Thread (3 messages) 3 messages, 3 authors, 2017-07-28

Fix dma unmap direction in iMX sahara aes calculation

From: Mogens Lauridsen <hidden>
Date: 2017-07-16 21:21:05
Subsystem: crypto api, the rest · Maintainers: Herbert Xu, "David S. Miller", Linus Torvalds

Hi,

The direction used in dma_unmap_sg in aes calc in sahara.c is wrong.
This result in the cache not being invalidated correct when aes
calculation is done and result is dma'ed to memory.
This is seen as sporadic wrong result from aes calc.

Thanks,
Mogens

Signed-off-by: Mogens Lauridsen <redacted>
diff --git a/drivers/crypto/sahara.c b/drivers/crypto/sahara.c
index 1d9ecd3..9538c52 100644ae
--- a/drivers/crypto/sahara.c
+++ b/drivers/crypto/sahara.c
@@ -543,10 +543,10 @@ static int sahara_hw_descriptor_create(struct sahara_dev *

 unmap_out:
        dma_unmap_sg(dev->device, dev->out_sg, dev->nb_out_sg,
-               DMA_TO_DEVICE);
+               DMA_FROM_DEVICE);
 unmap_in:
        dma_unmap_sg(dev->device, dev->in_sg, dev->nb_in_sg,
-               DMA_FROM_DEVICE);
+               DMA_TO_DEVICE);

        return -EINVAL;
 }
@@ -594,9 +594,9 @@ static int sahara_aes_process(struct ablkcipher_request *req
        }

        dma_unmap_sg(dev->device, dev->out_sg, dev->nb_out_sg,
-               DMA_TO_DEVICE);
-       dma_unmap_sg(dev->device, dev->in_sg, dev->nb_in_sg,
                DMA_FROM_DEVICE);
+       dma_unmap_sg(dev->device, dev->in_sg, dev->nb_in_sg,
+               DMA_TO_DEVICE);

        return 0;
 }
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help