Skip to contentSkip to conversation

arm64: cpufeature: Don't treat granule sizes as strict

PATCH Applied Will Deacon 5 participants 5 messages Aug 12, 2019 – Aug 13, 2019

Subsystems (2)
lore
Share & tools
lore

Will Deacon

author maintainer
#0 lore raw

PATCH 1/1 arm64: cpufeature: Don't treat granule sizes as strict

If a CPU doesn't support the page size for which the kernel is configured, then we will complain and refuse to bring it online. For secondary CPUs (and the boot CPU on a system booting with EFI), we will also print an error identifying the mismatch.

Consequently, the only time that the cpufeature code can detect a granule size mismatch is for a granule other than the one that is currently being used. Although we would rather such systems didn't exist, we've unfortunately lost that battle and Kevin reports that on his amlogic S922X (odroid-n2 board) we end up warning and taining with defconfig because 16k pages are not supported by all of the CPUs.

In such a situation, we don't actually care about the feature mismatch, particularly now that KVM only exposes the sanitised view of the CPU registers. Treat the granule fields as non-strict and let Kevin run without a tainted kernel.

Cc Mark Rutland <[email protected]>
Cc Marc Zyngier <[email protected]>
Cc Suzuki Poulose <[email protected]>
Cc Catalin Marinas <[email protected]>
---
 arch/arm64/kernel/cpufeature.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)
arch/arm64/kernel/cpufeature.c +11 -3
index d19d14ba9ae4..b1fdc486aed8 100644
@@ -184,9 +184,17 @@ static const struct arm64_ftr_bits ftr_id_aa64zfr0[] = {
 };
 
 static const struct arm64_ftr_bits ftr_id_aa64mmfr0[] = {
-	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN4_SHIFT, 4, ID_AA64MMFR0_TGRAN4_NI),
-	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN64_SHIFT, 4, ID_AA64MMFR0_TGRAN64_NI),
-	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN16_SHIFT, 4, ID_AA64MMFR0_TGRAN16_NI),
+	/*
+	 * We already refuse to boot CPUs that don't support our configured
+	 * page size, so we can only detect mismatches for a page size other
+	 * than the one we're currently using. Unfortunately, SoCs like this
+	 * exist in the wild so, even though we don't like it, we'll have to go
+	 * along with it and treat them as non-strict.
+	 */
+	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN4_SHIFT, 4, ID_AA64MMFR0_TGRAN4_NI),
+	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN64_SHIFT, 4, ID_AA64MMFR0_TGRAN64_NI),
+	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN16_SHIFT, 4, ID_AA64MMFR0_TGRAN16_NI),
+
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_BIGENDEL0_SHIFT, 4, 0),
 	/* Linux shouldn't care about secure memory */
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_SNSMEM_SHIFT, 4, 0),
Signature · 7 lines
2.11.0


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
reviewer Acked-by on v1 1/1
#1 lore raw
Quoted 17 lines · Will Deacon

If a CPU doesn't support the page size for which the kernel is configured, then we will complain and refuse to bring it online. For secondary CPUs (and the boot CPU on a system booting with EFI), we will also print an error identifying the mismatch.

Consequently, the only time that the cpufeature code can detect a granule size mismatch is for a granule other than the one that is currently being used. Although we would rather such systems didn't exist, we've unfortunately lost that battle and Kevin reports that on his amlogic S922X (odroid-n2 board) we end up warning and taining with defconfig because 16k pages are not supported by all of the CPUs.

In such a situation, we don't actually care about the feature mismatch, particularly now that KVM only exposes the sanitised view of the CPU registers. Treat the granule fields as non-strict and let Kevin run without a tainted kernel.

Quoted 6 lines
Cc Mark Rutland <[email protected]>
Cc Marc Zyngier <[email protected]>
Cc Suzuki Poulose <[email protected]>
Cc Catalin Marinas <[email protected]>

FWIW:

Mark.

Quoted 32 lines
---
 arch/arm64/kernel/cpufeature.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)
arch/arm64/kernel/cpufeature.c +11 -3
index d19d14ba9ae4..b1fdc486aed8 100644
@@ -184,9 +184,17 @@ static const struct arm64_ftr_bits ftr_id_aa64zfr0[] = {
 };
 
 static const struct arm64_ftr_bits ftr_id_aa64mmfr0[] = {
-	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN4_SHIFT, 4, ID_AA64MMFR0_TGRAN4_NI),
-	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN64_SHIFT, 4, ID_AA64MMFR0_TGRAN64_NI),
-	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN16_SHIFT, 4, ID_AA64MMFR0_TGRAN16_NI),
+	/*
+	 * We already refuse to boot CPUs that don't support our configured
+	 * page size, so we can only detect mismatches for a page size other
+	 * than the one we're currently using. Unfortunately, SoCs like this
+	 * exist in the wild so, even though we don't like it, we'll have to go
+	 * along with it and treat them as non-strict.
+	 */
+	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN4_SHIFT, 4, ID_AA64MMFR0_TGRAN4_NI),
+	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN64_SHIFT, 4, ID_AA64MMFR0_TGRAN64_NI),
+	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN16_SHIFT, 4, ID_AA64MMFR0_TGRAN16_NI),
+
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_BIGENDEL0_SHIFT, 4, 0),
 	/* Linux shouldn't care about secure memory */
 	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_SNSMEM_SHIFT, 4, 0),
Signature · 1 line
2.11.0

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Acked-by on v1 1/1
#2 lore raw
In reply to Will Deacon

Hi Will,

Quoted 12 lines · Will Deacon

If a CPU doesn't support the page size for which the kernel is configured, then we will complain and refuse to bring it online. For secondary CPUs (and the boot CPU on a system booting with EFI), we will also print an error identifying the mismatch.

