wasm: add flag to dump modules

wasm: add flag to dump modules

Flags --dump_wasm_module and --dump_wasm_module_path=/path/to/folder allow us to run a bunch of tests and capture all of the wasm module files including the ones that come from the .js and .cc tests which are built on the fly, as well as the asm2wasm tests.

The files are all uniquely named `HASH.{ok,failed}.wasm`.

This will be especilly useful for fuzz testing, but could also be used for other tests including non-V8 tests.

For now I manually hacked tools/testrunner/local/execution.py so that tools/run-tests.py can output the modules. We may want to ad a flag to run-tests.py proper if this turns out to be useful.

R=bradnelson@chromium.org, titzer@chromium.org, kcc@chromium.org

Committed: https://crrev.com/ed76b66aabf3693dcea19277d6eedfa792e6c2dd
Cr-Commit-Position: refs/heads/master@{#34968}

[JF, 2016-03-19 00:47:59]

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

[commit-bot: I haz the power, 2016-03-19 00:48:04]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1816583003/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1816583003/1

[commit-bot: I haz the power, 2016-03-19 00:56:14]

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

[commit-bot: I haz the power, 2016-03-19 00:56:15]

Dry run: Try jobs failed on following builders:
  v8_mac_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel/builds/17308)

[bradnelson, 2016-03-19 00:57:11]

lgtm

https://codereview.chromium.org/1816583003/diff/1/src/base/platform/platform-...
File src/base/platform/platform-posix.cc (right):

https://codereview.chromium.org/1816583003/diff/1/src/base/platform/platform-...
src/base/platform/platform-posix.cc:435: return ch == DirectorySeparator();
Your trust in the compiler astounds me :-)

[JF, 2016-03-19 01:04:48]

Why does Mac build with C++0x?

FAILED: /b/build/goma/gomacc
/b/build/slave/mac/build/v8/third_party/llvm-build/Release+Asserts/bin/clang++
-MMD -MF obj/src/wasm/v8_base.module-decoder.o.d -DCR_CLANG_REVISION=263324-1
-DV8_TARGET_ARCH_IA32 -DENABLE_GDB_JIT_INTERFACE -DV8_DEPRECATION_WARNINGS
-DV8_IMMINENT_DEPRECATION_WARNINGS -DV8_I18N_SUPPORT
-DV8_USE_EXTERNAL_STARTUP_DATA -DDEBUG -DICU_UTIL_DATA_IMPL=ICU_UTIL_DATA_STATIC
-DU_USING_ICU_NAMESPACE=0 -DU_ENABLE_DYLOAD=0 -DU_STATIC_IMPLEMENTATION
-DENABLE_HANDLE_ZAPPING -I../.. -I../../.. -I../../third_party/icu/source/i18n
-I../../third_party/icu/source/common -isysroot
/Applications/Xcode5.1.1.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.10.sdk
-O3 -gdwarf-2 -fstrict-aliasing -fvisibility=hidden -Werror
-mmacosx-version-min=10.5 -arch i386 -Wall -Wendif-labels -Wno-unused-parameter
-Wno-missing-field-initializers -Wno-gnu-zero-variadic-macro-arguments
-std=gnu++0x -fno-rtti -fno-exceptions -fvisibility-inlines-hidden
-fno-threadsafe-statics -fno-strict-aliasing -fcolor-diagnostics  -c
../../src/wasm/module-decoder.cc -o obj/src/wasm/v8_base.module-decoder.o
../../src/wasm/module-decoder.cc:62:65: error: no member named 'back' in
'std::basic_string<char>'
        if (path.size() && !base::OS::isDirectorySeparator(path.back())) {
                                                           ~~~~ ^
../../src/wasm/module-decoder.cc:67:24: error: no member named 'to_string' in
namespace 'std'
      auto name = std::to_string(base::hash_range(module->module_start,
                  ~~~~~^

https://codereview.chromium.org/1816583003/diff/1/src/base/platform/platform-...
File src/base/platform/platform-posix.cc (right):

https://codereview.chromium.org/1816583003/diff/1/src/base/platform/platform-...
src/base/platform/platform-posix.cc:435: return ch == DirectorySeparator();
On 2016/03/19 00:57:11, bradnelson wrote:
> Your trust in the compiler astounds me :-)

?

Comparing two chars?

[JF, 2016-03-19 01:12:03]

On 2016/03/19 01:04:48, JF wrote:
> Why does Mac build with C++0x?

Trying to answer my own question: https://codereview.chromium.org/1820583002

Seems like it was just there for a while, there are a few comments in the source
that seem out of date. We'll see how that works out.

[titzer, 2016-03-21 08:14:37]

lgtm other than comment below

https://codereview.chromium.org/1816583003/diff/1/src/wasm/module-decoder.cc
File src/wasm/module-decoder.cc (right):

https://codereview.chromium.org/1816583003/diff/1/src/wasm/module-decoder.cc#...
src/wasm/module-decoder.cc:58: if (FLAG_dump_wasm_module) {
Can you factor out the dumping logic into its own method and inline
DecodeModuleImpl back into its caller (i.e. just introduce the dumping method)?

