[Qemu-devel] Re: [Qemu-commits] [COMMIT aba800a] qemu/virtio: MSI-X support in virtio PCI

[Qemu-devel] Re: [Qemu-commits] [COMMIT aba800a] qemu/virtio: MSI-X support in virtio PCI

[Laurent Desnogues]

On Tue, Jun 30, 2009 at 2:57 AM, Anthony Liguori<aliguori@us.ibm.com> wrote:
> From: Michael S. Tsirkin <mst@redhat.com>
>
> This enables actual support for MSI-X in virtio PCI.
> First user will be virtio-net.
>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
[...]
> diff --git a/rules.mak b/rules.mak
> index 8d6d96e..defee1d 100644
> --- a/rules.mak
> +++ b/rules.mak
> @@ -1,6 +1,6 @@
>
>  %.o: %.c
> -       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -c -o $@ $<,"  CC    $(TARGET_DIR)$@")
> +       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -Werror -Wno-error=uninitialized -c -o $@ $<,"  CC    $(TARGET_DIR)$@")

-Wno-error=uninitialized is a rather recent addition to gcc and
will probably break compilation on many systems.


Laurent

Re: [Qemu-devel] Re: [Qemu-commits] [COMMIT aba800a] qemu/virtio: MSI-X support in virtio PCI

[Michael S. Tsirkin]

On Tue, Jun 30, 2009 at 08:34:50AM +0200, Laurent Desnogues wrote:
> On Tue, Jun 30, 2009 at 2:57 AM, Anthony Liguori<aliguori@us.ibm.com> wrote:
> > From: Michael S. Tsirkin <mst@redhat.com>
> >
> > This enables actual support for MSI-X in virtio PCI.
> > First user will be virtio-net.
> >
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
> [...]
> > diff --git a/rules.mak b/rules.mak
> > index 8d6d96e..defee1d 100644
> > --- a/rules.mak
> > +++ b/rules.mak
> > @@ -1,6 +1,6 @@
> >
> >  %.o: %.c
> > -       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -c -o $@ $<,"  CC    $(TARGET_DIR)$@")
> > +       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -Werror -Wno-error=uninitialized -c -o $@ $<,"  CC    $(TARGET_DIR)$@")
> 
> -Wno-error=uninitialized is a rather recent addition to gcc and
> will probably break compilation on many systems.
> 
> 
> Laurent

Yes, apparently, this was added with gcc 4.2.0, released in May 2007.
Maybe we need configure to figure this out.

-- 
MST

Re: [Qemu-devel] Re: [Qemu-commits] [COMMIT aba800a] qemu/virtio: MSI-X support in virtio PCI

[Anthony Liguori]

Michael S. Tsirkin wrote:
> On Tue, Jun 30, 2009 at 08:34:50AM +0200, Laurent Desnogues wrote:
>   
>> On Tue, Jun 30, 2009 at 2:57 AM, Anthony Liguori<aliguori@us.ibm.com> wrote:
>>     
>>> From: Michael S. Tsirkin <mst@redhat.com>
>>>
>>> This enables actual support for MSI-X in virtio PCI.
>>> First user will be virtio-net.
>>>
>>> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
>>> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
>>>       
>> [...]
>>     
>>> diff --git a/rules.mak b/rules.mak
>>> index 8d6d96e..defee1d 100644
>>> --- a/rules.mak
>>> +++ b/rules.mak
>>> @@ -1,6 +1,6 @@
>>>
>>>  %.o: %.c
>>> -       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -c -o $@ $<,"  CC    $(TARGET_DIR)$@")
>>> +       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -Werror -Wno-error=uninitialized -c -o $@ $<,"  CC    $(TARGET_DIR)$@")
>>>       
>> -Wno-error=uninitialized is a rather recent addition to gcc and
>> will probably break compilation on many systems.
>>
>>     

Adding something like this in the middle of a patch, in the middle of a 
big series, is also a really bad thing to do.  If you want to change 
build arguments, it should be a separate patch.

I removed this.  There are no warnings with my GCC.  You can submit 
follow up patches trying to make a case for different ways to handle 
this but please separate this sort of stuff out in the future.

Regards,

Anthony Liguori

Re: [Qemu-devel] Re: [Qemu-commits] [COMMIT aba800a] qemu/virtio: MSI-X support in virtio PCI

[Anthony Liguori]

Anthony Liguori wrote:
> Adding something like this in the middle of a patch, in the middle of 
> a big series, is also a really bad thing to do.  If you want to change 
> build arguments, it should be a separate patch.
>
> I removed this.  There are no warnings with my GCC.  You can submit 
> follow up patches trying to make a case for different ways to handle 
> this but please separate this sort of stuff out in the future.

Looks like Paul beat me to it.

-- 
Regards,

Anthony Liguori

Re: [Qemu-devel] Re: [Qemu-commits] [COMMIT aba800a] qemu/virtio: MSI-X support in virtio PCI

[Michael S. Tsirkin]

On Tue, Jun 30, 2009 at 08:10:55AM -0500, Anthony Liguori wrote:
> Michael S. Tsirkin wrote:
>> On Tue, Jun 30, 2009 at 08:34:50AM +0200, Laurent Desnogues wrote:
>>   
>>> On Tue, Jun 30, 2009 at 2:57 AM, Anthony Liguori<aliguori@us.ibm.com> wrote:
>>>     
>>>> From: Michael S. Tsirkin <mst@redhat.com>
>>>>
>>>> This enables actual support for MSI-X in virtio PCI.
>>>> First user will be virtio-net.
>>>>
>>>> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
>>>> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
>>>>       
>>> [...]
>>>     
>>>> diff --git a/rules.mak b/rules.mak
>>>> index 8d6d96e..defee1d 100644
>>>> --- a/rules.mak
>>>> +++ b/rules.mak
>>>> @@ -1,6 +1,6 @@
>>>>
>>>>  %.o: %.c
>>>> -       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -c -o $@ $<,"  CC    $(TARGET_DIR)$@")
>>>> +       $(call quiet-command,$(CC) $(CPPFLAGS) $(CFLAGS) -Werror -Wno-error=uninitialized -c -o $@ $<,"  CC    $(TARGET_DIR)$@")
>>>>       
>>> -Wno-error=uninitialized is a rather recent addition to gcc and
>>> will probably break compilation on many systems.
>>>
>>>     
>
> Adding something like this in the middle of a patch, in the middle of a  
> big series, is also a really bad thing to do.

I know. I think I just mistakenly did "git commit -a" at some point,
and it slipped through review.

> If you want to change  
> build arguments, it should be a separate patch.
>
> I removed this.  There are no warnings with my GCC.  You can submit  
> follow up patches trying to make a case for different ways to handle  
> this but please separate this sort of stuff out in the future.
>
> Regards,
>
> Anthony Liguori