Skip to contentSkip to conversation

KVM: x86/xsave: Remove 'return void' expression for 'void function'

PATCH Applied Like Xu 3 participants 8 messages Oct 7, 2023 – Dec 1, 2023

Subsystems (3)
lore
Share & tools
lore

Like Xu

author
#0 lore raw

PATCH 1/1 KVM: x86/xsave: Remove 'return void' expression for 'void function'

From: Like Xu <[email protected]>

The requested info will be stored in 'guest_xsave->region' referenced by the incoming pointer "struct kvm_xsave *guest_xsave", thus there is no need to explicitly use return void expression for a void function "static void kvm_vcpu_ioctl_x86_get_xsave(...)". The issue is caught with [-Wpedantic].

Fixes 2d287ec65e79 ("x86/fpu: Allow caller to constrain xfeatures when copying to uabi buffer")
---
 arch/x86/kvm/x86.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/kvm/x86.c +2 -2
index fdb2b0e61c43..2571466a317f 100644
@@ -5503,8 +5503,8 @@ static void kvm_vcpu_ioctl_x86_get_xsave2(struct kvm_vcpu *vcpu,
 static void kvm_vcpu_ioctl_x86_get_xsave(struct kvm_vcpu *vcpu,
 					 struct kvm_xsave *guest_xsave)
 {
-	return kvm_vcpu_ioctl_x86_get_xsave2(vcpu, (void *)guest_xsave->region,
-					     sizeof(guest_xsave->region));
+	kvm_vcpu_ioctl_x86_get_xsave2(vcpu, (void *)guest_xsave->region,
+				      sizeof(guest_xsave->region));
 }
 
 static int kvm_vcpu_ioctl_x86_set_xsave(struct kvm_vcpu *vcpu,

base-commit: 86701e115030e020a052216baa942e8547e0b487

Signature · 1 line
2.42.0
Reviewed-by on v1 1/1
#1 lore raw

У сб, 2023-10-07 у 14:40 +0800, Like Xu пише:

Quoted 29 lines

From: Like Xu <[email protected]>

The requested info will be stored in 'guest_xsave->region' referenced by the incoming pointer "struct kvm_xsave *guest_xsave", thus there is no need to explicitly use return void expression for a void function "static void kvm_vcpu_ioctl_x86_get_xsave(...)". The issue is caught with [-Wpedantic].

Fixes 2d287ec65e79 ("x86/fpu: Allow caller to constrain xfeatures when copying to uabi buffer")
---
 arch/x86/kvm/x86.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/kvm/x86.c +2 -2
index fdb2b0e61c43..2571466a317f 100644
@@ -5503,8 +5503,8 @@ static void kvm_vcpu_ioctl_x86_get_xsave2(struct kvm_vcpu *vcpu,
 static void kvm_vcpu_ioctl_x86_get_xsave(struct kvm_vcpu *vcpu,
 					 struct kvm_xsave *guest_xsave)
 {
-	return kvm_vcpu_ioctl_x86_get_xsave2(vcpu, (void *)guest_xsave->region,
-					     sizeof(guest_xsave->region));
+	kvm_vcpu_ioctl_x86_get_xsave2(vcpu, (void *)guest_xsave->region,
+				      sizeof(guest_xsave->region));
 }
 
 static int kvm_vcpu_ioctl_x86_set_xsave(struct kvm_vcpu *vcpu,
Quoted 1 line

base-commit: 86701e115030e020a052216baa942e8547e0b487

Best regards,
Maxim Levitsky

author
#2 lore raw

Emm, did we miss this little fix ?

Quoted 36 lines · Maxim Levitsky

У сб, 2023-10-07 у 14:40 +0800, Like Xu пише:

Quoted 29 lines

From: Like Xu <[email protected]>

The requested info will be stored in 'guest_xsave->region' referenced by the incoming pointer "struct kvm_xsave *guest_xsave", thus there is no need to explicitly use return void expression for a void function "static void kvm_vcpu_ioctl_x86_get_xsave(...)". The issue is caught with [-Wpedantic].

Fixes 2d287ec65e79 ("x86/fpu: Allow caller to constrain xfeatures when copying to uabi buffer")
---
  arch/x86/kvm/x86.c | 4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/kvm/x86.c +2 -2
index fdb2b0e61c43..2571466a317f 100644
@@ -5503,8 +5503,8 @@ static void kvm_vcpu_ioctl_x86_get_xsave2(struct kvm_vcpu *vcpu,
  static void kvm_vcpu_ioctl_x86_get_xsave(struct kvm_vcpu *vcpu,
  					 struct kvm_xsave *guest_xsave)
  {
-	return kvm_vcpu_ioctl_x86_get_xsave2(vcpu, (void *)guest_xsave->region,
-					     sizeof(guest_xsave->region));
+	kvm_vcpu_ioctl_x86_get_xsave2(vcpu, (void *)guest_xsave->region,
+				      sizeof(guest_xsave->region));
  }
 
  static int kvm_vcpu_ioctl_x86_set_xsave(struct kvm_vcpu *vcpu,
Quoted 1 line

base-commit: 86701e115030e020a052216baa942e8547e0b487

Best regards,
Maxim Levitsky

#3 lore raw
Quoted 1 line · Like Xu

Emm, did we miss this little fix ?

No, I have it earmarked, it's just not a priority because it doesn't truly fix anything. Though I suppose it probably makes to apply it for 6.8, waiting one more day to send PULL requests to Paolo isn't a problem.

#4 lore raw
Quoted 6 lines · Sean Christopherson
Quoted 1 line · Like Xu

Emm, did we miss this little fix ?

No, I have it earmarked, it's just not a priority because it doesn't truly fix anything. Though I suppose it probably makes to apply it for 6.8, waiting one more day to send PULL requests to Paolo isn't a problem.

Heh, when I tried to apply this I got reminded of why I held it for later. I want to apply it to kvm-x86/misc, but that's based on ~6.6-rc2 (plus a few KVM patches), i.e. doesn't have the "buggy" commit. I don't want to rebase "misc", nor do I want to create a branch and PULL request for a single trivial commit.

So for logistical reasons, I'm not going apply this right away, but I will make sure it gets into v6.7.

author
#5 lore raw
Quoted 8 lines · Sean Christopherson
Quoted 6 lines · Sean Christopherson
Quoted 1 line · Like Xu

Emm, did we miss this little fix ?

No, I have it earmarked, it's just not a priority because it doesn't truly fix anything. Though I suppose it probably makes to apply it for 6.8, waiting one more day to send PULL requests to Paolo isn't a problem.

Quoted 7 lines

Heh, when I tried to apply this I got reminded of why I held it for later. I want to apply it to kvm-x86/misc, but that's based on ~6.6-rc2 (plus a few KVM patches), i.e. doesn't have the "buggy" commit. I don't want to rebase "misc", nor do I want to create a branch and PULL request for a single trivial commit.

So for logistical reasons, I'm not going apply this right away, but I will make sure it gets into v6.7.

Thanks, and a similar pattern occurs with these functions:

'write_register_operand'
'account_shadowed'
'unaccount_shadowed'
'mtrr_lookup_fixed_next'
'pre_svm_run'
'svm_vcpu_deliver_sipi_vector'

Although the compiler will do the right thing, use 'return void' expression deliberately without grounds for exemption may annoy some CI pipelines.

If you need more cleanup or a new version to cover all these cases above, just let me know.

#6 lore raw
Quoted 26 lines · Like Xu
Quoted 8 lines · Sean Christopherson
Quoted 6 lines · Sean Christopherson
Quoted 1 line · Like Xu

Emm, did we miss this little fix ?

No, I have it earmarked, it's just not a priority because it doesn't truly fix anything. Though I suppose it probably makes to apply it for 6.8, waiting one more day to send PULL requests to Paolo isn't a problem.

Quoted 7 lines

Heh, when I tried to apply this I got reminded of why I held it for later. I want to apply it to kvm-x86/misc, but that's based on ~6.6-rc2 (plus a few KVM patches), i.e. doesn't have the "buggy" commit. I don't want to rebase "misc", nor do I want to create a branch and PULL request for a single trivial commit.

So for logistical reasons, I'm not going apply this right away, but I will make sure it gets into v6.7.

Thanks, and a similar pattern occurs with these functions:

'write_register_operand'
'account_shadowed'
'unaccount_shadowed'
'mtrr_lookup_fixed_next'
'pre_svm_run'
'svm_vcpu_deliver_sipi_vector'

Quoted 5 lines

Although the compiler will do the right thing, use 'return void' expression deliberately without grounds for exemption may annoy some CI pipelines.

If you need more cleanup or a new version to cover all these cases above, just let me know.

I'd rather update the CI pipelines to turn off -Wpedantic. There is zero chance that -Wpedantic will ever get enabled for kernel builds, the kernel is deliberately not ISO C compliant. I have no objection to cleaning up kvm_vcpu_ioctl_x86_get_xsave() because it's an obvious goof and a recent change, but like checkpatch warnings, I don't want to go around "fixing" warnings unless they are actively problematic for humans.

maintainer Applied
#7 lore raw
In reply to Like Xu
Quoted 6 lines · Like Xu

The requested info will be stored in 'guest_xsave->region' referenced by the incoming pointer "struct kvm_xsave *guest_xsave", thus there is no need to explicitly use return void expression for a void function "static void kvm_vcpu_ioctl_x86_get_xsave(...)". The issue is caught with [-Wpedantic].

Applied to kvm-x86 fixes, thanks!

[1/1] KVM: x86/xsave: Remove 'return void' expression for 'void function'

      https://github.com/kvm-x86/linux/commit/ef8d89033c3f

--
https://github.com/kvm-x86/linux/tree/next