Skip to contentSkip to conversation

x86/mce: Enable HSD131, HSM142, HSW131, BDM48, and HSM142

PATCH In review Prarit Bhargava 2 participants 6 messages Feb 5, 2020 – Feb 6, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 x86/mce: Enable HSD131, HSM142, HSW131, BDM48, and HSM142

Intel Errata HSD131, HSM142, HSW131, and BDM48 report that "spurious corrected errors may be logged in the IA32_MC0_STATUS register with the valid field (bit 63) set, the uncorrected error field (bit 61) not set, a Model Specific Error Code (bits [31:16]) of 0x000F, and an MCA Error Code (bits [15:0]) of 0x0005."

Block these spurious errors from the console and logs.

Links to Intel Specification updates:
HSD131: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-desktop-specification-update.html
HSM142: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-mobile-specification-update.html
HSW131: https://www.intel.com/content/www/us/en/processors/xeon/xeon-e3-1200v3-spec-update.html
BDM48: https://www.intel.com/content/www/us/en/products/docs/processors/core/5th-gen-core-family-spec-update.html

---
 arch/x86/kernel/cpu/mce/core.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
arch/x86/kernel/cpu/mce/core.c +21 -0
index 2c4f949611e4..d893cc764a06 100644
@@ -121,6 +121,8 @@ static struct irq_work mce_irq_work;
 
 static void (*quirk_no_way_out)(int bank, struct mce *m, struct pt_regs *regs);
 
+static int (*quirk_noprint)(struct mce *m);
+
 /*
  * CPU/chipset specific EDAC code can register a notifier call here to print
  * MCE errors in a human-readable form.
@@ -232,6 +234,9 @@ struct mca_msr_regs msr_ops = {
 
 static void __print_mce(struct mce *m)
 {
+	if (quirk_noprint && quirk_noprint(m))
+		return;
+
 	pr_emerg(HW_ERR "CPU %d: Machine Check%s: %Lx Bank %d: %016Lx\n",
 		 m->extcpu,
 		 (m->mcgstatus & MCG_STATUS_MCIP ? " Exception" : ""),
@@ -1622,6 +1627,15 @@ static void quirk_sandybridge_ifu(int bank, struct mce *m, struct pt_regs *regs)
 	m->cs = regs->cs;
 }
 
+static int quirk_spurious_ce_noprint(struct mce *m)
+{
+	if (m->bank == 0 &&
+	    (m->status & 0xa0000000ffffffff) == 0x80000000000f0005)
+		return 1;
+
+	return 0;
+}
+
 /* Add per CPU specific workarounds here */
 static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c)
 {
@@ -1696,6 +1710,13 @@ static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c)
 
 		if (c->x86 == 6 && c->x86_model == 45)
 			quirk_no_way_out = quirk_sandybridge_ifu;
+
+		if ((c->x86 == 6) &&
+		    ((c->x86_model == 0x3c) || (c->x86_model == 0x3d) ||
+		     (c->x86_model == 0x45) || (c->x86_model == 46))) {
+			pr_info("MCE errata HSD131, HSM142, HSW131, BDM48, or HSM142 enabled.\n");
+			quirk_noprint = quirk_spurious_ce_noprint;
+		}
 	}
 
 	if (c->x86_vendor == X86_VENDOR_ZHAOXIN) {
Signature · 1 line
2.21.1
maintainer
#1 lore raw
Quoted 1 line · Prarit Bhargava

Subject: Re: [PATCH] x86/mce: Enable HSD131, HSM142, HSW131, BDM48, and HSM142

That subject is unreadable for humans.

Quoted 7 lines

Intel Errata HSD131, HSM142, HSW131, and BDM48 report that "spurious corrected errors may be logged in the IA32_MC0_STATUS register with the valid field (bit 63) set, the uncorrected error field (bit 61) not set, a Model Specific Error Code (bits [31:16]) of 0x000F, and an MCA Error Code (bits [15:0]) of 0x0005."

Block these spurious errors from the console and logs.

Are they being hit in the wild or why do we need this?

Quoted 5 lines

Links to Intel Specification updates:
HSD131: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-desktop-specification-update.html
HSM142: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-mobile-specification-update.html
HSW131: https://www.intel.com/content/www/us/en/processors/xeon/xeon-e3-1200v3-spec-update.html
BDM48: https://www.intel.com/content/www/us/en/products/docs/processors/core/5th-gen-core-family-spec-update.html

Those links tend to get stale with time. If you really want to refer to the PDFs, add a new bugzilla entry on https://bugzilla.kernel.org/, add them there as an attachment and add the link to the entry to the commit message.

Quoted 1 line

What's that Signed-off-by: tag supposed to mean?

Quoted 11 lines
Cc Tony Luck <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc "H. Peter Anvin" <[email protected]>
---
 arch/x86/kernel/cpu/mce/core.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

If at all, this should be done by adding an intel_filter_mce() function and called from filter_mce() so that such errors don't get logged.

Thx.

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette
#2 lore raw
Quoted 5 lines · Borislav Petkov
Quoted 1 line · Prarit Bhargava

Subject: Re: [PATCH] x86/mce: Enable HSD131, HSM142, HSW131, BDM48, and HSM142

That subject is unreadable for humans.

Yeah :/ I couldn't think of a better one. Maybe "Block spurious corrected errors on some Intel processors"? Any other suggestion?

Quoted 10 lines
Quoted 8 lines

Intel Errata HSD131, HSM142, HSW131, and BDM48 report that "spurious corrected errors may be logged in the IA32_MC0_STATUS register with the valid field (bit 63) set, the uncorrected error field (bit 61) not set, a Model Specific Error Code (bits [31:16]) of 0x000F, and an MCA Error Code (bits [15:0]) of 0x0005."

Block these spurious errors from the console and logs.

Are they being hit in the wild or why do we need this?

Alexander, cc'd, is being hit by this in the wild.

Quoted 28 lines
Quoted 6 lines

Links to Intel Specification updates:
HSD131: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-desktop-specification-update.html
HSM142: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-mobile-specification-update.html
HSW131: https://www.intel.com/content/www/us/en/processors/xeon/xeon-e3-1200v3-spec-update.html
BDM48: https://www.intel.com/content/www/us/en/products/docs/processors/core/5th-gen-core-family-spec-update.html

Those links tend to get stale with time. If you really want to refer to the PDFs, add a new bugzilla entry on https://bugzilla.kernel.org/, add them there as an attachment and add the link to the entry to the commit message.

Quoted 1 line

What's that Signed-off-by: tag supposed to mean?

Quoted 11 lines
Cc Tony Luck <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc "H. Peter Anvin" <[email protected]>
---
 arch/x86/kernel/cpu/mce/core.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
Quoted 2 lines

If at all, this should be done by adding an intel_filter_mce() function and called from filter_mce() so that such errors don't get logged.

I'll take a look over there.

P.

Quoted 3 lines

Thx.

maintainer
#3 lore raw
Quoted 2 lines · Prarit Bhargava

Yeah :/ I couldn't think of a better one. Maybe "Block spurious corrected errors on some Intel processors"? Any other suggestion?

"Do not log ..."

Quoted 1 line

Alexander, cc'd, is being hit by this in the wild.

Do say that in the commit message.

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

What's that Signed-off-by: tag supposed to mean?

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

You missed this one.

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette
#4 lore raw
In reply to Prarit Bhargava
Quoted 19 lines · Prarit Bhargava
Quoted 5 lines · Borislav Petkov
Quoted 1 line · Prarit Bhargava

Subject: Re: [PATCH] x86/mce: Enable HSD131, HSM142, HSW131, BDM48, and HSM142

That subject is unreadable for humans.

Yeah :/ I couldn't think of a better one. Maybe "Block spurious corrected errors on some Intel processors"? Any other suggestion?

Quoted 7 lines
Quoted 7 lines

Intel Errata HSD131, HSM142, HSW131, and BDM48 report that "spurious corrected errors may be logged in the IA32_MC0_STATUS register with the valid field (bit 63) set, the uncorrected error field (bit 61) not set, a Model Specific Error Code (bits [31:16]) of 0x000F, and an MCA Error Code (bits [15:0]) of 0x0005."

Quoted 21 lines
Quoted 3 lines
Quoted 1 line

Block these spurious errors from the console and logs.

Are they being hit in the wild or why do we need this?

Alexander, cc'd, is being hit by this in the wild.

Quoted 15 lines
Quoted 6 lines

Links to Intel Specification updates:
HSD131: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-desktop-specification-update.html
HSM142: https://www.intel.com/content/www/us/en/products/docs/processors/core/4th-gen-core-family-mobile-specification-update.html
HSW131: https://www.intel.com/content/www/us/en/processors/xeon/xeon-e3-1200v3-spec-update.html
BDM48: https://www.intel.com/content/www/us/en/products/docs/processors/core/5th-gen-core-family-spec-update.html

Those links tend to get stale with time. If you really want to refer to the PDFs, add a new bugzilla entry on https://bugzilla.kernel.org/, add them there as an attachment and add the link to the entry to the commit message.

Quoted 1 line

What's that Signed-off-by: tag supposed to mean?

Sorry. I missed this question, but I really don't understand the question. Alexander posted a patch in a kernel bugzilla @ Red Hat and I modified the patch with some additional changes. I don't want him to lose credit for the work so he's got a proper Signed-off-by tag for this patch.

P.

maintainer
#5 lore raw
Quoted 4 lines · Prarit Bhargava

Sorry. I missed this question, but I really don't understand the question. Alexander posted a patch in a kernel bugzilla @ Red Hat and I modified the patch with some additional changes. I don't want him to lose credit for the work so he's got a proper Signed-off-by tag for this patch.

This is not how this is expressed. Either you write that in free text in the commit message or you use Co-developed-by. More details in

Documentation/process/submitting-patches.rst

Signature · 4 lines
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette