Skip to contentSkip to conversation

spi: spi-davinci: Fix direction in dma_map_single()

PATCH In review Christian Eggers 2 participants 2 messages Jul 29, 2013 SPI

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 spi: spi-davinci: Fix direction in dma_map_single()

The following patch I already sent on 2013-04-18 and 2013-06-06. Unfortunately it hasn't been committed yet.

Commit 048177ce3b3962852fd34a7e04938959271c7e70 (spi: spi-davinci:
convert to DMA engine API) introduced a regression: dma_map_single() is called with direction DMA_FROM_DEVICE for rx and for tx.

---

spi/spi-davinci.c renamed from spi/spi-davinci.c.orig +1 -1
@@ -608,7 +608,7 @@ static int davinci_spi_bufs(struct spi_d
 		else
 			buf = (void *)t->tx_buf;
 		t->tx_dma = dma_map_single(&spi->dev, buf,
-				t->len, DMA_FROM_DEVICE);
+				t->len, DMA_TO_DEVICE);
 		if (!t->tx_dma) {
 			ret = -EFAULT;
 			goto err_tx_map;
maintainer
#1 lore raw
Quoted 2 lines · Christian Eggers

The following patch I already sent on 2013-04-18 and 2013-06-06. Unfortunately it hasn't been committed yet.

I'm not sure where you submitted this to but it doesn't appear to have been to either me or to the mailing list. If you want a patch applying you need to send it to the maintainer as covered in SubmittingPatches. Google and your current post suggest it went to lkml and the retired sourceforge list...

As also covered in SubmittingPatches you should also not include things like this in the body of the commit message, put them after the --- where they will be stripped automatically by tools like git am.

Quoted 3 lines

Commit 048177ce3b3962852fd34a7e04938959271c7e70 (spi: spi-davinci:
convert to DMA engine API) introduced a regression: dma_map_single() is called with direction DMA_FROM_DEVICE for rx and for tx.

This doesn't apply automatically since:

spi/spi-davinci.c 2013-04-18 20:54:51.900623956 +0200

Quoted 3 lines

---

spi/spi-davinci.c 2013-04-18 20:54:51.900623956 +0200 renamed from spi/spi-davinci.c.orig 2013-04-18 20:54:02.728719412 +0200 +0 -0

you've generated the patch in -p0 format not -p1 format as covered in SubmittingPatches. I fixed this up manually but please do follow the process in SubmittingPatches, it makes life easier.