Consequently, the only time that the cpufeature code can detect a granule size mismatch is for a granule other than the one that is currently being used. Although we would rather such systems didn't exist, we've unfortunately lost that battle and Kevin reports that on his amlogic S922X (odroid-n2 board) we end up warning and taining with defconfig because 16k pages are not supported by all of the CPUs.

Quoted 4 lines

In such a situation, we don't actually care about the feature mismatch, particularly now that KVM only exposes the sanitised view of the CPU registers. Treat the granule fields as non-strict and let Kevin run without a tainted kernel.

I think it may be a good idea to mention the "KVM sanitised feature reg" feature as a pre-req for KVM safety in the comment section to alert people backporting this commit to older kernels.

Quoted 22 lines
Cc Mark Rutland <[email protected]>
Cc Marc Zyngier <[email protected]>
Cc Suzuki Poulose <[email protected]>
Cc Catalin Marinas <[email protected]>
---
  arch/arm64/kernel/cpufeature.c | 14 +++++++++++---
  1 file changed, 11 insertions(+), 3 deletions(-)
arch/arm64/kernel/cpufeature.c +0 -3
index d19d14ba9ae4..b1fdc486aed8 100644
@@ -184,9 +184,17 @@ static const struct arm64_ftr_bits ftr_id_aa64zfr0[] = {
  };
 
  static const struct arm64_ftr_bits ftr_id_aa64mmfr0[] = {
-	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN4_SHIFT, 4, ID_AA64MMFR0_TGRAN4_NI),
-	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN64_SHIFT, 4, ID_AA64MMFR0_TGRAN64_NI),
-	ARM64_FTR_BITS(FTR_HIDDEN, FTR_STRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN16_SHIFT, 4, ID_AA64MMFR0_TGRAN16_NI),

arch/arm64/kernel/cpufeature.c static const struct arm64_ftr_bits ftr_id_aa64zfr0[] = { · line 196

Quoted 10 lines
arch/arm64/kernel/cpufeature.c +10 -0
+	/*
+	 * We already refuse to boot CPUs that don't support our configured
+	 * page size, so we can only detect mismatches for a page size other
+	 * than the one we're currently using. Unfortunately, SoCs like this
+	 * exist in the wild so, even though we don't like it, we'll have to go
+	 * along with it and treat them as non-strict.
+	 */
+	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN4_SHIFT, 4, ID_AA64MMFR0_TGRAN4_NI),
+	S_ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN64_SHIFT, 4, ID_AA64MMFR0_TGRAN64_NI),
+	ARM64_FTR_BITS(FTR_HIDDEN, FTR_NONSTRICT, FTR_LOWER_SAFE, ID_AA64MMFR0_TGRAN16_SHIFT, 4, ID_AA64MMFR0_TGRAN16_NI),

Eitherway,

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

maintainer Tested-by on v1 1/1
#3 lore raw
In reply to Will Deacon
Quoted 17 lines · Will Deacon

If a CPU doesn't support the page size for which the kernel is configured, then we will complain and refuse to bring it online. For secondary CPUs (and the boot CPU on a system booting with EFI), we will also print an error identifying the mismatch.

Consequently, the only time that the cpufeature code can detect a granule size mismatch is for a granule other than the one that is currently being used. Although we would rather such systems didn't exist, we've unfortunately lost that battle and Kevin reports that on his amlogic S922X (odroid-n2 board) we end up warning and taining with defconfig because 16k pages are not supported by all of the CPUs.

In such a situation, we don't actually care about the feature mismatch, particularly now that KVM only exposes the sanitised view of the CPU registers. Treat the granule fields as non-strict and let Kevin run without a tainted kernel.

Quoted 6 lines
Cc Mark Rutland <[email protected]>
Cc Marc Zyngier <[email protected]>
Cc Suzuki Poulose <[email protected]>
Cc Catalin Marinas <[email protected]>

Thanks for the fix Will!

Kevin

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

maintainer Applied
#4 lore raw
In reply to Will Deacon
Quoted 17 lines · Will Deacon

If a CPU doesn't support the page size for which the kernel is configured, then we will complain and refuse to bring it online. For secondary CPUs (and the boot CPU on a system booting with EFI), we will also print an error identifying the mismatch.

Consequently, the only time that the cpufeature code can detect a granule size mismatch is for a granule other than the one that is currently being used. Although we would rather such systems didn't exist, we've unfortunately lost that battle and Kevin reports that on his amlogic S922X (odroid-n2 board) we end up warning and taining with defconfig because 16k pages are not supported by all of the CPUs.

In such a situation, we don't actually care about the feature mismatch, particularly now that KVM only exposes the sanitised view of the CPU registers. Treat the granule fields as non-strict and let Kevin run without a tainted kernel.

Quoted 6 lines
Cc Mark Rutland <[email protected]>
Cc Marc Zyngier <[email protected]>
Cc Suzuki Poulose <[email protected]>
Cc Catalin Marinas <[email protected]>

Queued for 5.3-rc5. Thanks.

Signature · 6 lines
Catalin

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel