x86/mce: Switch to new Intel CPU model defines
Subsystems (2)
The following commit has been merged into the x86/cpu branch of tip:
Commit-ID: 68b4c6d168c547b19ba5697f9398c3f4aa8b4276 Gitweb: https://git.kernel.org/tip/68b4c6d168c547b19ba5697f9398c3f4aa8b4276 Author: Tony Luck <[email protected]> AuthorDate: Wed, 24 Apr 2024 11:15:13 -07:00 Committer: Dave Hansen <[email protected]> CommitterDate: Fri, 26 Apr 2024 08:49:24 -07:00
x86/mce: Switch to new Intel CPU model defines
New CPU #defines encode vendor and family as well as model.
--- arch/x86/kernel/cpu/mce/severity.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
arch/x86/kernel/cpu/mce/severity.c +5 -5
@@ -12,7 +12,7 @@#include <linux/uaccess.h> #include <asm/mce.h> -#include <asm/intel-family.h> +#include <asm/cpu_device_id.h> #include <asm/traps.h> #include <asm/insn.h> #include <asm/insn-eval.h>
@@ -45,14 +45,14 @@ static struct severity {unsigned char context;unsigned char excp;unsigned char covered; -unsigned char cpu_model; +unsigned int cpu_vfm;unsigned char cpu_minstepping;unsigned char bank_lo, bank_hi;char *msg; } severities[] = { #define MCESEV(s, m, c...) { .sev = MCE_ ## s ## _SEVERITY, .msg = m, ## c }#define BANK_RANGE (l, h) .bank_lo = l, .bank_hi = h -#define MODEL_STEPPING (m, s) .cpu_model = m, .cpu_minstepping = s +#define VFM_STEPPING (m, s) .cpu_vfm = m, .cpu_minstepping = s#define KERNEL .context = IN_KERNEL#define USER .context = IN_USER#define KERNEL_RECOV .context = IN_KERNEL_RECOV
@@ -128,7 +128,7 @@ static struct severity {MCESEV (AO ,"Uncorrected Patrol Scrub Error" ,SER ,MASK (MCI_STATUS_UC |MCI_ADDR |0xffffeff0 ,MCI_ADDR |0x001000c0 ), -MODEL_STEPPING (INTEL_FAM6_SKYLAKE_X ,4 ),BANK_RANGE (13 ,18 ) +VFM_STEPPING (INTEL_SKYLAKE_X ,4 ),BANK_RANGE (13 ,18 ) ),/* ignore OVER for UCNA */
@@ -398,7 +398,7 @@ static noinstr int mce_severity_intel(struct mce *m, struct pt_regs *regs, charcontinue ;if (s->excp && excp != s->excp)continue ; -if (s->cpu_model && boot_cpu_data.x86_model != s->cpu_model) +if (s->cpu_vfm && boot_cpu_data.x86_model != s->cpu_vfm)continue ;if (s->cpu_minstepping && boot_cpu_data.x86_stepping < s->cpu_minstepping)continue ;
Borislav Petkov
maintainerarch/x86/kernel/cpu/mce/severity.c static noinstr int mce_severity_intel(struct mce *m, struct pt_regs *regs, char · line 401
Quoted 6 lines · On Fri, Apr 26, 2024 at 06:12:11PM -0000, tip-bot2 for Tony Luck wrote:
arch/x86/kernel/cpu/mce/severity.c +1
-1@@ -398,7 +398,7 @@ static noinstr int mce_severity_intel(struct mce *m, struct pt_regs *regs, charcontinue ;if (s->excp && excp != s->excp)continue ; -if (s->cpu_model && boot_cpu_data.x86_model != s->cpu_model) +if (s->cpu_vfm && boot_cpu_data.x86_model != s->cpu_vfm)
Hold on, isn't this supposed to be:
if (s->cpu_vfm && boot_cpu_data.cpu_vfm != s->cpu_vfm)
?
You're getting rid of the ->x86_model checking altogether...
Signature · 4 lines
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquetteTony Luck
maintainerPATCH 1/1 x86/cpu: Fix thinko comparing x86_vfm with x86_model
This is obviously nonsense. Fix it.
