[all-commits] [llvm/llvm-project] 7905ec: [libc] Use UMAXV.4S to reduce bcmp result.

[Peter Collingbourne via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 7905ec387e7a4470255b7856c71b2ec654ac397a
      https://github.com/llvm/llvm-project/commit/7905ec387e7a4470255b7856c71b2ec654ac397a
  Author: Peter Collingbourne <pcc at google.com>
  Date:   2025-10-13 (Mon, 13 Oct 2025)

  Changed paths:
    M libc/src/string/memory_utils/op_aarch64.h

  Log Message:
  -----------
  [libc] Use UMAXV.4S to reduce bcmp result.

We can use UMAXV.4S to reduce the comparison result in a single
instruction. This improves performance by roughly 4% on Apple M1:

Summary
  bin/libc.src.string.bcmp_benchmark3 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10 ran
    1.01 ± 0.02 times faster than bin/libc.src.string.bcmp_benchmark3 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.01 ± 0.03 times faster than bin/libc.src.string.bcmp_benchmark3 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.01 ± 0.03 times faster than bin/libc.src.string.bcmp_benchmark3 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.01 ± 0.02 times faster than bin/libc.src.string.bcmp_benchmark2 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.02 ± 0.03 times faster than bin/libc.src.string.bcmp_benchmark2 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.03 ± 0.03 times faster than bin/libc.src.string.bcmp_benchmark2 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.03 ± 0.03 times faster than bin/libc.src.string.bcmp_benchmark2 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.05 ± 0.02 times faster than bin/libc.src.string.bcmp_benchmark1 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.05 ± 0.02 times faster than bin/libc.src.string.bcmp_benchmark1 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.05 ± 0.03 times faster than bin/libc.src.string.bcmp_benchmark1 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10
    1.05 ± 0.02 times faster than bin/libc.src.string.bcmp_benchmark1 --study-name="new bcmp" --sweep-mode --sweep-max-size=128 --output=/dev/null --num-trials=10

(1 = original, 2 = a variant of this patch that uses UMAXV.16B, 3 = this patch)

Reviewers: michaelrj-google, gchatelet, overmighty, SchrodingerZhu

Pull Request: https://github.com/llvm/llvm-project/pull/99260



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications