Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Tejas Upadhyay <tejas.upadhyay@intel.com>
To: intel-xe@lists.freedesktop.org
Cc: himal.prasad.ghimiray@intel.com, rodrigo.vivi@intel.com,
	michal.wajdeczko@intel.com,
	Tejas Upadhyay <tejas.upadhyay@intel.com>
Subject: [PATCH V18 05/14] drm/xe: Guard teardown paths against purged BOs
Date: Wed, 26 Aug 2026 19:21:40 +0530	[thread overview]
Message-ID: <20260826135136.204044-21-tejas.upadhyay@intel.com> (raw)
In-Reply-To: <20260826135136.204044-16-tejas.upadhyay@intel.com>

VRAM page offlining can purge BOs that are still referenced by page
tables, exec queues, and DMA-buf exports. Add xe_bo_is_purged()
guards in the teardown paths to prevent unpinning or mapping an
already-purged BO:

- xe_bo_unpin_map_no_vm(): skip unpin if purged
- xe_dma_buf_map(): return -ENOENT early if purged
- xe_exec_queue_update_run_ticks(): skip LRC timestamp read if purged
- xe_pt_destroy(): skip unpin if purged

v3(Sashiko):
- Remove dma_resv is already held
v2(Himal):
- take dma_resv lock before calling xe_bo_is_purged()

Signed-off-by: Tejas Upadhyay <tejas.upadhyay@intel.com>
---
 drivers/gpu/drm/xe/xe_bo.h         | 3 ++-
 drivers/gpu/drm/xe/xe_dma_buf.c    | 3 +++
 drivers/gpu/drm/xe/xe_exec_queue.c | 8 ++++++--
 drivers/gpu/drm/xe/xe_pt.c         | 3 ++-
 4 files changed, 13 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_bo.h b/drivers/gpu/drm/xe/xe_bo.h
index eede678ad303..dfcd0e57073b 100644
--- a/drivers/gpu/drm/xe/xe_bo.h
+++ b/drivers/gpu/drm/xe/xe_bo.h
@@ -364,7 +364,8 @@ static inline void xe_bo_unpin_map_no_vm(struct xe_bo *bo)
 {
 	if (likely(bo)) {
 		xe_bo_lock(bo, false);
-		xe_bo_unpin(bo);
+		if (!xe_bo_is_purged(bo))
+			xe_bo_unpin(bo);
 		xe_bo_unlock(bo);
 
 		xe_bo_put(bo);
diff --git a/drivers/gpu/drm/xe/xe_dma_buf.c b/drivers/gpu/drm/xe/xe_dma_buf.c
index bf0728838ead..5d9f1cd24b7f 100644
--- a/drivers/gpu/drm/xe/xe_dma_buf.c
+++ b/drivers/gpu/drm/xe/xe_dma_buf.c
@@ -104,6 +104,9 @@ static struct sg_table *xe_dma_buf_map(struct dma_buf_attachment *attach,
 	struct sg_table *sgt;
 	int r = 0;
 
+	if (xe_bo_is_purged(bo))
+		return ERR_PTR(-ENOENT);
+
 	if (!attach->peer2peer && !xe_bo_can_migrate(bo, XE_PL_TT))
 		return ERR_PTR(-EOPNOTSUPP);
 
diff --git a/drivers/gpu/drm/xe/xe_exec_queue.c b/drivers/gpu/drm/xe/xe_exec_queue.c
index 91ed6c0fac84..91e4f3cb5617 100644
--- a/drivers/gpu/drm/xe/xe_exec_queue.c
+++ b/drivers/gpu/drm/xe/xe_exec_queue.c
@@ -1572,8 +1572,12 @@ void xe_exec_queue_update_run_ticks(struct xe_exec_queue *q)
 	 * errors.
 	 */
 	lrc = q->lrc[0];
-	new_ts = xe_lrc_update_timestamp(lrc, &old_ts);
-	q->xef->run_ticks[q->class] += (new_ts - old_ts) * q->width;
+	xe_bo_lock(lrc->bo, false);
+	if (!xe_bo_is_purged(lrc->bo)) {
+		new_ts = xe_lrc_update_timestamp(lrc, &old_ts);
+		q->xef->run_ticks[q->class] += (new_ts - old_ts) * q->width;
+	}
+	xe_bo_unlock(lrc->bo);
 
 	drm_dev_exit(idx);
 }
diff --git a/drivers/gpu/drm/xe/xe_pt.c b/drivers/gpu/drm/xe/xe_pt.c
index 5d990c1c3740..dbf1aa26a21b 100644
--- a/drivers/gpu/drm/xe/xe_pt.c
+++ b/drivers/gpu/drm/xe/xe_pt.c
@@ -212,7 +212,8 @@ void xe_pt_destroy(struct xe_pt *pt, u32 flags, struct llist_head *deferred)
 		return;
 
 	XE_WARN_ON(!list_empty(&pt->bo->ttm.base.gpuva.list));
-	xe_bo_unpin(pt->bo);
+	if (!xe_bo_is_purged(pt->bo))
+		xe_bo_unpin(pt->bo);
 	xe_bo_put_deferred(pt->bo, deferred);
 
 	if (pt->level > 0 && pt->num_live) {
-- 
2.52.0


  parent reply	other threads:[~2026-08-26 13:52 UTC|newest]

Thread overview: 54+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-26 13:51 [PATCH V18 00/14] Add memory page offlining support Tejas Upadhyay
2026-08-26 13:51 ` [PATCH V18 01/14] drm/xe: Link VRAM object with gpu buddy Tejas Upadhyay
2026-08-26 22:31   ` Andi Shyti
2026-08-26 13:51 ` [PATCH V18 02/14] drm/xe: Link LRC BO and its execution Queue Tejas Upadhyay
2026-08-26 22:34   ` Andi Shyti
2026-08-26 13:51 ` [PATCH V18 03/14] drm/xe: Extend BO purge to handle vram pages as well Tejas Upadhyay
2026-08-26 14:07   ` sashiko-bot
2026-08-26 22:42   ` Andi Shyti
2026-08-27  6:17     ` Upadhyay, Tejas
2026-08-27 14:40       ` Andi Shyti
2026-08-27 14:48         ` Upadhyay, Tejas
2026-08-28  5:25         ` Upadhyay, Tejas
2026-08-28  7:39           ` Andi Shyti
2026-08-28 17:29             ` Upadhyay, Tejas
2026-08-26 13:51 ` [PATCH V18 04/14] drm/xe/bo: Make xe_bo_is_user() public Tejas Upadhyay
2026-08-26 22:44   ` Andi Shyti
2026-08-26 13:51 ` Tejas Upadhyay [this message]
2026-08-26 14:12   ` [PATCH V18 05/14] drm/xe: Guard teardown paths against purged BOs sashiko-bot
2026-08-27  6:08   ` Ghimiray, Himal Prasad
2026-08-27  8:27     ` Upadhyay, Tejas
2026-08-26 13:51 ` [PATCH V18 06/14] drm/xe/vram: Extract buddy alloc and free helpers Tejas Upadhyay
2026-08-26 22:50   ` Andi Shyti
2026-08-26 13:51 ` [PATCH V18 07/14] drm/xe/vram: Add page offline data structures and lifecycle Tejas Upadhyay
2026-08-26 23:09   ` Andi Shyti
2026-08-27  6:19   ` Ghimiray, Himal Prasad
2026-08-26 13:51 ` [PATCH V18 08/14] drm/xe/vram: Add VRAM page offline fault handler Tejas Upadhyay
2026-08-26 14:05   ` sashiko-bot
2026-08-26 13:51 ` [PATCH V18 09/14] drm/xe/configfs: Add bad_page_reservation attribute Tejas Upadhyay
2026-08-27  6:42   ` Ghimiray, Himal Prasad
2026-08-27 15:00   ` Michal Wajdeczko
2026-08-28 17:48     ` Upadhyay, Tejas
2026-08-26 13:51 ` [PATCH V18 10/14] drm/xe/ras: Cache bad_page_reservation policy at init Tejas Upadhyay
2026-08-26 14:11   ` sashiko-bot
2026-08-27  6:45   ` Ghimiray, Himal Prasad
2026-08-26 13:51 ` [PATCH V18 11/14] drm/xe/vram: Check bad_page_reservation policy in fault handler Tejas Upadhyay
2026-08-26 14:08   ` sashiko-bot
2026-08-27  6:46   ` Ghimiray, Himal Prasad
2026-08-27 15:04   ` Michal Wajdeczko
2026-08-26 13:51 ` [PATCH V18 12/14] drm/xe: Expose bad VRAM pages via debugfs Tejas Upadhyay
2026-08-26 14:13   ` sashiko-bot
2026-08-27 15:16   ` Michal Wajdeczko
2026-08-28 19:06     ` Upadhyay, Tejas
2026-08-28 15:04   ` Rodrigo Vivi
2026-08-26 13:51 ` [PATCH V18 13/14] drm/xe/uapi: Expose ban reason in EXEC_QUEUE_GET_PROPERTY_BAN Tejas Upadhyay
2026-08-26 14:20   ` sashiko-bot
2026-08-27 18:26   ` Andi Shyti
2026-08-28  5:31     ` Upadhyay, Tejas
2026-08-26 13:51 ` [PATCH V18 14/14] drm/xe: Add fault-inject based VRAM page offline injection Tejas Upadhyay
2026-08-27  7:10   ` Ghimiray, Himal Prasad
2026-08-27  8:23     ` Upadhyay, Tejas
2026-08-26 14:37 ` ✗ CI.checkpatch: warning for Add memory page offlining support (rev21) Patchwork
2026-08-26 14:39 ` ✓ CI.KUnit: success " Patchwork
2026-08-26 15:21 ` ✓ Xe.CI.BAT: " Patchwork
2026-08-26 19:01 ` ✓ Xe.CI.FULL: " Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260826135136.204044-21-tejas.upadhyay@intel.com \
    --to=tejas.upadhyay@intel.com \
    --cc=himal.prasad.ghimiray@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=michal.wajdeczko@intel.com \
    --cc=rodrigo.vivi@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox