Skip to contentSkip to conversation

arm(64)/kvm: improve the documentation about HVC calls

PATCH In review Pingfan Liu 2 participants 4 messages Aug 12, 2020 – Aug 28, 2020

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

PATCH 1/1 arm(64)/kvm: improve the documentation about HVC calls

Both arm and arm64 kernel entry point have the following prerequisite:
MMU = off, D-cache = off, I-cache = dont care.

HVC_SOFT_RESTART call should meet this prerequisite before jumping to the new kernel.

Furthermore, on arm64, el2_setup doesn't set I+C bits and keeps EL2 MMU off, and KVM resets them when its unload. These are achieved by HVC_RESET_VECTORS call.

Improve the document.

Cc James Morse <[email protected]>
Cc Russell King <[email protected]>
Cc Marc Zyngier <[email protected]>
Cc Julien Thierry <[email protected]>
Cc Suzuki K Poulose <[email protected]>

To: [email protected]

---
 Documentation/virt/kvm/arm/hyp-abi.rst | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
Documentation/virt/kvm/arm/hyp-abi.rst +3 -3
index d9eba93..a95bc30 100644
@@ -40,9 +40,9 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
 
 * ::
 
-    r0/x0 = HVC_RESET_VECTORS
+    x0 = HVC_RESET_VECTORS (arm64 only)
 
-  Turn HYP/EL2 MMU off, and reset HVBAR/VBAR_EL2 to the initials
+  Disable HYP/EL2 MMU and D-cache, and reset HVBAR/VBAR_EL2 to the initials
   stubs' exception vector value. This effectively disables an existing
   hypervisor.
 
@@ -54,7 +54,7 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
     x3 = x1's value when entering the next payload (arm64)
     x4 = x2's value when entering the next payload (arm64)
 
-  Mask all exceptions, disable the MMU, move the arguments into place
+  Mask all exceptions, disable the MMU and D-cache, move the arguments into place
   (arm64 only), and jump to the restart address while at HYP/EL2. This
   hypercall is not expected to return to its caller.
 
Signature · 1 line
2.7.5
#1 lore raw

This issue is detected by Morse
(https://lore.kernel.org/linux-arm-kernel/[email protected]/)
during the discussion for my patch.

I am not quite sure about the arm, but based on the note at the head of arm/kernel/head.S, things should go that way. If any mistake, please correct me.

Thanks,
Pingfan

Quoted 54 lines · Pingfan Liu

Both arm and arm64 kernel entry point have the following prerequisite:
MMU = off, D-cache = off, I-cache = dont care.

HVC_SOFT_RESTART call should meet this prerequisite before jumping to the new kernel.

Furthermore, on arm64, el2_setup doesn't set I+C bits and keeps EL2 MMU off, and KVM resets them when its unload. These are achieved by HVC_RESET_VECTORS call.

Improve the document.

Cc James Morse <[email protected]>
Cc Russell King <[email protected]>
Cc Marc Zyngier <[email protected]>
Cc Julien Thierry <[email protected]>
Cc Suzuki K Poulose <[email protected]>

To: [email protected]

---
 Documentation/virt/kvm/arm/hyp-abi.rst | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
Documentation/virt/kvm/arm/hyp-abi.rst +3 -4
index d9eba93..a95bc30 100644
@@ -40,9 +40,9 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
 
 * ::
 
-    r0/x0 = HVC_RESET_VECTORS
+    x0 = HVC_RESET_VECTORS (arm64 only)
 
-  Turn HYP/EL2 MMU off, and reset HVBAR/VBAR_EL2 to the initials
+  Disable HYP/EL2 MMU and D-cache, and reset HVBAR/VBAR_EL2 to the initials
   stubs' exception vector value. This effectively disables an existing
   hypervisor.
 
@@ -54,7 +54,7 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
     x3 = x1's value when entering the next payload (arm64)
     x4 = x2's value when entering the next payload (arm64)
 
-  Mask all exceptions, disable the MMU, move the arguments into place
+  Mask all exceptions, disable the MMU and D-cache, move the arguments into place
   (arm64 only), and jump to the restart address while at HYP/EL2. This
   hypercall is not expected to return to its caller.
 
--

2.7.5

#2 lore raw
In reply to Pingfan Liu

Hi Pingfan,

Quoted 5 lines · Pingfan Liu

Both arm and arm64 kernel entry point have the following prerequisite:
MMU = off, D-cache = off, I-cache = dont care.

HVC_SOFT_RESTART call should meet this prerequisite before jumping to the new kernel.

I think you have this the wrong way up. This should describe what HVC_SOFT_RESTART does.

We want to remove some extra work kexec does on arm64, and both implementations of HVC_SOFT_RESTART on arm64 already do what we need. The change here should be to document that the D/I bits are cleared after a HVC_SOFT_RESTART on arm64.

