dri-devel.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Alex Deucher <alexdeucher@gmail.com>
To: dri-devel@lists.freedesktop.org
Cc: Alex Deucher <alexander.deucher@amd.com>, "monk.liu" <monk.liu@amd.com>
Subject: [PATCH 50/88] drm/amdgpu: fix userptr BO unpin bug (v2)
Date: Tue, 26 May 2015 23:19:49 -0400	[thread overview]
Message-ID: <1432696827-3752-20-git-send-email-alexander.deucher@amd.com> (raw)
In-Reply-To: <1432696827-3752-1-git-send-email-alexander.deucher@amd.com>

From: "monk.liu" <monk.liu@amd.com>

sg could point to array of contigiouse page*, only free page could lead
to memory leak.

v2: use iterator

Signed-off-by: monk.liu <monk.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 120e6e7..729e0bb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -555,8 +555,7 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_get_adev(ttm->bdev);
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
-	struct scatterlist *sg;
-	int i;
+	struct sg_page_iter sg_iter;
 
 	int write = !(gtt->userflags & AMDGPU_GEM_USERPTR_READONLY);
 	enum dma_data_direction direction = write ?
@@ -569,9 +568,8 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 	/* free the sg table and pages again */
 	dma_unmap_sg(adev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
 
-	for_each_sg(ttm->sg->sgl, sg, ttm->sg->nents, i) {
-		struct page *page = sg_page(sg);
-
+	for_each_sg_page(ttm->sg->sgl, &sg_iter, ttm->sg->nents, 0) {
+		struct page *page = sg_page_iter_page(&sg_iter);
 		if (!(gtt->userflags & AMDGPU_GEM_USERPTR_READONLY))
 			set_page_dirty(page);
 
-- 
1.8.3.1

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

  parent reply	other threads:[~2015-05-27  3:20 UTC|newest]

