On Tue, Jul 12, 2005 at 10:35:09AM -0400, James.Smart@xxxxxxxxxx wrote: > I really hated these patches. The issue that was causing >90% of the > problems we encountered was that pci_alloc_consistent was overextending > when no_iommu was set by ignoring the dma masks of the adapter and > forcing the use of the dma map pool when it wasn't necessary. IMHO, fixing > this issue seems like a much better approach than having to make the > dma pool or swiotlb larger. It's a matter of timing here though. This *was* fixed in the 2.6 kernel series, and thus in vendor kernels based on 2.6, it just wasn't ever backported fixed in 2.4, or in particular, the individual 2.4 vendor kernels. At least with RHEL, there are kABI constraints which make "proper" fixes more difficult, but I do have requests in with Red Hat to fix this. Their engineers are bright and know the constraints. They can choose to ignore my patches come up with something better. Thanks, Matt -- Matt Domsch Software Architect Dell Linux Solutions linux.dell.com & www.dell.com/linux Linux on Dell mailing lists @ http://lists.us.dell.com - : send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