[JF, 2016-03-21 20:11:30]

https://codereview.chromium.org/1816583003/diff/1/src/wasm/module-decoder.cc
File src/wasm/module-decoder.cc (right):

https://codereview.chromium.org/1816583003/diff/1/src/wasm/module-decoder.cc#...
src/wasm/module-decoder.cc:58: if (FLAG_dump_wasm_module) {
On 2016/03/21 08:14:37, titzer wrote:
> Can you factor out the dumping logic into its own method and inline
> DecodeModuleImpl back into its caller (i.e. just introduce the dumping
method)?

Done.

Inlining DecodeModuleImpl is slightly ugly because there are 3 exit points, but
it works. I just figured outline was more natural than copy/paste or goto.

[JF, 2016-03-21 20:11:40]

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

[commit-bot: I haz the power, 2016-03-21 20:11:44]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1816583003/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1816583003/60001

[commit-bot: I haz the power, 2016-03-21 20:16:54]

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

[commit-bot: I haz the power, 2016-03-21 20:16:55]

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

[JF, 2016-03-21 20:25:48]

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

[commit-bot: I haz the power, 2016-03-21 20:25:54]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1816583003/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1816583003/80001

[commit-bot: I haz the power, 2016-03-21 20:30:03]

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

[commit-bot: I haz the power, 2016-03-21 20:30:04]

Dry run: Try jobs failed on following builders:
  v8_win_nosnap_shared_compile_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_nosnap_shared_compil...)

[JF, 2016-03-21 20:34:25]

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

[commit-bot: I haz the power, 2016-03-21 20:34:33]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1816583003/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1816583003/100001

[commit-bot: I haz the power, 2016-03-21 21:01:24]

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

[commit-bot: I haz the power, 2016-03-21 21:01:25]

Dry run: This issue passed the CQ dry run.

[JF, 2016-03-21 21:39:01]

The CQ bit was checked by jfb@chromium.org

[JF, 2016-03-21 21:39:02]

The patchset sent to the CQ was uploaded after l-g-t-m from titzer@chromium.org,
bradnelson@chromium.org
Link to the patchset: https://codereview.chromium.org/1816583003/#ps100001
(title: "Windows has sprintf_s, same as in ostream.cc")

[commit-bot: I haz the power, 2016-03-21 21:39:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1816583003/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1816583003/100001

[JF, 2016-03-21 21:39:28]

I'll commit for now, titzer let me know if you'd want to to do any follow-ups.

[commit-bot: I haz the power, 2016-03-21 21:41:31]

Committed patchset #6 (id:100001)

[commit-bot: I haz the power, 2016-03-21 21:43:09]

Description was changed from

==========
wasm: add flag to dump modules

Flags --dump_wasm_module and --dump_wasm_module_path=/path/to/folder allow us to
run a bunch of tests and capture all of the wasm module files including the ones
that come from the .js and .cc tests which are built on the fly, as well as the
asm2wasm tests.

The files are all uniquely named `HASH.{ok,failed}.wasm`.

This will be especilly useful for fuzz testing, but could also be used for other
tests including non-V8 tests.

For now I manually hacked tools/testrunner/local/execution.py so that
tools/run-tests.py can output the modules. We may want to ad a flag to
run-tests.py proper if this turns out to be useful.

R=bradnelson@chromium.org, titzer@chromium.org, kcc@chromium.org
==========

to

==========
wasm: add flag to dump modules

Flags --dump_wasm_module and --dump_wasm_module_path=/path/to/folder allow us to
run a bunch of tests and capture all of the wasm module files including the ones
that come from the .js and .cc tests which are built on the fly, as well as the
asm2wasm tests.

The files are all uniquely named `HASH.{ok,failed}.wasm`.

This will be especilly useful for fuzz testing, but could also be used for other
tests including non-V8 tests.

For now I manually hacked tools/testrunner/local/execution.py so that
tools/run-tests.py can output the modules. We may want to ad a flag to
run-tests.py proper if this turns out to be useful.

R=bradnelson@chromium.org, titzer@chromium.org, kcc@chromium.org

Committed: https://crrev.com/ed76b66aabf3693dcea19277d6eedfa792e6c2dd
Cr-Commit-Position: refs/heads/master@{#34968}
==========

[commit-bot: I haz the power, 2016-03-21 21:43:10]

Patchset 6 (id:??) landed as
https://crrev.com/ed76b66aabf3693dcea19277d6eedfa792e6c2dd
Cr-Commit-Position: refs/heads/master@{#34968}

[JF, 2016-03-22 18:02:04]

Follow-up to hook this to run-tests.py:
  https://codereview.chromium.org/1828433002