[PATCH] x86: Use ioremap_cache() for crash dump copies

[Andi Kleen]

During code review I noticed that when reading the previous kernel
memory from a kdump kernel ioremap() is used. This was ok before
the PAT changes, but with PAT ioremap() this results in an uncached
mapping, which is very slow.

Change it to ioremap_cache(). This is fine because this code
only reads real memory, no IO mappings.

I haven't run numbers, but I suspect this will speed up crash dumping
somewhat.

Only affects 64bit, 32bit uses a different method.

Might be a stable candidate.

Signed-off-by: Andi Kleen <ak@xxxxxxxxxxxxxxx>
---
 arch/x86/kernel/crash_dump_64.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/kernel/crash_dump_64.c b/arch/x86/kernel/crash_dump_64.c
index 045b36c..bf43188 100644
--- a/arch/x86/kernel/crash_dump_64.c
+++ b/arch/x86/kernel/crash_dump_64.c
@@ -34,7 +34,7 @@ ssize_t copy_oldmem_page(unsigned long pfn, char *buf,
 	if (!csize)
 		return 0;
 
-	vaddr = ioremap(pfn << PAGE_SHIFT, PAGE_SIZE);
+	vaddr = ioremap_cache(pfn << PAGE_SHIFT, PAGE_SIZE);
 	if (!vaddr)
 		return -ENOMEM;
 
-- 
1.6.0.2


-- 
ak@xxxxxxxxxxxxxxx -- Speaking for myself only.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/