Re: [Xen-devel] Debootstrapping a rootfs for Xen on FastModels A15

[Sander Bogaert <sander.bogaert@xxxxxxxxx>]

On 2 April 2013 15:47, Stefano Stabellini
<stefano.stabellini@xxxxxxxxxxxxx> wrote:
> On Tue, 2 Apr 2013, Sander Bogaert wrote:
>> I did however try the branches you linked but without much progress.
>> Using the xen and kernel branches you linked together with this config
>> ( http://xenbits.xen.org/people/sstabellini/config where I selected
>> the defaults for newer kernel options ) Xen crashes. I attached a log;
>> log.txt. Isn't this the exact same setup as you are using?
>
> I am attaching the kernel config I am using with that branch.
> However I tried the one linked above, using the default for all the
> newer kernel options, and I can still boot Dom0 to a bash prompt :-S
> For the sake of completeness I am also going to attach my DTS (that is
> SMP ready).
>

Thanks a lot for all your help!

I used the files you supplied, build myself a A15x2 model and still
get a crash booting dom0, it is one I never seen before. I attached
the Xen and dom0 logs. The isue seems to be
"Unhandled fault: external abort on non-linefetch (0x1008) at 0xc8820084
Internal error: : 1008 [#1] SMP ARM"

Any idea what could still be different between our setups? Maybe the
compilers? I'm using arm-linux-gnueabi-gcc (Debian 4.7.2-4) 4.7.2 for
crosscompilation. The model was built with gcc (Debian 4.4.7-2) 4.4.7.

Kernel is the branch you indicated with the .config you supplied.
Xen is the branch you indicated.
I'm using the dts you supplied and compile it to a dtb with DTC 1.3.0.

Since everything starts I'm assuming I invoke the model_shell correct.
./model_shell ../resources/Cortex-A15x$3/cadi_system_Linux-Release-GCC-4.4.so
$1 -C motherboard.flashloader0.fname=$2 -C
motherboard.smsc_91c111.enabled=1 -C
motherboard.hostbridge.userNetworking=1 -C
motherboard.pl011_uart0.out_file=./log.txt -C
motherboard.mmc.p_mmc_file="$4"

>> About the 'regular' branches: The problem I'm having with a
>> non-ramdisk kernel and my debootstrapped rootfs I'm also facing
>> without Xen on the FastModel simulator. I know this isn't really the
>> place to ask but if you have any idea on how to fix it I would really,
>> really appreciate it! I've been stuck for quite some time on it and I
>> know others have it working. All the information from my initial mail
>> is still relevant.
>>
>> <NUL>[   12.340651] VFS: Cannot open root device "mmcblk0" or
>> unknown-block(0,0): error -6
>> [   12.340714] Please append a correct "root=" boot option; here are
>> the available partitions:
>> [   12.340805] Kernel panic - not syncing: VFS: Unable to mount root
>> fs on unknown-block(0,0)
>> [   12.340947] [<c000d3a8>] (unwind_backtrace+0x0/0xe0) from
>> [<c02ac5a0>] (panic+0x70/0x1c0)
>> ....
>>
>> I included logs from this kernel and my rootfs booting on the model (
>> log_linux.txt ) and with Xen ( log_xen.txt ). I also tried with having
>> root mmcblk0p0 and some other things. I just debootstrapped a rootfs (
>> wheezy ) the way I described so I'd think it should be mmcblk0 or
>> mmcblk0p0.
>>
>> I really hope someone could help with getting this setup working,
>> either the 'regular' one or the uncommited branches. I've been at this
>> for days and really need a working setup.
>
> It looks like an interrupt issue: Linux doesn't receive any interrupts
> from the MMC controller. But I don't know what the exact cause of the
> issue could be. I would start investigating by adding some printk's to
> xen/arch/arm/gic.c.

I'm focusing on the first scenario since that one most resembles a
working setup ( yours ). I'll add some debug prints and test this one
later.

Attachment: dom0_log.txt
Description: Text document

Attachment: xen_log.txt
Description: Text document

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