[Qemu-devel] virtio-pci msix error

[Qemu-devel] virtio-pci msix error

[Richard Henderson]

Re patch c7ff5482.  What's the point of this error?
It's going to always appear for older targets that
predate such new fangled things as msix.

Obviously the patch has been there a while, and it's
not actually causing any problems, but today it got
on my nerves...


r~

Re: [Qemu-devel] virtio-pci msix error

[Michael S. Tsirkin]

On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
> Re patch c7ff5482.  What's the point of this error?
> It's going to always appear for older targets that
> predate such new fangled things as msix.
> 
> Obviously the patch has been there a while, and it's
> not actually causing any problems, but today it got
> on my nerves...
> 
> 
> r~

So don't specify nvectors > 0 for these platforms then?

-- 
MST

Re: [Qemu-devel] virtio-pci msix error

[Peter Maydell]

On 19 May 2015 at 20:16, Michael S. Tsirkin <mst@redhat.com> wrote:
> On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
>> Re patch c7ff5482.  What's the point of this error?
>> It's going to always appear for older targets that
>> predate such new fangled things as msix.
>>
>> Obviously the patch has been there a while, and it's
>> not actually causing any problems, but today it got
>> on my nerves...

> So don't specify nvectors > 0 for these platforms then?

How do you do that? I did a quick 'git grep' for nvectors
and none of the hits are in platform-dependent code...

Why can't the virtio-pci device automatically detect
whether the PCI bus it's plugged into supports MSIx
and just do the right thing?

-- PMM

[Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Richard Henderson]

And do not issue an error_report in that case.

Signed-off-by: Richard Henderson <rth@twiddle.net>
---
 hw/virtio/virtio-pci.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)
--

This seems to be about the only sane way to test the value of
msi_supported from here.  Thoughts?


r~



diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
index 867c9d1..6763872 100644
--- a/hw/virtio/virtio-pci.c
+++ b/hw/virtio/virtio-pci.c
@@ -934,11 +934,16 @@ static void virtio_pci_device_plugged(DeviceState *d)
     pci_set_word(config + PCI_SUBSYSTEM_ID, virtio_bus_get_vdev_id(bus));
     config[PCI_INTERRUPT_PIN] = 1;
 
-    if (proxy->nvectors &&
-        msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors, 1)) {
-        error_report("unable to init msix vectors to %" PRIu32,
-                     proxy->nvectors);
-        proxy->nvectors = 0;
+    if (proxy->nvectors) {
+        int err = msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors, 1);
+        if (err) {
+            /* Notice when a system that supports MSIx can't initialize it.  */
+            if (err != -ENOTSUP) {
+                error_report("unable to init msix vectors to %" PRIu32,
+                             proxy->nvectors);
+            }
+            proxy->nvectors = 0;
+        }
     }
 
     proxy->pci_dev.config_write = virtio_write_config;
-- 
2.1.0

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Michael S. Tsirkin]

On Tue, May 19, 2015 at 01:29:51PM -0700, Richard Henderson wrote:
> And do not issue an error_report in that case.
> 
> Signed-off-by: Richard Henderson <rth@twiddle.net>
> ---
>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>  1 file changed, 10 insertions(+), 5 deletions(-)
> --
> 
> This seems to be about the only sane way to test the value of
> msi_supported from here.  Thoughts?
> 
> 
> r~

Can you please include the qemu command-line that triggers the error
in the commit log?

> 
> 
> diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
> index 867c9d1..6763872 100644
> --- a/hw/virtio/virtio-pci.c
> +++ b/hw/virtio/virtio-pci.c
> @@ -934,11 +934,16 @@ static void virtio_pci_device_plugged(DeviceState *d)
>      pci_set_word(config + PCI_SUBSYSTEM_ID, virtio_bus_get_vdev_id(bus));
>      config[PCI_INTERRUPT_PIN] = 1;
>  
> -    if (proxy->nvectors &&
> -        msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors, 1)) {
> -        error_report("unable to init msix vectors to %" PRIu32,
> -                     proxy->nvectors);
> -        proxy->nvectors = 0;
> +    if (proxy->nvectors) {
> +        int err = msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors, 1);
> +        if (err) {
> +            /* Notice when a system that supports MSIx can't initialize it.  */
> +            if (err != -ENOTSUP) {
> +                error_report("unable to init msix vectors to %" PRIu32,
> +                             proxy->nvectors);
> +            }
> +            proxy->nvectors = 0;
> +        }
>      }
>  
>      proxy->pci_dev.config_write = virtio_write_config;
> -- 
> 2.1.0

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Richard Henderson]

On 05/20/2015 02:59 AM, Michael S. Tsirkin wrote:
> On Tue, May 19, 2015 at 01:29:51PM -0700, Richard Henderson wrote:
>> And do not issue an error_report in that case.
>>
>> Signed-off-by: Richard Henderson <rth@twiddle.net>
>> ---
>>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>>  1 file changed, 10 insertions(+), 5 deletions(-)
>> --
>>
>> This seems to be about the only sane way to test the value of
>> msi_supported from here.  Thoughts?
>>
>>
>> r~
> 
> Can you please include the qemu command-line that triggers the error
> in the commit log?

$ ../run/bin/qemu-system-alpha -drive file=foo.img,if=virtio,format=raw
qemu-system-alpha: -drive file=foo.img,if=virtio,format=raw: unable to init
msix vectors to 2


r~

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Mark Cave-Ayland]

On 20/05/15 16:10, Richard Henderson wrote:

> On 05/20/2015 02:59 AM, Michael S. Tsirkin wrote:
>> On Tue, May 19, 2015 at 01:29:51PM -0700, Richard Henderson wrote:
>>> And do not issue an error_report in that case.
>>>
>>> Signed-off-by: Richard Henderson <rth@twiddle.net>
>>> ---
>>>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>>>  1 file changed, 10 insertions(+), 5 deletions(-)
>>> --
>>>
>>> This seems to be about the only sane way to test the value of
>>> msi_supported from here.  Thoughts?
>>>
>>>
>>> r~
>>
>> Can you please include the qemu command-line that triggers the error
>> in the commit log?
> 
> $ ../run/bin/qemu-system-alpha -drive file=foo.img,if=virtio,format=raw
> qemu-system-alpha: -drive file=foo.img,if=virtio,format=raw: unable to init
> msix vectors to 2

I also get the same message on qemu-system-sparc64 as below (FWIW I did
raise this on-list months ago but no-one replied to my post at the time):

./qemu-system-sparc64 -drive
file=debian-7.7.0-sparc-netinst.iso,if=virtio,readonly=on,index=1 -nographic


ATB,

Mark.

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Mark Cave-Ayland]

On 19/05/15 21:29, Richard Henderson wrote:

> And do not issue an error_report in that case.
> 
> Signed-off-by: Richard Henderson <rth@twiddle.net>
> ---
>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>  1 file changed, 10 insertions(+), 5 deletions(-)
> --
> 
> This seems to be about the only sane way to test the value of
> msi_supported from here.  Thoughts?
> 
> 
> r~
> 
> 
> 
> diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
> index 867c9d1..6763872 100644
> --- a/hw/virtio/virtio-pci.c
> +++ b/hw/virtio/virtio-pci.c
> @@ -934,11 +934,16 @@ static void virtio_pci_device_plugged(DeviceState *d)
>      pci_set_word(config + PCI_SUBSYSTEM_ID, virtio_bus_get_vdev_id(bus));
>      config[PCI_INTERRUPT_PIN] = 1;
>  
> -    if (proxy->nvectors &&
> -        msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors, 1)) {
> -        error_report("unable to init msix vectors to %" PRIu32,
> -                     proxy->nvectors);
> -        proxy->nvectors = 0;
> +    if (proxy->nvectors) {
> +        int err = msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors, 1);
> +        if (err) {
> +            /* Notice when a system that supports MSIx can't initialize it.  */
> +            if (err != -ENOTSUP) {
> +                error_report("unable to init msix vectors to %" PRIu32,
> +                             proxy->nvectors);
> +            }
> +            proxy->nvectors = 0;
> +        }
>      }
>  
>      proxy->pci_dev.config_write = virtio_write_config;
> 

Ping? Any chance of getting this applied for 2.4 since it's something
that I do get emails about for SPARC64?


ATB,

Mark.

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Peter Maydell]

On 3 July 2015 at 09:22, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> wrote:
> On 19/05/15 21:29, Richard Henderson wrote:
>
>> And do not issue an error_report in that case.
>>
>> Signed-off-by: Richard Henderson <rth@twiddle.net>
>> ---
>>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>>  1 file changed, 10 insertions(+), 5 deletions(-)
>> --
>>
>> This seems to be about the only sane way to test the value of
>> msi_supported from here.  Thoughts?

> Ping? Any chance of getting this applied for 2.4 since it's something
> that I do get emails about for SPARC64?

Michael would like a respin with the commit message correction
he suggested. RTH -- could you send out a fixed up version, please?

thanks
-- PMM

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Peter Maydell]

Ping^2 -- this is still confusing users, cf bug LP:1500175.

thanks
-- PMM

On 6 July 2015 at 11:38, Peter Maydell <peter.maydell@linaro.org> wrote:
> On 3 July 2015 at 09:22, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> wrote:
>> On 19/05/15 21:29, Richard Henderson wrote:
>>
>>> And do not issue an error_report in that case.
>>>
>>> Signed-off-by: Richard Henderson <rth@twiddle.net>
>>> ---
>>>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>>>  1 file changed, 10 insertions(+), 5 deletions(-)
>>> --
>>>
>>> This seems to be about the only sane way to test the value of
>>> msi_supported from here.  Thoughts?
>
>> Ping? Any chance of getting this applied for 2.4 since it's something
>> that I do get emails about for SPARC64?
>
> Michael would like a respin with the commit message correction
> he suggested. RTH -- could you send out a fixed up version, please?
>
> thanks
> -- PMM

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Mark Cave-Ayland]

On 27/09/15 15:40, Peter Maydell wrote:

> Ping^2 -- this is still confusing users, cf bug LP:1500175.
> 
> thanks
> -- PMM
> 
> On 6 July 2015 at 11:38, Peter Maydell <peter.maydell@linaro.org> wrote:
>> On 3 July 2015 at 09:22, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> wrote:
>>> On 19/05/15 21:29, Richard Henderson wrote:
>>>
>>>> And do not issue an error_report in that case.
>>>>
>>>> Signed-off-by: Richard Henderson <rth@twiddle.net>
>>>> ---
>>>>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>>>>  1 file changed, 10 insertions(+), 5 deletions(-)
>>>> --
>>>>
>>>> This seems to be about the only sane way to test the value of
>>>> msi_supported from here.  Thoughts?
>>
>>> Ping? Any chance of getting this applied for 2.4 since it's something
>>> that I do get emails about for SPARC64?
>>
>> Michael would like a respin with the commit message correction
>> he suggested. RTH -- could you send out a fixed up version, please?
>>
>> thanks
>> -- PMM

Another ping on this patch - please can we have it for 2.5? It has been
around since before 2.4 and I'd hate for it to miss another release :(


ATB,

Mark.

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Michael S. Tsirkin]

On Mon, Oct 26, 2015 at 09:44:47AM +0000, Mark Cave-Ayland wrote:
> On 27/09/15 15:40, Peter Maydell wrote:
> 
> > Ping^2 -- this is still confusing users, cf bug LP:1500175.
> > 
> > thanks
> > -- PMM
> > 
> > On 6 July 2015 at 11:38, Peter Maydell <peter.maydell@linaro.org> wrote:
> >> On 3 July 2015 at 09:22, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> wrote:
> >>> On 19/05/15 21:29, Richard Henderson wrote:
> >>>
> >>>> And do not issue an error_report in that case.
> >>>>
> >>>> Signed-off-by: Richard Henderson <rth@twiddle.net>
> >>>> ---
> >>>>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
> >>>>  1 file changed, 10 insertions(+), 5 deletions(-)
> >>>> --
> >>>>
> >>>> This seems to be about the only sane way to test the value of
> >>>> msi_supported from here.  Thoughts?
> >>
> >>> Ping? Any chance of getting this applied for 2.4 since it's something
> >>> that I do get emails about for SPARC64?
> >>
> >> Michael would like a respin with the commit message correction
> >> he suggested. RTH -- could you send out a fixed up version, please?

I don't think there was a reply to this, and by now the patch doesn't apply.

> >>
> >> thanks
> >> -- PMM
> 
> Another ping on this patch - please can we have it for 2.5? It has been
> around since before 2.4 and I'd hate for it to miss another release :(
> 
> 
> ATB,
> 
> Mark.

Pls address the issues if you want this applied.

-- 
MST

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Mark Cave-Ayland]

On 26/10/15 10:03, Michael S. Tsirkin wrote:

> On Mon, Oct 26, 2015 at 09:44:47AM +0000, Mark Cave-Ayland wrote:
>> On 27/09/15 15:40, Peter Maydell wrote:
>>
>>> Ping^2 -- this is still confusing users, cf bug LP:1500175.
>>>
>>> thanks
>>> -- PMM
>>>
>>> On 6 July 2015 at 11:38, Peter Maydell <peter.maydell@linaro.org> wrote:
>>>> On 3 July 2015 at 09:22, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> wrote:
>>>>> On 19/05/15 21:29, Richard Henderson wrote:
>>>>>
>>>>>> And do not issue an error_report in that case.
>>>>>>
>>>>>> Signed-off-by: Richard Henderson <rth@twiddle.net>
>>>>>> ---
>>>>>>  hw/virtio/virtio-pci.c | 15 ++++++++++-----
>>>>>>  1 file changed, 10 insertions(+), 5 deletions(-)
>>>>>> --
>>>>>>
>>>>>> This seems to be about the only sane way to test the value of
>>>>>> msi_supported from here.  Thoughts?
>>>>
>>>>> Ping? Any chance of getting this applied for 2.4 since it's something
>>>>> that I do get emails about for SPARC64?
>>>>
>>>> Michael would like a respin with the commit message correction
>>>> he suggested. RTH -- could you send out a fixed up version, please?
> 
> I don't think there was a reply to this, and by now the patch doesn't apply.

Hmmm that's a shame. I'm not overly familiar with the PCI parts of QEMU,
so is it much work to rebase?

>>>>
>>>> thanks
>>>> -- PMM
>>
>> Another ping on this patch - please can we have it for 2.5? It has been
>> around since before 2.4 and I'd hate for it to miss another release :(
>>
>>
>> ATB,
>>
>> Mark.
> 
> Pls address the issues if you want this applied.

The patch came from Richard rather than myself - my only involvement has
been to champion the patch since both myself and Peter have had reports
of this message confusing users.


ATB,

Mark.

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Peter Maydell]

On 26 October 2015 at 10:09, Mark Cave-Ayland
<mark.cave-ayland@ilande.co.uk> wrote:
> On 26/10/15 10:03, Michael S. Tsirkin wrote:
>> Pls address the issues if you want this applied.
>
> The patch came from Richard rather than myself - my only involvement has
> been to champion the patch since both myself and Peter have had reports
> of this message confusing users.

Isn't this patch already in master as commit 0d583647a7fc73f6 ?

thanks
-- PMM

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Michael S. Tsirkin]

On Mon, Oct 26, 2015 at 10:14:20AM +0000, Peter Maydell wrote:
> On 26 October 2015 at 10:09, Mark Cave-Ayland
> <mark.cave-ayland@ilande.co.uk> wrote:
> > On 26/10/15 10:03, Michael S. Tsirkin wrote:
> >> Pls address the issues if you want this applied.
> >
> > The patch came from Richard rather than myself - my only involvement has
> > been to champion the patch since both myself and Peter have had reports
> > of this message confusing users.
> 
> Isn't this patch already in master as commit 0d583647a7fc73f6 ?
> 
> thanks
> -- PMM

So it is.
That explains why it doesn't apply.

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Mark Cave-Ayland]

On 26/10/15 10:14, Peter Maydell wrote:

> On 26 October 2015 at 10:09, Mark Cave-Ayland
> <mark.cave-ayland@ilande.co.uk> wrote:
>> On 26/10/15 10:03, Michael S. Tsirkin wrote:
>>> Pls address the issues if you want this applied.
>>
>> The patch came from Richard rather than myself - my only involvement has
>> been to champion the patch since both myself and Peter have had reports
>> of this message confusing users.
> 
> Isn't this patch already in master as commit 0d583647a7fc73f6 ?

Oh I think you're right! I had that thread tagged and hadn't seen a
reply come through, but I guess Richard must have attached it to one of
his recent pull requests.

In that case apologies to all for the noise.


ATB,

Mark.

Re: [Qemu-devel] [PATCH] virtio: Notice when the system doesn't support MSIx at all

[Michael S. Tsirkin]

On Mon, Oct 26, 2015 at 10:47:48AM +0000, Mark Cave-Ayland wrote:
> On 26/10/15 10:14, Peter Maydell wrote:
> 
> > On 26 October 2015 at 10:09, Mark Cave-Ayland
> > <mark.cave-ayland@ilande.co.uk> wrote:
> >> On 26/10/15 10:03, Michael S. Tsirkin wrote:
> >>> Pls address the issues if you want this applied.
> >>
> >> The patch came from Richard rather than myself - my only involvement has
> >> been to champion the patch since both myself and Peter have had reports
> >> of this message confusing users.
> > 
> > Isn't this patch already in master as commit 0d583647a7fc73f6 ?
> 
> Oh I think you're right! I had that thread tagged and hadn't seen a
> reply come through, but I guess Richard must have attached it to one of
> his recent pull requests.

I did. I don't do replies, I merge patches then change my mind and drop
them too many times to make "applied thanks" work.  As I send pulls
about once a week, and Cc contributors, you only need to watch these to
know whether stuff is merged.


> In that case apologies to all for the noise.
> 
> 
> ATB,
> 
> Mark.

Re: [Qemu-devel] virtio-pci msix error

[Michael S. Tsirkin]

On Tue, May 19, 2015 at 09:11:26PM +0100, Peter Maydell wrote:
> On 19 May 2015 at 20:16, Michael S. Tsirkin <mst@redhat.com> wrote:
> > On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
> >> Re patch c7ff5482.  What's the point of this error?
> >> It's going to always appear for older targets that
> >> predate such new fangled things as msix.
> >>
> >> Obviously the patch has been there a while, and it's
> >> not actually causing any problems, but today it got
> >> on my nerves...
> 
> > So don't specify nvectors > 0 for these platforms then?
> 
> How do you do that? I did a quick 'git grep' for nvectors
> and none of the hits are in platform-dependent code...
> 
> Why can't the virtio-pci device automatically detect
> whether the PCI bus it's plugged into supports MSIx
> and just do the right thing?
> 
> -- PMM

I mean why does *user* specify nvectors > 0?

-- 
MST

Re: [Qemu-devel] virtio-pci msix error

[Peter Maydell]

On 20 May 2015 at 10:58, Michael S. Tsirkin <mst@redhat.com> wrote:
> On Tue, May 19, 2015 at 09:11:26PM +0100, Peter Maydell wrote:
>> On 19 May 2015 at 20:16, Michael S. Tsirkin <mst@redhat.com> wrote:
>> > On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
>> >> Re patch c7ff5482.  What's the point of this error?
>> >> It's going to always appear for older targets that
>> >> predate such new fangled things as msix.
>> >>
>> >> Obviously the patch has been there a while, and it's
>> >> not actually causing any problems, but today it got
>> >> on my nerves...
>>
>> > So don't specify nvectors > 0 for these platforms then?
>>
>> How do you do that? I did a quick 'git grep' for nvectors
>> and none of the hits are in platform-dependent code...
>>
>> Why can't the virtio-pci device automatically detect
>> whether the PCI bus it's plugged into supports MSIx
>> and just do the right thing?

> I mean why does *user* specify nvectors > 0?

The user isn't specifying nvectors at all. That's why
the message is annoying...

-- PMM

Re: [Qemu-devel] virtio-pci msix error

[Fam Zheng]

On Wed, 05/20 11:41, Peter Maydell wrote:
> On 20 May 2015 at 10:58, Michael S. Tsirkin <mst@redhat.com> wrote:
> > On Tue, May 19, 2015 at 09:11:26PM +0100, Peter Maydell wrote:
> >> On 19 May 2015 at 20:16, Michael S. Tsirkin <mst@redhat.com> wrote:
> >> > On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
> >> >> Re patch c7ff5482.  What's the point of this error?
> >> >> It's going to always appear for older targets that
> >> >> predate such new fangled things as msix.
> >> >>
> >> >> Obviously the patch has been there a while, and it's
> >> >> not actually causing any problems, but today it got
> >> >> on my nerves...
> >>
> >> > So don't specify nvectors > 0 for these platforms then?
> >>
> >> How do you do that? I did a quick 'git grep' for nvectors
> >> and none of the hits are in platform-dependent code...
> >>
> >> Why can't the virtio-pci device automatically detect
> >> whether the PCI bus it's plugged into supports MSIx
> >> and just do the right thing?
> 
> > I mean why does *user* specify nvectors > 0?
> 
> The user isn't specifying nvectors at all. That's why
> the message is annoying...

So I think it's better to fix the default for old targets?

Fam

Re: [Qemu-devel] virtio-pci msix error

[Peter Maydell]

On 20 May 2015 at 12:55, Fam Zheng <famz@redhat.com> wrote:
> On Wed, 05/20 11:41, Peter Maydell wrote:
>> On 20 May 2015 at 10:58, Michael S. Tsirkin <mst@redhat.com> wrote:
>> > On Tue, May 19, 2015 at 09:11:26PM +0100, Peter Maydell wrote:
>> >> On 19 May 2015 at 20:16, Michael S. Tsirkin <mst@redhat.com> wrote:
>> >> > On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
>> >> >> Re patch c7ff5482.  What's the point of this error?
>> >> >> It's going to always appear for older targets that
>> >> >> predate such new fangled things as msix.
>> >> >>
>> >> >> Obviously the patch has been there a while, and it's
>> >> >> not actually causing any problems, but today it got
>> >> >> on my nerves...
>> >>
>> >> > So don't specify nvectors > 0 for these platforms then?
>> >>
>> >> How do you do that? I did a quick 'git grep' for nvectors
>> >> and none of the hits are in platform-dependent code...
>> >>
>> >> Why can't the virtio-pci device automatically detect
>> >> whether the PCI bus it's plugged into supports MSIx
>> >> and just do the right thing?
>>
>> > I mean why does *user* specify nvectors > 0?
>>
>> The user isn't specifying nvectors at all. That's why
>> the message is annoying...
>
> So I think it's better to fix the default for old targets?

What default? No platform or PCI controller code specifies any
default value for nvectors, and the user doesn't specify a
value for nvectors.

The only thing that tries to specify a value for nvectors
is the virtio-*-pci device itself, which it then pointlessly
complains that it can't set.

-- PMM

Re: [Qemu-devel] virtio-pci msix error

[Mark Cave-Ayland]

On 20/05/15 13:10, Peter Maydell wrote:

> On 20 May 2015 at 12:55, Fam Zheng <famz@redhat.com> wrote:
>> On Wed, 05/20 11:41, Peter Maydell wrote:
>>> On 20 May 2015 at 10:58, Michael S. Tsirkin <mst@redhat.com> wrote:
>>>> On Tue, May 19, 2015 at 09:11:26PM +0100, Peter Maydell wrote:
>>>>> On 19 May 2015 at 20:16, Michael S. Tsirkin <mst@redhat.com> wrote:
>>>>>> On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
>>>>>>> Re patch c7ff5482.  What's the point of this error?
>>>>>>> It's going to always appear for older targets that
>>>>>>> predate such new fangled things as msix.
>>>>>>>
>>>>>>> Obviously the patch has been there a while, and it's
>>>>>>> not actually causing any problems, but today it got
>>>>>>> on my nerves...
>>>>>
>>>>>> So don't specify nvectors > 0 for these platforms then?
>>>>>
>>>>> How do you do that? I did a quick 'git grep' for nvectors
>>>>> and none of the hits are in platform-dependent code...
>>>>>
>>>>> Why can't the virtio-pci device automatically detect
>>>>> whether the PCI bus it's plugged into supports MSIx
>>>>> and just do the right thing?
>>>
>>>> I mean why does *user* specify nvectors > 0?
>>>
>>> The user isn't specifying nvectors at all. That's why
>>> the message is annoying...
>>
>> So I think it's better to fix the default for old targets?
> 
> What default? No platform or PCI controller code specifies any
> default value for nvectors, and the user doesn't specify a
> value for nvectors.
> 
> The only thing that tries to specify a value for nvectors
> is the virtio-*-pci device itself, which it then pointlessly
> complains that it can't set.

Ping? I've just done an OpenBIOS test run on qemu-system-sparc64 and I'm
still seeing this message appear on the console. Any chance we can get
this fixed for 2.4?


ATB,

Mark.

Re: [Qemu-devel] virtio-pci msix error

[Mark Cave-Ayland]

On 13/06/15 16:06, Mark Cave-Ayland wrote:

> On 20/05/15 13:10, Peter Maydell wrote:
> 
>> On 20 May 2015 at 12:55, Fam Zheng <famz@redhat.com> wrote:
>>> On Wed, 05/20 11:41, Peter Maydell wrote:
>>>> On 20 May 2015 at 10:58, Michael S. Tsirkin <mst@redhat.com> wrote:
>>>>> On Tue, May 19, 2015 at 09:11:26PM +0100, Peter Maydell wrote:
>>>>>> On 19 May 2015 at 20:16, Michael S. Tsirkin <mst@redhat.com> wrote:
>>>>>>> On Tue, May 19, 2015 at 08:54:36AM -0700, Richard Henderson wrote:
>>>>>>>> Re patch c7ff5482.  What's the point of this error?
>>>>>>>> It's going to always appear for older targets that
>>>>>>>> predate such new fangled things as msix.
>>>>>>>>
>>>>>>>> Obviously the patch has been there a while, and it's
>>>>>>>> not actually causing any problems, but today it got
>>>>>>>> on my nerves...
>>>>>>
>>>>>>> So don't specify nvectors > 0 for these platforms then?
>>>>>>
>>>>>> How do you do that? I did a quick 'git grep' for nvectors
>>>>>> and none of the hits are in platform-dependent code...
>>>>>>
>>>>>> Why can't the virtio-pci device automatically detect
>>>>>> whether the PCI bus it's plugged into supports MSIx
>>>>>> and just do the right thing?
>>>>
>>>>> I mean why does *user* specify nvectors > 0?
>>>>
>>>> The user isn't specifying nvectors at all. That's why
>>>> the message is annoying...
>>>
>>> So I think it's better to fix the default for old targets?
>>
>> What default? No platform or PCI controller code specifies any
>> default value for nvectors, and the user doesn't specify a
>> value for nvectors.
>>
>> The only thing that tries to specify a value for nvectors
>> is the virtio-*-pci device itself, which it then pointlessly
>> complains that it can't set.
> 
> Ping? I've just done an OpenBIOS test run on qemu-system-sparc64 and I'm
> still seeing this message appear on the console. Any chance we can get
> this fixed for 2.4?

Ping again?


ATB,

Mark.

Re: [Qemu-devel] virtio-pci msix error

[Peter Maydell]

On 3 July 2015 at 09:13, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> wrote:
> On 13/06/15 16:06, Mark Cave-Ayland wrote:
>> Ping? I've just done an OpenBIOS test run on qemu-system-sparc64 and I'm
>> still seeing this message appear on the console. Any chance we can get
>> this fixed for 2.4?
>
> Ping again?

rth submitted a patch for this -- you want to be pinging
that, I suspect. It looks like the only issue with it was
with the commit message...

-- PMM

Re: [Qemu-devel] virtio-pci msix error

[Mark Cave-Ayland]

On 03/07/15 09:16, Peter Maydell wrote:

> On 3 July 2015 at 09:13, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> wrote:
>> On 13/06/15 16:06, Mark Cave-Ayland wrote:
>>> Ping? I've just done an OpenBIOS test run on qemu-system-sparc64 and I'm
>>> still seeing this message appear on the console. Any chance we can get
>>> this fixed for 2.4?
>>
>> Ping again?
> 
> rth submitted a patch for this -- you want to be pinging
> that, I suspect. It looks like the only issue with it was
> with the commit message...

Got it thanks. Will do so.


ATB,

Mark.