Skip to contentSkip to conversation

arm64: KVM: Add XXX UAPI notes for swapped registers

PATCH In review Andrew Jones 2 participants 4 messages Jan 20, 2020

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 arm64: KVM: Add XXX UAPI notes for swapped registers

Two UAPI system register IDs do not derive their values from the ARM system register encodings. This is because their values were accidentally swapped. As the IDs are API, they cannot be changed. Add XXX notes to point them out.

---
 Documentation/virt/kvm/api.txt    |  8 ++++++++
 arch/arm64/include/uapi/asm/kvm.h | 11 +++++++++--
 2 files changed, 17 insertions(+), 2 deletions(-)
Documentation/virt/kvm/api.txt +8 -0
index ebb37b34dcfc..11556fc457c3 100644
@@ -2196,6 +2196,14 @@ arm64 CCSIDR registers are demultiplexed by CSSELR value:
 arm64 system registers have the following id bit patterns:
   0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>
 
+XXX: Two system register IDs do not follow the specified pattern.  These
+     are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to
+     system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively.  These
+     two had their values accidentally swapped, which means TIMER_CVAL is
+     derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is
+     derived from the register encoding for CNTV_CVAL_EL0.  As this is
+     API, it must remain this way.
+
 arm64 firmware pseudo-registers have the following bit pattern:
   0x6030 0000 0014 <regno:16>
 
arch/arm64/include/uapi/asm/kvm.h +9 -2
index 820e5751ada7..c72387e48851 100644
@@ -220,10 +220,17 @@ struct kvm_vcpu_events {
 #define KVM_REG_ARM_PTIMER_CVAL		ARM64_SYS_REG(3, 3, 14, 2, 2)
 #define KVM_REG_ARM_PTIMER_CNT		ARM64_SYS_REG(3, 3, 14, 0, 1)
 
-/* EL0 Virtual Timer Registers */
+/*
+ * EL0 Virtual Timer Registers
+ *
+ * XXX: KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT are not defined
+ *      with the appropriate register encodings.  Their values have been
+ *      accidentally swapped.  As this is set API, the definitions here
+ *      must be used, rather than ones derived from the encodings.
+ */
 #define KVM_REG_ARM_TIMER_CTL		ARM64_SYS_REG(3, 3, 14, 3, 1)
-#define KVM_REG_ARM_TIMER_CNT		ARM64_SYS_REG(3, 3, 14, 3, 2)
 #define KVM_REG_ARM_TIMER_CVAL		ARM64_SYS_REG(3, 3, 14, 0, 2)
+#define KVM_REG_ARM_TIMER_CNT		ARM64_SYS_REG(3, 3, 14, 3, 2)
 
 /* KVM-as-firmware specific pseudo-registers */
 #define KVM_REG_ARM_FW			(0x0014 << KVM_REG_ARM_COPROC_SHIFT)
Signature · 1 line
2.21.1
maintainer
#1 lore raw

Hi Andrew,

Many thanks for this. Comments below.

Quoted 31 lines · Andrew Jones

Two UAPI system register IDs do not derive their values from the ARM system register encodings. This is because their values were accidentally swapped. As the IDs are API, they cannot be changed. Add XXX notes to point them out.

---
 Documentation/virt/kvm/api.txt    |  8 ++++++++
 arch/arm64/include/uapi/asm/kvm.h | 11 +++++++++--
 2 files changed, 17 insertions(+), 2 deletions(-)
Documentation/virt/kvm/api.txt +0 -0
index ebb37b34dcfc..11556fc457c3 100644
@@ -2196,6 +2196,14 @@ arm64 CCSIDR registers are demultiplexed by 

CSSELR value:
arm64 system registers have the following id bit patterns:

   0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>

+XXX: Two system register IDs do not follow the specified pattern. These
+ are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map
to
+ system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively.
These
+ two had their values accidentally swapped, which means TIMER_CVAL
is
+ derived from the register encoding for CNTVCT_EL0 and TIMER_CNT
is

Documentation/virt/kvm/api.txt arm64 CCSIDR registers are demultiplexed by CSSELR value: · line 2205

Quoted 2 lines
Documentation/virt/kvm/api.txt +2 -0
+     derived from the register encoding for CNTV_CVAL_EL0.  As this is
+     API, it must remain this way.

Is 'XXX' an establiched way of documenting this kind of misfeature? I couldn't find any other occurrence in Documentation, but I haven't searched very hard.

If nobody has a better idea, I'll queue it as is.

