Re: [Xen-devel] libxl: crash fails to load vmcore when the guest memory greater than 4G

[Daniel Kiper <dkiper@xxxxxxxxxxxx>]

On Thu, Jun 05, 2014 at 08:11:28AM +0000, Zhangwei (FF) wrote:
> Description of problem:
>
> A crash dump of a redhat5.5(32 bit) pae xen guest when the guest memory 
> greater than 4G is not traceable, since crash can't figure out the phys_addr. 
> But it??s ok when the guest memory less than 4G.
> And so is it in the case of a suse11sp1(32 bit) pae xen guest.
>
> Xen version: 4.1.2
> dom0 version: suse11sp2(64bit)
> How reproducible: Always
>
> Steps to Reproduce:
> 1. Start redhat5.5??32bit?? pae guest with memory greater than 4G
> 2. Take a crash dump of guest: xl dump-core <domid> <path-to-corefile-to-save>
> 3. crash vmlinux  <path-to-corefile-to-save>
>     -- where vmlinux  is the unstripped vmlinux image that was used
>        to generate the bootable vmlinuz of the redhat5.5(32bit) pae guest.
>
>
> Actual results:
> [root@localhost sdb]# crash vmlinux redhat_5.5_32_hvm_8G.core
>
> crash 4.1.2-4.el5

Oh... Please do not use such ancient version of crash. Download latest
version from http://people.redhat.com/anderson/ build it and voila...
If it does not work drop me a line.

Daniel

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel