From: Andrew Davis <afd@ti.com>
To: Simon Glass <sjg@chromium.org>, Tom Rini <trini@konsulko.com>,
Vignesh Raghavendra <vigneshr@ti.com>, <u-boot@lists.denx.de>
Cc: Andrew Davis <afd@ti.com>
Subject: [PATCH 1/3] dma: Use dma-mapping for cache ops and sync after write
Date: Fri, 7 Oct 2022 12:11:11 -0500 [thread overview]
Message-ID: <20221007171113.31577-1-afd@ti.com> (raw)
The DMA'd memory area needs cleaned and invalidated after the DMA
write so that any stale cache lines do not mask new data.
Signed-off-by: Andrew Davis <afd@ti.com>
---
drivers/dma/dma-uclass.c | 17 +++++++++++++----
1 file changed, 13 insertions(+), 4 deletions(-)
diff --git a/drivers/dma/dma-uclass.c b/drivers/dma/dma-uclass.c
index 012609bb53..70f06f1f09 100644
--- a/drivers/dma/dma-uclass.c
+++ b/drivers/dma/dma-uclass.c
@@ -19,6 +19,7 @@
#include <asm/cache.h>
#include <dm/read.h>
#include <dma-uclass.h>
+#include <linux/dma-mapping.h>
#include <dt-structs.h>
#include <errno.h>
@@ -235,6 +236,8 @@ int dma_memcpy(void *dst, void *src, size_t len)
{
struct udevice *dev;
const struct dma_ops *ops;
+ dma_addr_t destination;
+ dma_addr_t source;
int ret;
ret = dma_get_device(DMA_SUPPORTS_MEM_TO_MEM, &dev);
@@ -245,11 +248,17 @@ int dma_memcpy(void *dst, void *src, size_t len)
if (!ops->transfer)
return -ENOSYS;
- /* Invalidate the area, so no writeback into the RAM races with DMA */
- invalidate_dcache_range((unsigned long)dst, (unsigned long)dst +
- roundup(len, ARCH_DMA_MINALIGN));
+ /* Clean the areas, so no writeback into the RAM races with DMA */
+ destination = dma_map_single(dst, len, DMA_FROM_DEVICE);
+ source = dma_map_single(src, len, DMA_TO_DEVICE);
- return ops->transfer(dev, DMA_MEM_TO_MEM, dst, src, len);
+ ret = ops->transfer(dev, DMA_MEM_TO_MEM, dst, src, len);
+
+ /* Clean+Invalidate the areas after, so we can see DMA'd data */
+ dma_unmap_single(destination, len, DMA_FROM_DEVICE);
+ dma_unmap_single(source, len, DMA_TO_DEVICE);
+
+ return ret;
}
UCLASS_DRIVER(dma) = {
--
2.37.3
next reply other threads:[~2022-10-07 17:11 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-07 17:11 Andrew Davis [this message]
2022-10-07 17:11 ` [PATCH 2/3] dma: ti-edma3: Add DMA map operations before and after transfers Andrew Davis
2022-10-18 22:15 ` Tom Rini
2022-10-07 17:11 ` [PATCH 3/3] dma: Transfer dma_ops should use DMA address types Andrew Davis
2022-10-18 22:15 ` Tom Rini
2022-10-18 22:15 ` [PATCH 1/3] dma: Use dma-mapping for cache ops and sync after write Tom Rini
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20221007171113.31577-1-afd@ti.com \
--to=afd@ti.com \
--cc=sjg@chromium.org \
--cc=trini@konsulko.com \
--cc=u-boot@lists.denx.de \
--cc=vigneshr@ti.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox