2.6.33.9-rt31 boot problem

4 messages, 2 authors, 2011-06-15 · open the first message on its own page

2.6.33.9-rt31 boot problem

From: dashesy <hidden>
Date: 2011-06-12 16:21:35

First, sorry for double post, I did not know if my email went through
first time (maybe for large config file attached)

Same machine 2.6.33.7.2-rt30 boots fine with Ubuntu and no problem at
all, today I tried the new version with Fedora 15 but it crashes early
in the boot process.
I have attached boot log I get from Qemu which is
the same (apart from the CPU number for example). I also have attached
the patch needed for successful build; V4l example in the
documentation was removed (did not compile with new API), also some
gcc4.6 related issues with anonymous union and superfluous warnings
during build was resolved. The patch is just to have the build so it
is not fully documented.
Please let me know if any further information is needed, if I should
submit the crash to another mailing list, or if I should enable some
option in the config file for more elaborate crash log. Since crash is
very early I could not get a crashdump.

I will be happy to look into the source code too (I am learning), if
you can give me some hint, thanks

dashesy

Re: 2.6.33.9-rt31 boot problem

From: Fernando Lopez-Lezcano <hidden>
Date: 2011-06-15 17:54:08

On 06/12/2011 09:21 AM, dashesy wrote:
First, sorry for double post, I did not know if my email went through
first time (maybe for large config file attached)

Same machine 2.6.33.7.2-rt30 boots fine with Ubuntu and no problem at
all, today I tried the new version with Fedora 15 but it crashes early
in the boot process.
A workaround in Fedora 15 (posted in another message to another list so 
you may not have seen it) is to use upstart instead of systemd for 
startup (systemd requires features only found on newer kernels).

In the kernel boot line add "init=/sbin/upstart" and of course make sure 
you have the upstart package installed.

BTW, thanks for the patch!
-- Fernando

I have attached boot log I get from Qemu which is
the same (apart from the CPU number for example). I also have attached
the patch needed for successful build; V4l example in the
documentation was removed (did not compile with new API), also some
gcc4.6 related issues with anonymous union and superfluous warnings
during build was resolved. The patch is just to have the build so it
is not fully documented.
Please let me know if any further information is needed, if I should
submit the crash to another mailing list, or if I should enable some
option in the config file for more elaborate crash log. Since crash is
very early I could not get a crashdump.

I will be happy to look into the source code too (I am learning), if
you can give me some hint, thanks

dashesy

Re: 2.6.33.9-rt31 boot problem

From: dashesy <hidden>
Date: 2011-06-15 19:18:02

On Wed, Jun 15, 2011 at 11:48 AM, Fernando Lopez-Lezcano
[off-list ref] wrote:
On 06/12/2011 09:21 AM, dashesy wrote:
quoted
First, sorry for double post, I did not know if my email went through
first time (maybe for large config file attached)

Same machine 2.6.33.7.2-rt30 boots fine with Ubuntu and no problem at
all, today I tried the new version with Fedora 15 but it crashes early
in the boot process.
A workaround in Fedora 15 (posted in another message to another list so you
may not have seen it) is to use upstart instead of systemd for startup
(systemd requires features only found on newer kernels).

In the kernel boot line add "init=/sbin/upstart" and of course make sure you
have the upstart package installed.
Thanks for the reply,
Actually oops is much earlier than init, as you can see in the
attached boot.log I have "init=/bin/sh"
[    0.000000] Linux version 2.6.33.9-rt31 (ehsan@ehsan) (gcc version
4.6.0 20110530 (Red Hat 4.6.0-9) (GCC) ) #3 SMP PREEMPT RT Wed Jun 8
16:16:46 MDT 2011

[    0.000000] Command line: console=ttyS0 root=/dev/sda rw init=/bin/sh
BTW, thanks for the patch!
-- Fernando

quoted
I have attached boot log I get from Qemu which is
the same (apart from the CPU number for example). I also have attached
the patch needed for successful build; V4l example in the
documentation was removed (did not compile with new API), also some
gcc4.6 related issues with anonymous union and superfluous warnings
during build was resolved. The patch is just to have the build so it
is not fully documented.
Please let me know if any further information is needed, if I should
submit the crash to another mailing list, or if I should enable some
option in the config file for more elaborate crash log. Since crash is
very early I could not get a crashdump.

I will be happy to look into the source code too (I am learning), if
you can give me some hint, thanks

dashesy
dashesy

Re: 2.6.33.9-rt31 boot problem

From: Fernando Lopez-Lezcano <hidden>
Date: 2011-06-15 19:28:24

On 06/15/2011 12:18 PM, dashesy wrote:
On Wed, Jun 15, 2011 at 11:48 AM, Fernando Lopez-Lezcano
[off-list ref]  wrote:
quoted
On 06/12/2011 09:21 AM, dashesy wrote:
quoted
First, sorry for double post, I did not know if my email went through
first time (maybe for large config file attached)

Same machine 2.6.33.7.2-rt30 boots fine with Ubuntu and no problem at
all, today I tried the new version with Fedora 15 but it crashes early
in the boot process.
A workaround in Fedora 15 (posted in another message to another list so you
may not have seen it) is to use upstart instead of systemd for startup
(systemd requires features only found on newer kernels).

In the kernel boot line add "init=/sbin/upstart" and of course make sure you
have the upstart package installed.
Thanks for the reply,
Actually oops is much earlier than init, as you can see in the
attached boot.log I have "init=/bin/sh"
[    0.000000] Linux version 2.6.33.9-rt31 (ehsan@ehsan) (gcc version
4.6.0 20110530 (Red Hat 4.6.0-9) (GCC) ) #3 SMP PREEMPT RT Wed Jun 8
16:16:46 MDT 2011

[    0.000000] Command line: console=ttyS0 root=/dev/sda rw init=/bin/sh
I see, yes (sorry I did not look at the log before), very very early... 
don't know what to make of it, sorry...

-- Fernando
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help