Replace DumpBacktrace with Chromium's StackTrace implementation.

Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4: v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
  /lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]

Committed: https://crrev.com/49c14f63ef1ea94b8d7b5a9dfe939b2dbc02e42e
Cr-Commit-Position: refs/heads/master@{#38717}

[rmcilroy, 2016-08-17 16:30:08]

Description was changed from

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
./out/x64.debug/d8() [0x1c67538]
./out/x64.debug/d8() [0x1ac80b5]
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)+0x3cb)
[0x1ac323b]
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)+0x40)
[0x1ac2570]

You may also see two dumps (one with symbols and the second without) on
an Abort as the second dump will be printed when handling the abort signal
handler.

BUG=
==========

to

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
./out/x64.debug/d8() [0x1c67538]
./out/x64.debug/d8() [0x1ac80b5]
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)+0x3cb)
[0x1ac323b]
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)+0x40)
[0x1ac2570]
==========

[rmcilroy, 2016-08-17 16:30:13]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 16:30:22]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 16:32:43]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 16:32:44]

Dry run: Try jobs failed on following builders:
  v8_android_arm_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_android_arm_compile_rel/...)
  v8_linux64_asan_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_asan_rel_ng/buil...)

[rmcilroy, 2016-08-17 19:47:39]

Description was changed from

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
./out/x64.debug/d8() [0x1c67538]
./out/x64.debug/d8() [0x1ac80b5]
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)+0x3cb)
[0x1ac323b]
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)+0x40)
[0x1ac2570]
==========

to

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
 
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)+0x3cb)
[0x1ac323b]
 
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)+0x40)
[0x1ac2570]
==========

[rmcilroy, 2016-08-17 19:48:30]

Description was changed from

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
 
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)+0x3cb)
[0x1ac323b]
 
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)+0x40)
[0x1ac2570]
==========

to

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
 
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)+0x3cb)
[0x1ac323b]

./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)+0x40)
[0x1ac2570]
==========

[rmcilroy, 2016-08-17 19:48:54]

Description was changed from

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
 
./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)+0x3cb)
[0x1ac323b]

./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)+0x40)
[0x1ac2570]
==========

to

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]
==========

[rmcilroy, 2016-08-17 19:54:12]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 19:54:23]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 19:58:12]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 19:58:13]

Dry run: Try jobs failed on following builders:
  v8_win64_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win64_rel_ng/builds/12610)
  v8_win_compile_dbg on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/2...)
  v8_win_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel_ng/builds/12625)

[rmcilroy, 2016-08-17 20:08:54]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 20:08:58]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 20:10:28]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 20:10:29]

Dry run: Try jobs failed on following builders:
  v8_linux64_avx2_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_avx2_rel_ng/buil...)
  v8_linux_dbg_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_dbg_ng/builds/11087)
  v8_linux_gcc_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_gcc_compile_rel/bu...)
  v8_linux_nodcheck_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel_ng/bu...)
  v8_linux_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_rel_ng/builds/11032)

[rmcilroy, 2016-08-17 20:57:19]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 20:57:24]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 21:00:15]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 21:00:16]

Dry run: Try jobs failed on following builders:
  v8_win_nosnap_shared_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_rel_ng...)
  v8_win_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel_ng/builds/12628)

[rmcilroy, 2016-08-17 21:04:52]

Patchset #7 (id:120001) has been deleted

[rmcilroy, 2016-08-17 21:06:10]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 21:06:21]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[rmcilroy, 2016-08-17 21:08:05]

Description was changed from

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]
==========

to

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
  /lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]
==========

[commit-bot: I haz the power, 2016-08-17 21:09:29]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 21:09:30]

Dry run: Try jobs failed on following builders:
  v8_mac_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/7124)

[rmcilroy, 2016-08-17 21:19:58]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 21:20:08]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 21:23:39]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 21:23:40]

Dry run: Try jobs failed on following builders:
  v8_mac_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/7126)

[rmcilroy, 2016-08-17 21:43:04]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 21:43:16]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 22:17:11]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 22:17:12]

Dry run: This issue passed the CQ dry run.

