9:22 p.m.
Hi Experts:
I got vmcore file on powerpc board(P2020 or P2040), but failed to do crash analysis on x86
host, it showed we need to build a powerpc supported crash binary.
On the README of the latest crash-7.1.3, it state to get crash binary which can support
powerpc32 vmcore analysis, we need to build with `make target=PPC` on PPC64 host, but
currently, it's hard to find this kind of host to do build, if there are some other
switch/method can let us to build powerpc supported binary on x86/x86_64 host?
Thanks a lot for your time!!
Best Regards,
James
Liu Jianbo | WIND RIVER | Senior Engineer - Technical Support
Tel 86 28 65318098 | Cell 86 13558641588 | Fax 86 28 65319983
[Description: Description: Wind River Support
Network]<https://support.windriver.com/>
[Description: Description: Wind River Knowledge Forum]<https://ask.windriver.com/>
[Description: Description: Submit a Technical Service
Request]<https://windriver.force.com/support>
[Description: Description: WIND]<http://www.windriver.com/>
9:40 p.m.
----- Original Message -----
Hi Experts:
I got vmcore file on powerpc board(P2020 or P2040), but failed to do crash
analysis on x86 host, it showed we need to build a powerpc supported crash
binary.
On the README of the latest crash-7.1.3, it state to get crash binary which
can support powerpc32 vmcore analysis, we need to build with `make
target=PPC` on PPC64 host, but currently, it’s hard to find this kind of
host to do build, if there are some other switch/method can let us to build
powerpc supported binary on x86/x86_64 host?
Thanks a lot for your time!!
Sorry, but no. The two architectures need to have the same endian-ness.
Dave
Best Regards,
James
Liu Jianbo | WIND RIVER | Senior Engineer - Technical Support
Tel 86 28 65318098 | Cell 86 13558641588 | Fax 86 28 65319983
--
Crash-utility mailing list
Crash-utility(a)redhat.com
https://www.redhat.com/mailman/listinfo/crash-utility
11:30 p.m.
Hi Dave:
Thanks for your quick reply.
Even if we build with cross compiler, we still can not meet this request?
Best Regards,
James
Liu Jianbo | WIND RIVER | Senior Engineer - Technical Support
Tel 86 28 65318098 | Cell 86 13558641588 | Fax 86 28 65319983
-----Original Message-----
From: crash-utility-bounces(a)redhat.com [mailto:crash-utility-bounces@redhat.com] On Behalf
Of Dave Anderson
Sent: Friday, November 27, 2015 11:41 AM
To: Discussion list for crash utility usage, maintenance and development
Subject: Re: [Crash-utility] On X86 host, how to build crash tool which can analysis
pwoerpc vmcore
----- Original Message -----
Hi Experts:
I got vmcore file on powerpc board(P2020 or P2040), but failed to do
crash analysis on x86 host, it showed we need to build a powerpc
supported crash binary.
On the README of the latest crash-7.1.3, it state to get crash binary
which can support powerpc32 vmcore analysis, we need to build with
`make target=PPC` on PPC64 host, but currently, it’s hard to find this
kind of host to do build, if there are some other switch/method can
let us to build powerpc supported binary on x86/x86_64 host?
Thanks a lot for your time!!
Sorry, but no. The two architectures need to have the same endian-ness.
Dave
Best Regards,
James
Liu Jianbo | WIND RIVER | Senior Engineer - Technical Support Tel 86
28 65318098 | Cell 86 13558641588 | Fax 86 28 65319983
--
Crash-utility mailing list
Crash-utility(a)redhat.com
https://www.redhat.com/mailman/listinfo/crash-utility
--
Crash-utility mailing list
Crash-utility(a)redhat.com
https://www.redhat.com/mailman/listinfo/crash-utility
6:05 a.m.
On Fri, Nov 27, 2015 at 05:30:30AM +0000, Liu, Jianbo (James) wrote:
Hi Dave:
Thanks for your quick reply.
Even if we build with cross compiler, we still can not meet this request?
Even with a cross compiler you'd need an environment to execute the
resulting powerpc build - about the only way that you can do that on
an x86 host is to use something like Qemu's powerpc support:
http://www.ibm.com/developerworks/library/l-qemu-development/
It's slow compared to real hardware but if using a real power host is
not an option this may be worth looking into.
Regards,
Bryn.
11:44 a.m.
On Fri, Nov 27, 2015 at 01:05:44PM +0100, Bryn M. Reeves wrote:
On Fri, Nov 27, 2015 at 05:30:30AM +0000, Liu, Jianbo (James) wrote:
> Even if we build with cross compiler, we still can not meet this request?
Even with a cross compiler you'd need an environment to execute the
resulting powerpc build - about the only way that you can do that on
an x86 host is to use something like Qemu's powerpc support:
http://www.ibm.com/developerworks/library/l-qemu-development/
It's slow compared to real hardware but if using a real power host is
not an option this may be worth looking into.
I tested this earlier for analysis of a big-endian MIPS dump on an
x86-64 host, with success. See these for more info:
https://www.redhat.com/archives/crash-utility/2015-March/msg00001.html
https://www.redhat.com/archives/crash-utility/2015-January/msg00028.html
Rabin
6:36 a.m.
On Fri, 27 Nov 2015 18:44:28 +0100
Rabin Vincent <rabin.vincent(a)axis.com> wrote:
On Fri, Nov 27, 2015 at 01:05:44PM +0100, Bryn M. Reeves wrote:
> On Fri, Nov 27, 2015 at 05:30:30AM +0000, Liu, Jianbo (James) wrote:
> > Even if we build with cross compiler, we still can not meet this request?
>
> Even with a cross compiler you'd need an environment to execute the
> resulting powerpc build - about the only way that you can do that on
> an x86 host is to use something like Qemu's powerpc support:
>
> http://www.ibm.com/developerworks/library/l-qemu-development/
>
> It's slow compared to real hardware but if using a real power host is
> not an option this may be worth looking into.
I tested this earlier for analysis of a big-endian MIPS dump on an
x86-64 host, with success. See these for more info:
https://www.redhat.com/archives/crash-utility/2015-March/msg00001.html
https://www.redhat.com/archives/crash-utility/2015-January/msg00028.html
If you're able (and willing) to invest some time into development, you
may find this project interesting:
https://www.sourceware.org/ml/gdb/2015-09/msg00014.html
Support for powerpc is currently missing, but it could be added with
relatively low effort, and there is a so-called HackWeek¹ at SUSE in the
week of Dec 7, so I could help with that. There are a few other SUSE
people interested in this project.
Petr Tesarik
¹ SUSE HackWeek is a week that can be used by all SUSE R&D employees for
innovative projects of their choice (instead of their regular work).
3215
days inactive
3216
days old
devel@lists.crash-utility.osci.io
5 comments
5 participants
participants (5)
-
Bryn M. Reeves -
Dave Anderson -
Liu, Jianbo (James) -
Petr Tesarik -
Rabin Vincent