[lkml]   [2009]   [Jul]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Patch in this message
Subject[PATCH] Fix pci_unmap_addr() et al on i386.
We can run a 32-bit kernel on boxes with an IOMMU, so we need
pci_unmap_addr() etc. to work -- without it, drivers will leak mappings.

To be honest, this whole thing looks like it's more pain than it's
worth; I'm half inclined to remove the no-op #else case altogether.

But this is the minimal fix, which just does the right thing if

Signed-off-by: David Woodhouse <>
Needed in 2.6.30, but not earlier because you can't enable CONFIG_DMAR
before that. Jörg, if you ever enable the AMD IOMMU in 32-bit mode you
may have to fix this again (unless we just make it unconditionally do
the right thing, which probably gets my vote).

diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h
index 927958d..1ff685c 100644
--- a/arch/x86/include/asm/pci.h
+++ b/arch/x86/include/asm/pci.h
@@ -91,7 +91,7 @@ extern void pci_iommu_alloc(void);

#define PCI_DMA_BUS_IS_PHYS (dma_ops->is_phys)

-#if defined(CONFIG_X86_64) || defined(CONFIG_DMA_API_DEBUG)
+#if defined(CONFIG_X86_64) || defined(CONFIG_DMAR) || defined(CONFIG_DMA_API_DEBUG)

dma_addr_t ADDR_NAME;
David Woodhouse Open Source Technology Centre Intel Corporation

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

 \ /
  Last update: 2009-07-01 19:37    [W:0.047 / U:0.288 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site