[rmcilroy, 2016-08-18 07:51:39]

rmcilroy@chromium.org changed reviewers:
+ mstarzinger@chromium.org

[rmcilroy, 2016-08-18 07:51:40]

Michi: could you take a look. This is mostly a copy of the Chromium
implementation, with some unnecessary things removed and then tweaked to avoid
pulling in more of Chromium base library. It gives us Windows dumps and also
signal and uncaught C++ exception dumps on d8. I'll send out a PSA before
landing this.

[Michael Starzinger, 2016-08-18 08:01:51]

mstarzinger@chromium.org changed reviewers:
+ jkummerow@chromium.org, rossberg@chromium.org

[Michael Starzinger, 2016-08-18 08:01:51]

I would be fine with this (integration looks good, didn't look at the actual
implementation within base::debug yet). But please do send out the PSA (maybe
even an RFC) as you suggested. Also I would like to hear Andreas' opinion, as he
wrote the old rudimentary stack trace dumping. Also Jakob might have an opinion
about the build changes.

One question I have: Does the mocking with signals in d8 somehow interfere with
debugging via GDB and friends in any way (i.e. does GDB still pause at the at
exactly the same places when signals happen)?

Andreas: WDYT?
Jakob: WDYT in general? PTAL at build system changes.

[rmcilroy, 2016-08-18 09:23:44]

On 2016/08/18 08:01:51, Michael Starzinger wrote:
> I would be fine with this (integration looks good, didn't look at the actual
> implementation within base::debug yet). But please do send out the PSA (maybe
> even an RFC) as you suggested. Also I would like to hear Andreas' opinion, as
he
> wrote the old rudimentary stack trace dumping. Also Jakob might have an
opinion
> about the build changes.
> 
> One question I have: Does the mocking with signals in d8 somehow interfere
with
> debugging via GDB and friends in any way (i.e. does GDB still pause at the at
> exactly the same places when signals happen)?

It doesn't seem to interfere. Breakpoints are completely unaffected (SIGBREAK
isn't handled by the stack dumper). For signals which are handled, the GDB
signal handlers end up running first, so you get the (e.g.) SEGV at the same
place you would see it otherwise. If you hit continue at this point (which would
usually just kill the program), it then runs the stack dump signal handler which
prints out the (unsymbolized) stack trace, and then re-throws the signal, which
GDB catches again (so you get a second chance at the same SEGV point).

Example gdb backtrace from a induced SEGV in VisitObjectLiteral:

Program received signal SIGSEGV, Segmentation fault.
v8::internal::interpreter::BytecodeGenerator::VisitObjectLiteral (
    this=0x7fffffffbf50, expr=0x243ff70)
    at ../src/interpreter/bytecode-generator.cc:1696
1696	  (*((int*)0))++;
(gdb) bt
#0  v8::internal::interpreter::BytecodeGenerator::VisitObjectLiteral (
    this=0x7fffffffbf50, expr=0x243ff70)
    at ../src/interpreter/bytecode-generator.cc:1696
#1  0x0000000001ac324b in v8::internal::interpreter::BytecodeGenerator::Visit (
    this=0x7fffffffbf50, node=0x243ff70)

...

(gdb) c
Continuing.
Received signal 11 SEGV_MAPERR 000000000000

==== C stack trace ===============================

 [0x000001c7009e]
 [0x000001c70013]

...

 [0x0ee9e0a063a7]
[end of stack trace]

Program received signal SIGSEGV, Segmentation fault.
v8::internal::interpreter::BytecodeGenerator::VisitObjectLiteral (
    this=0x7fffffffbf50, expr=0x243ff70)
    at ../src/interpreter/bytecode-generator.cc:1696
1696	  (*((int*)0))++;
(gdb) c
Continuing.
[Thread 0x7ffff5a20700 (LWP 22778) exited]

...

