Re: [PATCH 3/3] x86: correct fencing around CLFLUSH

2022-02-24 Thread Jan Beulich
On 23.02.2022 13:33, Andrew Cooper wrote: > On 23/02/2022 10:13, Jan Beulich wrote: >> --- a/xen/arch/x86/cpu/common.c >> +++ b/xen/arch/x86/cpu/common.c >> @@ -346,9 +346,14 @@ void __init early_cpu_init(void) >> c->x86_model, c->x86_model, c->x86_mask, eax); >> >> if (c->cpuid_

Re: [PATCH 3/3] x86: correct fencing around CLFLUSH

2022-02-23 Thread Andrew Cooper
On 23/02/2022 10:13, Jan Beulich wrote: > As noted in the context of 3330013e6739 ("VT-d / x86: re-arrange cache > syncing"): While cache_writeback() has the SFENCE on the correct side of > CLFLUSHOPT, flush_area_local() doesn't. While I can't prove it due to > lacking a copy of the old SDM version

[PATCH 3/3] x86: correct fencing around CLFLUSH

2022-02-23 Thread Jan Beulich
As noted in the context of 3330013e6739 ("VT-d / x86: re-arrange cache syncing"): While cache_writeback() has the SFENCE on the correct side of CLFLUSHOPT, flush_area_local() doesn't. While I can't prove it due to lacking a copy of the old SDM version, I can only assume this placement was a result