Here's another recent failure: https://uberchromegw.corp.google.com/i/chromium.linux/builders/Deterministic%20Linux/builds/2280
ppapi_nacl_tests_pnacl_newlib_x32_nonsfi.nexe    : DIFFERENT (unexpected): different size: 1853980 != 1977468

This is still occurring -- yyanagisawa, are you responsible for the deterministic build, or should this be assigned to the ppapi team?

Two recent failures:
https://build.chromium.org/p/chromium.linux/builders/Deterministic%20Linux/builds/5899
https://build.chromium.org/p/chromium.linux/builders/Deterministic%20Linux/builds/5892

I think this is caused by non-deterministic compile by pnacl-clang, and I hope PPAPI folks take care of this.

However, since PPAPI is deprecating (https://bugs.chromium.org/p/chromium/issues/detail?id=702990), nobody might want to invest to it.

I suggest to suppress the warning and we won't guarantee the deterministic build, instead of fixing the non-deterministic compile.

What do you think?

I'm ambivalent about this, since this likely has real impact on cache hit rate even if it's deprecated, it's going to be there for a while.

By the way, we might have seen the same behavior with
https://bugs.chromium.org/p/chromium/issues/detail?id=429358  ?

This failed again today (popping up in the Chromium Sheriffing):

https://uberchromegw.corp.google.com/i/chromium.linux/builders/Deterministic%20Linux/builds/6244:

ppapi_nacl_tests_pnacl_newlib_x64.nexe           : DIFFERENT (unexpected): different size: 2610552 != 2808840


and

https://uberchromegw.corp.google.com/i/chromium.linux/builders/Deterministic%20Linux/builds/6243:

ppapi_nacl_tests_pnacl_newlib_x64.nexe           : DIFFERENT (unexpected): different size: 2610552 != 2808840

 Issue 761480  has been merged into this issue.

FYI, I've added ppapi_nacl_tests_pnacl_newlib_x64.nexe back to the linux whitelist, so the errors should stop bothering sheriffs, but the underlying issue still remains.

Thanks!

https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Deterministic%20Linux/15169

Expected diffs:   0
Unexpected diffs: 2
Unexpected files with diffs:
  cctest
  mksnapshot
Unexpected files with no diffs:
  ppapi_nacl_tests_pnacl_newlib_x64.nexe
Checking cctest difference: (1197 deps)
  obj/v8/torque_generated_initializers/builtins-array-from-dsl-gen.o       : 6415 out of 89352 bytes are different (7.18%)
  obj/v8/torque_generated_initializers/builtins-typed-array-from-dsl-gen.o : different size: 102176 != 102192
  obj/v8/torque_generated_initializers/builtins-data-view-from-dsl-gen.o   : 506 out of 164248 bytes are different (0.31%)
  obj/v8/torque_generated_initializers/builtins-test-from-dsl-gen.o        : 270 out of 85256 bytes are different (0.32%)
Checking mksnapshot difference: (1009 deps)
  obj/v8/torque_generated_initializers/builtins-array-from-dsl-gen.o      : 6415 out of 89352 bytes are different (7.18%)
  obj/v8/torque_generated_initializers/builtins-typed-array-from-dsl-gen.o: different size: 102176 != 102192
  obj/v8/torque_generated_initializers/builtins-data-view-from-dsl-gen.o  : 506 out of 164248 bytes are different (0.31%)
  obj/v8/torque_generated_initializers/builtins-test-from-dsl-gen.o       : 270 out of 85256 bytes are different (0.32%)
step returned non-zero exit code: 1

The test fails again?

That looks like a real regression but that's unrelated to this bug, mksnapshot is v8, I don't know about cctest. I'd recommend filing a separate bug and actually marking this bug as WontFix.

The failure occured after https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Deterministic%20Linux/15168
on https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Deterministic%20Linux/15169.

Filed https://bugs.chromium.org/p/chromium/issues/detail?id=867021.