diff options
| author | Nick Desaulniers <ndesaulniers@google.com> | 2023-08-28 11:53:57 -0700 |
|---|---|---|
| committer | Ingo Molnar <mingo@kernel.org> | 2023-09-07 00:05:50 +0200 |
| commit | 3dae5c43badf285e22f6d88388e8a232a83bdfec (patch) | |
| tree | 82670ebc99ec4553ca8b2b3d2e4d04271761f898 | |
| parent | 4accdb9895349026d85e526036ff28c07921d7cf (diff) | |
x86/asm/bitops: Use __builtin_clz{l|ll} to evaluate constant expressions
Micro-optimize the bitops code some more, similar to commits:
fdb6649ab7c1 ("x86/asm/bitops: Use __builtin_ctzl() to evaluate constant expressions")
2fcff790dcb4 ("powerpc: Use builtin functions for fls()/__fls()/fls64()")
From a recent discussion, I noticed that x86 is lacking an optimization
that appears in arch/powerpc/include/asm/bitops.h related to constant
folding. If you add a BUILD_BUG_ON(__builtin_constant_p(param)) to
these functions, you'll find that there were cases where the use of
inline asm pessimized the compiler's ability to perform constant folding
resulting in runtime calculation of a value that could have been
computed at compile time.
Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20230828-x86_fls-v1-1-e6a31b9f79c3@google.com
| -rw-r--r-- | arch/x86/include/asm/bitops.h | 9 |
1 files changed, 9 insertions, 0 deletions
diff --git a/arch/x86/include/asm/bitops.h b/arch/x86/include/asm/bitops.h index 2edf68475fec..50e5ebf9d0a0 100644 --- a/arch/x86/include/asm/bitops.h +++ b/arch/x86/include/asm/bitops.h @@ -293,6 +293,9 @@ static __always_inline unsigned long variable_ffz(unsigned long word) */ static __always_inline unsigned long __fls(unsigned long word) { + if (__builtin_constant_p(word)) + return BITS_PER_LONG - 1 - __builtin_clzl(word); + asm("bsr %1,%0" : "=r" (word) : "rm" (word)); @@ -360,6 +363,9 @@ static __always_inline int fls(unsigned int x) { int r; + if (__builtin_constant_p(x)) + return x ? 32 - __builtin_clz(x) : 0; + #ifdef CONFIG_X86_64 /* * AMD64 says BSRL won't clobber the dest reg if x==0; Intel64 says the @@ -401,6 +407,9 @@ static __always_inline int fls(unsigned int x) static __always_inline int fls64(__u64 x) { int bitpos = -1; + + if (__builtin_constant_p(x)) + return x ? 64 - __builtin_clzll(x) : 0; /* * AMD64 says BSRQ won't clobber the dest reg if x==0; Intel64 says the * dest reg is undefined if x==0, but their CPU architect says its |