Michael Chan wrote:
[BNX2]: Add 40-bit DMA workaround for 5708.
The internal PCIE-to-PCIX bridge of the 5708 has the same 40-bit DMA
limitation as some of the tg3 chips. Use the same workaround used in
tg3. On 64-bit systems without IOMMU, linearize the SKB if any
address is 40-bit.
From: Jeff Garzik [EMAIL PROTECTED]
Date: Wed, 02 May 2007 03:06:43 -0400
Second, CONFIG_HIGHMEM is a bad test for IOMMU presence. I don't
think you /can/ test for IOMMU presence. Maybe DaveM knows a way
that I do not?
You really can't.
We have platforms that are both IOMMU and non-IOMMU
David Miller wrote:
From: Jeff Garzik [EMAIL PROTECTED]
Date: Wed, 02 May 2007 03:06:43 -0400
Second, CONFIG_HIGHMEM is a bad test for IOMMU presence. I don't
think you /can/ test for IOMMU presence. Maybe DaveM knows a way
that I do not?
You really can't.
We have platforms that
On Wed, May 02, 2007 at 08:23:40AM -0700, Michael Chan wrote:
A non-IOMMU system using 64-bit dma_addr_t will always set
CONFIG_HIGHMEM, right?
No.
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
Michael Chan wrote:
A non-IOMMU system using 64-bit dma_addr_t will always set
CONFIG_HIGHMEM, right?
No.
Jeff
-
To unsubscribe from this list: send the line unsubscribe netdev in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
On Wed, 2007-05-02 at 11:29 -0400, Jeff Garzik wrote:
Michael Chan wrote:
A non-IOMMU system using 64-bit dma_addr_t will always set
CONFIG_HIGHMEM, right?
No.
May be I misunderstood the code in illegal_highdma() in net/core/dev.c
where it is checking to see if it needs to linearize the
Michael Chan wrote:
On Wed, 2007-05-02 at 11:29 -0400, Jeff Garzik wrote:
Michael Chan wrote:
A non-IOMMU system using 64-bit dma_addr_t will always set
CONFIG_HIGHMEM, right?
No.
May be I misunderstood the code in illegal_highdma() in net/core/dev.c
where it is checking to see if it needs
From: Michael Chan [EMAIL PROTECTED]
Date: Wed, 2 May 2007 08:23:40 -0700
A non-IOMMU system using 64-bit dma_addr_t will always set
CONFIG_HIGHMEM, right?
Nope, IA-64 is at least one example.
IA-64 has both IOMMU and non-IOMMU configurations, and
never sets HIGHMEM.
-
To unsubscribe from
On Wed, 2007-05-02 at 14:24 -0400, Jeff Garzik wrote:
Michael Chan wrote:
On Wed, 2007-05-02 at 11:29 -0400, Jeff Garzik wrote:
Michael Chan wrote:
A non-IOMMU system using 64-bit dma_addr_t will always set
CONFIG_HIGHMEM, right?
No.
May be I misunderstood the code in
Michael Chan wrote:
If CONFIG_HIGHMEM is not set,
wouldn't the device get 32-bit DMA addresses that it cannot handle?
No -- presuming your PCI device's DMA mask is set correctly.
I know that there are 64-bit systems without IOMMUs. That's why we need
[...]
needed or not. These systems
On Wed, 2007-05-02 at 12:40 -0700, David Miller wrote:
From: Michael Chan [EMAIL PROTECTED]
Date: Wed, 2 May 2007 08:23:40 -0700
A non-IOMMU system using 64-bit dma_addr_t will always set
CONFIG_HIGHMEM, right?
Nope, IA-64 is at least one example.
IA-64 has both IOMMU and non-IOMMU
From: Michael Chan [EMAIL PROTECTED]
Date: Wed, 02 May 2007 14:27:49 -0700
I see. So IA64 always uses the SWIOTLB when it doesn't have IOMMU then?
I'm a bit confused. Is it enough to just set the DMA mask to 40-bit and
forget about all this checking? I thought that wasn't enough. A tx
On Wed, 2007-05-02 at 14:45 -0700, David Miller wrote:
From: Michael Chan [EMAIL PROTECTED]
Date: Wed, 02 May 2007 14:27:49 -0700
I see. So IA64 always uses the SWIOTLB when it doesn't have IOMMU then?
I'm a bit confused. Is it enough to just set the DMA mask to 40-bit and
forget
From: Michael Chan [EMAIL PROTECTED]
Date: Wed, 02 May 2007 16:28:22 -0700
Is it correct to assume that all 64-bit systems using 64-bit
dma_addr_t will either have real IOMMUs or will use SWIOTLB? If
this is true, we can just set the DMA mask to what is supported by
the hardware and let
On Wed, May 02, 2007 at 01:02:41PM -0700, Michael Chan wrote:
Let's say I have a 32-bit card that cannot do dual address cycle on a
64-bit dma_addr_t system without IOMMU. If CONFIG_HIGHMEM is not set,
wouldn't the device get 32-bit DMA addresses that it cannot handle?
It needs to set it's
From: Christoph Hellwig [EMAIL PROTECTED]
Date: Wed, 2 May 2007 23:48:58 +0100
On Wed, May 02, 2007 at 01:02:41PM -0700, Michael Chan wrote:
Let's say I have a 32-bit card that cannot do dual address cycle on a
64-bit dma_addr_t system without IOMMU. If CONFIG_HIGHMEM is not set,
wouldn't
[BNX2]: Add 40-bit DMA workaround for 5708.
The internal PCIE-to-PCIX bridge of the 5708 has the same 40-bit DMA
limitation as some of the tg3 chips. Use the same workaround used in
tg3. On 64-bit systems without IOMMU, linearize the SKB if any
address is 40-bit.
Signed-off-by: Michael Chan
17 matches
Mail list logo