[Michael Starzinger, 2016-08-18 09:36:07]

On 2016/08/18 09:23:44, rmcilroy wrote:
> On 2016/08/18 08:01:51, Michael Starzinger wrote:
> > I would be fine with this (integration looks good, didn't look at the actual
> > implementation within base::debug yet). But please do send out the PSA
(maybe
> > even an RFC) as you suggested. Also I would like to hear Andreas' opinion,
as
> he
> > wrote the old rudimentary stack trace dumping. Also Jakob might have an
> opinion
> > about the build changes.
> > 
> > One question I have: Does the mocking with signals in d8 somehow interfere
> with
> > debugging via GDB and friends in any way (i.e. does GDB still pause at the
at
> > exactly the same places when signals happen)?
> 
> It doesn't seem to interfere. Breakpoints are completely unaffected (SIGBREAK
> isn't handled by the stack dumper). For signals which are handled, the GDB
> signal handlers end up running first, so you get the (e.g.) SEGV at the same
> place you would see it otherwise. If you hit continue at this point (which
would
> usually just kill the program), it then runs the stack dump signal handler
which
> prints out the (unsymbolized) stack trace, and then re-throws the signal,
which
> GDB catches again (so you get a second chance at the same SEGV point).
> 
> Example gdb backtrace from a induced SEGV in VisitObjectLiteral:
> 
> Program received signal SIGSEGV, Segmentation fault.
> v8::internal::interpreter::BytecodeGenerator::VisitObjectLiteral (
>     this=0x7fffffffbf50, expr=0x243ff70)
>     at ../src/interpreter/bytecode-generator.cc:1696
> 1696	  (*((int*)0))++;
> (gdb) bt
> #0  v8::internal::interpreter::BytecodeGenerator::VisitObjectLiteral (
>     this=0x7fffffffbf50, expr=0x243ff70)
>     at ../src/interpreter/bytecode-generator.cc:1696
> #1  0x0000000001ac324b in v8::internal::interpreter::BytecodeGenerator::Visit
(
>     this=0x7fffffffbf50, node=0x243ff70)
> 
> ...
> 
> (gdb) c
> Continuing.
> Received signal 11 SEGV_MAPERR 000000000000
> 
> ==== C stack trace ===============================
> 
>  [0x000001c7009e]
>  [0x000001c70013]
> 
> ...
> 
>  [0x0ee9e0a063a7]
> [end of stack trace]
> 
> Program received signal SIGSEGV, Segmentation fault.
> v8::internal::interpreter::BytecodeGenerator::VisitObjectLiteral (
>     this=0x7fffffffbf50, expr=0x243ff70)
>     at ../src/interpreter/bytecode-generator.cc:1696
> 1696	  (*((int*)0))++;
> (gdb) c
> Continuing.
> [Thread 0x7ffff5a20700 (LWP 22778) exited]
> 
> ...

Cool. Thanks for elaborating/investigating! This behavior works for me.

[Jakob Kummerow, 2016-08-18 09:36:14]

LGTM with nits.

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn
File BUILD.gn (right):

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2047
BUILD.gn:2047: "src/base/functional.cc",
why is "src/base/free_deleter.h" not required here?

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2092
BUILD.gn:2092: "src/base/debug/stack_trace_posix.cc",
nit: "d" alpha-sorts before "p"

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2114
BUILD.gn:2114: "src/base/debug/stack_trace_android.cc",
again

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2120
BUILD.gn:2120: "src/base/debug/stack_trace_posix.cc",
again

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
File src/base/debug/stack_trace.h (right):

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
src/base/debug/stack_trace.h:1: // Copyright 2016 the V8 project authors. All
rights reserved.
If this is copied from Chromium, it should probably retain its original
copyright header, maybe with an added remark (see src/base/safe_math.h for an
existing example).

https://codereview.chromium.org/2248393002/diff/180001/src/v8.gyp
File src/v8.gyp (right):

https://codereview.chromium.org/2248393002/diff/180001/src/v8.gyp#newcode1825
src/v8.gyp:1825: 'base/debug/stack_trace_posix.cc'
nit: "d" alpha-sorts before "p", again several times below

[rossberg, 2016-08-18 09:45:00]

If this makes stack traces work on Windows I'm all for it (even though I find
the new format unnecessarily verbose).

[Michael Starzinger, 2016-08-18 10:39:19]

LGTM from my side. Just nits.

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
File src/base/debug/stack_trace.h (right):

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
src/base/debug/stack_trace.h:1: // Copyright 2016 the V8 project authors. All
rights reserved.
On 2016/08/18 09:36:13, Jakob wrote:
> If this is copied from Chromium, it should probably retain its original
> copyright header, maybe with an added remark (see src/base/safe_math.h for an
> existing example).

+1. Applies to all five stack-trace files I suppose.

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
src/base/debug/stack_trace.h:5: #ifndef BASE_DEBUG_STACK_TRACE_H_
nit: V8_BASE_DEBUG_STACK_TRACE_H_

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
src/base/debug/stack_trace.h:93: #endif  // BASE_DEBUG_STACK_TRACE_H_
nit: V8_BASE_DEBUG_STACK_TRACE_H_

https://codereview.chromium.org/2248393002/diff/180001/src/base/free_deleter.h
File src/base/free_deleter.h (right):

https://codereview.chromium.org/2248393002/diff/180001/src/base/free_deleter....
src/base/free_deleter.h:5: #ifndef BASE_MEMORY_FREE_DELETER_H_
nit: V8_BASE_FREE_DELETER_H_

https://codereview.chromium.org/2248393002/diff/180001/src/base/free_deleter....
src/base/free_deleter.h:25: #endif  // BASE_MEMORY_FREE_DELETER_H_
nit: V8_BASE_FREE_DELETER_H_

[rmcilroy, 2016-08-18 11:37:05]

The CQ bit was checked by rmcilroy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-18 11:37:13]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[rmcilroy, 2016-08-18 11:39:48]

> If this makes stack traces work on Windows I'm all for it (even though I find
> the new format unnecessarily verbose).

Yeah I'm not super keen on the new format either, but I wanted to keep the
changes from Chromium minimal, and there is some info in the new format that can
be useful (e.g., code offset in the function).

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn
File BUILD.gn (right):

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2047
BUILD.gn:2047: "src/base/functional.cc",
On 2016/08/18 09:36:13, Jakob wrote:
> why is "src/base/free_deleter.h" not required here?

Oops, good catch, thanks. Done

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2092
BUILD.gn:2092: "src/base/debug/stack_trace_posix.cc",
On 2016/08/18 09:36:13, Jakob wrote:
> nit: "d" alpha-sorts before "p"

Done

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2114
BUILD.gn:2114: "src/base/debug/stack_trace_android.cc",
On 2016/08/18 09:36:13, Jakob wrote:
> again

Done.

https://codereview.chromium.org/2248393002/diff/180001/BUILD.gn#newcode2120
BUILD.gn:2120: "src/base/debug/stack_trace_posix.cc",
On 2016/08/18 09:36:13, Jakob wrote:
> again

Done.

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
File src/base/debug/stack_trace.h (right):

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
src/base/debug/stack_trace.h:1: // Copyright 2016 the V8 project authors. All
rights reserved.
On 2016/08/18 10:39:18, Michael Starzinger wrote:
> On 2016/08/18 09:36:13, Jakob wrote:
> > If this is copied from Chromium, it should probably retain its original
> > copyright header, maybe with an added remark (see src/base/safe_math.h for
an
> > existing example).
> 
> +1. Applies to all five stack-trace files I suppose.

Thanks. I changed this since presumbit hooks complained, but the approach in
safe_math.h works. Done.

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
src/base/debug/stack_trace.h:5: #ifndef BASE_DEBUG_STACK_TRACE_H_
On 2016/08/18 10:39:19, Michael Starzinger wrote:
> nit: V8_BASE_DEBUG_STACK_TRACE_H_

Done.

https://codereview.chromium.org/2248393002/diff/180001/src/base/debug/stack_t...
src/base/debug/stack_trace.h:93: #endif  // BASE_DEBUG_STACK_TRACE_H_
On 2016/08/18 10:39:18, Michael Starzinger wrote:
> nit: V8_BASE_DEBUG_STACK_TRACE_H_

Done.

https://codereview.chromium.org/2248393002/diff/180001/src/base/free_deleter.h
File src/base/free_deleter.h (right):

https://codereview.chromium.org/2248393002/diff/180001/src/base/free_deleter....
src/base/free_deleter.h:5: #ifndef BASE_MEMORY_FREE_DELETER_H_
On 2016/08/18 10:39:19, Michael Starzinger wrote:
> nit: V8_BASE_FREE_DELETER_H_

Done.

https://codereview.chromium.org/2248393002/diff/180001/src/base/free_deleter....
src/base/free_deleter.h:25: #endif  // BASE_MEMORY_FREE_DELETER_H_
On 2016/08/18 10:39:19, Michael Starzinger wrote:
> nit: V8_BASE_FREE_DELETER_H_

Done.

https://codereview.chromium.org/2248393002/diff/180001/src/v8.gyp
File src/v8.gyp (right):

https://codereview.chromium.org/2248393002/diff/180001/src/v8.gyp#newcode1825
src/v8.gyp:1825: 'base/debug/stack_trace_posix.cc'
On 2016/08/18 09:36:13, Jakob wrote:
> nit: "d" alpha-sorts before "p", again several times below

Done.

[commit-bot: I haz the power, 2016-08-18 12:00:29]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-18 12:00:30]

Dry run: This issue passed the CQ dry run.

[rmcilroy, 2016-08-18 14:18:12]

The CQ bit was checked by rmcilroy@chromium.org

[rmcilroy, 2016-08-18 14:18:13]

The patchset sent to the CQ was uploaded after l-g-t-m from
mstarzinger@chromium.org, jkummerow@chromium.org
Link to the patchset: https://codereview.chromium.org/2248393002/#ps200001
(title: "Address Comments")

[commit-bot: I haz the power, 2016-08-18 14:18:24]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-18 14:25:26]

Description was changed from

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
  /lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]
==========

to

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
  /lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]
==========

[commit-bot: I haz the power, 2016-08-18 14:25:27]

Committed patchset #10 (id:200001)

[commit-bot: I haz the power, 2016-08-18 14:26:00]

Description was changed from

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
  /lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]
==========

to

==========
Replace DumpBacktrace with Chromium's StackTrace implementation.

Adds support for dumping the stack on Windows. Also enables in-process
stack dumping in d8 to dump the stack on exceptions and signals.

This CL changes the format of stack dumps from:
 1: V8_Fatal
 2: 0x1ac6ba5
 3: v8::internal::interpreter::BytecodeGenerator::Visit(v8::internal::AstNode*)
 4:
v8::internal::interpreter::BytecodeGenerator::VisitForAccumulatorValue(v8::internal::Expression*)
 ...

To:
  ./out/x64.debug/d8(v8::base::debug::StackTrace::StackTrace()+0x1e) [0x1c6ee5e]
  ./out/x64.debug/d8() [0x1c6ede5]
  /lib/x86_64-linux-gnu/libpthread.so.0(+0x10330) [0x7fa01193e330]
  ./out/x64.debug/d8(v8::base::OS::Abort()+0x12) [0x1c6cea2]
  ./out/x64.debug/d8() [0x1c67538]
  ./out/x64.debug/d8() [0x1ac80b5]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::Visit(v8::internal::AstNode*)+0x3cb) [0x1ac323b]
  ./out/x64.debug/d8(v8::internal::interpreter::BytecodeGenerator
::VisitForAccumulatorValue(v8::internal::Expression*)+0x40) [0x1ac2570]

Committed: https://crrev.com/49c14f63ef1ea94b8d7b5a9dfe939b2dbc02e42e
Cr-Commit-Position: refs/heads/master@{#38717}
==========

[commit-bot: I haz the power, 2016-08-18 14:26:02]

Patchset 10 (id:??) landed as
https://crrev.com/49c14f63ef1ea94b8d7b5a9dfe939b2dbc02e42e
Cr-Commit-Position: refs/heads/master@{#38717}