2.6.32.12: Build warning due to 78ce64a384 / missing in 2.6.33?

2.6.32.12: Build warning due to 78ce64a384 / missing in 2.6.33?

[Jan Kiszka]

[-- Attachment #1: Type: text/plain, Size: 397 bytes --]

Gleb,

I'm getting a build warning with latest 2.6.32.12 due to "Fix segment
descriptor loading". load_segment_descriptor_to_kvm_desct is unused
after that patch. I assume it's just forgotten code and did not
accidentally become unused, right?

The fact that 2.6.33.3 does not generate this makes me wonder why it
obviously lacks the above patch. Not required or not yet queued?

Jan


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 257 bytes --]

Re: 2.6.32.12: Build warning due to 78ce64a384 / missing in 2.6.33?

[Gleb Natapov]

On Mon, Apr 26, 2010 at 07:54:22PM +0200, Jan Kiszka wrote:
> Gleb,
> 
> I'm getting a build warning with latest 2.6.32.12 due to "Fix segment
> descriptor loading". load_segment_descriptor_to_kvm_desct is unused
> after that patch. I assume it's just forgotten code and did not
> accidentally become unused, right?
> 
Yes, the function was use only in one place and was removed by upstream
commit e01c2426149.

> The fact that 2.6.33.3 does not generate this makes me wonder why it
> obviously lacks the above patch. Not required or not yet queued?
> 
Doesn't make much sense to include it in stable-32, but not stable-33. I
think stable-32 just has more attention since it is used by many
distribution as base kernel.

--
			Gleb.

Re: 2.6.32.12: Build warning due to 78ce64a384 / missing in 2.6.33?

[Avi Kivity]

On 04/27/2010 09:17 AM, Gleb Natapov wrote:
>
>> The fact that 2.6.33.3 does not generate this makes me wonder why it
>> obviously lacks the above patch. Not required or not yet queued?
>>
>>      
> Doesn't make much sense to include it in stable-32, but not stable-33. I
> think stable-32 just has more attention since it is used by many
> distribution as base kernel.
>    

This was due to the previous process of mindlessly applying patches to 
stable without testing/looking at them first.  Hopefully the new process 
will deliver better results.

Gleb, can you look at 2.6.33 and see if it is missing something, or 
maybe the warning is gone for another reason?

-- 
Do not meddle in the internals of kernels, for they are subtle and quick to panic.

Re: 2.6.32.12: Build warning due to 78ce64a384 / missing in 2.6.33?

[Gleb Natapov]

On Tue, Apr 27, 2010 at 10:41:21AM +0300, Avi Kivity wrote:
> On 04/27/2010 09:17 AM, Gleb Natapov wrote:
> >
> >>The fact that 2.6.33.3 does not generate this makes me wonder why it
> >>obviously lacks the above patch. Not required or not yet queued?
> >>
> >Doesn't make much sense to include it in stable-32, but not stable-33. I
> >think stable-32 just has more attention since it is used by many
> >distribution as base kernel.
> 
> This was due to the previous process of mindlessly applying patches
> to stable without testing/looking at them first.  Hopefully the new
> process will deliver better results.
> 
> Gleb, can you look at 2.6.33 and see if it is missing something, or
> maybe the warning is gone for another reason?
> 
Already looked at it. 2.6.33-stable misses the patch that cause warning in
2.6.32-stable.

--
			Gleb.

[PATCH 2.6.32.12] KVM: remove unused load_segment_descriptor_to_kvm_desct

[Marcelo Tosatti]

Function is now unused.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 389fc55..2782509 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -4155,18 +4155,6 @@ static u16 get_segment_selector(struct kvm_vcpu *vcpu, int seg)
 	return kvm_seg.selector;
 }
 
-static int load_segment_descriptor_to_kvm_desct(struct kvm_vcpu *vcpu,
-						u16 selector,
-						struct kvm_segment *kvm_seg)
-{
-	struct desc_struct seg_desc;
-
-	if (load_guest_segment_descriptor(vcpu, selector, &seg_desc))
-		return 1;
-	seg_desct_to_kvm_desct(&seg_desc, selector, kvm_seg);
-	return 0;
-}
-
 static int kvm_load_realmode_segment(struct kvm_vcpu *vcpu, u16 selector, int seg)
 {
 	struct kvm_segment segvar = {