Thanks,

        M.
Signature · 1 line
Jazz is not dead. It just smells funny...
#2 lore raw
Quoted 27 lines · Marc Zyngier

Hi Andrew,

Many thanks for this. Comments below.

Quoted 22 lines · Andrew Jones

Two UAPI system register IDs do not derive their values from the ARM system register encodings. This is because their values were accidentally swapped. As the IDs are API, they cannot be changed. Add XXX notes to point them out.

---
 Documentation/virt/kvm/api.txt    |  8 ++++++++
 arch/arm64/include/uapi/asm/kvm.h | 11 +++++++++--
 2 files changed, 17 insertions(+), 2 deletions(-)
Documentation/virt/kvm/api.txt +0 -0
index ebb37b34dcfc..11556fc457c3 100644
@@ -2196,6 +2196,14 @@ arm64 CCSIDR registers are demultiplexed by

CSSELR value:
arm64 system registers have the following id bit patterns:

   0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>
Quoted 13 lines

Documentation/virt/kvm/api.txt arm64 CCSIDR registers are demultiplexed by CSSELR value: · line 2205

Quoted 9 lines

+XXX: Two system register IDs do not follow the specified pattern. These

Documentation/virt/kvm/api.txt +3 -0
+     are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to
+     system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively.  These
+     two had their values accidentally swapped, which means TIMER_CVAL

is

Documentation/virt/kvm/api.txt +3 -0
+     derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is
+     derived from the register encoding for CNTV_CVAL_EL0.  As this is
+     API, it must remain this way.

Is 'XXX' an establiched way of documenting this kind of misfeature? I couldn't find any other occurrence in Documentation, but I haven't searched very hard.

I didn't find anything claiming it was the standard way of doing it. I also considered using 'NOTE', but I was afraid it wouldn't stand out as a "problem" enough. And, even though 'BUG' would certainly stand out, I felt it implied we should be posting a fix.

Anyway, I'd be happy to change it to whatever the consensus is.

Thanks,
drew

Quoted 9 lines

If nobody has a better idea, I'll queue it as is.

Thanks,

        M.
Signature · 1 line
Jazz is not dead. It just smells funny...
maintainer
#3 lore raw
Quoted 42 lines · Andrew Jones
Quoted 27 lines · Marc Zyngier

Hi Andrew,

Many thanks for this. Comments below.

Quoted 22 lines · Andrew Jones

Two UAPI system register IDs do not derive their values from the ARM system register encodings. This is because their values were accidentally swapped. As the IDs are API, they cannot be changed. Add XXX notes to point them out.

---
 Documentation/virt/kvm/api.txt    |  8 ++++++++
 arch/arm64/include/uapi/asm/kvm.h | 11 +++++++++--
 2 files changed, 17 insertions(+), 2 deletions(-)
Documentation/virt/kvm/api.txt +0 -0
index ebb37b34dcfc..11556fc457c3 100644
@@ -2196,6 +2196,14 @@ arm64 CCSIDR registers are demultiplexed by

CSSELR value:
arm64 system registers have the following id bit patterns:

   0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>
Quoted 13 lines

Documentation/virt/kvm/api.txt arm64 CCSIDR registers are demultiplexed by CSSELR value: · line 2205

Quoted 9 lines

+XXX: Two system register IDs do not follow the specified pattern. These

Documentation/virt/kvm/api.txt +3 -0
+     are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to
+     system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively.  These
+     two had their values accidentally swapped, which means TIMER_CVAL

is

Documentation/virt/kvm/api.txt +3 -0
+     derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is
+     derived from the register encoding for CNTV_CVAL_EL0.  As this is
+     API, it must remain this way.

Is 'XXX' an establiched way of documenting this kind of misfeature? I couldn't find any other occurrence in Documentation, but I haven't searched very hard.

Quoted 6 lines

I didn't find anything claiming it was the standard way of doing it. I also considered using 'NOTE', but I was afraid it wouldn't stand out as a "problem" enough. And, even though 'BUG' would certainly stand out, I felt it implied we should be posting a fix.

Anyway, I'd be happy to change it to whatever the consensus is.

My personal preference would be for a big fat "*WARNING*" (with an optional <blink> attribute), and a link to a picture of me wearing a brown paper bag on my head.

But maybe "WARNING" is enough. Dunno. Anyway, if nobody comes up with a better idea by tomorrow, I'll take it as is.

Thanks,

        M.
Signature · 1 line
Jazz is not dead. It just smells funny...