Thread overview: 67+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-05-27  3:19 [PATCH 00/88] amdgpu driver updates Alex Deucher
2015-05-27  3:19 ` [PATCH 29/88] drm/amdgpu: add amdgpu uapi header (v4) Alex Deucher
2015-06-05 18:57   ` Jerome Glisse
2015-06-09  2:51     ` Alex Deucher
2015-06-09  9:13       ` Marek Olšák
2015-05-27  3:19 ` [PATCH 30/88] drm/amdgpu: add atombios headers Alex Deucher
2015-05-27  3:19 ` [PATCH 31/88] drm/amdgpu: add clearstate_defs.h Alex Deucher
2015-05-27  3:19 ` [PATCH 32/88] drm/amdgpu: add ppsmc.h Alex Deucher
2015-05-27  3:19 ` [PATCH 33/88] drm/amdgpu: add amdgpu_family.h Alex Deucher
2015-05-27  3:19 ` [PATCH 34/88] drm/amdgpu: add amdgpu.h (v2) Alex Deucher
2015-05-27  3:19 ` [PATCH 36/88] drm/amdgpu: fix const warnings in amdgpu_connectors.c Alex Deucher
2015-05-27  3:19 ` [PATCH 37/88] drm/amdgpu: Do not directly dereference pointers to BIOS area Alex Deucher
2015-05-27  3:19 ` [PATCH 40/88] drm/amdgpu: add CIK pci ids Alex Deucher
2015-05-27  3:19 ` [PATCH 41/88] drm/amdgpu: add VI " Alex Deucher
2015-05-27  3:19 ` [PATCH 42/88] drm/amdgpu: drop ttm two ended allocation Alex Deucher
2015-05-27  3:19 ` [PATCH 43/88] drm/amdgpu: fix error check issue in amdgpu_mn_invalidate_range_start Alex Deucher
2015-05-27  3:19 ` [PATCH 44/88] drm/amdgpu: fix bug occurs when bo_list is NULL Alex Deucher
2015-05-27  3:19 ` [PATCH 45/88] drm/amdgpu: let bo_list handler start from 1 Alex Deucher
2015-05-27  3:19 ` [PATCH 46/88] drm/amdgpu: fix error handling in cz_dpm_hw_fini/cz_dpm_suspend Alex Deucher
2015-05-27  3:19 ` [PATCH 47/88] drm/amdgpu: memset gds_info struct in info ioctl Alex Deucher
2015-05-27  3:19 ` [PATCH 48/88] drm/amdgpu fix amdgpu.dpm=0 (v2) Alex Deucher
2015-05-27  3:19 ` [PATCH 49/88] drm/amdgpu: remove AMDGPU_GEM_CREATE_CPU_GTT_UC Alex Deucher
2015-06-06 17:08   ` Oded Gabbay
2015-06-06 19:09     ` Christian König
2015-06-06 19:32       ` Oded Gabbay
2015-06-12 15:47   ` Emil Velikov
2015-06-12 16:41     ` Bridgman, John
2015-06-12 16:57       ` Emil Velikov
2015-05-27  3:19 ` Alex Deucher [this message]
2015-05-27  3:19 ` [PATCH 51/88] drm/amdgpu: fix userptr lockup Alex Deucher
2015-05-27  3:19 ` [PATCH 52/88] drm/amdgpu: remove unsafe context releasing Alex Deucher
2015-05-27  3:19 ` [PATCH 53/88] drm/amdgpu: make the CTX ioctl thread-safe Alex Deucher
2015-05-27  3:19 ` [PATCH 54/88] drm/amdgpu: allow unaligned memory access (v2) Alex Deucher
2015-05-27  3:19 ` [PATCH 55/88] drm/amdgpu: add ctx_id to the WAIT_CS IOCTL (v4) Alex Deucher
2015-05-27  3:19 ` [PATCH 56/88] drm/amdgpu: check context id for context switching (v2) Alex Deucher
2015-05-27  3:19 ` [PATCH 57/88] drm/amdgpu: add flags for amdgpu_ib structure Alex Deucher
2015-05-27  3:19 ` [PATCH 58/88] drm/amdgpu: add CE preamble flag v3 Alex Deucher
2015-05-27  3:19 ` [PATCH 59/88] drm/amdgpu: always emit GDS switch Alex Deucher
2015-05-27  3:19 ` [PATCH 60/88] drm/amdgpu: cleanup HDP flush handling Alex Deucher
2015-05-27  3:20 ` [PATCH 61/88] drm/amdgpu: fix dereference before check Alex Deucher
2015-05-27  3:20 ` [PATCH 62/88] drm/amdgpu: fix context switch Alex Deucher
2015-05-27  3:20 ` [PATCH 63/88] drm/amdgpu: expose the max virtual address Alex Deucher
2015-05-27  3:20 ` [PATCH 64/88] drm/amdgpu: do necessary NULL check Alex Deucher
2015-05-27  3:20 ` [PATCH 65/88] drm/amdgpu: switch to amdgpu folder for firmware files v2 Alex Deucher
2015-05-27  3:20 ` [PATCH 66/88] drm/amdgpu: rewording some left radeons Alex Deucher
2015-05-27  3:20 ` [PATCH 67/88] drm/amdgpu: add new bonaire pci id Alex Deucher
2015-05-27  3:20 ` [PATCH 68/88] drm/amdgpu: add some new tonga pci ids Alex Deucher
2015-05-27  3:20 ` [PATCH 69/88] drm/amdgpu: add and implement the GPU reset status query Alex Deucher
2015-05-27  3:20 ` [PATCH 70/88] drm/amdgpu: take the mode_config mutex when handling hpds Alex Deucher
2015-05-27  3:20 ` [PATCH 71/88] drm/amdgpu: make some DP parameters const Alex Deucher
2015-05-27  3:20 ` [PATCH 72/88] drm/amdgpu: simplify DPCD debug output Alex Deucher
2015-05-27  3:20 ` [PATCH 73/88] drm/amdgpu: retry dcpd fetch Alex Deucher
2015-05-27  3:20 ` [PATCH 74/88] drm/amdgpu: fix VM_CONTEXT*_PAGE_TABLE_END_ADDR handling Alex Deucher
2015-05-27  3:20 ` [PATCH 75/88] drm/amdgpu: enforce AMDGPU_GEM_CREATE_NO_CPU_ACCESS Alex Deucher
2015-05-27  3:20 ` [PATCH 76/88] drm/amdgpu: validate amdgpu_vm_bo_map parameters Alex Deucher
2015-05-27  3:20 ` [PATCH 77/88] drm/amdgpu: actually use the VM map parameters Alex Deucher
2015-05-27  3:20 ` [PATCH 78/88] drm/amdgpu: don't set unused tiling flags Alex Deucher
2015-05-27  3:20 ` [PATCH 79/88] drm/amdgpu: rework " Alex Deucher
2015-05-27  3:20 ` [PATCH 80/88] drm/amdgpu: remove unused TRACE_SYSTEM_STRING define Alex Deucher
2015-05-27  3:20 ` [PATCH 81/88] drm/amdgpu: recalculate VCE firmware BO size Alex Deucher
2015-05-27  3:20 ` [PATCH 82/88] drm/amdgpu: implement VCE two instances support Alex Deucher
2015-05-27  3:20 ` [PATCH 83/88] drm/amdgpu: enable uvd dpm and powergating Alex Deucher
2015-05-27  3:20 ` [PATCH 84/88] drm/amdgpu: port fault_reserve_notify changes from radeon Alex Deucher
2015-05-27  3:20 ` [PATCH 85/88] drm/amdgpu: drop AMDGPU_FENCE_SIGNALED_SEQ Alex Deucher
2015-05-27  3:20 ` [PATCH 86/88] drm/amdgpu: drop allocation flag masks Alex Deucher
2015-05-27  3:20 ` [PATCH 87/88] drm/amdgpu: rename amdgpu_ip_funcs to amd_ip_funcs (v2) Alex Deucher
2015-05-27  3:20 ` [PATCH 88/88] drm/amdgpu: implement the allocation range (v2) Alex Deucher

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=1432696827-3752-20-git-send-email-alexander.deucher@amd.com \
    --to=alexdeucher@gmail.com \
    --cc=alexander.deucher@amd.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=monk.liu@amd.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;
as well as URLs for NNTP newsgroup(s).