Re: [PATCH 2.6.32.12] KVM: remove unused load_segment_descriptor_to_kvm_desct

[Greg KH]

On Tue, Apr 27, 2010 at 11:14:14AM -0300, Marcelo Tosatti wrote:
> 
> Function is now unused.
> 
> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>

Was this patch also upstream?  Why does stable need it?

thanks,

greg k-h

Re: [PATCH 2.6.32.12] KVM: remove unused load_segment_descriptor_to_kvm_desct

[Gleb Natapov]

On Tue, Apr 27, 2010 at 07:43:55AM -0700, Greg KH wrote:
> On Tue, Apr 27, 2010 at 11:14:14AM -0300, Marcelo Tosatti wrote:
> > 
> > Function is now unused.
> > 
> > Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
> 
> Was this patch also upstream?  Why does stable need it?
> 
It is upstream, but as part of another commit. My guess is that when
patches were ported to stable this chunk was lost due to wrong merge
resolving.

--
			Gleb.

Re: [PATCH 2.6.32.12] KVM: remove unused load_segment_descriptor_to_kvm_desct

[Greg KH]

On Tue, Apr 27, 2010 at 05:52:20PM +0300, Gleb Natapov wrote:
> On Tue, Apr 27, 2010 at 07:43:55AM -0700, Greg KH wrote:
> > On Tue, Apr 27, 2010 at 11:14:14AM -0300, Marcelo Tosatti wrote:
> > > 
> > > Function is now unused.
> > > 
> > > Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
> > 
> > Was this patch also upstream?  Why does stable need it?
> > 
> It is upstream, but as part of another commit. My guess is that when
> patches were ported to stable this chunk was lost due to wrong merge
> resolving.

Ok, but as this patch doesn't do anything, why is it needed in -stable?

Also, this type of information (why it isn't upstream), is REQUIRED in
order to be accepted.  Please include it.

thanks,

greg k-h

Re: [PATCH 2.6.32.12] KVM: remove unused load_segment_descriptor_to_kvm_desct

[Marcelo Tosatti]

On Tue, Apr 27, 2010 at 07:56:41AM -0700, Greg KH wrote:
> On Tue, Apr 27, 2010 at 05:52:20PM +0300, Gleb Natapov wrote:
> > On Tue, Apr 27, 2010 at 07:43:55AM -0700, Greg KH wrote:
> > > On Tue, Apr 27, 2010 at 11:14:14AM -0300, Marcelo Tosatti wrote:
> > > > 
> > > > Function is now unused.
> > > > 
> > > > Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
> > > 
> > > Was this patch also upstream?  Why does stable need it?
> > > 
> > It is upstream, but as part of another commit. My guess is that when
> > patches were ported to stable this chunk was lost due to wrong merge
> > resolving.
> 
> Ok, but as this patch doesn't do anything, why is it needed in -stable?

Because it fixes a warning reported on the mailing list.

> Also, this type of information (why it isn't upstream), is REQUIRED in
> order to be accepted.  Please include it.

Sorry about that, will include proper changelog in the future.

Re: [PATCH 2.6.32.12] KVM: remove unused load_segment_descriptor_to_kvm_desct

[Greg KH]

On Tue, Apr 27, 2010 at 01:16:49PM -0300, Marcelo Tosatti wrote:
> On Tue, Apr 27, 2010 at 07:56:41AM -0700, Greg KH wrote:
> > On Tue, Apr 27, 2010 at 05:52:20PM +0300, Gleb Natapov wrote:
> > > On Tue, Apr 27, 2010 at 07:43:55AM -0700, Greg KH wrote:
> > > > On Tue, Apr 27, 2010 at 11:14:14AM -0300, Marcelo Tosatti wrote:
> > > > > 
> > > > > Function is now unused.
> > > > > 
> > > > > Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
> > > > 
> > > > Was this patch also upstream?  Why does stable need it?
> > > > 
> > > It is upstream, but as part of another commit. My guess is that when
> > > patches were ported to stable this chunk was lost due to wrong merge
> > > resolving.
> > 
> > Ok, but as this patch doesn't do anything, why is it needed in -stable?
> 
> Because it fixes a warning reported on the mailing list.

What mailing list?

> > Also, this type of information (why it isn't upstream), is REQUIRED in
> > order to be accepted.  Please include it.
> 
> Sorry about that, will include proper changelog in the future.

Can you resend this one, with such changelog information, so I can apply
it?

thanks,

greg k-h

[PATCH 2.6.32.12] KVM: remove unused load_segment_descriptor_to_kvm_desct

[Marcelo Tosatti]

Commit 78ce64a384 in v2.6.32.12 introduced a warning due to unused
load_segment_descriptor_to_kvm_desct helper, which has been opencoded by
this commit.

On upstream, the helper was removed as part of a different commit.

Remove the now unused function.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 389fc55..2782509 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -4155,18 +4155,6 @@ static u16 get_segment_selector(struct kvm_vcpu *vcpu, int seg)
 	return kvm_seg.selector;
 }
 
-static int load_segment_descriptor_to_kvm_desct(struct kvm_vcpu *vcpu,
-						u16 selector,
-						struct kvm_segment *kvm_seg)
-{
-	struct desc_struct seg_desc;
-
-	if (load_guest_segment_descriptor(vcpu, selector, &seg_desc))
-		return 1;
-	seg_desct_to_kvm_desct(&seg_desc, selector, kvm_seg);
-	return 0;
-}
-
 static int kvm_load_realmode_segment(struct kvm_vcpu *vcpu, u16 selector, int seg)
 {
 	struct kvm_segment segvar = {

patch kvm-remove-unused-load_segment_descriptor_to_kvm_desct.patch added to 2.6.32-stable tree

[gregkh]

This is a note to let you know that we have just queued up the patch titled

    Subject: KVM: remove unused load_segment_descriptor_to_kvm_desct

to the 2.6.32-stable tree.  Its filename is

    kvm-remove-unused-load_segment_descriptor_to_kvm_desct.patch

A git repo of this tree can be found at 
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary


>From mtosatti@redhat.com  Fri May  7 15:13:09 2010
From: Marcelo Tosatti <mtosatti@redhat.com>
Date: Tue, 27 Apr 2010 13:35:26 -0300
Subject: KVM: remove unused load_segment_descriptor_to_kvm_desct
To: Greg KH <gregkh@suse.de>
Cc: kvm <kvm@vger.kernel.org>, Jan Kiszka <jan.kiszka@web.de>, stable@kernel.org, Gleb Natapov <gleb@redhat.com>, Avi Kivity <avi@redhat.com>
Message-ID: <20100427163526.GA25766@amt.cnet>
Content-Disposition: inline

From: Marcelo Tosatti <mtosatti@redhat.com>

Commit 78ce64a384 in v2.6.32.12 introduced a warning due to unused
load_segment_descriptor_to_kvm_desct helper, which has been opencoded by
this commit.

On upstream, the helper was removed as part of a different commit.

Remove the now unused function.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

---
 arch/x86/kvm/x86.c |   12 ------------
 1 file changed, 12 deletions(-)

--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -4155,18 +4155,6 @@ static u16 get_segment_selector(struct k
 	return kvm_seg.selector;
 }
 
-static int load_segment_descriptor_to_kvm_desct(struct kvm_vcpu *vcpu,
-						u16 selector,
-						struct kvm_segment *kvm_seg)
-{
-	struct desc_struct seg_desc;
-
-	if (load_guest_segment_descriptor(vcpu, selector, &seg_desc))
-		return 1;
-	seg_desct_to_kvm_desct(&seg_desc, selector, kvm_seg);
-	return 0;
-}
-
 static int kvm_load_realmode_segment(struct kvm_vcpu *vcpu, u16 selector, int seg)
 {
 	struct kvm_segment segvar = {


Patches currently in stable-queue which might be from mtosatti@redhat.com are

queue-2.6.32/kvm-remove-unused-load_segment_descriptor_to_kvm_desct.patch