Skip to contentSkip to conversation

net: davinci_cpdma: use dma_addr_t for DMA address

PATCH New Pavel Machek 2 participants 3 messages Dec 13, 2019 – Dec 16, 2019

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: davinci_cpdma: use dma_addr_t for DMA address

Hi!

I have this in 4.4-rt-cip, and it would make my life easier if I applied it to 4.4-cip. Would that be ok to do?

Thanks and best regards,
Pavel

commit 3af83ff3e56ec9b96ef15a2e249e3019cd2e5245

Author: Arnd Bergmann <[email protected]>
Date:   Wed Sep 18 17:00:45 2019 +0900

net: davinci_cpdma: use dma_addr_t for DMA address

commit 84092996673211f16ef3b942a191d7952e9dfea9 upstream.

The davinci_cpdma mixes up physical addresses as seen from the CPU and DMA addresses as seen from a DMA master, since it can operate on both normal memory or an on-chip buffer. If dma_addr_t is different from phys_addr_t, this means we get a compile-time warning about the type mismatch:

ethernet/ti/davinci_cpdma.c: In function 'cpdma_desc_pool_create':
ethernet/ti/davinci_cpdma.c:182:48: error: passing argument 3 of 'dma_alloc_coherent' from incompatible pointer type [-Werror=incompatible-pointer-types]
pool->cpumap = dma_alloc_coherent(dev, size, &pool->phys,
In file included from ethernet/ti/davinci_cpdma.c:21:0:
dma-mapping.h:398:21: note: expected 'dma_addr_t * {aka long long unsigned int *}' but argument is of type 'phys_addr_t * {aka unsigned int *}'
static inline void *dma_alloc_coherent(struct device *dev, size_t size,

This slightly restructures the code so the address we use for mapping RAM into a DMA address is always a dma_addr_t, avoiding the warning. The code is correct even if both types are 32-bit because the DMA master in this device only supports 32-bit addressing anyway, independent of the types that are used.

We still assign this value to pool->phys, and that is wrong if the driver is ever used with an IOMMU, but that value appears to be never used, so there is no problem really. I've added a couple of comments about where we do things that are slightly violating the API.

drivers/net/ethernet/ti/davinci_cpdma.c +6 -6
index 657b65bf5cac..18bf3a8fdc50 100644
@@ -82,7 +82,7 @@ struct cpdma_desc {
 
 struct cpdma_desc_pool {
 	phys_addr_t		phys;
-	u32			hw_addr;
+	dma_addr_t		hw_addr;
 	void __iomem		*iomap;		/* ioremap map */
 	void			*cpumap;	/* dma_alloc map */
 	int			desc_size, mem_size;
@@ -152,7 +152,7 @@ struct cpdma_chan {
  * abstract out these details
  */
 static struct cpdma_desc_pool *
-cpdma_desc_pool_create(struct device *dev, u32 phys, u32 hw_addr,
+cpdma_desc_pool_create(struct device *dev, u32 phys, dma_addr_t hw_addr,
 				int size, int align)
 {
 	int bitmap_size;
@@ -176,13 +176,13 @@ cpdma_desc_pool_create(struct device *dev, u32 phys, u32 hw_addr,
 
 	if (phys) {
 		pool->phys  = phys;
-		pool->iomap = ioremap(phys, size);
+		pool->iomap = ioremap(phys, size); /* should be memremap? */
 		pool->hw_addr = hw_addr;
 	} else {
-		pool->cpumap = dma_alloc_coherent(dev, size, &pool->phys,
+		pool->cpumap = dma_alloc_coherent(dev, size, &pool->hw_addr,
 						  GFP_KERNEL);
-		pool->iomap = pool->cpumap;
-		pool->hw_addr = pool->phys;
+		pool->iomap = (void __iomem __force *)pool->cpumap;
+		pool->phys = pool->hw_addr; /* assumes no IOMMU, don't use this value */
 	}
 
 	if (pool->iomap)
Signature · 9 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: not available
URL: <http://lists.cip-project.org/pipermail/cip-dev/attachments/20191213/07e7d11a/attachment.sig>
#1 lore raw

PATCH 1/1 net: davinci_cpdma: use dma_addr_t for DMA address

Hi Pavel,

Quoted 8 lines

-----Original Message-----
From: Pavel Machek [mailto:pavel at ucw.cz]
Sent: Saturday, December 14, 2019 5:57 AM
To: iwamatsu nobuhiro(?? ?? ????????) <[email protected]>; cip-dev at lists.cip-project.org
Subject: [PATCH 4.4.y-cip] net: davinci_cpdma: use dma_addr_t for DMA address

Quoted 4 lines

Hi!

I have this in 4.4-rt-cip, and it would make my life easier if I applied it to 4.4-cip. Would that be ok to do?

I think your suggestion is good because the cip-rt branch is based on the cip branch.

Or this is not a bug but a warning, but it may be possible to include it with LTS.

https://gitlab.com/cip-project/cip-kernel/linux-cip/-/jobs/379209573#L8482

Best regards,
Nobuhiro

#2 lore raw

PATCH 1/1 net: davinci_cpdma: use dma_addr_t for DMA address

Hi!

Quoted 5 lines
Quoted 2 lines

I have this in 4.4-rt-cip, and it would make my life easier if I applied it to 4.4-cip. Would that be ok to do?

I think your suggestion is good because the cip-rt branch is based on the cip branch.

I applied the patch to 4.4-cip, and builds should be ok. I'd prefer not to push it to 4.4-stable, but it might make sense for 4.4-rt. Let me investigate.

Best regards,
Pavel

Signature · 9 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: Digital signature
URL: <http://lists.cip-project.org/pipermail/cip-dev/attachments/20191216/410e8f6b/attachment.sig>