arm64: kernel: fix PMUv3 registers unconditional access
PATCH 1/1 arm64: kernel: fix PMUv3 registers unconditional access
The Performance Monitors extension is an optional feature of the AArch64 architecture, therefore, in order to access Performance Monitors registers safely, the kernel should detect the PMUv3 unit presence through the ID_AA64DFR0_EL1 register PMUVer field before accessing them.
This patch implements a guard by reading the ID_AA64DFR0_EL1 register PMUVer field to detect the PMUv3 presence and prevent accessing PMUv3 system registers if the Performance Monitors extension is not implemented in the core.
---
Based on arm64 for-next/perf branch.
Tested on QEMU and Juno, I checked that the reported PMUVer field is correct on both A57 and A53 (ie == 0x1), it should leave behaviour unchanged on platforms implementing PMUv3.
arch/arm64/kernel/head.S | 5 +++++ arch/arm64/mm/proc-macros.S | 12 ++++++++++++ arch/arm64/mm/proc.S | 4 ++-- 3 files changed, 19 insertions(+), 2 deletions(-)
arch/arm64/kernel/head.S +5 -0
@@ -512,9 +512,14 @@ CPU_LE( movk x0, #0x30d0, lsl #16 ) // Clear EE and E0E on LE systems #endif /* EL2 debug */ + mrs x0, id_aa64dfr0_el1 // Check ID_AA64DFR0_EL1 PMUVer + ubfx x0, x0, #8, #4 + cmp x0, #1 + b.ne 4f // Skip if no PMUv3 present mrs x0, pmcr_el0 // Disable debug access traps ubfx x0, x0, #11, #5 // to EL2 and allow access to msr mdcr_el2, x0 // all PMU counters from EL1 +4: /* Stage-2 translation */ msr vttbr_el2, xzr
arch/arm64/mm/proc-macros.S +12 -0
@@ -62,3 +62,15 @@ bfi \valreg, \tmpreg, #TCR_T0SZ_OFFSET, #TCR_TxSZ_WIDTH #endif .endm + +/* + * reset_pmuserenr_el0 - reset PMUSERENR_EL0 if PMUv3 present + */ + .macro reset_pmuserenr_el0, tmpreg + mrs \tmpreg, id_aa64dfr0_el1 // Check ID_AA64DFR0_EL1 PMUVer + ubfx \tmpreg, \tmpreg, #8, #4 + cmp \tmpreg, #1 // Skip if no PMUv3 present + b.ne 9000f + msr pmuserenr_el0, xzr // Disable PMU access from EL0 +9000: + .endm
arch/arm64/mm/proc.S +2 -2
@@ -117,7 +117,7 @@ ENTRY(cpu_do_resume) */ ubfx x11, x11, #1, #1 msr oslar_el1, x11 - msr pmuserenr_el0, xzr // Disable PMU access from EL0 + reset_pmuserenr_el0 x0 // Disable PMU access from EL0 mov x0, x12 dsb nsh // Make sure local tlb invalidation completed isb
@@ -156,7 +156,7 @@ ENTRY(__cpu_setup) msr cpacr_el1, x0 // Enable FP/ASIMD mov x0, #1 << 12 // Reset mdscr_el1 and disable msr mdscr_el1, x0 // access to the DCC from EL0 - msr pmuserenr_el0, xzr // Disable PMU access from EL0 + reset_pmuserenr_el0 x0 // Disable PMU access from EL0 /* * Memory region attributes for LPAE: *
Signature · 1 line
2.5.1
PATCH 1/1 arm64: kernel: fix PMUv3 registers unconditional access
Hi Lorenzo,
Quoted 33 lines · Lorenzo Pieralisi
The Performance Monitors extension is an optional feature of the AArch64 architecture, therefore, in order to access Performance Monitors registers safely, the kernel should detect the PMUv3 unit presence through the
ID_AA64DFR0_EL1register PMUVer field before accessing them.This patch implements a guard by reading the
ID_AA64DFR0_EL1register PMUVer field to detect the PMUv3 presence and prevent accessing PMUv3 system registers if the Performance Monitors extension is not implemented in the core.Signed-off-by Lorenzo Pieralisi <[email protected]>Reported-by Guenter Roeck <[email protected]>Cc Will Deacon <[email protected]>Cc Peter Maydell <[email protected]>Cc Mark Rutland <[email protected]>---
Based on arm64 for-next/perf branch.Tested on QEMU and Juno, I checked that the reported PMUVer field is correct on both A57 and A53 (ie ==
0x1), it should leave behaviour unchanged on platforms implementing PMUv3.arch/arm64/kernel/head.S | 5 +++++ arch/arm64/mm/proc-macros.S | 12 ++++++++++++ arch/arm64/mm/proc.S | 4 ++-- 3 files changed, 19 insertions(+), 2 deletions(-)arch/arm64/kernel/head.S +0
-0
arch/arm64/kernel/head.S CPU_LE( movk x0, #0x30d0, lsl #16 ) // Clear EE and E0E on LE systems · line 518
Quoted 7 lines
arch/arm64/kernel/head.S +4
-0@@ -512,9 +512,14 @@ CPU_LE( movk x0, #0x30d0, lsl #16 ) // Clear EE and E0E on LE systems #endif /* EL2 debug */ + mrs x0, id_aa64dfr0_el1 // Check ID_AA64DFR0_EL1 PMUVer + ubfx x0, x0, #8, #4 + cmp x0, #1 + b.ne 4f // Skip if no PMUv3 present
This will fail if and when PMUVer gets newer revisions of the PMU architecture (e.g. value 2 to indicate some extended PMU). It looks like we should be treating it as a signed 4-bit field, so we can use sbfx to extract a signed value and then we know the PMU is not present if the value is (signed) less than 1.
Will
PATCH 1/1 arm64: kernel: fix PMUv3 registers unconditional access
On Fri, Jan 08, 2016 at 03:33:00PM +0000, Will Deacon wrote:
[...]
Quoted 11 lines
Review
Quoted 5 lines
(quoted diff) +4
-0/* EL2 debug */ + mrs x0, id_aa64dfr0_el1 // Check ID_AA64DFR0_EL1 PMUVer + ubfx x0, x0, #8, #4 + cmp x0, #1 + b.ne 4f // Skip if no PMUv3 presentThis will fail if and when PMUVer gets newer revisions of the PMU architecture (e.g. value 2 to indicate some extended PMU). It looks like we should be treating it as a signed 4-bit field, so we can use sbfx to extract a signed value and then we know the PMU is not present if the value is (signed) less than 1.
Sorry about that. Updated patch below, hopefully this time I have got it right (I restested on Juno and QEMU and kept Guenter tested-by tag).
Please let me know.
Thanks,
Lorenzo
-- >8 --
Subject: [PATCH] arm64: kernel: fix PMUv3 registers unconditional access
The Performance Monitors extension is an optional feature of the AArch64 architecture, therefore, in order to access Performance Monitors registers safely, the kernel should detect the PMUv3 unit presence through the ID_AA64DFR0_EL1 register PMUVer field before accessing them.
This patch implements a guard by reading the ID_AA64DFR0_EL1 register PMUVer field to detect the PMUv3 presence and prevent accessing PMUv3 system registers if the Performance Monitors extension is not implemented in the core.
--- arch/arm64/kernel/head.S | 5 +++++ arch/arm64/mm/proc-macros.S | 12 ++++++++++++ arch/arm64/mm/proc.S | 4 ++-- 3 files changed, 19 insertions(+), 2 deletions(-)
arch/arm64/kernel/head.S +5 -0
@@ -512,9 +512,14 @@ CPU_LE( movk x0, #0x30d0, lsl #16 ) // Clear EE and E0E on LE systems #endif /* EL2 debug */ + mrs x0, id_aa64dfr0_el1 // Check ID_AA64DFR0_EL1 PMUVer + sbfx x0, x0, #8, #4 + cmp x0, #1 + b.lt 4f // Skip if no PMU present mrs x0, pmcr_el0 // Disable debug access traps ubfx x0, x0, #11, #5 // to EL2 and allow access to msr mdcr_el2, x0 // all PMU counters from EL1 +4: /* Stage-2 translation */ msr vttbr_el2, xzr
arch/arm64/mm/proc-macros.S +12 -0
@@ -62,3 +62,15 @@ bfi \valreg, \tmpreg, #TCR_T0SZ_OFFSET, #TCR_TxSZ_WIDTH #endif .endm + +/* + * reset_pmuserenr_el0 - reset PMUSERENR_EL0 if PMUv3 present + */ + .macro reset_pmuserenr_el0, tmpreg + mrs \tmpreg, id_aa64dfr0_el1 // Check ID_AA64DFR0_EL1 PMUVer + sbfx \tmpreg, \tmpreg, #8, #4 + cmp \tmpreg, #1 // Skip if no PMU present + b.lt 9000f + msr pmuserenr_el0, xzr // Disable PMU access from EL0 +9000: + .endm
arch/arm64/mm/proc.S +2 -2
@@ -117,7 +117,7 @@ ENTRY(cpu_do_resume) */ ubfx x11, x11, #1, #1 msr oslar_el1, x11 - msr pmuserenr_el0, xzr // Disable PMU access from EL0 + reset_pmuserenr_el0 x0 // Disable PMU access from EL0 mov x0, x12 dsb nsh // Make sure local tlb invalidation completed isb
@@ -156,7 +156,7 @@ ENTRY(__cpu_setup) msr cpacr_el1, x0 // Enable FP/ASIMD mov x0, #1 << 12 // Reset mdscr_el1 and disable msr mdscr_el1, x0 // access to the DCC from EL0 - msr pmuserenr_el0, xzr // Disable PMU access from EL0 + reset_pmuserenr_el0 x0 // Disable PMU access from EL0 /* * Memory region attributes for LPAE: *
Signature · 1 line
2.5.1