If an exception is thrown and the wasm interpreter entry frame is
unwound, also the internal frames in the interpreter need to be unwound.
We did not do so before, leaving a corrupted internal state of the wasm
interpreter. Thus reusing it would fail.
This CL fixes this and adds a test which reenters a previously unwound
wasm interpreter. It checks that this works and the correct stack is
returned.
This test also requires support for calling an imported function which
throws, so this change is also included here.
R=ahaas@chromium.org, titzer@chromium.org
BUG=v8:5822
Change-Id: I12fb843f7a371a4e618b4ac63ed3299667a03a82
Reviewed-on: https://chromium-review.googlesource.com/453938
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43937}
This behaviour was missing before. If a trap is encountered in the
interpreter, we now throw the right error. With test.
R=titzer@chromium.org, ahaas@chromium.org
BUG=v8:5822
Change-Id: I09c23d15fcde32ec586fb6d3094a5ec49155a9a2
Reviewed-on: https://chromium-review.googlesource.com/453839
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43859}
When instantiating the wasm interpreter for debugging, we unwrap all
wasm-to-js wrappers and store the callable objects. The handles are
stored in a DeferredHandleScope and deleted when the InterpreterHandle
(store in WasmDebugInfo) is freed.
A call to an imported function reads the arguments from the stack,
converts them to JS objects, calls the callable, converts back the
return value and pushes it onto the stack.
Reentering the interpreter from the calles JS code is not permitted
yet, but will be in a follow-up CL.
Also, indirect calls to imported functions will have to follow.
R=titzer@chromium.org, ahaas@chromium.org
BUG=v8:5822
Change-Id: I66c35053bccb6cf8d416606e4f840d888ccb3b65
Reviewed-on: https://chromium-review.googlesource.com/453838
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43855}
This is a cleanup in preparation to implement calling imported
functions via the wasm interpreter.
For imported functions, we do not create entries in the
interpreter_code_ vector any more.
I also simplified the interface and removed unused or redundant return
values. More things are now DCHECKed instead of bailing out.
Also, we previously had two PushFrame methods: One is supposed to
initialize the interpreter from external code (i.e. adds the first
frame to the stack), the other one is used to push new frames on the
frame stack for called functions. This CL renames the first to
InitFrame, and makes it use the second one. The other remaining user is
the DoCall method.
R=titzer@chromium.org
BUG=v8:5822
Change-Id: Id09ff1e3256428fbd8c955e4664507a0c3167e53
Reviewed-on: https://chromium-review.googlesource.com/453482
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43793}
Instead of storing call_pc and ret_pc, store just one pc. This
simplifies frame inspection (no distinction between top frame and other
frames). The previous ret_pc can easily be computed from the stored pc,
since we know that we must be at a indirect or direct call site when
returning to a previous frame.
It also slightly simplifies the upcoming CL to call imported functions,
which would also have to set the call_pc.
R=titzer@chromium.org, ahaas@chromium.org
BUG=v8:5822
Change-Id: I5876c09ec36450dc1474a760282fd5e41eab38be
Reviewed-on: https://chromium-review.googlesource.com/453159
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43777}
Test the wasm interpreter entry stub by creating two wasm functions A
and B, make A pass arguments to B, then redirect B to be executed in the
interpreter.
Test different number and types or arguments.
BUG=v8:5822
R=titzer@chromium.org
Review-Url: https://codereview.chromium.org/2651793003
Cr-Commit-Position: refs/heads/master@{#43353}
Apparently it happens quite easily that different NaNs are produced in
the interpreter than in the execution of the compiled code. This
non-determinism caused problems for the fuzzer which compares the
equality of the results of the interpreter and the compiled code.
I decided therefore to refactor the detection of non-determinism in the
interpreter. Instead of tracking whether potentially non-deterministic
NaNs were produced, I track now whether potentially non-deterministic
NaNs could have been observed. The only way the NaN non-determinism can
be observed is by observing the non-deterministic bit pattern of the
NaN. AFAICT the only way to observe the bit pattern is with a
I(32|64)_REINTERPRET_F(32|64) instruction or with a F(32|64)_STORE
followed by a load. Therefore I flag an execution as potentially
non-deterministic when either a NaN is reinterpreted to an int, or when
a NaN is stored to memory.
R=titzer@chromium.org, eholk@chromium.org
BUG=682180
Review-Url: https://codereview.chromium.org/2671803002
Cr-Commit-Position: refs/heads/master@{#42917}
Always execute the implicit return if we fell off the function bytes.
This is not considered an additional "step" as it is not executing a
wasm instruction.
Otherwise, we might pause at an invalid position (one after the
function bytes).
R=titzer@chromium.org
BUG=v8:5822
Review-Url: https://codereview.chromium.org/2650293003
Cr-Commit-Position: refs/heads/master@{#42730}
Implement stepping by remembering the current step action in the wasm
interpreter handle in WasmDebugInfo, and using it when continuing
execution in the interpreter.
The control flow is as follows: After module compilation, the user sets
a breakpoint in wasm. The respective function is redirected to the
interpreter and the breakpoint is set on the interpreter. When it is
hit, we notify all debug event listeners, which might prepare stepping.
When returning from these listeners, before continuing execution, we
check whether stepping was requested and continue execution in the
interpreter accordingly.
Stepping from Wasm to JS and vice versa will be implemented and tested
in a follow-up CL. Testing this requires breakpoints and stepping in
Wasm to be exposed via the inspector interface, such that we can write
an inspector test. This mixed JS-Wasm-execution is hard to set up in a
cctest.
R=titzer@chromium.org, yangguo@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/2649533002
Cr-Commit-Position: refs/heads/master@{#42624}
The hardcoded constant caused a problem for the wasm fuzzer because
when the maximum memory was allocated in a test case, clusterfuzz ran
out of memory. with the command line flag we can set a lower limit
for the fuzzer.
The flag has the value of the constant as its default value, so that
for everything but the fuzzers nothing should change.
R=titzer@chromium.org
BUG=chromium:676888
Review-Url: https://codereview.chromium.org/2626313003
Cr-Commit-Position: refs/heads/master@{#42599}
As required by C++11, this CL changes the zone allocator to be able to
construct and destroy arbitrary types, and accept arbitrary arguments
for construct, passing them via perfect forwarding.
I also change some push_back to emplace_back. Some of those did not
compile before.
R=ishell@chromium.org, titzer@chromium.org
Review-Url: https://codereview.chromium.org/2646873004
Cr-Commit-Position: refs/heads/master@{#42597}
Frame inspection is currently limited to locations of execution.
Further details like local variables or stack content will follow later.
The FrameInspector now stores a pointer to the interpreted wasm frame,
and redirects certain requests there, just as for deoptimized frames.
Hitting breakpoints is now also supported for wasm frames.
R=yangguo@chromium.org, titzer@chromium.org
BUG=v8:5822
Review-Url: https://codereview.chromium.org/2629823003
Cr-Commit-Position: refs/heads/master@{#42551}
Reason for revert:
compilation problems on mips
Original issue's description:
> [wasm] Fix I32ReinterpretF32 and I64ReinterpretF64 on ia32.
>
> On ia32 return statements in C++ automatically convert signalling NaNs
> to quiet NaNs, even when bit_cast is used. This CL removes all uses of
> bit_cast<float> and bit_cast<double> in the wasm compiler and wasm
> interpreter.
>
> R=titzer@chromium.org, clemensh@chromium.org
>
> Review-Url: https://codereview.chromium.org/2639353002
> Cr-Commit-Position: refs/heads/master@{#42512}
> Committed: 7739affa5bTBR=clemensh@chromium.org,titzer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review-Url: https://codereview.chromium.org/2645693003
Cr-Commit-Position: refs/heads/master@{#42514}
On ia32 return statements in C++ automatically convert signalling NaNs
to quiet NaNs, even when bit_cast is used. This CL removes all uses of
bit_cast<float> and bit_cast<double> in the wasm compiler and wasm
interpreter.
R=titzer@chromium.org, clemensh@chromium.org
Review-Url: https://codereview.chromium.org/2639353002
Cr-Commit-Position: refs/heads/master@{#42512}
As no one will ever try to allocate a Thread directly, we can just make
Thread a proxy of ThreadImpl by reinterpret_casting between both types.
This allows to not mention ThreadImpl in the header at all, and to
define it in an anonymous namespace in the implementation, allowing for
more optimizations.
It also saves runtime, as no memory load is needed to forward from
Thread to ThreadImpl, and we do not need to allocate ThreadImpl objects
on the heap.
R=titzer@chromium.org
Review-Url: https://codereview.chromium.org/2643443002
Cr-Commit-Position: refs/heads/master@{#42450}
We used dynamic dispatch before, which has considerable runtime
overhead. Pimpl (a.k.a. opaque pointer) has the same memory overhead
of one additional pointer, but less runtime overhead.
R=titzer@chromium.org
Review-Url: https://codereview.chromium.org/2639013002
Cr-Commit-Position: refs/heads/master@{#42442}
The special treatment is not required anymore because the wasm spec
changed and NaNs do not have to be preserved anymore. In addition, it
may introduce non-determinism which confused the fuzzers.
R=titzer@chromium.org
BUG=chromium:681033
Review-Url: https://codereview.chromium.org/2630293002
Cr-Commit-Position: refs/heads/master@{#42359}
Separated:
- decoding of locals
- loop assignment analysis
- determination of opcode length
as statics that work on a Decoder. Neither need the context of a
Module, and were used in scenarios where one wasn't available either.
Changed BodyLocalDecls to match the usecases for the type. In all but
one (a printer), we want the list (in order of declaration, with
repetitions) of types of locals.
Removed a now-unnecessary constructor for the WasmFullDecoder.
BUG=
Review-Url: https://codereview.chromium.org/2610813009
Cr-Commit-Position: refs/heads/master@{#42115}
This is more renaming work to comply with the naming in the public
design repository. E.g. types are called "value types" and we no longer
refer to ASTs.
R=clemensh@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/2594993002
Cr-Commit-Position: refs/heads/master@{#41891}
The current CHECK/DCHECK implementation fails statically if a signed
value is compared against an unsigned value. The common solution is to
cast on each caller, which is tedious and error-prone (might hide bugs).
This CL implements signed vs. unsigned comparisons by executing up to
two comparisons. For example, if i is int32_t and u is uint_32_t, a
DCHECK_LE(i, u) would create the check
i <= 0 || static_cast<uint32_t>(i) <= u.
For checks against constants, at least one of the checks can be removed
by compiler optimizations.
The tradeoff we have to make is to sometimes silently execute an
additional comparison. And we increase code complexity of course, even
though the usage is just as easy (or even easier) as before.
The compile time impact seems to be minimal:
I ran 3 full compilations for Optdebug on my local machine, one time on
the current ToT, one time with this CL plus http://crrev.com/2524093002.
Before: 143.72 +- 1.21 seconds
Now: 144.18 +- 0.67 seconds
In order to check that the new comparisons are working, I refactored
some DCHECKs in wasm to use the new magic, and added unit test cases.
R=ishell@chromium.org, titzer@chromium.orgCC=ahaas@chromium.org, bmeurer@chromium.org
Committed: https://crrev.com/5925074a9dab5a8577766545b91b62f2c531d3dc
Review-Url: https://codereview.chromium.org/2526783002
Cr-Original-Commit-Position: refs/heads/master@{#41275}
Cr-Commit-Position: refs/heads/master@{#41411}
These byte pointers (module_start and module_end) were only valid
during decoding. During instantiation or execution, they can get
invalidated by garbage collection.
This CL removes them from the WasmModule struct, and introduces a new
ModuleStorage struct as interface to the wasm wire bytes.
Since the storage is often needed together with the ModuleEnv, a new
ModuleStorageEnv struct holds both a ModuleEnv and a ModuleStorage.
The pointers in the ModuleStorage should never escape the live range of
this struct, as they might point into a SeqOneByteString or ArrayBuffer.
Therefore, the WasmInterpreter needs to create its own copy of the
whole module.
Runtime functions that previously used the raw pointers in WasmModule
(leading to memory errors) now have to use the SeqOneByteString in the
WasmCompiledModule.
R=titzer@chromium.org
BUG=chromium:669518
Review-Url: https://codereview.chromium.org/2540133002
Cr-Commit-Position: refs/heads/master@{#41388}
Reason for revert:
Need to revert previous CL because of Android compile error, and this one depends in it.
