On 1/25/21, Jessica Clarke <[email protected]> wrote: > On 25 Jan 2021, at 22:42, Mateusz Guzik <[email protected]> wrote: >> index a3548977b940..4c29d132c369 100644 >> --- a/sys/sys/atomic_common.h >> +++ b/sys/sys/atomic_common.h >> @@ -58,7 +58,7 @@ >> #define atomic_store_long(p, v) \ >> (*(volatile u_long *)(p) = (u_long)(v)) >> #define atomic_store_ptr(p, v) \ >> - (*(volatile uintptr_t *)(p) = (uintptr_t)(v)) >> + (*(volatile __typeof(*p) *)(p) = (v)) > > This is in fact general enough to be any of the atomic_store_*, and > won't give any errors if you use it for a uint16_t (where it will store > a uint16_t). >
Unclear if you are suggesting validating that the passed thing is a pointer or replacing the rest. The former is useful, the latter is significantly more work. -- Mateusz Guzik <mjguzik gmail.com> _______________________________________________ [email protected] mailing list https://lists.freebsd.org/mailman/listinfo/dev-commits-src-main To unsubscribe, send any mail to "[email protected]"
