[PATCH] drm/pagemap: Reset migration page count on eviction retry

[PATCH] drm/pagemap: Reset migration page count on eviction retry

[Arvind Yadav]

drm_pagemap_evict_to_ram() may retry eviction, but mpages retains
the count from the previous attempt. A retry can therefore continue
to the copy path even when no RAM pages were populated.

Reset mpages at the retry label so it reflects only the current
attempt.

Fixes: 99624bdff867 ("drm/gpusvm: Add support for GPU Shared Virtual Memory")
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Maxime Ripard <mripard@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: David Airlie <airlied@gmail.com>
Cc: Simona Vetter <simona@ffwll.ch>
Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
---
 drivers/gpu/drm/drm_pagemap.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_pagemap.c b/drivers/gpu/drm/drm_pagemap.c
index 4a794544b7dc..7faba4acd59f 100644
--- a/drivers/gpu/drm/drm_pagemap.c
+++ b/drivers/gpu/drm/drm_pagemap.c
@@ -1122,7 +1122,7 @@ int drm_pagemap_evict_to_ram(struct drm_pagemap_devmem *devmem_allocation)
 {
 	const struct drm_pagemap_devmem_ops *ops = devmem_allocation->ops;
 	struct drm_pagemap_iova_state state = {};
-	unsigned long npages, mpages = 0;
+	unsigned long npages, mpages;
 	struct page **pages;
 	unsigned long *src, *dst;
 	struct drm_pagemap_addr *pagemap_addr;
@@ -1133,6 +1133,7 @@ int drm_pagemap_evict_to_ram(struct drm_pagemap_devmem *devmem_allocation)
 	npages = devmem_allocation->size >> PAGE_SHIFT;
 
 retry:
+	mpages = 0;
 	if (!mmget_not_zero(devmem_allocation->mm))
 		return -EFAULT;
 
-- 
2.43.0

Re: [PATCH] drm/pagemap: Reset migration page count on eviction retry

[Matthew Brost]

On Tue, Jul 28, 2026 at 02:33:04PM +0530, Arvind Yadav wrote:
> drm_pagemap_evict_to_ram() may retry eviction, but mpages retains
> the count from the previous attempt. A retry can therefore continue
> to the copy path even when no RAM pages were populated.
> 
> Reset mpages at the retry label so it reflects only the current
> attempt.
> 
> Fixes: 99624bdff867 ("drm/gpusvm: Add support for GPU Shared Virtual Memory")

I think you also want the patch which moved drm_pagemap to its own file:

f86ad0ed620c drm/gpusvm, drm/pagemap: Move migration functionality to drm_pagemap

> Cc: Matthew Brost <matthew.brost@intel.com>

Fix itself LGTM:
Reviewed-by: Matthew Brost <matthew.brost@intel.com>

> Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
> Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
> Cc: Maxime Ripard <mripard@kernel.org>
> Cc: Thomas Zimmermann <tzimmermann@suse.de>
> Cc: David Airlie <airlied@gmail.com>
> Cc: Simona Vetter <simona@ffwll.ch>
> Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
> ---
>  drivers/gpu/drm/drm_pagemap.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/drm_pagemap.c b/drivers/gpu/drm/drm_pagemap.c
> index 4a794544b7dc..7faba4acd59f 100644
> --- a/drivers/gpu/drm/drm_pagemap.c
> +++ b/drivers/gpu/drm/drm_pagemap.c
> @@ -1122,7 +1122,7 @@ int drm_pagemap_evict_to_ram(struct drm_pagemap_devmem *devmem_allocation)
>  {
>  	const struct drm_pagemap_devmem_ops *ops = devmem_allocation->ops;
>  	struct drm_pagemap_iova_state state = {};
> -	unsigned long npages, mpages = 0;
> +	unsigned long npages, mpages;
>  	struct page **pages;
>  	unsigned long *src, *dst;
>  	struct drm_pagemap_addr *pagemap_addr;
> @@ -1133,6 +1133,7 @@ int drm_pagemap_evict_to_ram(struct drm_pagemap_devmem *devmem_allocation)
>  	npages = devmem_allocation->size >> PAGE_SHIFT;
>  
>  retry:
> +	mpages = 0;
>  	if (!mmget_not_zero(devmem_allocation->mm))
>  		return -EFAULT;
>  
> -- 
> 2.43.0
>