Description of problem: After running sysrq -c, capture kernel does not boot up on /dev/mdx. And I found the following line was commented out in /sbin/mkdumprd. ---/sbin/mkdumprd : emit "\#need code here to set up md dev : Version-Release number of selected component (if applicable): kexec-tools-1.101-164.el5 How reproducible: 1. install Red Hat EL 5 with Software Raid. 2. setup the kdump 3. run sysrq -c --> a capture kernel does not boot up. Actual results: capture kernel does not boot up. Expected results: capture kernel should boot up and make vmcore. Additional info: I have backported the raidautorun function from nash source code to busybox, and changed /sbin/mkdumprd as bellow, and then it boots up well. --- mkdumprd.org 2008-01-25 14:49:39.000000000 +0900 +++ mkdumprd 2008-01-25 14:46:54.000000000 +0900 @@ -1397,7 +1397,7 @@ emitdms if [ -n "$raiddevices" ]; then for dev in $raiddevices; do cp -a /dev/${dev} $MNTIMAGE/dev - emit "\#need code here to set up md devs" + emit "raidautorun /dev/${dev}" done fi
Created attachment 292932 [details] this is my proposal patch
no, we're not going to start putting in miscelaneous nash utility calls into busybox. We've already fixed this in a later release (kexec-tools 1.102pre-5 specifically using the nominal user space utilities). The latest kexec-tools package can be found here if you need this fixed before the 5.2 release. Thanks!
Thank you for your informations. Would you please give me the latest kexec-tools src package(1.102pre-5) ? I would like to test it. Thanks in advance, Atsushi Nakabayashi.
sorry, meant to include this url: http://people.redhat.com/nhorman/rpms/latest_kexec-tools.tbz2