[3.16.y-ckt stable] Patch "drm/ttm: Don't evict BOs outside of the requested placement range" has been added to staging queue

Luis Henriques luis.henriques at canonical.com
Mon Nov 17 11:16:03 UTC 2014


This is a note to let you know that I have just added a patch titled

    drm/ttm: Don't evict BOs outside of the requested placement range

to the linux-3.16.y-queue branch of the 3.16.y-ckt extended stable tree 
which can be found at:

 http://kernel.ubuntu.com/git?p=ubuntu/linux.git;a=shortlog;h=refs/heads/linux-3.16.y-queue

This patch is scheduled to be released in version 3.16.7-ckt2.

If you, or anyone else, feels it should not be added to this tree, please 
reply to this email.

For more information about the 3.16.y-ckt tree, see
https://wiki.ubuntu.com/Kernel/Dev/ExtendedStable

Thanks.
-Luis

------

>From ebc39080bfdbc1b1a109b44a45faf3690f240690 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Michel=20D=C3=A4nzer?= <michel.daenzer at amd.com>
Date: Thu, 9 Oct 2014 15:02:59 +0900
Subject: drm/ttm: Don't evict BOs outside of the requested placement range
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

commit e300180f71037fd9ed1ca967006fd9f3ee466bcd upstream.

The radeon driver uses placement range restrictions for several reasons,
in particular to make sure BOs in VRAM can be accessed by the CPU, e.g.
during a page fault.

Without this change, TTM could evict other BOs while trying to satisfy
the requested placement, even if the evicted BOs were outside of the
requested placement range. Doing so didn't free up any space in the
requested placement range, so the (potentially high) eviction cost was
incurred for no benefit.

Nominating for stable because radeon driver changes in 3.17 made this
much more noticeable than before.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84662
Signed-off-by: Michel Dänzer <michel.daenzer at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
[ luis: backported to 3.16: based on Michel's backport ]
Signed-off-by: Luis Henriques <luis.henriques at canonical.com>
---
 drivers/gpu/drm/ttm/ttm_bo.c | 21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index a13a10025ec7..f7befc4c0c49 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -716,6 +716,7 @@ out:

 static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 				uint32_t mem_type,
+				const struct ttm_placement *placement,
 				bool interruptible,
 				bool no_wait_gpu)
 {
@@ -727,8 +728,22 @@ static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 	spin_lock(&glob->lru_lock);
 	list_for_each_entry(bo, &man->lru, lru) {
 		ret = __ttm_bo_reserve(bo, false, true, false, 0);
-		if (!ret)
+		if (!ret) {
+			if (placement && (placement->fpfn || placement->lpfn)) {
+				/* Don't evict this BO if it's outside of the
+				 * requested placement range
+				 */
+				if (placement->fpfn >= (bo->mem.start + bo->mem.size) ||
+				    (placement->lpfn &&
+				     placement->lpfn <= bo->mem.start)) {
+					__ttm_bo_unreserve(bo);
+					ret = -EBUSY;
+					continue;
+				}
+			}
+
 			break;
+		}
 	}

 	if (ret) {
@@ -789,7 +804,7 @@ static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo,
 			return ret;
 		if (mem->mm_node)
 			break;
-		ret = ttm_mem_evict_first(bdev, mem_type,
+		ret = ttm_mem_evict_first(bdev, mem_type, placement,
 					  interruptible, no_wait_gpu);
 		if (unlikely(ret != 0))
 			return ret;
@@ -1245,7 +1260,7 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev,
 	spin_lock(&glob->lru_lock);
 	while (!list_empty(&man->lru)) {
 		spin_unlock(&glob->lru_lock);
-		ret = ttm_mem_evict_first(bdev, mem_type, false, false);
+		ret = ttm_mem_evict_first(bdev, mem_type, NULL, false, false);
 		if (ret) {
 			if (allow_errors) {
 				return ret;
--
2.1.0





More information about the kernel-team mailing list