Skip to contentSkip to conversation

cciss: change Cmd_sg_list.sg_chain_dma type to dma_addr_t

PATCH In review Alex Chiang 2 participants 2 messages Nov 18, 2009 – Nov 23, 2009 Block layer

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 cciss: change Cmd_sg_list.sg_chain_dma type to dma_addr_t

A recent commit broke the ia64 build:

Author: Don Brace <[email protected]>
Date: Thu Nov 12 12:50:01 2009 -0600

cciss: Add enhanced scatter-gather support.

because of this hunk:

	--- a/drivers/block/cciss.h
	+++ b/drivers/block/cciss.h
	+struct Cmd_sg_list {
	+       SGDescriptor_struct     *sgchain;
	+       dma64_addr_t            sg_chain_dma;
	+       int                     chain_block_size;
	+};

The issue is that dma64_addr_t isn't #define'd on ia64.

The way that we're using Cmd_sg_list.sg_chain_dma is to hold an address returned from pci_map_single().

+ temp64.val = pci_map_single(h->pdev,
+ h->cmd_sg_list[c->cmdindex]->sgchain,
+ len, dir);
+
+ h->cmd_sg_list[c->cmdindex]->sg_chain_dma = temp64.val;

pci_map_single() returns a dma_addr_t too.

This code will still work even on a 32-bit x86 build, where dma_addr_t is defined to be a u32 because it will simply be promoted to the __u64 that temp64.val is defined as.

Thus, declaring Cmd_sg_list.sg_chain_dma as dma_addr_t is safe.

Cc Don Brace <[email protected]>
Cc Stephen M. Cameron <[email protected]>

---
This changelog is probably an analysis overkill, but I was concerned about 32 vs 64 bit DMA addresses getting passed to the hardware. After doing a bunch more reading, I'm satisfied that a) we're dealing with CPU-side addresses and b) proper type promotion occurs anyway.

If anyone wants to double-check my work, I won't be offended. ;)

I build tested this on ia64, x86_64, and i386. No gcc nor sparse warnings. However, I didn't do any run-time testing on i386.
---

drivers/block/cciss.h +1 -1
index e5c63e5..1d95db2 100644
@@ -57,7 +57,7 @@ typedef struct _drive_info_struct
 
 struct Cmd_sg_list {
 	SGDescriptor_struct	*sgchain;
-	dma64_addr_t		sg_chain_dma;
+	dma_addr_t		sg_chain_dma;
 	int			chain_block_size;
 };
 
#1 lore raw
Quoted 28 lines · Alex Chiang

A recent commit broke the ia64 build:

Author: Don Brace <[email protected]>
Date: Thu Nov 12 12:50:01 2009 -0600

cciss: Add enhanced scatter-gather support.

because of this hunk:

	--- a/drivers/block/cciss.h
	+++ b/drivers/block/cciss.h
	+struct Cmd_sg_list {
	+       SGDescriptor_struct     *sgchain;
	+       dma64_addr_t            sg_chain_dma;
	+       int                     chain_block_size;
	+};

The issue is that dma64_addr_t isn't #define'd on ia64.

The way that we're using Cmd_sg_list.sg_chain_dma is to hold an address returned from pci_map_single().

+ temp64.val = pci_map_single(h->pdev,
+ h->cmd_sg_list[c->cmdindex]->sgchain,
+ len, dir);
+
+ h->cmd_sg_list[c->cmdindex]->sg_chain_dma = temp64.val;

Quoted 7 lines

pci_map_single() returns a dma_addr_t too.

This code will still work even on a 32-bit x86 build, where dma_addr_t is defined to be a u32 because it will simply be promoted to the __u64 that temp64.val is defined as.

Thus, declaring Cmd_sg_list.sg_chain_dma as dma_addr_t is safe.

Applied for 2.6.32.

Signature · 1 line
Jens Axboe