---
Boris: Do you rebase x86/cpu branch in TIP? If so this could be merged into the broken patch. If not then slap it on top. Risk of bisection breakage is tiny as this only affects a quirk handling patrol scrub errors on Skylake and Cascade lake systems.
arch/x86/kernel/cpu/mce/severity.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/cpu/mce/severity.c +1 -1
@@ -398,7 +398,7 @@ static noinstr int mce_severity_intel(struct mce *m, struct pt_regs *regs, charcontinue ;if (s->excp && excp != s->excp)continue ; -if (s->cpu_vfm && boot_cpu_data.x86_model != s->cpu_vfm) +if (s->cpu_vfm && boot_cpu_data.x86_vfm != s->cpu_vfm)continue ;if (s->cpu_minstepping && boot_cpu_data.x86_stepping < s->cpu_minstepping)continue ;
Signature · 1 line
2.44.0
Borislav Petkov
maintainerRe: [PATCH] x86/cpu: Fix thinko comparing x86_vfm with x86_model
Quoted 11 lines · Tony Luck
This is obviously nonsense. Fix it.
Fixes 68b4c6d168c5 ("x86/mce: Switch to new Intel CPU model defines")Reported-by Borislav Petkov <[email protected]>Signed-off-by Tony Luck <[email protected]>---
Boris: Do you rebase x86/cpu branch in TIP? If so this could be merged into the broken patch. If not then slap it on top. Risk of bisection breakage is tiny as this only affects a quirk handling patrol scrub errors on Skylake and Cascade lake systems.
Lemme try to rebase. And btw, you could've avoided the same name patches:
e063b531d4e8 (refs/remotes/tip/x86/cpu) x86/mm: Switch to new Intel CPU model defines 94eb882e9575 x86/tsc_msr: Switch to new Intel CPU model defines 95be3ccd711c x86/tsc: Switch to new Intel CPU model defines 7cd2e9a90299 x86/cpu: Switch to new Intel CPU model defines 343ec8beae30 x86/resctrl: Switch to new Intel CPU model defines 43e8e0a95a89 x86/resctrl: Switch to new Intel CPU model defines
<--
173e856a55d0 x86/microcode/intel: Switch to new Intel CPU model defines 68b4c6d168c5 x86/mce: Switch to new Intel CPU model defines 8a5943977900 x86/mce: Switch to new Intel CPU model defines 066f54e65e47 x86/mce: Switch to new Intel CPU model defines
<---
34304acb3290 x86/cpu: Switch to new Intel CPU model defines 77d0930acedf x86/cpu/intel_epb: Switch to new Intel CPU model defines a7c345216f88 x86/aperfmperf: Switch to new Intel CPU model defines 22878345b4c1 x86/apic: Switch to new Intel CPU model defines e8475a26a94f perf/x86/msr: Switch to new Intel CPU model defines 438731421a2f perf/x86/intel/uncore: Switch to new Intel CPU model defines d413a1955a8e perf/x86/intel/uncore: Switch to new Intel CPU model defines 9b2583d8103e perf/x86/intel/uncore: Switch to new Intel CPU model defines
<---
a7011b852a30 perf/x86/intel/pt: Switch to new Intel CPU model defines 0011a51d73d5 perf/x86/lbr: Switch to new Intel CPU model defines 5ee800945a34 perf/x86/intel/cstate: Switch to new Intel CPU model defines
and why in the hell are there 2 or 3 patches of the same name touching the same thing? Why can't they be a single patch?
Lemme rebase the lot.
Signature · 4 lines
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquetteLuck, Tony
maintainerRE: [PATCH] x86/cpu: Fix thinko comparing x86_vfm with x86_model
Quoted 2 lines
and why in the hell are there 2 or 3 patches of the same name touching the same thing? Why can't they be a single patch?
I made one patch per file touched (wherever possible ... there are a few patches where ".h" and ".c" files needed to be bundled together).
Quoted 1 line
Lemme rebase the lot.
Thanks. I just did a "git fetch" the x86/cpu branch. It all looks good.
-Tony
Borislav Petkov
maintainerRe: [PATCH] x86/cpu: Fix thinko comparing x86_vfm with x86_model
Quoted 2 lines · Tony
I made one patch per file touched (wherever possible ... there are a few patches where ".h" and ".c" files needed to be bundled together).
I figured. I think the next step would've been to group them all into a single patch based on the subject prefix. But we're doing that now anyway.
Quoted 1 line
Thanks. I just did a "git fetch" the x86/cpu branch. It all looks good.
Thx!
Signature · 4 lines
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquette