[libcxx-commits] [PATCH] D72160: Optimize / partially inline basic_string copy constructor

[Michał Górny via libcxx-commits]

On Mon, 2020-01-20 at 04:31 -0500, Eric Fiselier wrote:
> I don't see why you think the given error maps to this change.
> 
> This change isn't in the set of changes in the build you point to:
> http://lab.llvm.org:8014/builders/netbsd-amd64/builds/834/
> 
> And the build before it doesn't fail.
> http://lab.llvm.org:8014/builders/netbsd-amd64/builds/833/
> 
> Can you elucidate why this change is responsible?
> 

I'm not sure how you infer that.

Build 834 has got_revision a93aa5347641159aa0d2d48dda9e1a51b2273462.

a93aa534764 [lldb/Docs] Fix formatting for the variable formatting page
64c4dcb5eef [mlir][Linalg] Extend linalg vectorization to MatmulOp
a8a9c8e0a11 [libc++] Optimize / partially inline basic_string copy constructor
cd40bd0a32e hwasan: Move .note.hwasan.globals note to hwasan.module_ctor comdat.

So your commit is two commits before that build.  Build 833 has
cd40bd0a32e29fb62c89b120adbc89a847443da3 which confirms the given range.

I've further bisected this and it confirms a8a9c8e0a11 being
the culprit.

To be honest, I have no clue what the scheduler is doing there and why
it's running things out of order.

-- 
Best regards,
Michał Górny

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 618 bytes
Desc: This is a digitally signed message part
URL: <http://lists.llvm.org/pipermail/libcxx-commits/attachments/20200120/df2c6d74/attachment.sig>