On Mon, 15 Sep 2025, Gustavo Sousa <gustavo.so...@intel.com> wrote:
> Quoting Dnyaneshwar Bhadane (2025-09-11 17:55:40-03:00)
>>To form the WCL platform as a subplatform of PTL in definition,
>>WCL pci ids are splited into saparate group from PTL.
>>So update the pciidlist struct to cover all the pci ids.
>>
>>Signed-off-by: Dnyaneshwar Bhadane <dnyaneshwar.bhad...@intel.com>
>>---
>> drivers/gpu/drm/xe/xe_pci.c | 1 +
>> include/drm/intel/pciids.h | 4 +++-
>> 2 files changed, 4 insertions(+), 1 deletion(-)
>>
>>diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
>>index 701ba9baa9d7..fc2ea9132804 100644
>>--- a/drivers/gpu/drm/xe/xe_pci.c
>>+++ b/drivers/gpu/drm/xe/xe_pci.c
>>@@ -375,6 +375,7 @@ static const struct pci_device_id pciidlist[] = {
>> INTEL_LNL_IDS(INTEL_VGA_DEVICE, &lnl_desc),
>> INTEL_BMG_IDS(INTEL_VGA_DEVICE, &bmg_desc),
>> INTEL_PTL_IDS(INTEL_VGA_DEVICE, &ptl_desc),
>>+ INTEL_WCL_IDS(INTEL_VGA_DEVICE, &ptl_desc),
>> { }
>> };
>> MODULE_DEVICE_TABLE(pci, pciidlist);
>>diff --git a/include/drm/intel/pciids.h b/include/drm/intel/pciids.h
>>index da6301a6fcea..9d378c65be4b 100644
>>--- a/include/drm/intel/pciids.h
>>+++ b/include/drm/intel/pciids.h
>>@@ -877,7 +877,9 @@
>> MACRO__(0xB08F, ## __VA_ARGS__), \
>> MACRO__(0xB090, ## __VA_ARGS__), \
>> MACRO__(0xB0A0, ## __VA_ARGS__), \
>>- MACRO__(0xB0B0, ## __VA_ARGS__), \
>>+ MACRO__(0xB0B0, ## __VA_ARGS__)
>>+
>>+#define INTEL_WCL_IDS(MACRO__, ...) \
>> MACRO__(0xFD80, ## __VA_ARGS__), \
>> MACRO__(0xFD81, ## __VA_ARGS__)
>
> This patch, at its current state, will break the display part, because
> WCL will not be detected until the next patch. We should either:
>
> - bring the line "INTEL_WCL_IDS(INTEL_DISPLAY_DEVICE, &ptl_desc)" from
> path #2 into this one.
This. I've already replied to a newer version of this series to this
effect [1][2].
[1] https://lore.kernel.org/r/70fc412b47d9972ad2d1b6eca13bbdd9da992...@intel.com
[2] https://lore.kernel.org/r/84fc10ec3b82b3436b521811589067ad0850e...@intel.com
> - squash this and patch #2 together.
IMO cleaner with separate patches.
> That said, since we are defining WCL as a subplatform, I think we
> probably should make INTEL_WCL_IDS() be called from INTEL_PTL_IDS().
No. Please don't do that.
There are various consumers for the PCI ID macros, and they should be
kept independent. It's easier to deal with the platform/subplatform
relationships at the consumer side, instead of forcing it in the PCI ID
macros.
Just consider having to promote WCL to an independent platform later. It
would mean shuffling the macros again.
> Either that or make both separate platforms from the display point of
> view.
>
> Also, I'm not sure how having a prelimiary patch affects backporting
> fixes. So, I'm wondering if we should tag this patch somehow or if
> something else should be made here to make the backporting easier.
It's easy enough to ask for deps to be backported.
BR,
Jani.
--
Jani Nikula, Intel
