[v2,01/10] pci: move REDHAT_SDHCI device ID to make room for Rocker

From: Scott Feldman <sfeldma@gmail.com>

The rocker device uses same PCI device ID as sdhci.  Since rocker device driver
has already been accepted into Linux 3.18, and REDHAT_SDHCI device ID isn't
used by any drivers, it's safe to move REDHAT_SDHCI device ID, avoiding
conflict with rocker.

Signed-off-by: Scott Feldman <sfeldma@gmail.com>
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
---
 docs/specs/pci-ids.txt |    2 +-
 include/hw/pci/pci.h   |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

On 6 January 2015 at 02:24,  <sfeldma@gmail.com> wrote:
> From: Scott Feldman <sfeldma@gmail.com>
>
> The rocker device uses same PCI device ID as sdhci.  Since rocker device driver
> has already been accepted into Linux 3.18, and REDHAT_SDHCI device ID isn't
> used by any drivers, it's safe to move REDHAT_SDHCI device ID, avoiding
> conflict with rocker.

Same remarks apply as I made on v1 of this patch -- I don't want
to take this series until we have an answer for who's the
authoritative source for handing out IDs in this space and
why we ended up with this conflict.

thanks
-- PMM

On 06/01/2015 10:52, Peter Maydell wrote:
> On 6 January 2015 at 02:24,  <sfeldma@gmail.com> wrote:
>> From: Scott Feldman <sfeldma@gmail.com>
>>
>> The rocker device uses same PCI device ID as sdhci.  Since rocker device driver
>> has already been accepted into Linux 3.18, and REDHAT_SDHCI device ID isn't
>> used by any drivers, it's safe to move REDHAT_SDHCI device ID, avoiding
>> conflict with rocker.
> 
> Same remarks apply as I made on v1 of this patch -- I don't want
> to take this series until we have an answer for who's the
> authoritative source for handing out IDs in this space and
> why we ended up with this conflict.

Within the virt team, we have always considered the authoritative source
to be qemu.git and Gerd to be the maintainer.  Jiri is a Red Hatter but
not in the virt team, hence the confusion.

Paolo

On 7 January 2015 at 07:47, Paolo Bonzini <pbonzini@redhat.com> wrote:
>
>
> On 06/01/2015 10:52, Peter Maydell wrote:
>> On 6 January 2015 at 02:24,  <sfeldma@gmail.com> wrote:
>>> From: Scott Feldman <sfeldma@gmail.com>
>>>
>>> The rocker device uses same PCI device ID as sdhci.  Since rocker device driver
>>> has already been accepted into Linux 3.18, and REDHAT_SDHCI device ID isn't
>>> used by any drivers, it's safe to move REDHAT_SDHCI device ID, avoiding
>>> conflict with rocker.
>>
>> Same remarks apply as I made on v1 of this patch -- I don't want
>> to take this series until we have an answer for who's the
>> authoritative source for handing out IDs in this space and
>> why we ended up with this conflict.
>
> Within the virt team, we have always considered the authoritative source
> to be qemu.git and Gerd to be the maintainer.  Jiri is a Red Hatter but
> not in the virt team, hence the confusion.

OK, so do we:
 (1) say that the authoritative list says this ID is SDHCI,
 so rocker needs to renumber
 (2) as this patch suggests, renumber SDHCI as a one-off fixing
 of an error?

-- PMM

On 07/01/2015 11:39, Peter Maydell wrote:
>> > Within the virt team, we have always considered the authoritative source
>> > to be qemu.git and Gerd to be the maintainer.  Jiri is a Red Hatter but
>> > not in the virt team, hence the confusion.
> OK, so do we:
>  (1) say that the authoritative list says this ID is SDHCI,
>  so rocker needs to renumber
>  (2) as this patch suggests, renumber SDHCI as a one-off fixing
>  of an error?

I already sent a patch for (2) in a pull request.  SDHCI was never in a
released version, unlike rocker which was in Linux 3.18, so I don't
think there's even a choice. :)

I'll send a patch to Linux saying that the authoritative list of device
IDs for 0x1b36 resides in qemu.git.

Paolo

On Wed, Jan 7, 2015 at 2:55 AM, Paolo Bonzini <pbonzini@redhat.com> wrote:
>
>
> On 07/01/2015 11:39, Peter Maydell wrote:
>>> > Within the virt team, we have always considered the authoritative source
>>> > to be qemu.git and Gerd to be the maintainer.  Jiri is a Red Hatter but
>>> > not in the virt team, hence the confusion.
>> OK, so do we:
>>  (1) say that the authoritative list says this ID is SDHCI,
>>  so rocker needs to renumber
>>  (2) as this patch suggests, renumber SDHCI as a one-off fixing
>>  of an error?
>
> I already sent a patch for (2) in a pull request.  SDHCI was never in a
> released version, unlike rocker which was in Linux 3.18, so I don't
> think there's even a choice. :)
>
> I'll send a patch to Linux saying that the authoritative list of device
> IDs for 0x1b36 resides in qemu.git.

Thanks Paolo.  I'll resume the qemu rocker submission process assuming
rocker ID stays where it's at and SDHCI is moving.

-scott