Nouveau - Nov 2016 - [PATCH] gr: fallback to legacy paths during firmware lookup

On 11/02/2016 02:07 PM, Ilia Mirkin wrote:
> On Wed, Nov 2, 2016 at 12:54 AM, Alexandre Courbot <acourbot at
nvidia.com> wrote:
>> Look for firmware files using the legacy
("nouveau/nvxx_fucxxxx") path
>> if they cannot be found in the new, "official" path. User
setups were
>> broken by the switch, which is bad.
>>
>> There are only 4 firmware files we may want to look up that way, so
>> hardcode them into the lookup function. All new firmware files should
>> use the standard "nvidia/<chip>/gr/" path.
>>
>> Fixes: 8539b37acef7 ("drm/nouveau/gr: use NVIDIA-provided external
firmwares")
>> Signed-off-by: Alexandre Courbot <acourbot at nvidia.com>
>> ---
>>  drm/nouveau/nvkm/engine/gr/gf100.c | 56
++++++++++++++++++++++++++++++++++----
>>  1 file changed, 51 insertions(+), 5 deletions(-)
>>
>> diff --git a/drm/nouveau/nvkm/engine/gr/gf100.c
b/drm/nouveau/nvkm/engine/gr/gf100.c
>> index 157919c788e6..9e65adbab21c 100644
>> --- a/drm/nouveau/nvkm/engine/gr/gf100.c
>> +++ b/drm/nouveau/nvkm/engine/gr/gf100.c
>> @@ -1756,24 +1756,70 @@ gf100_gr_ = {
>>  };
>>
>>  int
>> +gf100_gr_ctor_fw_legacy(struct gf100_gr *gr, const char *fwname,
>> +                       struct gf100_gr_fuc *fuc)
>> +{
>> +       struct nvkm_subdev *subdev = &gr->base.engine.subdev;
>> +       struct nvkm_device *device = subdev->device;
>> +       const struct firmware *fw;
>> +       char f[32];
>> +       int ret;
>> +
>> +       snprintf(f, sizeof(f), "nouveau/nv%02x_%s",
device->chipset, fwname);
>> +       ret = request_firmware(&fw, f, device->dev);
>> +       if (ret) {
>> +               snprintf(f, sizeof(f), "nouveau/%s", fwname);
>> +               ret = request_firmware(&fw, f, device->dev);
>> +               if (ret) {
>> +                       nvkm_error(subdev, "failed to load
%s\n", fwname);
>> +                       return ret;
>> +               }
>> +       }
>> +
>> +       fuc->size = fw->size;
>> +       fuc->data = kmemdup(fw->data, fuc->size, GFP_KERNEL);
>> +       release_firmware(fw);
>> +       return (fuc->data != NULL) ? 0 : -ENOMEM;
>> +}
>> +
>> +int
>>  gf100_gr_ctor_fw(struct gf100_gr *gr, const char *fwname,
>>                  struct gf100_gr_fuc *fuc)
>>  {
>>         struct nvkm_subdev *subdev = &gr->base.engine.subdev;
>>         struct nvkm_device *device = subdev->device;
>> +       const char *legacy_fwname = NULL;
>>         const struct firmware *fw;
>>         int ret;
>>
>>         ret = nvkm_firmware_get(device, fwname, &fw);
>> -       if (ret) {
>> +       /* firmware found, success! */
>> +       if (!ret) {
>> +               fuc->size = fw->size;
>> +               fuc->data = kmemdup(fw->data, fuc->size,
GFP_KERNEL);
>> +               nvkm_firmware_put(fw);
>> +               return (fuc->data != NULL) ? 0 : -ENOMEM;
>> +       }
>> +
>> +       /* see if this firmware has a legacy path */
>> +       if (!strcmp(fwname, "fecs_inst"))
>> +               legacy_fwname = "fuc409c";
>> +       else if (!strcmp(fwname, "fecs_data"))
>> +               legacy_fwname = "fuc409d";
>> +       else if (!strcmp(fwname, "gpccs_inst"))
>> +               legacy_fwname = "fuc41ac";
>> +       else if (!strcmp(fwname, "gpccs_data"))
>> +               legacy_fwname = "fuc41ad";
> 
> As I mentioned on IRC, I find this strcmp thing a little jarring. It
> should be pretty easy to just pass the legacy fwname into
> gf100_gr_ctor_fw directly - there are only a handful of callers, and
> most would just pass NULL in as the legacy fwname - only the ones in
> gf100.c would pass the "old" names in.
Yeah, that was the original approach actually. I switched to this for
the following reasons:

- We don't want to encourage using this legacy path, hence hiding it
from callers
- There are only 4 possible legacy files - it's ugly but still
manageable and contained in a single function

Of course, if the general consensus is that this is too ugly, it would
be trivial to turn it the way you suggested, so I don't feel too
strongly for one approach over the other.

On 11/02/2016 03:52 PM, Alexandre Courbot wrote:
> On 11/02/2016 02:07 PM, Ilia Mirkin wrote:
>> On Wed, Nov 2, 2016 at 12:54 AM, Alexandre Courbot <acourbot at
nvidia.com> wrote:
>>> Look for firmware files using the legacy
("nouveau/nvxx_fucxxxx") path
>>> if they cannot be found in the new, "official" path. User
setups were
>>> broken by the switch, which is bad.
>>>
>>> There are only 4 firmware files we may want to look up that way, so
>>> hardcode them into the lookup function. All new firmware files
should
>>> use the standard "nvidia/<chip>/gr/" path.
>>>
>>> Fixes: 8539b37acef7 ("drm/nouveau/gr: use NVIDIA-provided
external firmwares")
>>> Signed-off-by: Alexandre Courbot <acourbot at nvidia.com>
>>> ---
>>>  drm/nouveau/nvkm/engine/gr/gf100.c | 56
++++++++++++++++++++++++++++++++++----
>>>  1 file changed, 51 insertions(+), 5 deletions(-)
>>>
>>> diff --git a/drm/nouveau/nvkm/engine/gr/gf100.c
b/drm/nouveau/nvkm/engine/gr/gf100.c
>>> index 157919c788e6..9e65adbab21c 100644
>>> --- a/drm/nouveau/nvkm/engine/gr/gf100.c
>>> +++ b/drm/nouveau/nvkm/engine/gr/gf100.c
>>> @@ -1756,24 +1756,70 @@ gf100_gr_ = {
>>>  };
>>>
>>>  int
>>> +gf100_gr_ctor_fw_legacy(struct gf100_gr *gr, const char *fwname,
>>> +                       struct gf100_gr_fuc *fuc)
>>> +{
>>> +       struct nvkm_subdev *subdev =
&gr->base.engine.subdev;
>>> +       struct nvkm_device *device = subdev->device;
>>> +       const struct firmware *fw;
>>> +       char f[32];
>>> +       int ret;
>>> +
>>> +       snprintf(f, sizeof(f), "nouveau/nv%02x_%s",
device->chipset, fwname);
>>> +       ret = request_firmware(&fw, f, device->dev);
>>> +       if (ret) {
>>> +               snprintf(f, sizeof(f), "nouveau/%s",
fwname);
>>> +               ret = request_firmware(&fw, f, device->dev);
>>> +               if (ret) {
>>> +                       nvkm_error(subdev, "failed to load
%s\n", fwname);
>>> +                       return ret;
>>> +               }
>>> +       }
>>> +
>>> +       fuc->size = fw->size;
>>> +       fuc->data = kmemdup(fw->data, fuc->size,
GFP_KERNEL);
>>> +       release_firmware(fw);
>>> +       return (fuc->data != NULL) ? 0 : -ENOMEM;
>>> +}
>>> +
>>> +int
>>>  gf100_gr_ctor_fw(struct gf100_gr *gr, const char *fwname,
>>>                  struct gf100_gr_fuc *fuc)
>>>  {
>>>         struct nvkm_subdev *subdev =
&gr->base.engine.subdev;
>>>         struct nvkm_device *device = subdev->device;
>>> +       const char *legacy_fwname = NULL;
>>>         const struct firmware *fw;
>>>         int ret;
>>>
>>>         ret = nvkm_firmware_get(device, fwname, &fw);
>>> -       if (ret) {
>>> +       /* firmware found, success! */
>>> +       if (!ret) {
>>> +               fuc->size = fw->size;
>>> +               fuc->data = kmemdup(fw->data, fuc->size,
GFP_KERNEL);
>>> +               nvkm_firmware_put(fw);
>>> +               return (fuc->data != NULL) ? 0 : -ENOMEM;
>>> +       }
>>> +
>>> +       /* see if this firmware has a legacy path */
>>> +       if (!strcmp(fwname, "fecs_inst"))
>>> +               legacy_fwname = "fuc409c";
>>> +       else if (!strcmp(fwname, "fecs_data"))
>>> +               legacy_fwname = "fuc409d";
>>> +       else if (!strcmp(fwname, "gpccs_inst"))
>>> +               legacy_fwname = "fuc41ac";
>>> +       else if (!strcmp(fwname, "gpccs_data"))
>>> +               legacy_fwname = "fuc41ad";
>>
>> As I mentioned on IRC, I find this strcmp thing a little jarring. It
>> should be pretty easy to just pass the legacy fwname into
>> gf100_gr_ctor_fw directly - there are only a handful of callers, and
>> most would just pass NULL in as the legacy fwname - only the ones in
>> gf100.c would pass the "old" names in.
> 
> Yeah, that was the original approach actually. I switched to this for
> the following reasons:
> 
> - We don't want to encourage using this legacy path, hence hiding it
> from callers
> - There are only 4 possible legacy files - it's ugly but still
> manageable and contained in a single function
I agree.
> 
> Of course, if the general consensus is that this is too ugly, it would
> be trivial to turn it the way you suggested, so I don't feel too
> strongly for one approach over the other.
As it is a legacy path, I'm more for hiding it away in ctor_legacy() as
Emil suggests.

Ben.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: OpenPGP digital signature
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20161104/0679027c/attachment.sig>

On 11/04/2016 07:29 AM, Ben Skeggs wrote:
> * PGP Signed by an unknown key
> 
> On 11/02/2016 03:52 PM, Alexandre Courbot wrote:
>> On 11/02/2016 02:07 PM, Ilia Mirkin wrote:
>>> On Wed, Nov 2, 2016 at 12:54 AM, Alexandre Courbot <acourbot at
nvidia.com> wrote:
>>>> Look for firmware files using the legacy
("nouveau/nvxx_fucxxxx") path
>>>> if they cannot be found in the new, "official" path.
User setups were
>>>> broken by the switch, which is bad.
>>>>
>>>> There are only 4 firmware files we may want to look up that
way, so
>>>> hardcode them into the lookup function. All new firmware files
should
>>>> use the standard "nvidia/<chip>/gr/" path.
>>>>
>>>> Fixes: 8539b37acef7 ("drm/nouveau/gr: use NVIDIA-provided
external firmwares")
>>>> Signed-off-by: Alexandre Courbot <acourbot at nvidia.com>
>>>> ---
>>>>  drm/nouveau/nvkm/engine/gr/gf100.c | 56
++++++++++++++++++++++++++++++++++----
>>>>  1 file changed, 51 insertions(+), 5 deletions(-)
>>>>
>>>> diff --git a/drm/nouveau/nvkm/engine/gr/gf100.c
b/drm/nouveau/nvkm/engine/gr/gf100.c
>>>> index 157919c788e6..9e65adbab21c 100644
>>>> --- a/drm/nouveau/nvkm/engine/gr/gf100.c
>>>> +++ b/drm/nouveau/nvkm/engine/gr/gf100.c
>>>> @@ -1756,24 +1756,70 @@ gf100_gr_ = {
>>>>  };
>>>>
>>>>  int
>>>> +gf100_gr_ctor_fw_legacy(struct gf100_gr *gr, const char
*fwname,
>>>> +                       struct gf100_gr_fuc *fuc)
>>>> +{
>>>> +       struct nvkm_subdev *subdev =
&gr->base.engine.subdev;
>>>> +       struct nvkm_device *device = subdev->device;
>>>> +       const struct firmware *fw;
>>>> +       char f[32];
>>>> +       int ret;
>>>> +
>>>> +       snprintf(f, sizeof(f), "nouveau/nv%02x_%s",
device->chipset, fwname);
>>>> +       ret = request_firmware(&fw, f, device->dev);
>>>> +       if (ret) {
>>>> +               snprintf(f, sizeof(f), "nouveau/%s",
fwname);
>>>> +               ret = request_firmware(&fw, f,
device->dev);
>>>> +               if (ret) {
>>>> +                       nvkm_error(subdev, "failed to load
%s\n", fwname);
>>>> +                       return ret;
>>>> +               }
>>>> +       }
>>>> +
>>>> +       fuc->size = fw->size;
>>>> +       fuc->data = kmemdup(fw->data, fuc->size,
GFP_KERNEL);
>>>> +       release_firmware(fw);
>>>> +       return (fuc->data != NULL) ? 0 : -ENOMEM;
>>>> +}
>>>> +
>>>> +int
>>>>  gf100_gr_ctor_fw(struct gf100_gr *gr, const char *fwname,
>>>>                  struct gf100_gr_fuc *fuc)
>>>>  {
>>>>         struct nvkm_subdev *subdev =
&gr->base.engine.subdev;
>>>>         struct nvkm_device *device = subdev->device;
>>>> +       const char *legacy_fwname = NULL;
>>>>         const struct firmware *fw;
>>>>         int ret;
>>>>
>>>>         ret = nvkm_firmware_get(device, fwname, &fw);
>>>> -       if (ret) {
>>>> +       /* firmware found, success! */
>>>> +       if (!ret) {
>>>> +               fuc->size = fw->size;
>>>> +               fuc->data = kmemdup(fw->data,
fuc->size, GFP_KERNEL);
>>>> +               nvkm_firmware_put(fw);
>>>> +               return (fuc->data != NULL) ? 0 : -ENOMEM;
>>>> +       }
>>>> +
>>>> +       /* see if this firmware has a legacy path */
>>>> +       if (!strcmp(fwname, "fecs_inst"))
>>>> +               legacy_fwname = "fuc409c";
>>>> +       else if (!strcmp(fwname, "fecs_data"))
>>>> +               legacy_fwname = "fuc409d";
>>>> +       else if (!strcmp(fwname, "gpccs_inst"))
>>>> +               legacy_fwname = "fuc41ac";
>>>> +       else if (!strcmp(fwname, "gpccs_data"))
>>>> +               legacy_fwname = "fuc41ad";
>>>
>>> As I mentioned on IRC, I find this strcmp thing a little jarring.
It
>>> should be pretty easy to just pass the legacy fwname into
>>> gf100_gr_ctor_fw directly - there are only a handful of callers,
and
>>> most would just pass NULL in as the legacy fwname - only the ones
in
>>> gf100.c would pass the "old" names in.
>>
>> Yeah, that was the original approach actually. I switched to this for
>> the following reasons:
>>
>> - We don't want to encourage using this legacy path, hence hiding
it
>> from callers
>> - There are only 4 possible legacy files - it's ugly but still
>> manageable and contained in a single function
> I agree.
> 
>>
>> Of course, if the general consensus is that this is too ugly, it would
>> be trivial to turn it the way you suggested, so I don't feel too
>> strongly for one approach over the other.
> As it is a legacy path, I'm more for hiding it away in ctor_legacy() as
> Emil suggests.
Makes sense - will respin later today.

Thanks!
Alex.