fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-06 13:48:06 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

softfloat: make sign bit an unsigned int
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details

Browse source 
According to ubsan shifting an int32_t by 31 bits to the left is undefined
behavior. So just declare it as uint32_t.

Backport-to: 26.1
Reviewed-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41252>
This commit is contained in:
Karol Herbst 2026-04-29 01:39:41 +02:00 • committed by Marge Bot
parent 7c69f31ecc
commit b2aa92b523
1 changed files with 5 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
src/util/softfloat.c
View file

@ -153,7 +153,7 @@ double _mesa_roundtozero_f64(int64_t s, int64_t e, int64_t m)
* \brief Extracted from softfloat_roundPackToF32()
*/
static inline
float _mesa_round_f32(int32_t s, int32_t e, int32_t m, bool rtz)
float _mesa_round_f32(uint32_t s, int32_t e, int32_t m, bool rtz)
{
fi_type result;
uint8_t round_increment = rtz ? 0 : 0x40;
@ -1192,7 +1192,8 @@ _mesa_float_fma_rtz(float a, float b, float c)
uint32_t c_flt_m = c_fi.u & 0x07fffff;
uint32_t c_flt_e = (c_fi.u >> 23) & 0xff;
uint32_t c_flt_s = (c_fi.u >> 31) & 0x1;
int32_t s, e, m = 0;
uint32_t s;
int32_t e, m = 0;
c_flt_s ^= 0;
s = a_flt_s ^ b_flt_s ^ 0;
@ -1371,7 +1372,8 @@ _mesa_double_to_f32(double val, bool rtz)
uint64_t flt_m = di.u & 0x0fffffffffffff;
uint64_t flt_e = (di.u >> 52) & 0x7ff;
uint64_t flt_s = (di.u >> 63) & 0x1;
int32_t s, e, m = 0;
uint32_t s;
int32_t e, m = 0;
s = flt_s;