-
Notifications
You must be signed in to change notification settings - Fork 11k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[X86] Worse code generation from patch 24780e13e5be1501e34330148137a10fa9965166 #61923
Labels
Comments
@RKSimon can you please take a look? |
Reduced the test (links updated). |
This will be fixed by https://reviews.llvm.org/D147452 |
@llvm/issue-subscribers-backend-x86 |
gysit
pushed a commit
to nextsilicon/llvm-project
that referenced
this issue
Apr 27, 2023
We see increased number of assembly instructions after patch 24780e1 for this test. See details at llvm#61923.
gysit
pushed a commit
to nextsilicon/llvm-project
that referenced
this issue
Apr 27, 2023
… bitcast(<X x i1> V)) canonicalization This already exists in InstCombine but was missing from the late stage ExpandReductions pass Fixes llvm#53419 Fixes llvm#61923 Differential Revision: https://reviews.llvm.org/D147452
DianQK
pushed a commit
to DianQK/llvm-project
that referenced
this issue
Oct 10, 2023
We see increased number of assembly instructions after patch 24780e1 for this test. See details at llvm#61923.
DianQK
pushed a commit
to DianQK/llvm-project
that referenced
this issue
Oct 10, 2023
… bitcast(<X x i1> V)) canonicalization This already exists in InstCombine but was missing from the late stage ExpandReductions pass Fixes llvm#53419 Fixes llvm#61923 Differential Revision: https://reviews.llvm.org/D147452
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
We see worse code generation on memcmp-like function caused by
Asm on last release: https://godbolt.org/z/Ebcd7P3jP
Current asm: https://godbolt.org/z/nxv1vc16e
Loop block changed from
to
I am not sure into how big performance regression this translates (build & benchmark runs are underway), but the assembly definitely looks worse in the loop block.
The text was updated successfully, but these errors were encountered: