From: "Ahmed S. Darwish" <darwi@linutronix.de>
To: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: Borislav Petkov <bp@alien8.de>, Ingo Molnar <mingo@redhat.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
Thomas Gleixner <tglx@linutronix.de>,
John Ogness <john.ogness@linutronix.de>,
"H. Peter Anvin" <hpa@zytor.com>,
x86@kernel.org, x86-cpuid@lists.linux.dev,
LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v1 37/40] x86/cacheinfo: Extract out cache self-snoop checks
Date: Wed, 5 Mar 2025 19:40:05 +0100 [thread overview]
Message-ID: <Z8iahUTYuq0gVXmP@lx-t490> (raw)
In-Reply-To: <01367a6b-8eeb-417e-9a66-e7a64d518319@citrix.com>
Hi Andrew,
On Tue, 04 Mar 2025, Andrew Cooper wrote:
>
> On 04/03/2025 8:51 am, Ahmed S. Darwish wrote:
> > The logic of not doing a cache flush if the CPU declares cache self
> > snooping support is repeated across the x86/cacheinfo code. Extract it
> > into its own function.
> >
> > Signed-off-by: Ahmed S. Darwish <darwi@linutronix.de>
>
> I know you're just refactoring code, but the SDM has basically reverted
> this statement about it being safe to skip WBINVD based on SELFSNOOP.
>
Still, thanks a lot for sharing :)
> It turns out not to be safe in cases where the underlying physical
> memory changes from cacheable to unchangeable. By skipping the WBINVD
> as part of changing the memory type, you end up with spurious writebacks
> at a later point when the memory is expected to be UC. Apparently this
> is a problem for CLX devices, hence the change in the SDM.
While writing that refactoring patch, I indeed noticed that there is an
errata list of CPUs where X86_FEATURE_SELFSNOOP is force disabled, thus
ensuring WBINVD is never skipped:
static void check_memory_type_self_snoop_errata(...)
{
switch (c->x86_vfm) {
case INTEL_CORE_YONAH:
case INTEL_CORE2_MEROM:
case INTEL_CORE2_MEROM_L:
case INTEL_CORE2_PENRYN:
case INTEL_CORE2_DUNNINGTON:
case INTEL_NEHALEM:
case INTEL_NEHALEM_G:
case INTEL_NEHALEM_EP:
case INTEL_NEHALEM_EX:
case INTEL_WESTMERE:
case INTEL_WESTMERE_EP:
case INTEL_SANDYBRIDGE:
setup_clear_cpu_cap(X86_FEATURE_SELFSNOOP);
}
}
That's why I added "CPUs without known erratas" in the comments:
/*
* Cache flushing is the most time-consuming step when programming
* the MTRRs. On many Intel CPUs without known erratas, it can be
* skipped if the CPU declares cache self-snooping support.
*/
static void maybe_flush_caches(void)
{
if (!static_cpu_has(X86_FEATURE_SELFSNOOP))
wbinvd();
}
But, interestingly, CLX devices (intel-family.h CASCADELAKE_X /
SKYLAKE_X) are not part of the kernel's Self Snoop errata list above.
@Thomas, @Ingo, any ideas?
Thanks,
--
Ahmed S. Darwish
Linutronix GmbH
next prev parent reply other threads:[~2025-03-05 18:40 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-04 8:51 [PATCH v1 00/40] x86: Leaf 0x2 and leaf 0x4 refactorings Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 01/40] x86/cacheinfo: Validate cpuid leaf 0x2 EDX output Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 02/40] x86/cpu: " Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 03/40] x86/cpu: Properly parse leaf 0x2 TLB descriptor 0x63 Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 04/40] x86/cpuid: Include linux/build_bug.h Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 05/40] x86/cpu: Remove unnecessary headers and reorder the rest Ahmed S. Darwish
2025-03-04 9:14 ` Ingo Molnar
2025-03-04 9:28 ` Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 06/40] x86/cpu: Use max() for leaf 0x2 TLB descriptors parsing Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 07/40] x86/cpu: Simplify TLB entry count storage Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 08/40] x86/cpu: Get rid of smp_store_cpu_info() indirection Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 09/40] x86/cpu: Remove unused TLB strings Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 10/40] x86/cpu: Remove leaf 0x2 parsing loop and add helpers Ahmed S. Darwish
2025-03-04 9:26 ` Ingo Molnar
2025-03-05 16:01 ` Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 11/40] x86/cacheinfo: Remove the P4 trace leftovers for real Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 12/40] x86/cacheinfo: Remove unnecessary headers and reorder the rest Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 13/40] x86/cacheinfo: Use cpuid leaf 0x2 parsing helpers Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 14/40] x86/cacheinfo: Refactor leaf 0x2 cache descriptor lookup Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 15/40] x86/cacheinfo: Properly name amd_cpuid4()'s first parameter Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 16/40] x86/cacheinfo: Use proper name for cacheinfo instances Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 17/40] x86/cacheinfo: Constify _cpuid4_info_regs instances Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 18/40] x86/cacheinfo: Align ci_info_init() assignment expressions Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 19/40] x86/cacheinfo: Standardize _cpuid4_info_regs instance naming Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 20/40] x86: treewide: Introduce x86_vendor_amd_or_hygon() Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 21/40] x86/cacheinfo: Consolidate AMD/Hygon leaf 0x8000001d calls Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 22/40] x86/cacheinfo: Separate amd_northbridge from _cpuid4_info_regs Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 23/40] x86/cacheinfo: Move AMD cache_disable_0/1 handling to separate file Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 24/40] x86/cacheinfo: Use sysfs_emit() for sysfs attributes show() Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 25/40] x86/cacheinfo: Separate Intel and AMD leaf 0x4 code paths Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 26/40] x86/cacheinfo: Rename _cpuid4_info_regs to _cpuid4_info Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 27/40] x86/cacheinfo: Clarify type markers for leaf 0x2 cache descriptors Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 28/40] x86/cacheinfo: Use enums for cache descriptor types Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 29/40] x86/cpu: Use enums for TLB " Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 30/40] sizes.h: Cover all possible x86 cpu cache sizes Ahmed S. Darwish
2025-03-04 9:35 ` Ingo Molnar
2025-03-05 16:18 ` Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 31/40] x86/cpu: Consolidate CPUID leaf 0x2 tables Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 32/40] x86/cacheinfo: Use consolidated leaf 0x2 descriptor table Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 33/40] x86/cpu: " Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 34/40] x86/cacheinfo: Separate leaf 0x2 handling and post-processing logic Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 35/40] x86/cacheinfo: Separate intel leaf 0x4 handling Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 36/40] x86/cacheinfo: Extract out cache level topology ID calculation Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 37/40] x86/cacheinfo: Extract out cache self-snoop checks Ahmed S. Darwish
2025-03-04 10:38 ` Andrew Cooper
2025-03-05 18:40 ` Ahmed S. Darwish [this message]
2025-03-05 18:42 ` Andrew Cooper
2025-03-05 18:58 ` Ahmed S. Darwish
2025-03-05 19:01 ` Andrew Cooper
2025-03-04 8:51 ` [PATCH v1 38/40] x86/cacheinfo: Relocate leaf 0x4 cache_type mapping Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 39/40] x86/cacheinfo: Introduce amd_hygon_cpu_has_l3_cache() Ahmed S. Darwish
2025-03-04 8:51 ` [PATCH v1 40/40] x86/cacheinfo: Apply maintainer-tip coding style fixes Ahmed S. Darwish
2025-03-04 9:19 ` [PATCH v1 00/40] x86: Leaf 0x2 and leaf 0x4 refactorings Ingo Molnar
2025-03-04 9:38 ` Ingo Molnar
2025-03-05 17:36 ` Ahmed S. Darwish
2025-03-04 9:33 ` Ingo Molnar
2025-03-05 16:38 ` Ahmed S. Darwish
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Z8iahUTYuq0gVXmP@lx-t490 \
--to=darwi@linutronix.de \
--cc=andrew.cooper3@citrix.com \
--cc=bp@alien8.de \
--cc=dave.hansen@linux.intel.com \
--cc=hpa@zytor.com \
--cc=john.ogness@linutronix.de \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=tglx@linutronix.de \
--cc=x86-cpuid@lists.linux.dev \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox