summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915/i915_gem.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/i915/i915_gem.c')
-rw-r--r--drivers/gpu/drm/i915/i915_gem.c97
1 files changed, 94 insertions, 3 deletions
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 6908123162d1..10777da73039 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1434,6 +1434,12 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
trace_i915_gem_object_pwrite(obj, args->offset, args->size);
+ ret = -ENODEV;
+ if (obj->ops->pwrite)
+ ret = obj->ops->pwrite(obj, args);
+ if (ret != -ENODEV)
+ goto err;
+
ret = i915_gem_object_wait(obj,
I915_WAIT_INTERRUPTIBLE |
I915_WAIT_ALL,
@@ -2119,6 +2125,7 @@ i915_gem_object_truncate(struct drm_i915_gem_object *obj)
*/
shmem_truncate_range(file_inode(obj->base.filp), 0, (loff_t)-1);
obj->mm.madv = __I915_MADV_PURGED;
+ obj->mm.pages = ERR_PTR(-EFAULT);
}
/* Try to discard unwanted pages */
@@ -2218,7 +2225,9 @@ void __i915_gem_object_put_pages(struct drm_i915_gem_object *obj,
__i915_gem_object_reset_page_iter(obj);
- obj->ops->put_pages(obj, pages);
+ if (!IS_ERR(pages))
+ obj->ops->put_pages(obj, pages);
+
unlock:
mutex_unlock(&obj->mm.lock);
}
@@ -2437,7 +2446,7 @@ int __i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
if (err)
return err;
- if (unlikely(!obj->mm.pages)) {
+ if (unlikely(IS_ERR_OR_NULL(obj->mm.pages))) {
err = ____i915_gem_object_get_pages(obj);
if (err)
goto unlock;
@@ -2515,7 +2524,7 @@ void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
pinned = true;
if (!atomic_inc_not_zero(&obj->mm.pages_pin_count)) {
- if (unlikely(!obj->mm.pages)) {
+ if (unlikely(IS_ERR_OR_NULL(obj->mm.pages))) {
ret = ____i915_gem_object_get_pages(obj);
if (ret)
goto err_unlock;
@@ -2563,6 +2572,75 @@ err_unlock:
goto out_unlock;
}
+static int
+i915_gem_object_pwrite_gtt(struct drm_i915_gem_object *obj,
+ const struct drm_i915_gem_pwrite *arg)
+{
+ struct address_space *mapping = obj->base.filp->f_mapping;
+ char __user *user_data = u64_to_user_ptr(arg->data_ptr);
+ u64 remain, offset;
+ unsigned int pg;
+
+ /* Before we instantiate/pin the backing store for our use, we
+ * can prepopulate the shmemfs filp efficiently using a write into
+ * the pagecache. We avoid the penalty of instantiating all the
+ * pages, important if the user is just writing to a few and never
+ * uses the object on the GPU, and using a direct write into shmemfs
+ * allows it to avoid the cost of retrieving a page (either swapin
+ * or clearing-before-use) before it is overwritten.
+ */
+ if (READ_ONCE(obj->mm.pages))
+ return -ENODEV;
+
+ /* Before the pages are instantiated the object is treated as being
+ * in the CPU domain. The pages will be clflushed as required before
+ * use, and we can freely write into the pages directly. If userspace
+ * races pwrite with any other operation; corruption will ensue -
+ * that is userspace's prerogative!
+ */
+
+ remain = arg->size;
+ offset = arg->offset;
+ pg = offset_in_page(offset);
+
+ do {
+ unsigned int len, unwritten;
+ struct page *page;
+ void *data, *vaddr;
+ int err;
+
+ len = PAGE_SIZE - pg;
+ if (len > remain)
+ len = remain;
+
+ err = pagecache_write_begin(obj->base.filp, mapping,
+ offset, len, 0,
+ &page, &data);
+ if (err < 0)
+ return err;
+
+ vaddr = kmap(page);
+ unwritten = copy_from_user(vaddr + pg, user_data, len);
+ kunmap(page);
+
+ err = pagecache_write_end(obj->base.filp, mapping,
+ offset, len, len - unwritten,
+ page, data);
+ if (err < 0)
+ return err;
+
+ if (unwritten)
+ return -EFAULT;
+
+ remain -= len;
+ user_data += len;
+ offset += len;
+ pg = 0;
+ } while (remain);
+
+ return 0;
+}
+
static bool ban_context(const struct i915_gem_context *ctx)
{
return (i915_gem_context_is_bannable(ctx) &&
@@ -3029,6 +3107,16 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
args->timeout_ns -= ktime_to_ns(ktime_sub(ktime_get(), start));
if (args->timeout_ns < 0)
args->timeout_ns = 0;
+
+ /*
+ * Apparently ktime isn't accurate enough and occasionally has a
+ * bit of mismatch in the jiffies<->nsecs<->ktime loop. So patch
+ * things up to make the test happy. We allow up to 1 jiffy.
+ *
+ * This is a regression from the timespec->ktime conversion.
+ */
+ if (ret == -ETIME && !nsecs_to_jiffies(args->timeout_ns))
+ args->timeout_ns = 0;
}
i915_gem_object_put(obj);
@@ -3974,8 +4062,11 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
static const struct drm_i915_gem_object_ops i915_gem_object_ops = {
.flags = I915_GEM_OBJECT_HAS_STRUCT_PAGE |
I915_GEM_OBJECT_IS_SHRINKABLE,
+
.get_pages = i915_gem_object_get_pages_gtt,
.put_pages = i915_gem_object_put_pages_gtt,
+
+ .pwrite = i915_gem_object_pwrite_gtt,
};
struct drm_i915_gem_object *