Add preliminary support for UBsan's vptr in Chromium (-fsanitize=vptr).

Add preliminary support for UBsan's vptr in Chromium (-fsanitize=vptr).

Can be enabled using ubsan_vptr=1, but not working right now due to
1) Clang's missing library and 2) Clang's link bug.

BUG=174801
TBR=kbr@chromium.org, mseaborn@chromium.org
R=glider@chromium.org, inferno@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=280162

[aarya, 2014-06-25 20:08:32]

+Ken for mesa gyp change.

[aarya, 2014-06-25 20:12:09]

And +cc Mark for Nacl change.

[inferno, 2014-06-26 15:36:53]

lgtm

[commit-bot: I haz the power, 2014-06-26 15:38:01]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/byoungyoung@google.com/357753002/1

[earthdok, 2014-06-26 16:48:31]

https://codereview.chromium.org/357753002/diff/1/build/common.gypi
File build/common.gypi (right):

https://codereview.chromium.org/357753002/diff/1/build/common.gypi#newcode2065
build/common.gypi:2065: 'use_allocator%': 'none',
UBSan doesn't replace the allocator. Why are we setting this?

https://codereview.chromium.org/357753002/diff/1/build/common.gypi#newcode2066
build/common.gypi:2066: 'clang_use_chrome_plugins%': 0,  # Temporally turn
plugins off as Ubsan's vptr now relies on custom Clang binaries.
Please either pass this explicitly in GYP_DEFINES (preferable) or at least put a
TODO here.

https://codereview.chromium.org/357753002/diff/1/build/common.gypi#newcode3879
build/common.gypi:3879: ['asan==1 or lsan==1 or tsan==1 or msan==1 or
ubsan_vptr==1', {
Again, why define MEMORY_TOOL_REPLACES_ALLOCATOR?

[aarya, 2014-06-26 16:49:58]

The CQ bit was unchecked by aarya@google.com

[chromium-reviews, 2014-06-26 17:13:55]

Uploaded the patch as you suggested, and could you please review earthdock@?


On Thu, Jun 26, 2014 at 9:48 AM, <earthdok@chromium.org> wrote:

>
> https://codereview.chromium.org/357753002/diff/1/build/common.gypi
> File build/common.gypi (right):
>
> https://codereview.chromium.org/357753002/diff/1/build/
> common.gypi#newcode2065
> build/common.gypi:2065: 'use_allocator%': 'none',
> UBSan doesn't replace the allocator. Why are we setting this?
>
> https://codereview.chromium.org/357753002/diff/1/build/
> common.gypi#newcode2066
> build/common.gypi:2066: 'clang_use_chrome_plugins%': 0,  # Temporally
> turn plugins off as Ubsan's vptr now relies on custom Clang binaries.
> Please either pass this explicitly in GYP_DEFINES (preferable) or at
> least put a TODO here.
>
> https://codereview.chromium.org/357753002/diff/1/build/
> common.gypi#newcode3879
> build/common.gypi:3879: ['asan==1 or lsan==1 or tsan==1 or msan==1 or
> ubsan_vptr==1', {
> Again, why define MEMORY_TOOL_REPLACES_ALLOCATOR?
>
> https://codereview.chromium.org/357753002/
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[earthdok, 2014-06-26 17:23:01]

On 2014/06/26 17:13:55, chromium-reviews wrote:
> Uploaded the patch as you suggested, and could you please review earthdock@?

Actually does ubsan_vptr only work when another sanitizer (ASan/MSan etc) is
also enabled, or can it be used separately?

If the former, then we can drop most of the changes to common.gypi (i.e. no need
to set clang=1 if we assume that e.g. asan=1 has already set it).

Otherwise LGTM.

[inferno, 2014-06-26 17:30:37]

The CQ bit was checked by inferno@chromium.org

[inferno, 2014-06-26 17:30:39]

On 2014/06/26 17:23:01, earthdok wrote:
> On 2014/06/26 17:13:55, chromium-reviews wrote:
> > Uploaded the patch as you suggested, and could you please review earthdock@?
> 
> Actually does ubsan_vptr only work when another sanitizer (ASan/MSan etc) is
> also enabled, or can it be used separately?
> 
> If the former, then we can drop most of the changes to common.gypi (i.e. no
need
> to set clang=1 if we assume that e.g. asan=1 has already set it).
> 
> Otherwise LGTM.

It is expected to be used seperately and that is how we have been using it in
custom build. When clubbed with ASAN, it is a bad perf and memory hit. I will cc
you on Byoungyoung's doc. cqing this one.

[commit-bot: I haz the power, 2014-06-26 17:33:07]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/byoungyoung@google.com/357753002/20001

[Ken Russell (switch to Gerrit), 2014-06-26 17:39:26]

Mesa change LGTM

[commit-bot: I haz the power, 2014-06-26 21:30:47]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  win_chromium_rel on tryserver.chromium
(http://build.chromium.org/p/tryserver.chromium/builders/win_chromium_rel/buil...)

[commit-bot: I haz the power, 2014-06-26 23:31:47]

Change committed as 280162