[PATCH 4.2.y-ckt 59/93] drm/i915/userptr: Hold mmref whilst calling get-user-pages

[Kamal Mostafa]

4.2.8-ckt9 -stable review patch.  If anyone has any objections, please let me know.

---8<------------------------------------------------------------

From: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>

commit db9f9203e27495b2d151b695504e286eec026e9b upstream.

Holding a reference to the containing task_struct is not sufficient to
prevent the mm_struct from being reaped under memory pressure. If this
happens whilst we are calling get_user_pages(), explosions erupt -
sometimes an immediate GPF, sometimes page flag corruption. To prevent
the target mm from being reaped as we are reading from it, acquire a
reference before we begin.

Testcase: igt/gem_shrink/*userptr
Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
Cc: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx>
Cc: MichaÅ Winiarski <michal.winiarski@xxxxxxxxx>
Reviewed-by: MichaÅ Winiarski <michal.winiarski@xxxxxxxxx>
Link: http://patchwork.freedesktop.org/patch/msgid/1459864801-28606-2-git-send-email-chris@xxxxxxxxxxxxxxxxxx
(cherry picked from commit 40313f0cd0b711a7a5905e5182422799e157d8aa)
Signed-off-by: Jani Nikula <jani.nikula@xxxxxxxxx>
[ kamal: backport to 4.2-stable: s/get_user_pages_remote/get_user_pages/ ]
Signed-off-by: Kamal Mostafa <kamal@xxxxxxxxxxxxx>
---
 drivers/gpu/drm/i915/i915_gem_userptr.c | 29 +++++++++++++++++------------
 1 file changed, 17 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c
index 4037f89..76d8290 100644
--- a/drivers/gpu/drm/i915/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/i915_gem_userptr.c
@@ -610,19 +610,24 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work)
 	if (pvec != NULL) {
 		struct mm_struct *mm = obj->userptr.mm->mm;
 
-		down_read(&mm->mmap_sem);
-		while (pinned < npages) {
-			ret = get_user_pages(work->task, mm,
-					     obj->userptr.ptr + pinned * PAGE_SIZE,
-					     npages - pinned,
-					     !obj->userptr.read_only, 0,
-					     pvec + pinned, NULL);
-			if (ret < 0)
-				break;
-
-			pinned += ret;
+		ret = -EFAULT;
+		if (atomic_inc_not_zero(&mm->mm_users)) {
+			down_read(&mm->mmap_sem);
+			while (pinned < npages) {
+				ret = get_user_pages
+					(work->task, mm,
+					 obj->userptr.ptr + pinned * PAGE_SIZE,
+					 npages - pinned,
+					 !obj->userptr.read_only, 0,
+					 pvec + pinned, NULL);
+				if (ret < 0)
+					break;
+
+				pinned += ret;
+			}
+			up_read(&mm->mmap_sem);
+			mmput(mm);
 		}
-		up_read(&mm->mmap_sem);
 	}
 
 	mutex_lock(&dev->struct_mutex);
-- 
2.7.4