s390/atomic: remove small optimization to fix clang build

With commit f0cbd3b83e ("s390/atomic: circumvent gcc 10 build
regression") there was an attempt to workaroud a gcc build bug,
however with the workaround a similar problem with clang appeared.
It was recommended to use a workaround which would fail again with
gcc. Therefore simply remove the optimization. It is just not worth
the effort.

Besides that all of this will be changed to use compiler atomic
builtins instead anyway.

See https://reviews.llvm.org/D90231
and https://reviews.llvm.org/D91786

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
This commit is contained in:
Heiko Carstens 2021-01-15 19:39:23 +01:00 committed by Vasily Gorbik
parent 4520a91a97
commit 6110ccecd3
1 changed files with 0 additions and 20 deletions

View File

@ -44,16 +44,6 @@ static inline int atomic_fetch_add(int i, atomic_t *v)
static inline void atomic_add(int i, atomic_t *v)
{
#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
/*
* Order of conditions is important to circumvent gcc 10 bug:
* https://gcc.gnu.org/pipermail/gcc-patches/2020-July/549318.html
*/
if ((i > -129) && (i < 128) && __builtin_constant_p(i)) {
__atomic_add_const(i, &v->counter);
return;
}
#endif
__atomic_add(i, &v->counter);
}
@ -115,16 +105,6 @@ static inline s64 atomic64_fetch_add(s64 i, atomic64_t *v)
static inline void atomic64_add(s64 i, atomic64_t *v)
{
#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
/*
* Order of conditions is important to circumvent gcc 10 bug:
* https://gcc.gnu.org/pipermail/gcc-patches/2020-July/549318.html
*/
if ((i > -129) && (i < 128) && __builtin_constant_p(i)) {
__atomic64_add_const(i, (long *)&v->counter);
return;
}
#endif
__atomic64_add(i, (long *)&v->counter);
}