Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Debian Versatile Express under QEMU
214 views
Skip to first unread message
Jerry Stuckle
Jan 22, 2014, 9:20:02 AM1/22/14
to
Hi, all,
I'm missing something here, and I don't know what it is. I'm trying to
get Debian's Versatile Express ARM version running under qemu. What I
did was:
Download initrd.gz and vmlinux-3.2.0-4-vexpress from
ftp://ftp.debian.org/debian/dists/wheezy/main/installer-armhf/current/images/vexpress/netboot/
(and renamed the former to initrd.install.gz).
I created a 4GB raw QEMU disk image:
qemu-img -f raw armhf.img 4G
I then ran the following command:
qemu-system-arm -M vexpress-a9 -m 512M -kernel vmlinuz -initrd
initrd.install.gz -sd armhf.img -append "root=/dev/ram" -no-reboot
Taking all of the defaults (except user and password, obviously), the
system installed with no problems.
I mounted the resulting disk and it looks good. I copied
initrd.img-3.2.0.4-vexpress from the first partition on the disk to my
working directory and gzipped it, with a symlink from initrd.gz.
vmlinux-3.2.0-4-vexpress is the same as what I downloaded.
Everything up to this point looked good. However, when I try to boot
the image with:
qemu-system-arm -M vexpress-a9 -m 512M -kernel vmlinuz -initrd initrd.gz
-sd armhf.img -append "root=/dev/mmcblk0p2"
it fails.
It starts out in my terminal with two messages to stderr:
l2x0_priv_write: Bad offset 900
l2x0_priv_write: Bad offset 904
Don't know whether these are pertinent or not.
The system tries to boot; in the console window, the last few lines are:
[ 0.859520] rtc-pl031 mb:rtc: setting system clock to 2014-01-22
02:00:39 UTC (1390356039)
[ 0.860180] Initializing network drop monitor service
[ 0.895033] mmc0: host does not support reading read-only switch.
assuming write-enable.
[ 0.895819] mmc0: new SDHC card at address 4567
[ 0.907223] input: AT Raw Set 2 keyboard as
/devices/mb:kmi0/serio0/input/input0
[ 0.910420] mmcblk0: mmc0:4567 QEMU! 4.00 GiB
[ 0.940223] mmcblk0: p1 p2 p3 < p5 >
[ 0.952258] List of all partitions:
[ 0.952648] b300 4194304 mmcblk0 driver: mmcblk
[ 0.953130] b301 229376 mmcblk0p1
00000000-0000-0000-0000-000000000000
[ 0.953715] b302 3727360 mmcblk0p2
00000000-0000-0000-0000-000000000000
[ 0.954218] b303 1 mmcblk0p3
00000000-0000-0000-0000-000000000000
[ 0.954602] b305 234496 mmcblk0p5
00000000-0000-0000-0000-000000000000
[ 0.955077] No filesystem could mount root, tried:
[ 0.955511] Kernel panic - not syncing: VFS: Unable to mount root fs
on unknown-block(179,2)
[ 0.957767] [<80013f40>] (unwind_backtrace+0x0/0xe0) from
[<8029d378>] (panic+0x50/0x190)
[ 0.958300] [<8029d378>] (panic+0x50/0x190) from [<803a0cb0>]
(mount_block_root+0x234/0x284)
[ 0.958776] [<803a0cb0>] (mount_block_root+0x234/0x284) from
[<803a0e84>] (prepare_namespace+0x124/0x184)
[ 0.959076] [<803a0e84>] (prepare_namespace+0x124/0x184) from
[<803a0940>] (kernel_init+0x170/0x1a8)
[ 0.959532] [<803a0940>] (kernel_init+0x170/0x1a8) from [<8000e798>]
(kernel_thread_exit+0x0/0x8)
So although mmcblk0p2 is available, it looks like the system can't mount
it for some reason.
Can anyone tell me what I'm missing?
TIA,
Jerry
--
To UNSUBSCRIBE, email to debian-ar...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: http://lists.debian.org/52DFD254...@attglobal.net
I'm missing something here, and I don't know what it is. I'm trying to
get Debian's Versatile Express ARM version running under qemu. What I
did was:
Download initrd.gz and vmlinux-3.2.0-4-vexpress from
ftp://ftp.debian.org/debian/dists/wheezy/main/installer-armhf/current/images/vexpress/netboot/
(and renamed the former to initrd.install.gz).
I created a 4GB raw QEMU disk image:
qemu-img -f raw armhf.img 4G
I then ran the following command:
qemu-system-arm -M vexpress-a9 -m 512M -kernel vmlinuz -initrd
initrd.install.gz -sd armhf.img -append "root=/dev/ram" -no-reboot
Taking all of the defaults (except user and password, obviously), the
system installed with no problems.
I mounted the resulting disk and it looks good. I copied
initrd.img-3.2.0.4-vexpress from the first partition on the disk to my
working directory and gzipped it, with a symlink from initrd.gz.
vmlinux-3.2.0-4-vexpress is the same as what I downloaded.
Everything up to this point looked good. However, when I try to boot
the image with:
qemu-system-arm -M vexpress-a9 -m 512M -kernel vmlinuz -initrd initrd.gz
-sd armhf.img -append "root=/dev/mmcblk0p2"
it fails.
It starts out in my terminal with two messages to stderr:
l2x0_priv_write: Bad offset 900
l2x0_priv_write: Bad offset 904
Don't know whether these are pertinent or not.
The system tries to boot; in the console window, the last few lines are:
[ 0.859520] rtc-pl031 mb:rtc: setting system clock to 2014-01-22
02:00:39 UTC (1390356039)
[ 0.860180] Initializing network drop monitor service
[ 0.895033] mmc0: host does not support reading read-only switch.
assuming write-enable.
[ 0.895819] mmc0: new SDHC card at address 4567
[ 0.907223] input: AT Raw Set 2 keyboard as
/devices/mb:kmi0/serio0/input/input0
[ 0.910420] mmcblk0: mmc0:4567 QEMU! 4.00 GiB
[ 0.940223] mmcblk0: p1 p2 p3 < p5 >
[ 0.952258] List of all partitions:
[ 0.952648] b300 4194304 mmcblk0 driver: mmcblk
[ 0.953130] b301 229376 mmcblk0p1
00000000-0000-0000-0000-000000000000
[ 0.953715] b302 3727360 mmcblk0p2
00000000-0000-0000-0000-000000000000
[ 0.954218] b303 1 mmcblk0p3
00000000-0000-0000-0000-000000000000
[ 0.954602] b305 234496 mmcblk0p5
00000000-0000-0000-0000-000000000000
[ 0.955077] No filesystem could mount root, tried:
[ 0.955511] Kernel panic - not syncing: VFS: Unable to mount root fs
on unknown-block(179,2)
[ 0.957767] [<80013f40>] (unwind_backtrace+0x0/0xe0) from
[<8029d378>] (panic+0x50/0x190)
[ 0.958300] [<8029d378>] (panic+0x50/0x190) from [<803a0cb0>]
(mount_block_root+0x234/0x284)
[ 0.958776] [<803a0cb0>] (mount_block_root+0x234/0x284) from
[<803a0e84>] (prepare_namespace+0x124/0x184)
[ 0.959076] [<803a0e84>] (prepare_namespace+0x124/0x184) from
[<803a0940>] (kernel_init+0x170/0x1a8)
[ 0.959532] [<803a0940>] (kernel_init+0x170/0x1a8) from [<8000e798>]
(kernel_thread_exit+0x0/0x8)
So although mmcblk0p2 is available, it looks like the system can't mount
it for some reason.
Can anyone tell me what I'm missing?
TIA,
Jerry
--
To UNSUBSCRIBE, email to debian-ar...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: http://lists.debian.org/52DFD254...@attglobal.net
Ian Campbell
Jan 22, 2014, 9:30:01 AM1/22/14
to
On Wed, 2014-01-22 at 09:14 -0500, Jerry Stuckle wrote:
> I mounted the resulting disk and it looks good. I copied
> initrd.img-3.2.0.4-vexpress from the first partition on the disk to my
> working directory and gzipped it, with a symlink from initrd.gz.
> vmlinux-3.2.0-4-vexpress is the same as what I downloaded.
I think initrd.img-3.2... would already be compressed, I don't know if
> I mounted the resulting disk and it looks good. I copied
> initrd.img-3.2.0.4-vexpress from the first partition on the disk to my
> working directory and gzipped it, with a symlink from initrd.gz.
> vmlinux-3.2.0-4-vexpress is the same as what I downloaded.
doing that again would have confused anything. Worth trying without
though it.
> So although mmcblk0p2 is available, it looks like the system can't mount
> it for some reason.
ended up on it? Are you sure it is the rootfs and not e.g. swap?
Ian.
--
To UNSUBSCRIBE, email to debian-ar...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Jerry Stuckle
Jan 22, 2014, 9:50:01 AM1/22/14
to
On 1/22/2014 9:22 AM, Ian Campbell wrote:
> On Wed, 2014-01-22 at 09:14 -0500, Jerry Stuckle wrote:
>> I mounted the resulting disk and it looks good. I copied
>> initrd.img-3.2.0.4-vexpress from the first partition on the disk to my
>> working directory and gzipped it, with a symlink from initrd.gz.
>> vmlinux-3.2.0-4-vexpress is the same as what I downloaded.
>
> I think initrd.img-3.2... would already be compressed, I don't know if
> doing that again would have confused anything. Worth trying without
> though it.
>
(Feeling VERY sheepish) You're right. Just copying the file over and
> On Wed, 2014-01-22 at 09:14 -0500, Jerry Stuckle wrote:
>> I mounted the resulting disk and it looks good. I copied
>> initrd.img-3.2.0.4-vexpress from the first partition on the disk to my
>> working directory and gzipped it, with a symlink from initrd.gz.
>> vmlinux-3.2.0-4-vexpress is the same as what I downloaded.
>
> I think initrd.img-3.2... would already be compressed, I don't know if
> doing that again would have confused anything. Worth trying without
> though it.
>
not gzipping it works fine. I knew it was something simple. I can't
believe how many days I spent on this before breaking down and asking here!
>> So although mmcblk0p2 is available, it looks like the system can't mount
>> it for some reason.
>
> Can you use kpartx on the image and mount in the host? What filesystem
> ended up on it? Are you sure it is the rootfs and not e.g. swap?
>
> Ian.
>
>
and verified all of the partitions were ok.
Thanks again for your quick response, Ian. I only wish I had asked last
week :)
Jerry
--
To UNSUBSCRIBE, email to debian-ar...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
0 new messages