Lithium currently supports 3 division-like operations on integral operands: "Normal" division (rounding towards zero), flooring division (rounding towards -Infinity) and modulus calculation (the counterpart for the "normal" division). For divisors which are a power of 2, one can efficiently use some bit fiddling to avoid the actual division for such operations. This CL cleanly splits off these operations into separate Lithium instructions, making the code much more maintainable and more consistent across platforms.
There are 2 basic variations of these bit fiddling algorithms: One involving branches and a seemingly more clever one without branches. Choosing between the two is not as easy as it seems: Benchmarks (and probably real-world) programs seem to favor positive dividends, registers and shifting units are sometimes scarce resources, and branch prediction is quite good in modern processors. Therefore only the "normal" division by a power of 2 is implemented in a branch-free manner, this seems to be the best approach in practice. If this turns out to be wrong, we can easily and locally change this.
R=bmeurer@chromium.org
Review URL: https://codereview.chromium.org/175143002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@19715 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Depending on what we know about the right operand, we basically do 3
different things (and the code is actually structured this way):
* If we statically know that the right operand is a power of 2, we do
some bit fiddling instead of doing a "real" modulus calculation.
This should actually be done on the Hydrogen level, not on the
Lithium level, but this will be a separate CL.
* If type feedback tells us that the right operand is a power of 2, we
do the same as above, but guarded by conditional deoptimization to
make sure that the assumption is still valid. In the long run, we
should make this guard visible on the Hydrogen level to make it
visible for GVN and other optimizations.
* In the general case we only do the minimum steps necessary and don't
try to be too clever, because cleverness actually slows us down on
real-world code.
If we look at the code gerators for LModI, we actually see that we
basically have 3 (4 on ARM) fundamentally different translations. I
don't really like lumping them together, they should probably be
different Lithium instructions. For the time being, I restructured the
generators to make this crystal-clear, at the cost of some duplication
regarding the power-of-2 cases. This will go away when we do the
strength reduction on the Hydrogen level, so I'd like to keep it as it
is for now.
Note that the MIPS part was only slightly restructured, there is still
some work to do there.
R=jkummerow@chromium.org
Review URL: https://codereview.chromium.org/15769010
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@15034 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
We iteratively remove all dead Hydrogen instruction until we reach a fixed point. We consider an instruction dead if it is unused, has no observable side effects and is deletable. The last part of the condition is currently not very nice: We basically have to whitelist "safe" instructions, because we are missing more detailed dependencies and/or more detailed tracking of side effects.
We disable dead code elimination for now in our test runners, because we have tons of poorly written tests which wouldn't test anymore what they are supposed to test with this phase enabled. To get test coverage for dead code elimination itself, we should enable it on a few build bots. This is not really a perfect state, but the best we can do for now.
This patch includes a few const-correctness fixes, most of them were necessary for this CL.
Review URL: https://codereview.chromium.org/11088027
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@12697 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This shaves 416+ KB, just under 1% off the size of the debug d8 executable
on Linux (mostly because the CheckHelper functions for assertions were
getting separate copies for each compilation unit). The difference in
release builds is negligible---a size reduction of 0.1%.
Also, change namespace-level 'static const' variables to remove the static
storage class as it's the default.
R=danno@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/8680013
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@10083 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This implementation extends the internal ObjectHashTable to be able to
hold arbitrary objects (e.g. Smis, Strings, ...) as keys by applying
specialized hashing functions to primitive types. Equality of keys is
defined using the internal SameValue function.
R=rossberg@chromium.org
BUG=v8:1622
TEST=mjsunit/harmony/collections
Review URL: http://codereview.chromium.org/8372027
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9777 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This is reapplying r9501 with this single change which seemed to be causing most (all) of the failures for r9501.
--- a/src/debug.cc
+++ b/src/debug.cc
@@ -2230,6 +2230,7 @@ Debugger::Debugger(Isolate* isolate)
compiling_natives_(false),
is_loading_debugger_(false),
never_unload_debugger_(false),
+ force_debugger_active_(true),
message_handler_(NULL),
debugger_unload_pending_(false),
host_dispatch_handler_(NULL),
R=kmillikin@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org//8337009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9684 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This change will ensure that full code with debug break slots is compiled and activated for all functions which already have activation frames.
This additional handling is only for functions which have activations on the stack, and that activation is of the full code compiled without debug break slots. In that case the full code is recompiled with debug break slots. It is ensured that the full code is compiled generating the exact same instructions - except for the additional debug break slots - as before. The return address on the stack is then patched to continue execution in the new code.
Also fixed SortedListBSearch to actually use the passed comparision function.
R=svenpanne@chromium.org, kmillikin@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org//8050010
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9489 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Use the BitField helper class for the code flags, so that we do not have to
define both a shift and a mask explicitly. This makes changing the flags
layout simpler.
Also, make the 'mask' and 'max' members of BitField into constants, because
they are constant and so that they can be used as constant expressions.
E.g., so they can be used in declaring other const members or in static
asserts.
R=fschneider@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/7787028
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9232 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Duplicate identifier detection must be an early syntax error in strict code,
so errors in otherwise lazily compiled functions must be caught in the
preparser.
Originally introduced in r8541 and reverted in r8542.
Now really compiles on Windows.
Review URL: http://codereview.chromium.org/7782023
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9172 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This is a fix and reapply of r8516 with some comments addressed and more
tests added.
The difference from r8516 is that canonicalization of number literals is
no performed using the same methods as in v8, to avoid false positives/negatives
when detecting duplicates.
Review URL: http://codereview.chromium.org/7193045
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8541 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This change makes the number of successors of a control instruction
configurable with a template parameter and changes the existing instructions
to use it.
To iterate over all successors I added an iterator instead of always calling
First- and SecondSuccessor.
Review URL: http://codereview.chromium.org/7114004
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8262 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Only IA32 version for now. I'll start porting.
Strict mode functions are to get 'undefined' as the receiver when
called with an implicit receiver. Modes are bad! It forces us to have
checks on all function calls.
This change attempts to limit the cost by passing information about
whether or not a call is with an implicit or explicit receiver in ecx
as part of the calling convention. The cost is setting ecx on all
calls and checking ecx on entry to strict mode functions.
Implicit/explicit receiver state has to be maintained by ICs. Various
stubs have to not clobber ecx or save and restore it.
CallFunction stub needs to check if the receiver is implicit when it
doesn't know from the context.
Review URL: http://codereview.chromium.org/7039036
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8040 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Switched to using binary low-level log instead of the textual log used
by the ticks processor. The binary log contains code-related events,
code object names, and their bodies. When writing to the log we ask
glibc to use a larger buffer. To avoid complex processing of the
snapshot log (which is still textual) the serializer emits final
snapshot position to code name mappings that can be quickly be read
without replaying the snapshot log. (This might be useful for the
ticks processor.)
Review URL: http://codereview.chromium.org/6904127
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@7729 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Approximation is done by building a dominators tree for the heap graph.
Dominator nodes and retained sizes are serialized into JSON.
Removed:
- reachable size (it is useless, after all);
- HeapEntryCalculatedData (size is now stored in the node, retaining
paths in a hash map);
Review URL: http://codereview.chromium.org/5154007
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5867 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
The collector class automatically expands to hold the values added to it,
like a List, but doesn't ensure that the backing store is contiguous, which
allows it to avoid copying back and forth as the buffer grows.
This is in preparation for identifyng identical symbols during preparsing.
Review URL: http://codereview.chromium.org/3181036
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5325 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Although algorithmically correct, the compiler would not allow to instantiate
a BitField that uses all 32 bits without warnings about a too large shift
count. As a consequence we were limited to 31 bit values when using BitField.
This happened when instantiating a bitfield BitField<T, shift, size> with
[shift=0, size=32] or [shift=31, size=1] or more general any
[shift=X, size=32-X]
As side-effect of the new implementation the compiler now warns if we ever
try instantiating a bitfield with size 0.
Review URL: http://codereview.chromium.org/606063
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3910 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
The different length string types was used to encode the string length and the hash in one field. This is now split into two fields one for length and one for hash. The hash field still encodes the array index of the string if it has one. If an array index is encoded in the hash field the string length is added to the top bits of the hash field to avoid a hash value of zero.
On 32-bit this causes an additional 4 bytes to be used for all string objects. On 64-bit this will be half on average dur to pointer alignment.
Review URL: http://codereview.chromium.org/436001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3350 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Two techniques are involved:
- compress repeated line ends (common stack beginnings) by using back references;
- do RLE compression of repeated tick events.
This gives only 5% size reduction on benchmarks run, but this is because tick events are only comprise 10% of file size. Under Chromium winnings are bigger because long repeated samples of idleness are now compressed into a single line.
Tickprocessor will be updated in the next patch.
Review URL: http://codereview.chromium.org/123012
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2147 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This issue was raised by Brett Wilson while reviewing my changelist for readability. Craig Silverstein (one of C++ SG maintainers) confirmed that we should declare one namespace per line. Our way of namespaces closing seems not violating style guides (there is no clear agreement on it), so I left it intact.
Review URL: http://codereview.chromium.org/115756
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@2038 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
When loaded scripts are requested this cache is filled with all the script objects in the heap. Hereafter its content is kept in sync with the active scripts in the heap through the notifications of new scripts compiled and by using weak handles to get notified when a script is collected.
Through the tracking of collected scripts the debugger event OnScriptCollected have been added to notify a debugger that a script previously returned through the scripts command is no longer in use.
Make the ComputeIntegerHash globally available.
Moved clearing of the mirror cache to when debugger is really left. Previously recursive invocations of the debugger cause the mirror cache to be cleared causing handles to become either stale or reference other objects.
Review URL: http://codereview.chromium.org/115462
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@1988 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
- TARGET, the architecture we will generate code for.
This is brought it from the build system.
- HOST, the architecture our C++ compiler is building for.
This is detected automatically based on compiler defines.
This adds macros for 32 or 64 bit, and cleans up some
include conditionals, etc.
Review URL: http://codereview.chromium.org/99355
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@1864 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
a reason to stack allocate large chunks of stack space.
- Runtime_GetCFrames used to allocate a frame size of 52040 bytes.
- PreallocatedMemoryThread::Run used to allocate 32784 bytes.
- Fixed StringStream overflow conditions.
Review URL: http://codereview.chromium.org/67197
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@1729 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Cleaned up ARM version by removing top of stack caching and by introducing push/pop elimination.
Cleaned up the way runtime functions are called to allow runtime calls with no arguments.
Changed Windows build options to make sure that exceptions are disabled and that optimization flags are enabled.
Added first version of Visual Studio project files.
git-svn-id: http://v8.googlecode.com/svn/trunk@13 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Fixed the handling of '>' and '<=' to use right-to-left conversion and left-to-right evaluation as specified by ECMA-262.
Fixed a branch elimination bug on the ARM platform where incorrect code was generated because of overly aggressive branch elimination.
Improved performance of code that repeatedly assigns the same function to the same property of different objects with the same map.
Untangled DEBUG and ENABLE_DISASSEMBLER defines. The disassembler no longer expects DEBUG to be defined.
Added platform-nullos.cc to serve as the basis for new platform implementations.
git-svn-id: http://v8.googlecode.com/svn/trunk@9 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
Added a few samples and support for building them. The samples include a simple shell that can be used to benchmark and test V8.
Changed V8::GetVersion to return the version as a string.
Added source for lazily loaded scripts to snapshots and made serialization non-destructive.
Improved ARM support by fixing the write barrier code to use aligned loads and stores and by removing premature locals optimization that relied on broken support for callee-saved registers (removed).
Refactored the code for marking live objects during garbage collection and the code for allocating objects in paged spaces. Introduced an abstraction for the map word of a heap-allocated object and changed the memory allocator to allocate executable memory only for spaces that may contain code objects.
Moved StringBuilder to utils.h and ScopedLock to platform.h, where they can be used by debugging and logging modules. Added thread-safe message queues for dealing with debugger events.
Fixed the source code reported by toString for certain builtin empty functions and made sure that the prototype property of a function is enumerable.
Improved performance of converting values to condition flags in generated code.
Merged disassembler-{arch} files.
git-svn-id: http://v8.googlecode.com/svn/trunk@8 ce2b1a6d-e550-0410-aec6-3dcde31c8c00