Description of problem: I have a system with 2 3ware 9550SX-12 boards, each running 11 disk hardware RAID 5s. I want to do a stripe across these 2 large (>2TiB) disks. I can create the array just fine, but the system Oopses every time I try to mke2fs the array. Running mke2fs on 1 of the "disks" works just fine. Version-Release number of selected component (if applicable): 2.6.9-34.ELsmp How reproducible: Every time Steps to Reproduce: 1. mdadm -C /dev/md0 -l 0 -c 512 -n 2 /dev/sd[bc]1 2. mke2fs -b 4096 -j -m 0 -E stride=128 -O dir_index -T largefile4 /dev/md0 3. Actual results: Oops (see attached serial console capture) Expected results: It should chug right along. Additional info: I asked on the linux-raid list, and Neil Brown pointed me at this patch <http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdif f;h=1eb29128c644581fa51f822545921394ad4f719f>. I created a custom kernel from the SRPM with this patch, and the problem seems to be resolved.
Created attachment 129202 [details] serial console capture of Oops
I've posted the patch internally for review/possible inclusion in the next RHEL update. Thanks for the consice report and pretested fix.
No problem. Thank *you* for reminding me that this fix isn't in u4. I was about to update that server. :)
*** This bug has been marked as a duplicate of 198495 ***