Quoted 12 lines

Furthermore, on arm64, el2_setup doesn't set I+C bits and keeps EL2 MMU off, and KVM resets them when its unload. These are achieved by HVC_RESET_VECTORS call.

Improve the document.

Documentation/virt/kvm/arm/hyp-abi.rst +0 -0
index d9eba93..a95bc30 100644

Documentation/virt/kvm/arm/hyp-abi.rst these functions (see arch/arm{,64}/include/asm/virt.h): · line 47

Quoted 10 lines
Documentation/virt/kvm/arm/hyp-abi.rst +2 -2
@@ -40,9 +40,9 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
 
 * ::
 
-    r0/x0 = HVC_RESET_VECTORS
+    x0 = HVC_RESET_VECTORS (arm64 only)
 
-  Turn HYP/EL2 MMU off, and reset HVBAR/VBAR_EL2 to the initials
+  Disable HYP/EL2 MMU and D-cache, and reset HVBAR/VBAR_EL2 to the initials
   stubs' exception vector value. This effectively disables an existing
   hypervisor.

I don't think we should remove this. KVM on 32bit was the only implementer, but if there ever is another, this is how it should work.

Documentation/virt/kvm/arm/hyp-abi.rst these functions (see arch/arm{,64}/include/asm/virt.h): · line 59

Quoted 8 lines
Documentation/virt/kvm/arm/hyp-abi.rst +1 -1
@@ -54,7 +54,7 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
     x3 = x1's value when entering the next payload (arm64)
     x4 = x2's value when entering the next payload (arm64)
 
-  Mask all exceptions, disable the MMU, move the arguments into place
+  Mask all exceptions, disable the MMU and D-cache, move the arguments into place
   (arm64 only), and jump to the restart address while at HYP/EL2. This
   hypercall is not expected to return to its caller.

(I don't think disable the D-cache is what the bit does, it forces the attributes that are used for a data access).

Please just describe this as the on arm64 the D and I bits are cleared. (it might be true on 32bit, I can't work the assembly out).

Thanks,

James

_______________________________________________
kvmarm mailing list
[email protected]
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

#3 lore raw
Quoted 11 lines · James Morse

Hi Pingfan,

Quoted 5 lines · Pingfan Liu

Both arm and arm64 kernel entry point have the following prerequisite:
MMU = off, D-cache = off, I-cache = dont care.

HVC_SOFT_RESTART call should meet this prerequisite before jumping to the new kernel.

I think you have this the wrong way up. This should describe what HVC_SOFT_RESTART does.

Yes, it is a wrong way.

Quoted 45 lines

We want to remove some extra work kexec does on arm64, and both implementations of HVC_SOFT_RESTART on arm64 already do what we need. The change here should be to document that the D/I bits are cleared after a HVC_SOFT_RESTART on arm64.

Quoted 12 lines

Furthermore, on arm64, el2_setup doesn't set I+C bits and keeps EL2 MMU off, and KVM resets them when its unload. These are achieved by HVC_RESET_VECTORS call.

Improve the document.

Documentation/virt/kvm/arm/hyp-abi.rst +0 -0
index d9eba93..a95bc30 100644

Documentation/virt/kvm/arm/hyp-abi.rst these functions (see arch/arm{,64}/include/asm/virt.h): · line 47

Quoted 10 lines
Documentation/virt/kvm/arm/hyp-abi.rst +2 -2
@@ -40,9 +40,9 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
 
 * ::
 
-    r0/x0 = HVC_RESET_VECTORS
+    x0 = HVC_RESET_VECTORS (arm64 only)
 
-  Turn HYP/EL2 MMU off, and reset HVBAR/VBAR_EL2 to the initials
+  Disable HYP/EL2 MMU and D-cache, and reset HVBAR/VBAR_EL2 to the initials
   stubs' exception vector value. This effectively disables an existing
   hypervisor.

I don't think we should remove this. KVM on 32bit was the only implementer, but if there ever is another, this is how it should work.

Documentation/virt/kvm/arm/hyp-abi.rst these functions (see arch/arm{,64}/include/asm/virt.h): · line 59

Quoted 8 lines
Documentation/virt/kvm/arm/hyp-abi.rst +1 -1
@@ -54,7 +54,7 @@ these functions (see arch/arm{,64}/include/asm/virt.h):
     x3 = x1's value when entering the next payload (arm64)
     x4 = x2's value when entering the next payload (arm64)
 
-  Mask all exceptions, disable the MMU, move the arguments into place
+  Mask all exceptions, disable the MMU and D-cache, move the arguments into place
   (arm64 only), and jump to the restart address while at HYP/EL2. This
   hypercall is not expected to return to its caller.

(I don't think disable the D-cache is what the bit does, it forces the attributes that are used for a data access).

Quoted 1 line

Please just describe this as the on arm64 the D and I bits are cleared.

OK, I will do it.

Quoted 1 line

(it might be true on 32bit, I can't work the assembly out).

I will leave 32bit as it is.

Thanks,
Pingfan