V4L/DVB (7023): Fix a regresion left by changeset 7e65d6e8f6df

Changeset 7e65d6e8f6df removed a very bad hack on mmap(). However, the fixes
weren't considering usermap and overlay memory models. This were breaking
direct reading from /dev/video?, used mostly by mpeg aware drivers.

Thanks to Steven Toth <stoth@linuxtv.org> for reporting the issue and
bissecting it.

Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>

diff --git a/drivers/media/video/videobuf-core.c b/drivers/media/video/videobuf-core.c
index c3adbd686ffe2ec1920691c8bde04c029534e5f8..80a14da9acefa939842e9d6806ca5dea99d199c1 100644 (file)
--- a/drivers/media/video/videobuf-core.c
+++ b/drivers/media/video/videobuf-core.c
@@ -102,10 +102,14 @@ int videobuf_iolock(struct videobuf_queue *q, struct videobuf_buffer *vb,
           since mmap_mapper() method should be called before _iolock.
           On some cases, the mmap_mapper() is called only after scheduling.
         */
-       wait_event_timeout(vb->done, q->is_mmapped, msecs_to_jiffies(100));
-       if (!q->is_mmapped) {
-               printk(KERN_ERR "Error: mmap_mapper() never called!\n");
-               return -EINVAL;
+       if (vb->memory == V4L2_MEMORY_MMAP) {
+               wait_event_timeout(vb->done, q->is_mmapped,
+                                  msecs_to_jiffies(100));
+               if (!q->is_mmapped) {
+                       printk(KERN_ERR
+                              "Error: mmap_mapper() never called!\n");
+                       return -EINVAL;
+               }
        }
 
        return CALL(q, iolock, q, vb, fbuf);