I have a custom linux kernel version 2.6.32 running as a VM on Hyper-V 2008 R2 with 2vCPUs.
When I am triggering kdump using the magic sysrq (sh#echo c > /proc/sysrq-trigger), the VM gets rebooted in the middle of starting the capture kernel. This happens intermittently with no reason for the reboot. The Windows eventlog shows it is a CPU triple fault issue.
But, when I trigger kdump after entering in to kernel kdb mode (kdb#kdump), kdump is working successfully for multiple times.
My kdump commandline is
Kernel command line: root=/dev/ram0 nmi_watchdog=0 irqpoll maxcpus=1 reset_devices no_timer_check debug ata_piix.prefer_ms_hyperv=0 console=ttyS0,9600n8 ip=none
So, what is the difference between the two methods? Is there any interrupt which is not properly handle in sysrq kdump ?
Please explain ?
Thanks in advance,