Re: [BUGFIX] [PATCH] kexec/x86: Do map crash kernel area

[Don Slutz <dslutz@verizon.com>]

On 01/02/14 06:49, David Vrabel wrote:
> On 02/01/14 10:48, Andrew Cooper wrote:
>> On 02/01/14 10:46, David Vrabel wrote:
>>> On 01/01/14 16:51, Don Slutz wrote:
>>>> Revert of commit 7113a45451a9f656deeff070e47672043ed83664
>>> Since this commit introduced a regression, a revert is the best thing to
>>> do here.
>>>
>>> Acked-by: David Vrabel <david.vrabel@citrix.com>
>>>
>>>> Using kexec commit 027413d822fd57dd39d2d2afab1484bc6b6b84f9
>>>>
>>>> With "crashkernel=256M@256M" ((XEN) Kdump: 256MB (262144kB) at 0x10000000)
>>> I guess Daniel tested a debug build without this crashkernel option.
>>> This would place the crash region above the direct mapping region and
>>> map_domain_page() would do the right thing.
>>>
>>>
>>>> +    map_pages_to_xen((unsigned long)__va(kexec_crash_area.start),
>>>> +                     kexec_crash_area.start >> PAGE_SHIFT,
>>>> +                     PFN_UP(kexec_crash_area.size), PAGE_HYPERVISOR);
>>>> +
>>> This should be made conditional on the location of the crash region --
>>> it is wrong to do this for portions of the crash region that are outside
>>> the crash region.
>> Presume you mean "outside the direct-map region"?
> Yes.
>
> David

I have no idea on how to even check for "outside the direct-map region" 
and how to test any additional changes.  Testing with "crashkernel=256M" 
((XEN) Kdump: 256MB (262144kB) at 0x82a1be000) does not work before or 
after this patch.  In both cases, I get an error out of kexec:

~/kexec/build/sbin/kexec -p '--command-line=placeholder 
root=/dev/mapper/vg_f17--xen-lv_root ro rd.md=0 rd.dm=0 
rd.lvm.lv=vg_f17-xen/lv_swap KEYTABLE=us SYSFONT=True rd.luks=0 
console=ttyS0,9600n8 rd.lvm.lv=vg_f17-xen/lv_root LANG=en_US.UTF-8 
earlyprintk=ttyS0 rd_NO_PLYMOUTH irqpoll nr_cpus=1 reset_devices 
cgroup_disable=memory mce=off' 
--initrd=/boot/initramfs-3.8.11-100.fc17.x86_64kdump.img 
/boot/vmlinuz-3.8.11-100.fc17.x86_64
Could not find a free area of memory of 0xa000 bytes...
locate_hole failed

I am currently busy with other tasks and so do not expect to get to 
learning about xen's direct-map region and/or finding out why kexec does 
not work in this case before xen 4.4.0 release date.

    -Don Slutz