Created attachment 396164 [details] ixgbe-fix-early-dma-unmapping-for-devices-using-RSC.patch Description of problem: The latest 5.5 kernel will throw the following errors when running a receive stress test with and ixgbe-based device that supports RSC: DRHD: handling fault status reg 402 DMAR:[DMA Write] Request device [04:00.0] fault addr ffe43000 DMAR:[fault reason 05] PTE Write access is not set DRHD: handling fault status reg 502 DMAR:[DMA Write] Request device [04:00.0] fault addr fff62000 DMAR:[fault reason 05] PTE Write access is not set DRHD: handling fault status reg 602 DMAR:[DMA Write] Request device [04:00.0] fault addr ffcf4000 DMAR:[fault reason 05] PTE Write access is not set DRHD: handling fault status reg 702 DMAR:[DMA Write] Request device [04:00.0] fault addr ffd04000 DMAR:[fault reason 05] PTE Write access is not set DRHD: handling fault status reg 2 DMAR:[DMA Write] Request device [04:00.0] fault addr ffbeb000 DMAR:[fault reason 05] PTE Write access is not set DRHD: handling fault status reg 102 DMAR:[DMA Write] Request device [04:00.0] fault addr ff7ac000 DMAR:[fault reason 05] PTE Write access is not set and performance (throughput) was terrible. It was determined that buffers were being unmapped too early when using RSC as the first data buffer needed to be left unmapped so it could be modified later. This attached patch from Intel appears to resolve the issue upstream and in the latest (2.6.18-190.el5) RHEL5.5 development kernels.
From: https://bugzilla.redhat.com/show_bug.cgi?id=525577#c50 "Sibai Li and Greg Rose from Intel LAD have tested Andy's latest kernel [containing the patch attached to bug 568153] for the RSC fix and the problem no longer exists in our tests."
This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux maintenance release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux Update release for currently deployed products. This request is not yet committed for inclusion in an Update release.
Andy, did this make 5.5? People here are asking. It looks like as the status is POST but I need to make sure.
John, I feel *quite* confident that it will. This close to the release bugs will probably remain in POST longer than normal.
in kernel-2.6.18-191.el5 You can download this test kernel from http://people.redhat.com/jwilson/el5 Please update the appropriate value in the Verified field (cf_verified) to indicate this fix has been successfully verified. Include a comment with verification details.
Any test status updates?
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHSA-2010-0178.html
*** Bug 566597 has been marked as a duplicate of this bug. ***
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days