提交 060810d7 编写于 作者: B Ben Skeggs

drm/nouveau: fix locking issues in page flipping paths

b580c9e2 introduced additional problems
while trying to solve issues that became apparent while porting to the
new reservation stuff.

The major problem was that the the previously mentioned patch took the
client mutex earlier than previously, but the pinning of new_bo can
can potentially cause a buffer move, which would result in attempting to
acquire the same mutex again.

This commit attempts to fix that "fix".

Thanks to Maarten for the tips on keeping lockdep happy and cooking :)
Reported-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
Acked-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
上级 06d5a24f
......@@ -973,7 +973,7 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr,
struct ttm_mem_reg *old_mem = &bo->mem;
int ret;
mutex_lock(&chan->cli->mutex);
mutex_lock_nested(&chan->cli->mutex, SINGLE_DEPTH_NESTING);
/* create temporary vmas for the transfer and attach them to the
* old nouveau_mem node, these will get cleaned up after ttm has
......
......@@ -524,9 +524,12 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
struct nouveau_page_flip_state *s;
struct nouveau_channel *chan = NULL;
struct nouveau_fence *fence;
struct list_head res;
struct ttm_validate_buffer res_val[2];
struct ttm_validate_buffer resv[2] = {
{ .bo = &old_bo->bo },
{ .bo = &new_bo->bo },
};
struct ww_acquire_ctx ticket;
LIST_HEAD(res);
int ret;
if (!drm->channel)
......@@ -545,27 +548,19 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
chan = drm->channel;
spin_unlock(&old_bo->bo.bdev->fence_lock);
mutex_lock(&chan->cli->mutex);
if (new_bo != old_bo) {
ret = nouveau_bo_pin(new_bo, TTM_PL_FLAG_VRAM);
if (likely(!ret)) {
res_val[0].bo = &old_bo->bo;
res_val[1].bo = &new_bo->bo;
INIT_LIST_HEAD(&res);
list_add_tail(&res_val[0].head, &res);
list_add_tail(&res_val[1].head, &res);
ret = ttm_eu_reserve_buffers(&ticket, &res);
if (ret)
nouveau_bo_unpin(new_bo);
}
} else
ret = ttm_bo_reserve(&new_bo->bo, false, false, false, 0);
if (ret) {
mutex_unlock(&chan->cli->mutex);
goto fail_free;
list_add(&resv[1].head, &res);
}
list_add(&resv[0].head, &res);
mutex_lock(&chan->cli->mutex);
ret = ttm_eu_reserve_buffers(&ticket, &res);
if (ret)
goto fail_unpin;
/* Initialize a page flip struct */
*s = (struct nouveau_page_flip_state)
......@@ -576,11 +571,9 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
/* Emit a page flip */
if (nv_device(drm->device)->card_type >= NV_50) {
ret = nv50_display_flip_next(crtc, fb, chan, 0);
if (ret) {
mutex_unlock(&chan->cli->mutex);
if (ret)
goto fail_unreserve;
}
}
ret = nouveau_page_flip_emit(chan, old_bo, new_bo, s, &fence);
mutex_unlock(&chan->cli->mutex);
......@@ -590,22 +583,18 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
/* Update the crtc struct and cleanup */
crtc->fb = fb;
if (old_bo != new_bo) {
ttm_eu_fence_buffer_objects(&ticket, &res, fence);
if (old_bo != new_bo)
nouveau_bo_unpin(old_bo);
} else {
nouveau_bo_fence(new_bo, fence);
ttm_bo_unreserve(&new_bo->bo);
}
nouveau_fence_unref(&fence);
return 0;
fail_unreserve:
if (old_bo != new_bo) {
ttm_eu_backoff_reservation(&ticket, &res);
fail_unpin:
mutex_unlock(&chan->cli->mutex);
if (old_bo != new_bo)
nouveau_bo_unpin(new_bo);
} else
ttm_bo_unreserve(&new_bo->bo);
fail_free:
kfree(s);
return ret;
......
......@@ -284,8 +284,6 @@ static int nouveau_drm_probe(struct pci_dev *pdev,
return 0;
}
static struct lock_class_key drm_client_lock_class_key;
static int
nouveau_drm_load(struct drm_device *dev, unsigned long flags)
{
......@@ -297,7 +295,6 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags)
ret = nouveau_cli_create(pdev, "DRM", sizeof(*drm), (void**)&drm);
if (ret)
return ret;
lockdep_set_class(&drm->client.mutex, &drm_client_lock_class_key);
dev->dev_private = drm;
drm->dev = dev;
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册