V4L/DVB (9726): cx18: Restore buffers that have fallen out of the transfer rotation

Restore buffers that have fallen out of the transfer rotation, and check
for coherent mailbox data when processing a stale mailbox.

Signed-off-by: Andy Walls <awalls@radix.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

1 files changed, 46 insertions, 13 deletions

diff --git a/drivers/media/video/cx18/cx18-mailbox.c b/drivers/media/video/cx18/cx18-mailbox.c
index e5d4f3112293..abd39aaa345c 100644
--- a/drivers/media/video/cx18/cx18-mailbox.c
+++ b/drivers/media/video/cx18/cx18-mailbox.c
@@ -120,7 +120,7 @@ static void dump_mb(struct cx18 *cx, struct cx18_mailbox *mb, char *name)
 
 static void epu_dma_done(struct cx18 *cx, struct cx18_epu_work_order *order)
 {
-        u32 handle, mdl_ack_count;
+        u32 handle, mdl_ack_count, id;
         struct cx18_mailbox *mb;
         struct cx18_mdl_ack *mdl_ack;
         struct cx18_stream *s;
@@ -133,19 +133,50 @@ static void epu_dma_done(struct cx18 *cx, struct cx18_epu_work_order *order)
 
         if (s == NULL) {
                 CX18_WARN("Got DMA done notification for unknown/inactive"
-                          " handle %d\n", handle);
+                          " handle %d, %s mailbox seq no %d\n", handle,
+                          (order->flags & CX18_F_EWO_MB_STALE_UPON_RECEIPT) ?
+                          "stale" : "good", mb->request);
                 return;
         }
 
         mdl_ack_count = mb->args[2];
         mdl_ack = order->mdl_ack;
         for (i = 0; i < mdl_ack_count; i++, mdl_ack++) {
-                buf = cx18_queue_get_buf(s, mdl_ack->id, mdl_ack->data_used);
-                CX18_DEBUG_HI_DMA("DMA DONE for %s (buffer %d)\n", s->name,
-                                  mdl_ack->id);
+                id = mdl_ack->id;
+                /*
+                 * Simple integrity check for processing a stale (and possibly
+                 * inconsistent mailbox): make sure the buffer id is in the
+                 * valid range for the stream.
+                 *
+                 * We go through the trouble of dealing with stale mailboxes
+                 * because most of the time, the mailbox data is still valid and
+                 * unchanged (and in practice the firmware ping-pongs the
+                 * two mdl_ack buffers so mdl_acks are not stale).
+                 *
+                 * There are occasions when we get a half changed mailbox,
+                 * which this check catches for a handle & id mismatch.  If the
+                 * handle and id do correspond, the worst case is that we
+                 * completely lost the old buffer, but pick up the new buffer
+                 * early (but the new mdl_ack is guaranteed to be good in this
+                 * case as the firmware wouldn't point us to a new mdl_ack until
+                 * it's filled in).
+                 *
+                 * cx18_queue_get buf() will detect the lost buffers
+                 * and put them back in rotation eventually.
+                 */
+                if ((order->flags & CX18_F_EWO_MB_STALE_UPON_RECEIPT) &&
+                    !(id >= s->mdl_offset &&
+                      id < (s->mdl_offset + s->buffers))) {
+                        CX18_WARN("Fell behind! Ignoring stale mailbox with "
+                                  " inconsistent data. Lost buffer for mailbox "
+                                  "seq no %d\n", mb->request);
+                        break;
+                }
+                buf = cx18_queue_get_buf(s, id, mdl_ack->data_used);
+                CX18_DEBUG_HI_DMA("DMA DONE for %s (buffer %d)\n", s->name, id);
                 if (buf == NULL) {
                         CX18_WARN("Could not find buf %d for stream %s\n",
-                                  mdl_ack->id, s->name);
+                                  id, s->name);
                         continue;
                 }
 
@@ -158,6 +189,7 @@ static void epu_dma_done(struct cx18 *cx, struct cx18_epu_work_order *order)
                                          buf->bytesused);
 
                         cx18_buf_sync_for_device(s, buf);
+                        cx18_enqueue(s, buf, &s->q_free);
 
                         if (s->handle != CX18_INVALID_TASK_HANDLE &&
                             test_bit(CX18_F_S_STREAMING, &s->s_flags))
@@ -257,10 +289,10 @@ static void mb_ack_irq(struct cx18 *cx, struct cx18_epu_work_order *order)
         /* Don't ack if the RPU has gotten impatient and timed us out */
         if (req != cx18_readl(cx, &ack_mb->request) ||
             req == cx18_readl(cx, &ack_mb->ack)) {
-                CX18_WARN("Possibly falling behind: %s self-ack'ed our incoming"
-                          " %s to EPU mailbox (sequence no. %u) while "
-                          "processing\n",
-                          rpu_str[order->rpu], rpu_str[order->rpu], req);
+                CX18_DEBUG_WARN("Possibly falling behind: %s self-ack'ed our "
+                                "incoming %s to EPU mailbox (sequence no. %u) "
+                                "while processing\n",
+                                rpu_str[order->rpu], rpu_str[order->rpu], req);
                 order->flags |= CX18_F_EWO_MB_STALE_WHILE_PROC;
                 return;
         }
@@ -407,9 +439,10 @@ void cx18_api_epu_cmd_irq(struct cx18 *cx, int rpu)
                            2 * sizeof(u32));
 
         if (order_mb->request == order_mb->ack) {
-                CX18_WARN("Possibly falling behind: %s self-ack'ed our incoming"
-                          " %s to EPU mailbox (sequence no. %u)\n",
-                          rpu_str[rpu], rpu_str[rpu], order_mb->request);
+                CX18_DEBUG_WARN("Possibly falling behind: %s self-ack'ed our "
+                                "incoming %s to EPU mailbox (sequence no. %u)"
+                                "\n",
+                                rpu_str[rpu], rpu_str[rpu], order_mb->request);
                 dump_mb(cx, order_mb, "incoming");
                 order->flags = CX18_F_EWO_MB_STALE_UPON_RECEIPT;
         }