Messages in this thread | ![/](/images/icornerl.gif) | | Date | Sat, 7 Jun 2003 10:44:34 +0100 | From | Russell King <> | Subject | Re: problem with blk_queue_bounce_limit() |
| |
On Sat, Jun 07, 2003 at 12:20:46AM -0700, David Mosberger wrote: > ./include/asm-arm/pci.h:#define PCI_DMA_BUS_IS_PHYS (0)
I suspect we probably set this incorrectly; we have some platforms where there is merely an offset between the phys address and the bus address. For these, I think we want to set this to 1.
Other platforms require the dma functions to allocate a new buffer and copy the data to work around buggy "wont fix" errata (eg, new buffer below 1MB) and for these I think we want to leave this at 0.
It is rather unfortunate that this got called "PCI_xxx" since it has been used in a non pci-bus manner in (eg) the scsi layer.
Also note that I have platforms where the dma_mask is a real mask not "a set of zeros followed by a set of ones from MSB to LSB." I can see this breaking the block layer if PCI_DMA_BUS_IS_PHYS is defined to one. 8/
-- Russell King (rmk@arm.linux.org.uk) The developer of ARM Linux http://www.arm.linux.org.uk/personal/aboutme.html
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| ![\](/images/icornerr.gif) |