Re: [PATCH v4 3/4] drm/xe/guc: Start use explicitly usable GuC IDs for for submission

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Direct dereference of xe_device_get_gt() return value without a NULL check
--

commit 115f41f646ebc2f7ba925f4f7342a88fd5ea0b31
Author: Piotr Piórkowski <piotr.piorkowski@intel.com>

drm/xe/guc: Start use explicitly usable GuC IDs for for submission

This commit updates GuC submission to explicitly allocate context IDs from
the usable ID pool instead of assuming all IDs are available. It configures
the GuC ID manager according to the device operating mode by using dedicated
initialization paths for native, PF, and VF modes.

> diff --git a/drivers/gpu/drm/xe/xe_guc.c b/drivers/gpu/drm/xe/xe_guc.c
> index 4286bd05c6861..14b12bcfaf548 100644
> --- a/drivers/gpu/drm/xe/xe_guc.c
> +++ b/drivers/gpu/drm/xe/xe_guc.c

[ ... ]

[Severity: High]
This is a pre-existing issue, but I noticed a potential NULL pointer
dereference during GuC initialization in guc_g2g_alloc():

drivers/gpu/drm/xe/xe_guc.c:guc_g2g_alloc() {
	...
	if (gt->info.id != 0) {
		struct xe_gt *root_gt = xe_device_get_gt(xe, 0);
		struct xe_guc *root_guc = &root_gt->uc.guc;
		struct xe_bo *bo;

		bo = xe_bo_get(root_guc->g2g.bo);
	...
}

The subsystem guidelines state that xe_device_get_gt() can return NULL for
invalid GT indices and its result must always be checked. Here, the return
value is immediately used to calculate a pointer (&root_gt->uc.guc) and then
dereferenced to access root_guc->g2g.bo.

Could this lead to a memory corruption or crash if xe_device_get_gt() returns
NULL? If we specifically need the primary GT of the root tile, would it be
safer to use xe_root_mmio_gt() instead to avoid the NULL check requirement?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260827105750.1248792-1-piotr.piorkowski@intel.com?part=3