[PATCH] drm/xe: Use managed BO in memirq

[Lucas De Marchi]

On Wed, Dec 04, 2024 at 01:16:55PM +0200, Ilia Levi wrote:
>When probe has an error, the current scheme causes a page fault during
>resource unwinding. This happens because GTTMMADR_BAR gets unmapped
>before the allocated BO is released via drmm. Switching to a managed BO
>that is released via devm solves this ordering issue.
>
>Signed-off-by: Ilia Levi <ilia.levi at intel.com>


Reviewed-by: Lucas De Marchi <lucas.demarchi at intel.com>

thanks
Lucas De Marchi

>---
> drivers/gpu/drm/xe/xe_memirq.c | 23 +++++++----------------
> 1 file changed, 7 insertions(+), 16 deletions(-)
>
>diff --git a/drivers/gpu/drm/xe/xe_memirq.c b/drivers/gpu/drm/xe/xe_memirq.c
>index 51dc90906003..404fa2a456d5 100644
>--- a/drivers/gpu/drm/xe/xe_memirq.c
>+++ b/drivers/gpu/drm/xe/xe_memirq.c
>@@ -155,13 +155,6 @@ static const char *guc_name(struct xe_guc *guc)
>  *
>  */
>
>-static void __release_xe_bo(struct drm_device *drm, void *arg)
>-{
>-	struct xe_bo *bo = arg;
>-
>-	xe_bo_unpin_map_no_vm(bo);
>-}
>-
> static inline bool hw_reports_to_instance_zero(struct xe_memirq *memirq)
> {
> 	/*
>@@ -184,14 +177,12 @@ static int memirq_alloc_pages(struct xe_memirq *memirq)
> 	BUILD_BUG_ON(!IS_ALIGNED(XE_MEMIRQ_SOURCE_OFFSET(0), SZ_64));
> 	BUILD_BUG_ON(!IS_ALIGNED(XE_MEMIRQ_STATUS_OFFSET(0), SZ_4K));
>
>-	/* XXX: convert to managed bo */
>-	bo = xe_bo_create_pin_map(xe, tile, NULL, bo_size,
>-				  ttm_bo_type_kernel,
>-				  XE_BO_FLAG_SYSTEM |
>-				  XE_BO_FLAG_GGTT |
>-				  XE_BO_FLAG_GGTT_INVALIDATE |
>-				  XE_BO_FLAG_NEEDS_UC |
>-				  XE_BO_FLAG_NEEDS_CPU_ACCESS);
>+	bo = xe_managed_bo_create_pin_map(xe, tile, bo_size,
>+					  XE_BO_FLAG_SYSTEM |
>+					  XE_BO_FLAG_GGTT |
>+					  XE_BO_FLAG_GGTT_INVALIDATE |
>+					  XE_BO_FLAG_NEEDS_UC |
>+					  XE_BO_FLAG_NEEDS_CPU_ACCESS);
> 	if (IS_ERR(bo)) {
> 		err = PTR_ERR(bo);
> 		goto out;
>@@ -215,7 +206,7 @@ static int memirq_alloc_pages(struct xe_memirq *memirq)
> 		     xe_bo_ggtt_addr(bo), bo_size, XE_MEMIRQ_SOURCE_OFFSET(0),
> 		     XE_MEMIRQ_STATUS_OFFSET(0));
>
>-	return drmm_add_action_or_reset(&xe->drm, __release_xe_bo, memirq->bo);
>+	return 0;
>
> out:
> 	memirq_err(memirq, "Failed to allocate memirq page (%pe)\n", ERR_PTR(err));
>-- 
>2.43.2
>