Original issue's description:
> [base] Define CHECK comparison for signed vs. unsigned
>
> The current CHECK/DCHECK implementation fails statically if a signed
> value is compared against an unsigned value. The common solution is to
> cast on each caller, which is tedious and error-prone (might hide bugs).
> This CL implements signed vs. unsigned comparisons by executing up to
> two comparisons. For example, if i is int32_t and u is uint_32_t, a
> DCHECK_LE(i, u) would create the check
> i <= 0 || static_cast<uint32_t>(i) <= u.
> For checks against constants, at least one of the checks can be removed
> by compiler optimizations.
>
> The tradeoff we have to make is to sometimes silently execute an
> additional comparison. And we increase code complexity of course, even
> though the usage is just as easy (or even easier) as before.
>
> The compile time impact seems to be minimal:
> I ran 3 full compilations for Optdebug on my local machine, one time on
> the current ToT, one time with this CL plus http://crrev.com/2524093002.
> Before: 143.72 +- 1.21 seconds
> Now: 144.18 +- 0.67 seconds
>
> In order to check that the new comparisons are working, I refactored
> some DCHECKs in wasm to use the new magic.
>
> R=bmeurer@chromium.org, titzer@chromium.org
>
> Committed: https://crrev.com/5925074a9dab5a8577766545b91b62f2c531d3dc
> Cr-Commit-Position: refs/heads/master@{#41275}
TBR=ishell@chromium.org,titzer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review-Url: https://codereview.chromium.org/2531533003
Cr-Commit-Position: refs/heads/master@{#41277}
The current CHECK/DCHECK implementation fails statically if a signed
value is compared against an unsigned value. The common solution is to
cast on each caller, which is tedious and error-prone (might hide bugs).
This CL implements signed vs. unsigned comparisons by executing up to
two comparisons. For example, if i is int32_t and u is uint_32_t, a
DCHECK_LE(i, u) would create the check
i <= 0 || static_cast<uint32_t>(i) <= u.
For checks against constants, at least one of the checks can be removed
by compiler optimizations.
The tradeoff we have to make is to sometimes silently execute an
additional comparison. And we increase code complexity of course, even
though the usage is just as easy (or even easier) as before.
The compile time impact seems to be minimal:
I ran 3 full compilations for Optdebug on my local machine, one time on
the current ToT, one time with this CL plus http://crrev.com/2524093002.
Before: 143.72 +- 1.21 seconds
Now: 144.18 +- 0.67 seconds
In order to check that the new comparisons are working, I refactored
some DCHECKs in wasm to use the new magic.
R=bmeurer@chromium.org, titzer@chromium.org
Review-Url: https://codereview.chromium.org/2526783002
Cr-Commit-Position: refs/heads/master@{#41275}
The wasm interpreter crashed because it interpreted the table of
br_table as a table of uint8, but according to the spec it is a table of
varint32. Therefore the wasm interpreter misinterpreted 0x80 0x00 as 128
and not as 0, which caused a crash.
R=tizer@chromium.org
BUG=chromium:660262
TEST=cctest/test-run-wasm/RunWasmInterpreted_Regression_660262
Review-Url: https://codereview.chromium.org/2463063002
Cr-Commit-Position: refs/heads/master@{#40708}
The wasm specification does not fully specify the binary representation
of NaN: the sign bit can be non-deterministic. The wasm-code fuzzer
found a test case where the wasm interpreter and the compiled code
produce a different sign bit for a NaN, and as a consequence they
produce different results.
With this CL the interpreter tracks whether it executed an instruction
which can produce a NaN, which are div and sqrt instructions. The
fuzzer uses this information and compares the result of the interpreter
with the result of the compiled code only if there was no instruction
which could have produced a NaN.
R=titzer@chromium.org
TEST=cctest/test-run-wasm-interpreter/TestMayProduceNaN
BUG=chromium:657481
Review-Url: https://chromiumcodereview.appspot.com/2438603003
Cr-Commit-Position: refs/heads/master@{#40474}
This adds more useful information to the v8-heap-stats tool.
BUG=v8:5489
Review-Url: https://codereview.chromium.org/2394213003
Cr-Commit-Position: refs/heads/master@{#40361}
[0xC] Convert to stack machine semantics.
[0xC] Use section codes instead of names.
[0xC] Add elements section decoding.
[0xC] Decoding of globals section.
[0xC] Decoding of memory section.
[0xC] Decoding of imports section.
[0xC] Decoding of exports section.
[0xC] Decoding of data section.
[0xC] Remove CallImport bytecode.
[0xC] Function bodies have an implicit block.
[0xC] Remove the bottom label from loops.
[0xC] Add signatures to blocks.
[0xC] Remove arities from branches.
Add tests for init expression decoding.
Rework compilation of import wrappers and how they are patched.
Rework function indices in debugging.
Fix ASM->WASM builder for stack machine.
Reorganize asm.js foreign functions due to import indices change.
R=ahaas@chromium.org,rossberg@chromium.org,bradnelson@chromium.org
BUG=chromium:575167
LOG=Y
Committed: https://crrev.com/76eb976a67273b8c03c744f64ad850b0432554b9
Review-Url: https://codereview.chromium.org/2345593003
Cr-Original-Commit-Position: refs/heads/master@{#39678}
Cr-Commit-Position: refs/heads/master@{#39795}
Reason for revert:
Main suspect for tsan:
https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20TSAN/builds/11893
Also changes layout tests:
https://build.chromium.org/p/client.v8.fyi/builders/V8-Blink%20Linux%2064/builds/10036
+mips builder:
https://build.chromium.org/p/client.v8.ports/builders/V8%20Mips%20-%20builder/builds/4032
Original issue's description:
> [wasm] Master CL for Binary 0xC changes.
>
> [0xC] Convert to stack machine semantics.
> [0xC] Use section codes instead of names.
> [0xC] Add elements section decoding.
> [0xC] Decoding of globals section.
> [0xC] Decoding of memory section.
> [0xC] Decoding of imports section.
> [0xC] Decoding of exports section.
> [0xC] Decoding of data section.
> [0xC] Remove CallImport bytecode.
> [0xC] Function bodies have an implicit block.
> [0xC] Remove the bottom label from loops.
> [0xC] Add signatures to blocks.
> [0xC] Remove arities from branches.
> Add tests for init expression decoding.
> Rework compilation of import wrappers and how they are patched.
> Rework function indices in debugging.
> Fix ASM->WASM builder for stack machine.
> Reorganize asm.js foreign functions due to import indices change.
>
> R=ahaas@chromium.org,rossberg@chromium.org,bradnelson@chromium.org
> BUG=chromium:575167
> LOG=Y
>
> Committed: https://crrev.com/76eb976a67273b8c03c744f64ad850b0432554b9
> Cr-Commit-Position: refs/heads/master@{#39678}
TBR=ahaas@chromium.org,bradnelson@chromium.org,mtrofin@chromium.org,rossberg@chromium.org,bradnelson@google.com,titzer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:575167
Review-Url: https://codereview.chromium.org/2361053004
Cr-Commit-Position: refs/heads/master@{#39685}
[0xC] Convert to stack machine semantics.
[0xC] Use section codes instead of names.
[0xC] Add elements section decoding.
[0xC] Decoding of globals section.
[0xC] Decoding of memory section.
[0xC] Decoding of imports section.
[0xC] Decoding of exports section.
[0xC] Decoding of data section.
[0xC] Remove CallImport bytecode.
[0xC] Function bodies have an implicit block.
[0xC] Remove the bottom label from loops.
[0xC] Add signatures to blocks.
[0xC] Remove arities from branches.
Add tests for init expression decoding.
Rework compilation of import wrappers and how they are patched.
Rework function indices in debugging.
Fix ASM->WASM builder for stack machine.
Reorganize asm.js foreign functions due to import indices change.
R=ahaas@chromium.org,rossberg@chromium.org,bradnelson@chromium.org
BUG=chromium:575167
LOG=Y
Review-Url: https://codereview.chromium.org/2345593003
Cr-Commit-Position: refs/heads/master@{#39678}
This is some initial cleanup to keep /src clean. The
AccountingAllocator is actually exclusively used by zones and this
common subfolder makes that more clear.
BUG=v8:5409
Review-Url: https://codereview.chromium.org/2344143003
Cr-Commit-Position: refs/heads/master@{#39558}
