Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-08-25 Thread Michael Ellerman
Joel Stanley writes: > On Thu, 24 Aug 2023 at 12:12, Michael Ellerman wrote: >> >> Michael Ellerman writes: >> > Michael Ellerman writes: >> >> "Nicholas Piggin" writes: >> >>> On Wed Jun 14, 2023 at 3:56 PM AEST, Michael Ellerman wrote: >> Michael Ellerman writes: >> > Nicholas

Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-08-24 Thread Joel Stanley
On Thu, 24 Aug 2023 at 12:12, Michael Ellerman wrote: > > Michael Ellerman writes: > > Michael Ellerman writes: > >> "Nicholas Piggin" writes: > >>> On Wed Jun 14, 2023 at 3:56 PM AEST, Michael Ellerman wrote: > Michael Ellerman writes: > > Nicholas Piggin writes: > >> The

Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-08-24 Thread Michael Ellerman
Michael Ellerman writes: > Michael Ellerman writes: >> "Nicholas Piggin" writes: >>> On Wed Jun 14, 2023 at 3:56 PM AEST, Michael Ellerman wrote: Michael Ellerman writes: > Nicholas Piggin writes: >> The most expensive ordering for hwsync to provide is the store-load >>

Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-08-24 Thread Michael Ellerman
Michael Ellerman writes: > "Nicholas Piggin" writes: >> On Wed Jun 14, 2023 at 3:56 PM AEST, Michael Ellerman wrote: >>> Michael Ellerman writes: >>> > Nicholas Piggin writes: >>> >> The most expensive ordering for hwsync to provide is the store-load >>> >> barrier, because all prior stores

Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-06-14 Thread Michael Ellerman
"Nicholas Piggin" writes: > On Wed Jun 14, 2023 at 3:56 PM AEST, Michael Ellerman wrote: >> Michael Ellerman writes: >> > Nicholas Piggin writes: >> >> The most expensive ordering for hwsync to provide is the store-load >> >> barrier, because all prior stores have to be drained to the caches >>

Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-06-14 Thread Nicholas Piggin
On Wed Jun 14, 2023 at 3:56 PM AEST, Michael Ellerman wrote: > Michael Ellerman writes: > > Nicholas Piggin writes: > >> The most expensive ordering for hwsync to provide is the store-load > >> barrier, because all prior stores have to be drained to the caches > >> before subsequent instructions

Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-06-13 Thread Michael Ellerman
Michael Ellerman writes: > Nicholas Piggin writes: >> The most expensive ordering for hwsync to provide is the store-load >> barrier, because all prior stores have to be drained to the caches >> before subsequent instructions can complete. >> >> stsync just orders stores which means it can just

Re: [PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-06-13 Thread Michael Ellerman
Nicholas Piggin writes: > The most expensive ordering for hwsync to provide is the store-load > barrier, because all prior stores have to be drained to the caches > before subsequent instructions can complete. > > stsync just orders stores which means it can just be a barrer that > goes down the

[PATCH 4/4] powerpc/64s: Use POWER10 stsync barrier for wmb()

2023-06-09 Thread Nicholas Piggin
The most expensive ordering for hwsync to provide is the store-load barrier, because all prior stores have to be drained to the caches before subsequent instructions can complete. stsync just orders stores which means it can just be a barrer that goes down the store queue and orders draining, and