|Summary:||megaraid_sas ioctls fail under load|
|Product:||Red Hat Enterprise Linux 4||Reporter:||Bryn M. Reeves <bmr>|
|Component:||kernel||Assignee:||Bryn M. Reeves <bmr>|
|Status:||CLOSED DUPLICATE||QA Contact:||Martin Jenner <mjenner>|
|Version:||4.5||CC:||andriusb, publicmailbox, suran.kularatna, tao|
|Fixed In Version:||Doc Type:||Bug Fix|
|Doc Text:||Story Points:||---|
|Last Closed:||2008-02-06 18:08:29 UTC||Type:||---|
|oVirt Team:||---||RHEL 7.3 requirements from Atomic Host:|
|Cloudforms Team:||---||Target Upstream Version:|
|Bug Depends On:|
|Bug Blocks:||422551, 430698|
Description Bryn M. Reeves 2007-07-11 11:29:12 UTC
Description of problem: The megaraid_sas driver uses pci_alloc_consistent to allocate an array of buffers to copy user arguments into within the ioctl path. This can fail under heavy load leading to messages like: megasas: Failed to alloc kernel SGL buffer for IOCTL Upstream fixed this here: commit 9f35fa8a14e6216a859e2dfbe50ade497f9603ef Author: Sumant Patro <firstname.lastname@example.org> Date: Wed Feb 14 12:55:45 2007 -0800 [SCSI] megaraid_sas: replace pci_alloc_consitent with dma_alloc_coherent in ioctl path Replaced pci_alloc_consistent with dma_alloc_coherent from the ioctl path. This is to avoid situations where ioctl fails for lack of memory (when system under heavy stress). Signed-off-by: Sumant Patro <email@example.com> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com> Appearing in dmesg. Version-Release number of selected component (if applicable): kernel-2.6.9-55.EL How reproducible: Moderate; problem was observed during pre-production stress testing of boxes using megaraid_sas as the primary storage. Steps to Reproduce: 1. Configure a system with megaraid_sas providing storage 2. Create memory pressure on the system while something is attempting ioctl commands on the megaraid Actual results: Appearance of: "megasas: Failed to alloc kernel SGL buffer for IOCTL" in dmesg. Expected results: Megaraid_sas ioctls succeed.
Comment 1 Bryn M. Reeves 2007-07-11 11:29:13 UTC
Created attachment 158935 [details] re-diff of upstream patch for RHEL4
Comment 3 Bryn M. Reeves 2008-02-06 18:08:29 UTC
This problem is addressed with the updated driver being tracked in bug 312061. *** This bug has been marked as a duplicate of 312061 ***
Comment 4 Suran Kularatna 2009-10-02 13:06:10 UTC
where is the patch for this? what is thenumber?
Comment 5 Bryn M. Reeves 2009-10-02 13:15:07 UTC
Attached in comment #4 (a backport of the git commit given in comment #0).
Comment 6 Ben 2009-12-15 15:39:51 UTC
This bug is a duplicate of bug 312061, but that bug is not viewable by me. Why is it being kept a secret?