Re: [RFC] URBs and buffer management
Pete Zaitcev <[email protected]>
| Newsgroups | gmane.linux.usb.devel |
|---|---|
| Organization | Red Hat, Inc. |
| Message-ID | <[email protected]> |
On Sun, 1 Jul 2007 21:02:11 -0700, David Brownell <[email protected]> wrote: > As I understand it the actual mechanism is more like: platforms > always set up a chunk of memory as uncached, and always map that > through the IOMMU. Call that the DMA-{coherent,consistent} region. > > Utilities like dma_alloc_coherent() -- and hence usb_buffer_alloc(), > dma_pool_*(), and so on -- return memory from that region. > > So the resources are *always* tied up; fixed overhead. What's visible > to drivers is that per-transfer buffer operations are not needed. No, they are normal pages, only uncached. They don't have to be in a single contiguous region virtually or physically. There's no fixed size allocation for it (at least, not on every architecture). Look at arch/sparc/kernel/ioport.c:pci_alloc_consistent for an example. On Intel, what you write is true in a way, in case of swiotlb. The swiotlb reserves a real chunk of memory (still cached though). That memory remains unavailable to normal users regardless of the number and size of I/Os currently active in the system. -- Pete ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ [email protected] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel