[llvm] r318323 - [InstCombine] trunc (binop X, C) --> binop (trunc X, C')

[Chandler Carruth via llvm-commits]

On Fri, Dec 8, 2017 at 4:01 PM Sanjay Patel <spatel at rotateright.com> wrote:

> On Fri, Dec 8, 2017 at 4:13 PM, Chandler Carruth <chandlerc at gmail.com>
> wrote:
>
>> On Fri, Dec 8, 2017 at 11:07 PM Sanjay Patel via llvm-commits <
>> llvm-commits at lists.llvm.org> wrote:
>>
>>> This is taking 'ready, fire, aim' too far IMO.
>>>
>>> Reverting this commit won't prevent hitting the bug.
>>>
>>
>> Er, in fact, it will in one important case: a particular version of
>> ffmpeg is impossible to compile with top-of-tree at the moment. Reverting
>> will allow it to compile.
>>
>>
>
> Unless I missed it, this is the first mention of anything that doesn't
> look like a fuzzer test?
>

Just to close the loop some, sorry if I didn't do it clearly enough, I
thought I had indicated that we hit this testing real software (rather than
a fuzzer thing) and were just reducing a test case. But I probably just
didn't spell that out clearly enough. Very sorry for that, it was
definitely the only reason we were considering revert.

The forward fix is great of course. Just wanted to circle back here.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20171212/ba3749ec/attachment.html>