Skip to contentSkip to conversation

arm64: errata: Add Cortex-A510 to the repeat tlbi list

PATCH In review Lucas Wei 2 participants 4 messages Aug 30, 2022

Subsystems (3)
lore
Share & tools
lore

Lucas Wei

author
#0 lore raw

PATCH 1/1 arm64: errata: Add Cortex-A510 to the repeat tlbi list

From: James Morse <[email protected]>

Cortex-A510 is affected by an erratum where in rare circumstances the CPUs may not handle a race between a break-before-make sequence on one CPU, and another CPU accessing the same page. This could allow a store to a page that has been unmapped.

Work around this by adding the affected CPUs to the list that needs TLB sequences to be done twice.

---
 Documentation/arm64/silicon-errata.rst |  2 ++
 arch/arm64/Kconfig                     | 17 +++++++++++++++++
 arch/arm64/kernel/cpu_errata.c         |  8 +++++++-
 3 files changed, 26 insertions(+), 1 deletion(-)
Documentation/arm64/silicon-errata.rst +2 -0
index 0b4235b1f8c4..33b04db8408f 100644
@@ -106,6 +106,8 @@ stable kernels.
 +----------------+-----------------+-----------------+-----------------------------+
 | ARM            | Cortex-A510     | #2077057        | ARM64_ERRATUM_2077057       |
 +----------------+-----------------+-----------------+-----------------------------+
+| ARM            | Cortex-A510     | #2441009        | ARM64_ERRATUM_2441009       |
++----------------+-----------------+-----------------+-----------------------------+
 | ARM            | Cortex-A710     | #2119858        | ARM64_ERRATUM_2119858       |
 +----------------+-----------------+-----------------+-----------------------------+
 | ARM            | Cortex-A710     | #2054223        | ARM64_ERRATUM_2054223       |
arch/arm64/Kconfig +17 -0
index a5d1b561ed53..001eaba5a6b4 100644
@@ -838,6 +838,23 @@ config ARM64_ERRATUM_2224489
 
 	  If unsure, say Y.
 
+config ARM64_ERRATUM_2441009
+	bool "Cortex-A510: Completion of affected memory accesses might not be guaranteed by completion of a TLBI"
+	default y
+	select ARM64_WORKAROUND_REPEAT_TLBI
+	help
+	  This option adds a workaround for ARM Cortex-A510 erratum #2441009.
+
+	  Under very rare circumstances, affected Cortex-A510 CPUs
+	  may not handle a race between a break-before-make sequence on one
+	  CPU, and another CPU accessing the same page. This could allow a
+	  store to a page that has been unmapped.
+
+	  Work around this by adding the affected CPUs to the list that needs
+	  TLB sequences to be done twice.
+
+	  If unsure, say Y.
+
 config ARM64_ERRATUM_2064142
 	bool "Cortex-A510: 2064142: workaround TRBE register writes while disabled"
 	depends on CORESIGHT_TRBE
arch/arm64/kernel/cpu_errata.c +7 -1
index 6b92989f4cc2..aa9609e6ca67 100644
@@ -211,6 +211,12 @@ static const struct arm64_cpu_capabilities arm64_repeat_tlbi_list[] = {
 		/* Kryo4xx Gold (rcpe to rfpe) => (r0p0 to r3p0) */
 		ERRATA_MIDR_RANGE(MIDR_QCOM_KRYO_4XX_GOLD, 0xc, 0xe, 0xf, 0xe),
 	},
+#endif
+#ifdef CONFIG_ARM64_ERRATUM_2441009
+	{
+		/* Cortex-A510 r0p0 -> r1p1. Fixed in r1p2 */
+		ERRATA_MIDR_RANGE(MIDR_CORTEX_A510, 0, 0, 1, 1),
+	},
 #endif
 	{},
 };
@@ -488,7 +494,7 @@ const struct arm64_cpu_capabilities arm64_errata[] = {
 #endif
 #ifdef CONFIG_ARM64_WORKAROUND_REPEAT_TLBI
 	{
-		.desc = "Qualcomm erratum 1009, or ARM erratum 1286807",
+		.desc = "Qualcomm erratum 1009, or ARM erratum 1286807, 2441009",
 		.capability = ARM64_WORKAROUND_REPEAT_TLBI,
 		.type = ARM64_CPUCAP_LOCAL_CPU_ERRATUM,
 		.matches = cpucap_multi_entry_cap_matches,
author
#1 lore raw

Hi Greg,

I send this mail through git send-email from kernel guide(https://www.kernel.org/doc/html/v4.10/process/email-clients.html). Would you help to confirm this could be applied on linux-5.15.y? Great thanks.

- Lucas

Quoted 90 lines · Lucas Wei

From: James Morse <[email protected]>

Cortex-A510 is affected by an erratum where in rare circumstances the CPUs may not handle a race between a break-before-make sequence on one CPU, and another CPU accessing the same page. This could allow a store to a page that has been unmapped.

Work around this by adding the affected CPUs to the list that needs TLB sequences to be done twice.

---
 Documentation/arm64/silicon-errata.rst |  2 ++
 arch/arm64/Kconfig                     | 17 +++++++++++++++++
 arch/arm64/kernel/cpu_errata.c         |  8 +++++++-
 3 files changed, 26 insertions(+), 1 deletion(-)
Documentation/arm64/silicon-errata.rst +2 -0
index 0b4235b1f8c4..33b04db8408f 100644
@@ -106,6 +106,8 @@ stable kernels.
 +----------------+-----------------+-----------------+-----------------------------+
 | ARM            | Cortex-A510     | #2077057        | ARM64_ERRATUM_2077057       |
 +----------------+-----------------+-----------------+-----------------------------+
+| ARM            | Cortex-A510     | #2441009        | ARM64_ERRATUM_2441009       |
++----------------+-----------------+-----------------+-----------------------------+
 | ARM            | Cortex-A710     | #2119858        | ARM64_ERRATUM_2119858       |
 +----------------+-----------------+-----------------+-----------------------------+
 | ARM            | Cortex-A710     | #2054223        | ARM64_ERRATUM_2054223       |
arch/arm64/Kconfig +17 -0
index a5d1b561ed53..001eaba5a6b4 100644
@@ -838,6 +838,23 @@ config ARM64_ERRATUM_2224489
 
          If unsure, say Y.
 
+config ARM64_ERRATUM_2441009
+       bool "Cortex-A510: Completion of affected memory accesses might not be guaranteed by completion of a TLBI"
+       default y
+       select ARM64_WORKAROUND_REPEAT_TLBI
+       help
+         This option adds a workaround for ARM Cortex-A510 erratum #2441009.
+
+         Under very rare circumstances, affected Cortex-A510 CPUs
+         may not handle a race between a break-before-make sequence on one
+         CPU, and another CPU accessing the same page. This could allow a
+         store to a page that has been unmapped.
+
+         Work around this by adding the affected CPUs to the list that needs
+         TLB sequences to be done twice.
+
+         If unsure, say Y.
+
 config ARM64_ERRATUM_2064142
        bool "Cortex-A510: 2064142: workaround TRBE register writes while disabled"
        depends on CORESIGHT_TRBE
arch/arm64/kernel/cpu_errata.c +7 -2
index 6b92989f4cc2..aa9609e6ca67 100644
@@ -211,6 +211,12 @@ static const struct arm64_cpu_capabilities arm64_repeat_tlbi_list[] = {
                /* Kryo4xx Gold (rcpe to rfpe) => (r0p0 to r3p0) */
                ERRATA_MIDR_RANGE(MIDR_QCOM_KRYO_4XX_GOLD, 0xc, 0xe, 0xf, 0xe),
        },
+#endif
+#ifdef CONFIG_ARM64_ERRATUM_2441009
+       {
+               /* Cortex-A510 r0p0 -> r1p1. Fixed in r1p2 */
+               ERRATA_MIDR_RANGE(MIDR_CORTEX_A510, 0, 0, 1, 1),
+       },
 #endif
        {},
 };
@@ -488,7 +494,7 @@ const struct arm64_cpu_capabilities arm64_errata[] = {
 #endif
 #ifdef CONFIG_ARM64_WORKAROUND_REPEAT_TLBI
        {
-               .desc = "Qualcomm erratum 1009, or ARM erratum 1286807",
+               .desc = "Qualcomm erratum 1009, or ARM erratum 1286807, 2441009",
                .capability = ARM64_WORKAROUND_REPEAT_TLBI,
                .type = ARM64_CPUCAP_LOCAL_CPU_ERRATUM,
                .matches = cpucap_multi_entry_cap_matches,
--
Signature · 4 lines
Lucas Wei
Embedded Software Engineer
[email protected]
0287260408
maintainer
#2 lore raw

A: http://en.wikipedia.org/wiki/Top_post
Q: Were do I find info about this thing called top-posting?
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

A: No.
Q: Should I include quotations after my reply?

http://daringfireball.net/2007/07/on_top

Quoted 5 lines · Lucas Wei

Hi Greg,

I send this mail through git send-email from kernel guide(https://www.kernel.org/doc/html/v4.10/process/email-clients.html). Would you help to confirm this could be applied on linux-5.15.y? Great thanks.

See my other response, what git id is this in Linus's tree?

Always give us context on what to do please.

thanks,

greg k-h

maintainer
#3 lore raw
In reply to Lucas Wei
Quoted 10 lines · Lucas Wei

From: James Morse <[email protected]>

Cortex-A510 is affected by an erratum where in rare circumstances the CPUs may not handle a race between a break-before-make sequence on one CPU, and another CPU accessing the same page. This could allow a store to a page that has been unmapped.

Work around this by adding the affected CPUs to the list that needs TLB sequences to be done twice.

Quoted 8 lines
---
 Documentation/arm64/silicon-errata.rst |  2 ++
 arch/arm64/Kconfig                     | 17 +++++++++++++++++
 arch/arm64/kernel/cpu_errata.c         |  8 +++++++-
 3 files changed, 26 insertions(+), 1 deletion(-)

What is the upstream commit id of this patch in Linus's tree, and what tree(s) do you want it applied to?

Always be specific, remember, some of us have to deal with over a thousand emails a day...

thanks,

greg k-h