[PATCH AUTOSEL 5.8 55/62] RDMA/efa: Add EFA 0xefa1 PCI ID

[PATCH AUTOSEL 5.8 55/62] RDMA/efa: Add EFA 0xefa1 PCI ID

[Sasha Levin]

From: Gal Pressman <galpress@amazon.com>

[ Upstream commit d4f9cb5c5b224dca3ff752c1bb854250bf114944 ]

Add support for 0xefa1 devices.

Link: https://lore.kernel.org/r/20200722140312.3651-5-galpress@amazon.com
Reviewed-by: Shadi Ammouri <sammouri@amazon.com>
Reviewed-by: Yossi Leybovich <sleybo@amazon.com>
Signed-off-by: Gal Pressman <galpress@amazon.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/infiniband/hw/efa/efa_main.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/efa/efa_main.c b/drivers/infiniband/hw/efa/efa_main.c
index 82145574c9286..92d7011463203 100644
--- a/drivers/infiniband/hw/efa/efa_main.c
+++ b/drivers/infiniband/hw/efa/efa_main.c
@@ -12,10 +12,12 @@
 
 #include "efa.h"
 
-#define PCI_DEV_ID_EFA_VF 0xefa0
+#define PCI_DEV_ID_EFA0_VF 0xefa0
+#define PCI_DEV_ID_EFA1_VF 0xefa1
 
 static const struct pci_device_id efa_pci_tbl[] = {
-	{ PCI_VDEVICE(AMAZON, PCI_DEV_ID_EFA_VF) },
+	{ PCI_VDEVICE(AMAZON, PCI_DEV_ID_EFA0_VF) },
+	{ PCI_VDEVICE(AMAZON, PCI_DEV_ID_EFA1_VF) },
 	{ }
 };
 
-- 
2.25.1

Re: [PATCH AUTOSEL 5.8 55/62] RDMA/efa: Add EFA 0xefa1 PCI ID

[Jason Gunthorpe]

On Fri, Aug 21, 2020 at 12:14:16PM -0400, Sasha Levin wrote:
> From: Gal Pressman <galpress@amazon.com>
> 
> [ Upstream commit d4f9cb5c5b224dca3ff752c1bb854250bf114944 ]
> 
> Add support for 0xefa1 devices.
> 
> Link: https://lore.kernel.org/r/20200722140312.3651-5-galpress@amazon.com
> Reviewed-by: Shadi Ammouri <sammouri@amazon.com>
> Reviewed-by: Yossi Leybovich <sleybo@amazon.com>
> Signed-off-by: Gal Pressman <galpress@amazon.com>
> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
> Signed-off-by: Sasha Levin <sashal@kernel.org>
> ---
>  drivers/infiniband/hw/efa/efa_main.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)

Wait, what? Why is this being autosel'd?

This needs to be the last patch in a series enabling support for this
chip, it will badly break this driver to pick it out of sequence!!

Jason

Re: [PATCH AUTOSEL 5.8 55/62] RDMA/efa: Add EFA 0xefa1 PCI ID

[Sasha Levin]

On Fri, Aug 21, 2020 at 04:40:36PM -0300, Jason Gunthorpe wrote:
>On Fri, Aug 21, 2020 at 12:14:16PM -0400, Sasha Levin wrote:
>> From: Gal Pressman <galpress@amazon.com>
>>
>> [ Upstream commit d4f9cb5c5b224dca3ff752c1bb854250bf114944 ]
>>
>> Add support for 0xefa1 devices.
>>
>> Link: https://lore.kernel.org/r/20200722140312.3651-5-galpress@amazon.com
>> Reviewed-by: Shadi Ammouri <sammouri@amazon.com>
>> Reviewed-by: Yossi Leybovich <sleybo@amazon.com>
>> Signed-off-by: Gal Pressman <galpress@amazon.com>
>> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
>> Signed-off-by: Sasha Levin <sashal@kernel.org>
>> ---
>>  drivers/infiniband/hw/efa/efa_main.c | 6 ++++--
>>  1 file changed, 4 insertions(+), 2 deletions(-)
>
>Wait, what? Why is this being autosel'd?

Stable trees try to pick up device enablement patches (such as patches
that add PCI IDs). I suppose that AUTOSEL get pretty eager to grab
those.

>This needs to be the last patch in a series enabling support for this
>chip, it will badly break this driver to pick it out of sequence!!

I'll drop it, thanks for letting me know!

-- 
Thanks,
Sasha

Re: [PATCH AUTOSEL 5.8 55/62] RDMA/efa: Add EFA 0xefa1 PCI ID

[Jason Gunthorpe]

On Fri, Aug 21, 2020 at 03:53:22PM -0400, Sasha Levin wrote:
> On Fri, Aug 21, 2020 at 04:40:36PM -0300, Jason Gunthorpe wrote:
> > On Fri, Aug 21, 2020 at 12:14:16PM -0400, Sasha Levin wrote:
> > > From: Gal Pressman <galpress@amazon.com>
> > > 
> > > [ Upstream commit d4f9cb5c5b224dca3ff752c1bb854250bf114944 ]
> > > 
> > > Add support for 0xefa1 devices.
> > > 
> > > Link: https://lore.kernel.org/r/20200722140312.3651-5-galpress@amazon.com
> > > Reviewed-by: Shadi Ammouri <sammouri@amazon.com>
> > > Reviewed-by: Yossi Leybovich <sleybo@amazon.com>
> > > Signed-off-by: Gal Pressman <galpress@amazon.com>
> > > Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
> > > Signed-off-by: Sasha Levin <sashal@kernel.org>
> > >  drivers/infiniband/hw/efa/efa_main.c | 6 ++++--
> > >  1 file changed, 4 insertions(+), 2 deletions(-)
> > 
> > Wait, what? Why is this being autosel'd?
> 
> Stable trees try to pick up device enablement patches (such as patches
> that add PCI IDs). I suppose that AUTOSEL get pretty eager to grab
> those.

Is it so common that old drivers will work with new HW with just a
PCI_ID update?

I would have guessed that is the minority situation

Jason

Re: [PATCH AUTOSEL 5.8 55/62] RDMA/efa: Add EFA 0xefa1 PCI ID

[Sasha Levin]

On Fri, Aug 21, 2020 at 05:19:52PM -0300, Jason Gunthorpe wrote:
>On Fri, Aug 21, 2020 at 03:53:22PM -0400, Sasha Levin wrote:
>> On Fri, Aug 21, 2020 at 04:40:36PM -0300, Jason Gunthorpe wrote:
>> > On Fri, Aug 21, 2020 at 12:14:16PM -0400, Sasha Levin wrote:
>> > > From: Gal Pressman <galpress@amazon.com>
>> > >
>> > > [ Upstream commit d4f9cb5c5b224dca3ff752c1bb854250bf114944 ]
>> > >
>> > > Add support for 0xefa1 devices.
>> > >
>> > > Link: https://lore.kernel.org/r/20200722140312.3651-5-galpress@amazon.com
>> > > Reviewed-by: Shadi Ammouri <sammouri@amazon.com>
>> > > Reviewed-by: Yossi Leybovich <sleybo@amazon.com>
>> > > Signed-off-by: Gal Pressman <galpress@amazon.com>
>> > > Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
>> > > Signed-off-by: Sasha Levin <sashal@kernel.org>
>> > >  drivers/infiniband/hw/efa/efa_main.c | 6 ++++--
>> > >  1 file changed, 4 insertions(+), 2 deletions(-)
>> >
>> > Wait, what? Why is this being autosel'd?
>>
>> Stable trees try to pick up device enablement patches (such as patches
>> that add PCI IDs). I suppose that AUTOSEL get pretty eager to grab
>> those.
>
>Is it so common that old drivers will work with new HW with just a
>PCI_ID update?
>
>I would have guessed that is the minority situation

So keep in mind that a lot of it is not brand new HW, but rather same
HW repackaged by a different vendor, or HW that received minor tweaks
but where the old driver still works.

I suppose it's more common in the USB ID world these days, so I guess
I'll give PCI IDs a closer look next time.

-- 
Thanks,
Sasha

Re: [PATCH AUTOSEL 5.8 55/62] RDMA/efa: Add EFA 0xefa1 PCI ID

[Gal Pressman]

On 21/08/2020 23:34, Sasha Levin wrote:
> On Fri, Aug 21, 2020 at 05:19:52PM -0300, Jason Gunthorpe wrote:
>> On Fri, Aug 21, 2020 at 03:53:22PM -0400, Sasha Levin wrote:
>>> On Fri, Aug 21, 2020 at 04:40:36PM -0300, Jason Gunthorpe wrote:
>>> > On Fri, Aug 21, 2020 at 12:14:16PM -0400, Sasha Levin wrote:
>>> > > From: Gal Pressman <galpress@amazon.com>
>>> > >
>>> > > [ Upstream commit d4f9cb5c5b224dca3ff752c1bb854250bf114944 ]
>>> > >
>>> > > Add support for 0xefa1 devices.
>>> > >
>>> > > Link: https://lore.kernel.org/r/20200722140312.3651-5-galpress@amazon.com
>>> > > Reviewed-by: Shadi Ammouri <sammouri@amazon.com>
>>> > > Reviewed-by: Yossi Leybovich <sleybo@amazon.com>
>>> > > Signed-off-by: Gal Pressman <galpress@amazon.com>
>>> > > Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
>>> > > Signed-off-by: Sasha Levin <sashal@kernel.org>
>>> > >  drivers/infiniband/hw/efa/efa_main.c | 6 ++++--
>>> > >  1 file changed, 4 insertions(+), 2 deletions(-)
>>> >
>>> > Wait, what? Why is this being autosel'd?
>>>
>>> Stable trees try to pick up device enablement patches (such as patches
>>> that add PCI IDs). I suppose that AUTOSEL get pretty eager to grab
>>> those.
>>
>> Is it so common that old drivers will work with new HW with just a
>> PCI_ID update?
>>
>> I would have guessed that is the minority situation
> 
> So keep in mind that a lot of it is not brand new HW, but rather same
> HW repackaged by a different vendor, or HW that received minor tweaks
> but where the old driver still works.
> 
> I suppose it's more common in the USB ID world these days, so I guess
> I'll give PCI IDs a closer look next time.

FWIW, Jason is right, this patch will break without taking the rest of the series:
https://lore.kernel.org/linux-rdma/20200722140312.3651-1-galpress@amazon.com/

Thanks Jason and Sasha.