net_dma: poll for a descriptor after allocation failure

Handle descriptor allocation failures by polling for a descriptor.  The
driver will force forward progress when polled.  In the best case this
polling interval will be the time it takes for one dma memcpy
transaction to complete.  In the worst case, channel hang, we will need
to wait 100ms for the cleanup watchdog to fire (ioatdma driver).

Signed-off-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>

diff --git a/drivers/dma/iovlock.c b/drivers/dma/iovlock.c
index 9f6fe46..c0a272c 100644 (file)
--- a/drivers/dma/iovlock.c
+++ b/drivers/dma/iovlock.c
@@ -183,6 +183,11 @@ dma_cookie_t dma_memcpy_to_iovec(struct dma_chan *chan, struct iovec *iov,
                                        iov_byte_offset,
                                        kdata,
                                        copy);
+                       /* poll for a descriptor slot */
+                       if (unlikely(dma_cookie < 0)) {
+                               dma_async_issue_pending(chan);
+                               continue;
+                       }
 
                        len -= copy;
                        iov[iovec_idx].iov_len -= copy;
@@ -248,6 +253,11 @@ dma_cookie_t dma_memcpy_pg_to_iovec(struct dma_chan *chan, struct iovec *iov,
                                        page,
                                        offset,
                                        copy);
+                       /* poll for a descriptor slot */
+                       if (unlikely(dma_cookie < 0)) {
+                               dma_async_issue_pending(chan);
+                               continue;
+                       }
 
                        len -= copy;
                        iov[iovec_idx].iov_len -= copy;