Issue metadata
Sign in to add a comment
|
0.6% regression in sizes at 452880:452892 |
||||||||||||||||||||
Issue descriptionSuspect changes (bisecting to determine which): 5187e38 Revert of build: Enable auto raw pointer deduction check on linux. (patchset #2 id:20001 of https://codereview.chromium.org/2697873004/ ) by vmpstr · 4 hours ago c702a81 Roll clang 289944:295762. by thakis · 4 hours ago 4075255 build: Enable auto raw pointer deduction check on linux. by vmpstr · 5 hours ago
,
Feb 24 2017
Actually, perf bots can't due bisects on sizes. Since vmpstr's change both landed and was reverted in the same range; I'm suspecting the clang roll could be the cause. Regardless, the regression is pretty small. So, if we're fine with minor codegen differences; I suppose we can just WontFix this. thakis: WDYT?
,
Feb 27 2017
Seems likely that this was the last clang roll. WontFix is probably the way to go, but +hans who's been tracking binary size.
,
Feb 28 2017
I had a look but couldn't see anything in my graph. Since the diff is so small, and the range of commits is so large, I don't think this is worth investigating more.
,
Mar 11 2017
,
May 9 2017
|
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by m...@chromium.org
, Feb 24 2017