kiper220/mimalloc
1
0
Fork 0
mirror of https://github.com/microsoft/mimalloc.git synced 2025-07-06 19:38:41 +03:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix 32 bit multiply in generic ctz/clz

Browse source
This commit is contained in:
daanx 2024-12-09 19:43:00 -08:00
parent f28d5c7029
commit 56a1bd7f9e
2 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

4
include/mimalloc/bits.h
View file

@ -220,7 +220,7 @@ static inline size_t mi_popcount(size_t x) {
// return false if `x==0` (with `*idx` undefined) and true otherwise,
// with the `idx` is set to the bit index (`0 <= *idx < MI_BFIELD_BITS`).
static inline bool mi_bsf(size_t x, size_t* idx) {
#if defined(__GNUC__) && MI_ARCH_X64 && defined(__BMI1__)
#if defined(__GNUC__) && MI_ARCH_X64 && defined(__BMI1__) && (!defined(__clang_major__) || __clang_major__ >= 9)
// on x64 the carry flag is set on zero which gives better codegen
bool is_zero;
__asm ( "tzcnt\t%2, %1" : "=@ccc"(is_zero), "=r"(*idx) : "r"(x) : "cc" );
@ -237,7 +237,7 @@ static inline bool mi_bsf(size_t x, size_t* idx) {
// return false if `x==0` (with `*idx` undefined) and true otherwise,
// with the `idx` is set to the bit index (`0 <= *idx < MI_BFIELD_BITS`).
static inline bool mi_bsr(size_t x, size_t* idx) {
#if defined(__GNUC__) && MI_ARCH_X64 && defined(__BMI1__)
#if defined(__GNUC__) && MI_ARCH_X64 && defined(__BMI1__) && (!defined(__clang_major__) || __clang_major__ >= 9)
// on x64 the carry flag is set on zero which gives better codegen
bool is_zero;
__asm ("lzcnt\t%2, %1" : "=@ccc"(is_zero), "=r"(*idx) : "r"(x) : "cc");