Skip to content

V
V8
Switch branch/tag
  1. 22 Jun, 2017 1 commit
  3. 21 Jun, 2017 1 commit
  5. 20 Jun, 2017 1 commit
  7. 26 May, 2017 1 commit
  9. 23 May, 2017 1 commit
  11. 20 Mar, 2017 1 commit
    • leszeks's avatar
      [disasm] Make jump target printing perf friendly · dc789377
      leszeks authored 
      Makes disassembly jump target printing look more like the output of
objdump, for compatibility with perf's jump arrows. This includes
swapping the order of address and offset, and making the offset and line
numbers hex.

As a drive-by, print comment lines in objdump-v8 so that they can be
shown/hidden as "source" lines by perf.

Review-Url: https://codereview.chromium.org/2757263002
Cr-Commit-Position: refs/heads/master@{#43940}
      dc789377
  13. 17 Mar, 2017 1 commit
    • neis's avatar
      Disentangle assembler from isolate. · 94b088ca
      neis authored 
      This is a first step towards moving Turbofan code generation off the main thread.

Summary of the changes:
- AssemblerBase no longer has a pointer to the isolate. Instead, its
  constructor receives the few things that it needs from the isolate (on most
  architectures this is just the serializer_enabled flag).
- RelocInfo no longer has a pointer to the isolate. Instead, the functions
  that need it take it as an argument.  (There are currently still a few that
  implicitly access the isolate through a HeapObject.)
- The MacroAssembler now explicitly holds a pointer to the isolate (before, it
  used to get it from the Assembler).
- The jit_cookie also moved from AssemblerBase to the MacroAssemblers, since
  it's not used at all in the Assemblers.
- A few architectures implemented parts of the Assembler with the help
  of a Codepatcher that is based on MacroAssembler.  Since the Assembler no
  longer has the isolate, but the MacroAssembler still needs it, this doesn't
  work anymore.  Instead, these Assemblers now use a new PatchingAssembler.

BUG=v8:6048

Review-Url: https://codereview.chromium.org/2732273003
Cr-Commit-Position: refs/heads/master@{#43890}
      94b088ca
  15. 15 Mar, 2017 1 commit
  17. 23 Feb, 2017 1 commit
  19. 07 Feb, 2017 1 commit
  21. 14 Nov, 2016 1 commit
    • tebbi's avatar
      This CL enables precise source positions for all V8 compilers. It merges... · c3a6ca68
      tebbi authored 
      This CL enables precise source positions for all V8 compilers. It merges compiler::SourcePosition and internal::SourcePosition to a single class used throughout the codebase. The new internal::SourcePosition instances store an id identifying an inlined function in addition to a script offset.
SourcePosition::InliningId() refers to a the new table DeoptimizationInputData::InliningPositions(), which provides the following data for every inlining id:
 - The inlined SharedFunctionInfo as an offset into DeoptimizationInfo::LiteralArray
 - The SourcePosition of the inlining. Recursively, this yields the full inlining stack.
Before the Code object is created, the same information can be found in CompilationInfo::inlined_functions().

If SourcePosition::InliningId() is SourcePosition::kNotInlined, it refers to the outer (non-inlined) function.
So every SourcePosition has full information about its inlining stack, as long as the corresponding Code object is known. The internal represenation of a source position is a positive 64bit integer.

All compilers create now appropriate source positions for inlined functions. In the case of Turbofan, this required using AstGraphBuilderWithPositions for inlined functions too. So this class is now moved to a header file.

At the moment, the additional information in source positions is only used in --trace-deopt and --code-comments. The profiler needs to be updated, at the moment it gets the correct script offsets from the deopt info, but the wrong script id from the reconstructed deopt stack, which can lead to wrong outputs. This should be resolved by making the profiler use the new inlining information for deopts.

I activated the inlined deoptimization tests in test-cpu-profiler.cc for Turbofan, changing them to a case where the deopt stack and the inlining position agree. It is currently still broken for other cases.

The following additional changes were necessary:
 - The source position table (internal::SourcePositionTableBuilder etc.) supports now 64bit source positions. Encoding source positions in a single 64bit int together with the difference encoding in the source position table results in very little overhead for the inlining id, since only 12% of the source positions in Octane have a changed inlining id.
 - The class HPositionInfo was effectively dead code and is now removed.
 - SourcePosition has new printing and information facilities, including computing a full inlining stack.
 - I had to rename compiler/source-position.{h,cc} to compiler/compiler-source-position-table.{h,cc} to avoid clashes with the new src/source-position.cc file.
 - I wrote the new wrapper PodArray for ByteArray. It is a template working with any POD-type. This is used in DeoptimizationInputData::InliningPositions().
 - I removed HInlinedFunctionInfo and HGraph::inlined_function_infos, because they were only used for the now obsolete Crankshaft inlining ids.
 - Crankshaft managed a list of inlined functions in Lithium: LChunk::inlined_functions. This is an analog structure to CompilationInfo::inlined_functions. So I removed LChunk::inlined_functions and made Crankshaft use CompilationInfo::inlined_functions instead, because this was necessary to register the offsets into the literal array in a uniform way. This is a safe change because LChunk::inlined_functions has no other uses and the functions in CompilationInfo::inlined_functions have a strictly longer lifespan, being created earlier (in Hydrogen already).

BUG=v8:5432

Review-Url: https://codereview.chromium.org/2451853002
Cr-Commit-Position: refs/heads/master@{#40975}
      c3a6ca68
  23. 25 Jul, 2016 1 commit
  25. 18 Jul, 2016 1 commit
    • bmeurer's avatar
      [turbofan] Add support for eager/soft deoptimization reasons. · db635d5b
      bmeurer authored 
      So far TurboFan wasn't adding the deoptimization reasons for eager/soft
deoptimization exits that can be used by either the DevTools profiler or
the --trace-deopt flag. This adds basic support for deopt reasons on
Deoptimize, DeoptimizeIf and DeoptimizeUnless nodes and threads through
the reasons to the code generation.

Also moves the DeoptReason to it's own file (to resolve include cycles)
and drops unused reasons.

R=jarin@chromium.org

Review-Url: https://codereview.chromium.org/2161543002
Cr-Commit-Position: refs/heads/master@{#37823}
      db635d5b
  27. 29 Jun, 2016 1 commit
  29. 16 Jun, 2016 1 commit
    • ishell's avatar
      [ic] LoadICState cleanup. · 5fcd3eb8
      ishell authored 
      LoadICState was used to hold the TypeofMode flag which is relevant only for LoadGlobalIC.
This CL removes usage of this state from LoadIC and KeyedLoadIC and renames the state
class to LoadGlobalICState.

BUG=chromium:576312
LOG=Y

Review-Url: https://codereview.chromium.org/2065373003
Cr-Commit-Position: refs/heads/master@{#37033}
      5fcd3eb8
  31. 14 Jun, 2016 1 commit
    • ishell's avatar
      [ic] Split LoadIC into LoadGlobalIC and LoadIC. · d9e8764f
      ishell authored 
      The former will handle loads of predeclared global variables (vars and
functions), lets, consts and undeclared variables. The latter will handle
named loads from explicit receiver. In addition, named loads does not
depend of the TypeofMode.

TypeofMode related cleanup will be done in the follow-up CL.

BUG=chromium:576312
LOG=Y
TBR=bmeurer@chromium.org

Review-Url: https://codereview.chromium.org/1912633002
Cr-Commit-Position: refs/heads/master@{#36965}
      d9e8764f
  33. 09 Jun, 2016 1 commit
    • ishell's avatar
      [ic] [stubs] Remove InlineCacheState field from the code flags. · 9dc62d27
      ishell authored 
      There are no ICs left that store their state in this field: vector based
ICs use feedback vector and the rest three (BinaryOpIC, CompareIC and
ToBooleanIC) reconstruct their state from the ExtraICState field.

This CL also removes unused InlineCacheState::DEBUG_STUB which was used
mostly in Code::is_debug_stub(). The latter now checks if the code is one
of the debug builtins instead.

BUG=chromium:618701
LOG=Y

Review-Url: https://codereview.chromium.org/2052763003
Cr-Commit-Position: refs/heads/master@{#36871}
      9dc62d27
  35. 24 May, 2016 1 commit
  37. 13 May, 2016 1 commit
  39. 11 May, 2016 1 commit
    • mstarzinger's avatar
      [compiler] Pass inlining_id via relocation info. · 32049620
      mstarzinger authored 
      This passes the inlining_id of deoptimization points via the relocation
info instead of via a side-channel to the CPU profiler. This is one step
towards deprecating the side-channel in question and avoid the need for
performing a lookup of the return address of the deopt point.

R=jarin@chromium.org

Review-Url: https://codereview.chromium.org/1956693002
Cr-Commit-Position: refs/heads/master@{#36177}
      32049620
  41. 12 Apr, 2016 1 commit
  43. 11 Apr, 2016 2 commits
  45. 08 Apr, 2016 2 commits
    • jfb's avatar
      Revert of Fix printf formats (patchset #8 id:140001 of... · 4c4fdc2d
      jfb authored 
      Revert of Fix printf formats (patchset #8 id:140001 of https://codereview.chromium.org/1869433004/ )

Reason for revert:
One small issue easily fixed here: https://codereview.chromium.org/1867333003/

But it looks like MSVS 2013 doesn't like some of the formats and exists with the unhelpful:
Stderr:
f:\dd\vctools\crt\crtw32\stdio\output.c(1125) : Assertion failed: ("Incorrect
format specifier", 0)

It's easier to revert for now, I'll dig more into the docs:
https://msdn.microsoft.com/en-us/library/56e442dc(v=vs.120).aspx
https://msdn.microsoft.com/en-us/library/tcxf1dw6(v=vs.120).aspx

And then resubmit, making sure I run these bots.

Original issue's description:
> Fix printf formats
>
> The usage of __attribute__((format(x, y)) was either wrong or missing from multiple functions, leading to erroneous formats. This CL:
>
>  - Imports PRINTF_FORMAT macro from Chrome's src/base/compiler-specific.h.
>  - Uses it appropriately.
>  - Imports Chrome's base/format_macros.h mainly to fix size_t formats (further cleanup could be done).
>  - Fixes a bunch of incorrect formats.
>
> R= jochen@chromium.org, bmeurer@chromium.org, yangguo@chromium.org, ahaas@chromium.org
>
> Committed: https://crrev.com/6ebf9fbb93d31f9be41156a3325d58704ed4933d
> Cr-Commit-Position: refs/heads/master@{#35365}

TBR=jochen@chromium.org,bmeurer@chromium.org,yangguo@chromium.org,ahaas@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/1867383002

Cr-Commit-Position: refs/heads/master@{#35366}
      4c4fdc2d
    • jfb's avatar
      Fix printf formats · 6ebf9fbb
      jfb authored 
      The usage of __attribute__((format(x, y)) was either wrong or missing from multiple functions, leading to erroneous formats. This CL:

 - Imports PRINTF_FORMAT macro from Chrome's src/base/compiler-specific.h.
 - Uses it appropriately.
 - Imports Chrome's base/format_macros.h mainly to fix size_t formats (further cleanup could be done).
 - Fixes a bunch of incorrect formats.

R= jochen@chromium.org, bmeurer@chromium.org, yangguo@chromium.org, ahaas@chromium.org

Review URL: https://codereview.chromium.org/1869433004

Cr-Commit-Position: refs/heads/master@{#35365}
      6ebf9fbb
  47. 06 Apr, 2016 1 commit
    • verwaest's avatar
      Use a dictionary-mode code cache on the map rather than a dual system. · d2eb555e
      verwaest authored 
      The previous code cache system required stubs to be marked with a StubType, causing them to be inserted either into a fixed array or into a dictionary-mode code cache. This could cause names to be in both cases, and lookup would just find the "fast" one first. Given that we clear out the caches on each GC, the memory overhead shouldn't be too bad. Additionally, the dictionary itself should just stay linear for small arrays; that's faster anyway.

This CL additionally deletes some dead IC code.

BUG=

Review URL: https://codereview.chromium.org/1846963002

Cr-Commit-Position: refs/heads/master@{#35291}
      d2eb555e
  49. 01 Mar, 2016 1 commit
  51. 30 Nov, 2015 1 commit
  53. 27 Nov, 2015 1 commit
  55. 09 Nov, 2015 1 commit
    • mtrofin's avatar
      [assembler] Introduce proper AssemblerBase::Print() for improved debuggability. · ab1d270a
      mtrofin authored 
      While working on frame elision, I wanted to disassemble codegen in the
debugger, as the code generation is progressing. I discovered we had a
 "Print" member on the x64 assembler, without any implementation. I
pulled it up to AssemblerBase and gave it an implementation that
should work for the other architectures.

Also checked that ia32, x87, arm and arm64 assemblers didn't have
such an implementation - free Print.

Arm64 has a naming conflict with the v8::internal::Disassembler. I
renamed the arm64 type with a more specific name.

Opportunistically fixed a bug in the name converter. This debug-time
printer doesn't provide a Code object, which should be OK with the
name converters, by the looks of other APIs there. All this means is that
when using the Print() API, we just get addresses dumped without any
context (like what this address may be - a stub maybe, etc). This seems
fine for the scenario.

There may be other places that assume a Code object. Since this is
a diagnostics-only scenario, for codegen developers, I feel it is
reasonable to fix such other places as we find them.

Review URL: https://codereview.chromium.org/1431933003

Cr-Commit-Position: refs/heads/master@{#31869}
      ab1d270a
  57. 24 Aug, 2015 1 commit
    • rmcilroy's avatar
      Add CompileInfo::GetDebugName() · 53ac9fe8
      rmcilroy authored 
      Replaces all instances of the code which computed the debug
name of a stub or function with calls to CompileInfo::GetDebugName instead.

Also:
  - Removes useless parameter on CodeStub::GetMajorName
  - Removes FakeStubForTesting since it is no longer required
  - Adds CompileInfo::ShouldEnsureSpaceForLazyDeopt() to replace unclear calls to IsStub().

Review URL: https://codereview.chromium.org/1297203002

Cr-Commit-Position: refs/heads/master@{#30324}
      53ac9fe8
  59. 14 Aug, 2015 1 commit
  61. 31 Jul, 2015 1 commit
  63. 13 Jul, 2015 2 commits
  65. 10 Jul, 2015 1 commit
    • yangguo's avatar
      Debugger: use debug break slot to break on call. · 8965b683
      yangguo authored 
      Break point at calls are currently set via IC. To change this, we
need to set debug break slots instead. We also need to distinguish
those debug break slots as calls to support step-in.

To implement this, we add a data field to debug break reloc info to
indicate non-call debug breaks or in case of call debug breaks, the
number of arguments. We can later use this to find the callee on the
evaluation stack in Debug::PrepareStep.

BUG=v8:4269
R=ulan@chromium.org
LOG=N

Review URL: https://codereview.chromium.org/1222093007

Cr-Commit-Position: refs/heads/master@{#29561}
      8965b683
  67. 03 Jul, 2015 1 commit
  69. 16 Jun, 2015 1 commit
    • mstarzinger's avatar
      Revert of Decompiler improvements. (patchset #2 id:20001 of... · e2ce4681
      mstarzinger authored 
      Revert of Decompiler improvements. (patchset #2 id:20001 of https://codereview.chromium.org/1177123002/)

Reason for revert:
Code printout has become unreadable. Offsets are printed in decimal numbers everywhere else. This is inconsistent with the rest of the code-base. Some examples are tables for deoptimization data, safepoints and exception handlers. I would be fine with this change if _all_ tracing would be adapted. But there are _many_ places to touch.

Original issue's description:
> Decompiler improvements.
>
> The main motivation is simplifying profiling activities:
>
> 1) Use hex instead of decimal for offsets, just like perf does. This
> affects --print-opt-code
>
> 2) When printing block information, indicate loop information: if
> block is header, where the end is; if block is in a loop, where the
> loop starts. This affects --code-comments.
>
> Using --print-opt-code --code-comments, and cross-referencing with data
> obtained from perf, one may now find the block a hotspot belongs to
> without needing to do hex2dec/dec2hex conversions. Once found, loop info
> is available locally, on the block.
>
> BUG=
>
> Committed: https://crrev.com/32f4bd659d38eb5485eedb1d0dd236ff1bdc01d5
> Cr-Commit-Position: refs/heads/master@{#28964}

TBR=jarin@chromium.org,stichnot@chromium.org,jvoung@chromium.org,mtrofin@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=

Review URL: https://codereview.chromium.org/1188093002

Cr-Commit-Position: refs/heads/master@{#29046}
      e2ce4681
  71. 11 Jun, 2015 1 commit
    • mtrofin's avatar
      Decompiler improvements. · 32f4bd65
      mtrofin authored 
      The main motivation is simplifying profiling activities:

1) Use hex instead of decimal for offsets, just like perf does. This
affects --print-opt-code

2) When printing block information, indicate loop information: if
block is header, where the end is; if block is in a loop, where the
loop starts. This affects --code-comments.

Using --print-opt-code --code-comments, and cross-referencing with data
obtained from perf, one may now find the block a hotspot belongs to
without needing to do hex2dec/dec2hex conversions. Once found, loop info
is available locally, on the block.

BUG=

Review URL: https://codereview.chromium.org/1177123002

Cr-Commit-Position: refs/heads/master@{#28964}
      32f4bd65
  73. 01 Jun, 2015 1 commit