l.a = 3;
l.b = 4;
f2();
return arguments.callee;
})();
}
"
[Compile] Move decision of whether an IIFE is oneshot into parser. The decision as to whether to optimize an IIFE as oneshot depends on whether it's outer scope is the script scope. During lazy compile, we might have discarded scopes which don't need a context between the IIFE and the script scope, which means we might treat an IIFE as oneshot, even though initial eager compile treated it as non-oneshot. Both bytecode flushing and lazy source positions rely on us generating the same bytecode during lazy compile as eager compile, so we move the decision into the parser where it happens once and is then stored in the SFI for any future lazy compiles. BUG=v8:8395,v8:8510 Change-Id: I88f1e74ad95d47a2636c393ceb1318d7d610055d Reviewed-on: https://chromium-review.googlesource.com/c/1421841 Reviewed-by: Toon Verwaest <verwaest@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#58996}
2019-01-21 10:13:52 +00:00
frame size: 2
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 42
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 79 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
/* 93 S> */ B(CreateEmptyObjectLiteral),
/* 95 E> */ B(StaGlobal), U8(0), U8(0),
/* 111 S> */ B(LdaGlobal), U8(0), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
B(LdaSmi), I8(3),
[Compile] Move decision of whether an IIFE is oneshot into parser. The decision as to whether to optimize an IIFE as oneshot depends on whether it's outer scope is the script scope. During lazy compile, we might have discarded scopes which don't need a context between the IIFE and the script scope, which means we might treat an IIFE as oneshot, even though initial eager compile treated it as non-oneshot. Both bytecode flushing and lazy source positions rely on us generating the same bytecode during lazy compile as eager compile, so we move the decision into the parser where it happens once and is then stored in the SFI for any future lazy compiles. BUG=v8:8395,v8:8510 Change-Id: I88f1e74ad95d47a2636c393ceb1318d7d610055d Reviewed-on: https://chromium-review.googlesource.com/c/1421841 Reviewed-by: Toon Verwaest <verwaest@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#58996}
2019-01-21 10:13:52 +00:00
/* 115 E> */ B(StaNamedProperty), R(1), U8(1), U8(4),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
/* 130 S> */ B(LdaGlobal), U8(0), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
B(LdaSmi), I8(4),
[Compile] Move decision of whether an IIFE is oneshot into parser. The decision as to whether to optimize an IIFE as oneshot depends on whether it's outer scope is the script scope. During lazy compile, we might have discarded scopes which don't need a context between the IIFE and the script scope, which means we might treat an IIFE as oneshot, even though initial eager compile treated it as non-oneshot. Both bytecode flushing and lazy source positions rely on us generating the same bytecode during lazy compile as eager compile, so we move the decision into the parser where it happens once and is then stored in the SFI for any future lazy compiles. BUG=v8:8395,v8:8510 Change-Id: I88f1e74ad95d47a2636c393ceb1318d7d610055d Reviewed-on: https://chromium-review.googlesource.com/c/1421841 Reviewed-by: Toon Verwaest <verwaest@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#58996}
2019-01-21 10:13:52 +00:00
/* 134 E> */ B(StaNamedProperty), R(1), U8(2), U8(6),
/* 149 S> */ B(LdaGlobal), U8(3), U8(8),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[Compile] Move decision of whether an IIFE is oneshot into parser. The decision as to whether to optimize an IIFE as oneshot depends on whether it's outer scope is the script scope. During lazy compile, we might have discarded scopes which don't need a context between the IIFE and the script scope, which means we might treat an IIFE as oneshot, even though initial eager compile treated it as non-oneshot. Both bytecode flushing and lazy source positions rely on us generating the same bytecode during lazy compile as eager compile, so we move the decision into the parser where it happens once and is then stored in the SFI for any future lazy compiles. BUG=v8:8395,v8:8510 Change-Id: I88f1e74ad95d47a2636c393ceb1318d7d610055d Reviewed-on: https://chromium-review.googlesource.com/c/1421841 Reviewed-by: Toon Verwaest <verwaest@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#58996}
2019-01-21 10:13:52 +00:00
/* 149 E> */ B(CallUndefinedReceiver0), R(1), U8(10),
/* 182 S> */ B(LdaNamedProperty), R(0), U8(4), U8(12),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
/* 189 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["l"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["b"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f2"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
function f2() {};
function f() {
return (function(){
l = {};
l.a = 3;
l.b = 4;
f2();
return arguments.callee;
})();
}
f();
"
frame size: 2
parameter count: 1
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 42
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 76 E> */ B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
/* 92 S> */ B(CreateEmptyObjectLiteral),
/* 94 E> */ B(StaGlobal), U8(0), U8(0),
/* 112 S> */ B(LdaGlobal), U8(0), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
B(LdaSmi), I8(3),
/* 116 E> */ B(StaNamedProperty), R(1), U8(1), U8(4),
/* 133 S> */ B(LdaGlobal), U8(0), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
B(LdaSmi), I8(4),
/* 137 E> */ B(StaNamedProperty), R(1), U8(2), U8(6),
/* 154 S> */ B(LdaGlobal), U8(3), U8(8),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(1),
[interpreter] Restrict one-shot optimizations to only top-level IIFEs IIFE`s within a function are not guaranteed to be executed only once. They can be called multiple times and compiler can inline them. Do the one-shot optimizations only for IIFE`s from top-level code. Bug: v8:8072, chromium:886580 Change-Id: I02370681cc3eab270edcc75ee120ca7ad768ed52 Reviewed-on: https://chromium-review.googlesource.com/1231174 Commit-Queue: Chandan Reddy <chandanreddy@google.com> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56024}
2018-09-19 09:30:52 +00:00
/* 154 E> */ B(CallUndefinedReceiver0), R(1), U8(10),
/* 189 S> */ B(LdaNamedProperty), R(0), U8(4), U8(12),
/* 196 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["l"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["b"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["f2"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
---
snippet: "
var f = function(l) { l.a = 3; return l; };
f({});
f;
"
frame size: 0
parameter count: 2
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
bytecode array length: 9
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
bytecodes: [
/* 32 S> */ B(LdaSmi), I8(3),
/* 36 E> */ B(StaNamedProperty), R(arg0), U8(0), U8(0),
/* 41 S> */ B(Ldar), R(arg0),
/* 50 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
]
handlers: [
]
---
snippet: "
var f = (function(l) { l.a = 3; return l; });
f;
"
frame size: 0
parameter count: 2
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
bytecode array length: 9
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
bytecodes: [
/* 33 S> */ B(LdaSmi), I8(3),
/* 37 E> */ B(StaNamedProperty), R(arg0), U8(0), U8(0),
/* 42 S> */ B(Ldar), R(arg0),
/* 51 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
]
handlers: [
]
---
snippet: "
var f = (function foo(l) { l.a = 3; return l; });
f;
"
frame size: 0
parameter count: 2
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
bytecode array length: 9
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
bytecodes: [
/* 37 S> */ B(LdaSmi), I8(3),
/* 41 E> */ B(StaNamedProperty), R(arg0), U8(0), U8(0),
/* 46 S> */ B(Ldar), R(arg0),
/* 55 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
]
handlers: [
]
---
snippet: "
var f = function foo(l) { l.a = 3; return l; };
f({});
f;
"
frame size: 0
parameter count: 2
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
bytecode array length: 9
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
bytecodes: [
/* 36 S> */ B(LdaSmi), I8(3),
/* 40 E> */ B(StaNamedProperty), R(arg0), U8(0), U8(0),
/* 45 S> */ B(Ldar), R(arg0),
/* 54 S> */ B(Return),
]
constant pool: [
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
]
handlers: [
]
---
snippet: "
l = {};
var f = (function foo(l) { l.a = 3; return arguments.callee; })(l);
f;
"
frame size: 3
parameter count: 2
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 26
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 46 E> */ B(CreateFunctionContext), U8(0), U8(1),
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
B(PushContext), R(1),
B(Ldar), R(arg0),
Reland x3 "[runtime] Remove extension slots from context objects" Original change's description: > [runtime] Remove extension slots from context objects > > Context objects have an extension slot, which contains further > additional data that depends on the type of the context. > > This CL removes the extension slot from contexts that don't need > them, hence reducing memory. > > The following contexts will still have an extension slot: native, > module, await, block and with contexts. See objects/contexts.h for > what the slot is used for. > The following contexts will not have an extension slot anymore (they > were not used before): script, catch and builtin contexts. > Eval and function contexts only have the extension slot if they > contain a sloppy eval. > > Bug: v8:9744 > Change-Id: I8ca56c22fa02437bbac392ea72174ebfca80e030 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1863191 > Commit-Queue: Victor Gomes <victorgomes@google.com> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Ulan Degenbaev <ulan@chromium.org> > Reviewed-by: Leszek Swirski <leszeks@chromium.org> > Reviewed-by: Peter Marshall <petermarshall@chromium.org> > Auto-Submit: Victor Gomes <victorgomes@google.com> > Cr-Commit-Position: refs/heads/master@{#64372} TBR=verwaest@chromium.org,jgruber@chromium.org,ulan@chromium.org,leszeks@chromium.org,petermarshall@chromium.org Bug: v8:9744 Change-Id: I8700ed2fa62c89e86c39bb16ac3167f38ea8d63f Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873695 Commit-Queue: Victor Gomes <victorgomes@chromium.org> Reviewed-by: Leszek Swirski <leszeks@chromium.org> Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Peter Marshall <petermarshall@chromium.org> Cr-Commit-Position: refs/heads/master@{#64477}
2019-10-22 12:59:24 +00:00
B(StaCurrentContextSlot), U8(2),
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
Reland x3 "[runtime] Remove extension slots from context objects" Original change's description: > [runtime] Remove extension slots from context objects > > Context objects have an extension slot, which contains further > additional data that depends on the type of the context. > > This CL removes the extension slot from contexts that don't need > them, hence reducing memory. > > The following contexts will still have an extension slot: native, > module, await, block and with contexts. See objects/contexts.h for > what the slot is used for. > The following contexts will not have an extension slot anymore (they > were not used before): script, catch and builtin contexts. > Eval and function contexts only have the extension slot if they > contain a sloppy eval. > > Bug: v8:9744 > Change-Id: I8ca56c22fa02437bbac392ea72174ebfca80e030 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1863191 > Commit-Queue: Victor Gomes <victorgomes@google.com> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Ulan Degenbaev <ulan@chromium.org> > Reviewed-by: Leszek Swirski <leszeks@chromium.org> > Reviewed-by: Peter Marshall <petermarshall@chromium.org> > Auto-Submit: Victor Gomes <victorgomes@google.com> > Cr-Commit-Position: refs/heads/master@{#64372} TBR=verwaest@chromium.org,jgruber@chromium.org,ulan@chromium.org,leszeks@chromium.org,petermarshall@chromium.org Bug: v8:9744 Change-Id: I8700ed2fa62c89e86c39bb16ac3167f38ea8d63f Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873695 Commit-Queue: Victor Gomes <victorgomes@chromium.org> Reviewed-by: Leszek Swirski <leszeks@chromium.org> Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Peter Marshall <petermarshall@chromium.org> Cr-Commit-Position: refs/heads/master@{#64477}
2019-10-22 12:59:24 +00:00
/* 53 S> */ B(LdaCurrentContextSlot), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
B(LdaSmi), I8(3),
/* 57 E> */ B(StaNamedPropertyNoFeedback), R(2), U8(1), U8(0),
/* 79 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(2),
/* 86 S> */ B(Return),
]
constant pool: [
SCOPE_INFO_TYPE,
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
---
snippet: "
var f = (function foo(l) { l.a = 3; return arguments.callee; })({});
f;
"
frame size: 3
parameter count: 2
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
bytecode array length: 26
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
bytecodes: [
[interpreter] Make FunctionEntry StackCheck bytecodes implicit FunctionEntry StackChecks is one of the two cases where we generate a StackCheck bytecode. In these cases, we do stack check against the js limit (not to be confused with the real js limit). Their purpose is to be able to interrupt the running code. We can omit the FunctionEntry StackCheck by embedding its code into the InterpreterEntryTrampoline builtin. We save one bytecode per interpreted function. This change has rippling effects for optimized code, as well as the deoptimizer. Bug: v8:10149, v8:9977, v8:9960 Change-Id: I6156de48b3bc0b519dd21190a8e6214fbe96c78d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1914218 Reviewed-by: Jakob Gruber <jgruber@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org> Cr-Commit-Position: refs/heads/master@{#66206}
2020-02-10 16:09:27 +00:00
/* 30 E> */ B(CreateFunctionContext), U8(0), U8(1),
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
B(PushContext), R(1),
B(Ldar), R(arg0),
Reland x3 "[runtime] Remove extension slots from context objects" Original change's description: > [runtime] Remove extension slots from context objects > > Context objects have an extension slot, which contains further > additional data that depends on the type of the context. > > This CL removes the extension slot from contexts that don't need > them, hence reducing memory. > > The following contexts will still have an extension slot: native, > module, await, block and with contexts. See objects/contexts.h for > what the slot is used for. > The following contexts will not have an extension slot anymore (they > were not used before): script, catch and builtin contexts. > Eval and function contexts only have the extension slot if they > contain a sloppy eval. > > Bug: v8:9744 > Change-Id: I8ca56c22fa02437bbac392ea72174ebfca80e030 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1863191 > Commit-Queue: Victor Gomes <victorgomes@google.com> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Ulan Degenbaev <ulan@chromium.org> > Reviewed-by: Leszek Swirski <leszeks@chromium.org> > Reviewed-by: Peter Marshall <petermarshall@chromium.org> > Auto-Submit: Victor Gomes <victorgomes@google.com> > Cr-Commit-Position: refs/heads/master@{#64372} TBR=verwaest@chromium.org,jgruber@chromium.org,ulan@chromium.org,leszeks@chromium.org,petermarshall@chromium.org Bug: v8:9744 Change-Id: I8700ed2fa62c89e86c39bb16ac3167f38ea8d63f Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873695 Commit-Queue: Victor Gomes <victorgomes@chromium.org> Reviewed-by: Leszek Swirski <leszeks@chromium.org> Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Peter Marshall <petermarshall@chromium.org> Cr-Commit-Position: refs/heads/master@{#64477}
2019-10-22 12:59:24 +00:00
B(StaCurrentContextSlot), U8(2),
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
B(CreateMappedArguments),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(0),
Reland x3 "[runtime] Remove extension slots from context objects" Original change's description: > [runtime] Remove extension slots from context objects > > Context objects have an extension slot, which contains further > additional data that depends on the type of the context. > > This CL removes the extension slot from contexts that don't need > them, hence reducing memory. > > The following contexts will still have an extension slot: native, > module, await, block and with contexts. See objects/contexts.h for > what the slot is used for. > The following contexts will not have an extension slot anymore (they > were not used before): script, catch and builtin contexts. > Eval and function contexts only have the extension slot if they > contain a sloppy eval. > > Bug: v8:9744 > Change-Id: I8ca56c22fa02437bbac392ea72174ebfca80e030 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1863191 > Commit-Queue: Victor Gomes <victorgomes@google.com> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Reviewed-by: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Ulan Degenbaev <ulan@chromium.org> > Reviewed-by: Leszek Swirski <leszeks@chromium.org> > Reviewed-by: Peter Marshall <petermarshall@chromium.org> > Auto-Submit: Victor Gomes <victorgomes@google.com> > Cr-Commit-Position: refs/heads/master@{#64372} TBR=verwaest@chromium.org,jgruber@chromium.org,ulan@chromium.org,leszeks@chromium.org,petermarshall@chromium.org Bug: v8:9744 Change-Id: I8700ed2fa62c89e86c39bb16ac3167f38ea8d63f Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873695 Commit-Queue: Victor Gomes <victorgomes@chromium.org> Reviewed-by: Leszek Swirski <leszeks@chromium.org> Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Peter Marshall <petermarshall@chromium.org> Cr-Commit-Position: refs/heads/master@{#64477}
2019-10-22 12:59:24 +00:00
/* 37 S> */ B(LdaCurrentContextSlot), U8(2),
Revert "[interpreter] Short Star bytecode" This reverts commit cf93071c91a932996a4df2e2d8880aca3cffbe28. Reason for revert: Speculative revert because of Mac4 GC stress failure: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac64%20GC%20Stress/16697/overview Original change's description: > [interpreter] Short Star bytecode > > Design doc: > https://docs.google.com/document/d/1g_NExMT78II_KnIYNa9MvyPYIj23qAiFUEsyemY5KRk/edit > > This change adds 16 new interpreter opcodes, kStar0 through kStar15, so > that we can use a single byte to represent the common operation of > storing to a low-numbered register. This generally reduces the quantity > of bytecode generated on web sites by 8-9%. > > In order to not degrade speed, a couple of other changes are required: > > The existing lookahead logic to check for Star after certain other > bytecode handlers is updated to check for these new short Star codes > instead. Furthermore, that lookahead logic is updated to contain its own > copy of the dispatch jump rather than merging control flow with the > lookahead-failed case, to improve branch prediction. > > A bunch of constants use bytecode size in bytes as a proxy for the size > or complexity of a function, and are adjusted downward proportionally to > the decrease in generated bytecode size. > > Other small drive-by fix: update generate-bytecode-expectations to emit > \n instead of \r\n on Windows. > > Change-Id: I6307c2b0f5794a3a1088bb0fb94f6e1615441ed5 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2641180 > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> > Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> > Cr-Commit-Position: refs/heads/master@{#72773} TBR=rmcilroy@chromium.org,mythria@chromium.org,seth.brenith@microsoft.com Change-Id: I0162b9400861b90bacef27cca9aebc8ab9d74c10 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2697350 Reviewed-by: Leszek Swirski <leszeks@chromium.org> Commit-Queue: Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72777}
2021-02-16 15:49:56 +00:00
B(Star), R(2),
Add bytecode-generator test for function in assignments Adding test to confirm that no one-shot optimizations are not done for functions enclosed in parentheses but not immediately invoked in an assignment. Bug: v8:8072 Change-Id: I282132a7cc570b59290f2ec314462be060d48e5a Reviewed-on: https://chromium-review.googlesource.com/c/1238576 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56763}
2018-10-02 15:02:21 +00:00
B(LdaSmi), I8(3),
/* 41 E> */ B(StaNamedPropertyNoFeedback), R(2), U8(1), U8(0),
/* 63 S> */ B(LdaNamedPropertyNoFeedback), R(0), U8(2),
/* 70 S> */ B(Return),
]
constant pool: [
SCOPE_INFO_TYPE,
ONE_BYTE_INTERNALIZED_STRING_TYPE ["a"],
ONE_BYTE_INTERNALIZED_STRING_TYPE ["callee"],
]
handlers: [
]
Reference in New Issue Copy Permalink