New: Buildroot's coreutils 'uname -p' reports 'Unknown' on recent ARM kernels
https://bugs.busybox.net/show_bug.cgi?id=6404
Summary: Buildroot's coreutils 'uname -p' reports 'Unknown' on
recent ARM kernels
Product: buildroot
Version: 2013.05
Platform: PC
OS/Version: Linux
Status: NEW
Severity: normal
Priority: P5
Component: Other
AssignedTo: unassigned at buildroot.uclibc.org
ReportedBy: daniel.price at gmail.com
CC: buildroot at uclibc.org
Estimated Hours: 0.0
Buildroot's coreutils uname(1) command has a special patch applied as part of buildroot. When you run 'uname -p', it tries to print out extended information about the processor you are running. It does this by parsing /proc/cpuinfo. I'm not sure of the provenance of this enhancement, but you can read the code in package/coreutils/coreutils-uname.patch.
However, on recent ARM kernels, the format of /proc/cpuinfo has changed incompatibly. Several packages subsequently break. I believe the commit in question is this:
http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=82b5df7bf27ca6cf642499bc3bcae1c93b9ac4ba
So, I recently discovered that when matched with recent ARM kernels, uname -p on buildroot is simply reporting 'Unknown'. I'm not really sure what the right answer might be. The code doesn't really seem set up to parse the /proc/cpuinfo file in more than one way.
I am presently using buildroot 2012.11, but the code seems to be the same in the latest release as well.
Signature · 3 lines
Configure bugmail: https://bugs.busybox.net/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
Quoted 5 lines · bugzilla at busybox.net
Buildroot's coreutils uname(1) command has a special patch applied as part of buildroot. When you run 'uname -p', it tries to print out extended information about the processor you are running. It does this by parsing
/proc/cpuinfo. I'm not sure of the provenance of this enhancement, but you can read the code inpackage/coreutils/coreutils-uname.patch.
Seems to me that this patch should be dropped since it's a feature
patch. Martin, you added it and Gustavo acked it in 2010. OK to drop it?
Regards,
Arnout
Signature · 6 lines
Arnout Vandecappelle arnout at mind be Senior Embedded Software Architect +32-16-286500 Essensium/Mind http://www.mind.be G.Geenslaan 9, 3001 Leuven, Belgium BE 872 984 063 RPR Leuven LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle GPG fingerprint: 7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F
This is a very long running problem that goes back to at least 2005. Google coreutils uname patch, to see the history of this problem. Uname will porbably need a different patch to work on ARM kernels.
Regards,
Martin
Quoted 19 lines · Arnout Vandecappelle
Quoted 10 lines · bugzilla at busybox.net
Buildroot's coreutils uname(1) command has a special patch applied as part of
buildroot. When you run 'uname -p', it tries to print out extended information
about the processor you are running. It does this by parsing/proc/cpuinfo.
I'm not sure of the provenance of this enhancement, but you can read the code
inpackage/coreutils/coreutils-**uname.patch.Seems to me that this patch should be dropped since it's a feature patch.
Martin, you added it and Gustavo acked it in 2010. OK to drop it?Regards,
Arnout
--
Quoted 7 lines
Arnout Vandecappelle arnout at mind be Senior Embedded Software Architect +32-16-286500 Essensium/Mind http://www.mind.be G.Geenslaan 9, 3001 Leuven, Belgium BE 872 984 063 RPR Leuven LinkedIn profile: http://www.linkedin.com/in/**arnoutvandecappelle<http://www.linkedin.com/in/arnoutvandecappelle> GPG fingerprint: 7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20130812/fb9aff5d/attachment.html>
Dear Martin Banky,
Quoted 3 lines · Martin Banky
This is a very long running problem that goes back to at least 2005. Google coreutils uname patch, to see the history of this problem. Uname will porbably need a different patch to work on ARM kernels.
If it's such a long-standing issue, why hasn't this never been fixed in upstream coreutils? Some disagreement on what the problem is?
Thomas
Signature · 4 lines
Thomas Petazzoni, Free Electrons Kernel, drivers, real-time and embedded Linux development, consulting, training and support. http://free-electrons.com
Thomas,
The issue is, who is responsible for the fix. Coreutils feels it's the kernel team's issue. They want the sysinfo or sysctl calls implemented. The kernel team thinks that it's not POSIX compliant, and so they won't implement it. This is my understanding of the problem. Take a look at the following thread for one of the discussions.
http://lists.gnu.org/archive/html/bug-coreutils/2005-09/msg00053.html
Martin
Quoted 11 lines · Thomas Petazzoni
Dear Martin Banky,
Quoted 1 line · Martin Banky
This is a very long running problem that goes back to at least 2005.
Quoted 2 lines
coreutils uname patch, to see the history of this problem. Uname will porbably need a different patch to work on ARM kernels.
If it's such a long-standing issue, why hasn't this never been fixed in upstream coreutils? Some disagreement on what the problem is?
Quoted 7 lines
Thomas
--
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20130813/b1a4f685/attachment.html>
Quoted 2 lines · Arnout Vandecappelle
Seems to me that this patch should be dropped since it's a feature
patch. Martin, you added it and Gustavo acked it in 2010. OK to drop it?
+1 it doesn't seem to be upstreamed and it's cosmetic. Regards.
Dear Gustavo Zacarias,
Quoted 6 lines · Gustavo Zacarias
Quoted 2 lines · Arnout Vandecappelle
Seems to me that this patch should be dropped since it's a feature
patch. Martin, you added it and Gustavo acked it in 2010. OK to drop it?+1 it doesn't seem to be upstreamed and it's cosmetic.
I also agree, so I'll drop the patch.
Thanks,
Thomas
Signature · 4 lines
Thomas Petazzoni, Free Electrons Kernel, drivers, real-time and embedded Linux development, consulting, training and support. http://free-electrons.com
[Buildroot] [Bug 6404] Buildroot's coreutils 'uname -p' reports 'Unknown' on recent ARM kernels
In reply to bugzilla at busybox.nethttps://bugs.busybox.net/show_bug.cgi?id=6404
Thomas Petazzoni <[email protected]> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |WONTFIX--- Comment #1 from Thomas Petazzoni <[email protected]> 2013-10-30 22:30:53 UTC --- This really is a coreutils upstream problem. Could you report this directly to
the coreutils developers? On Buildroot side, I believe we should drop coreutils-uname.patch since it's really a feature patch, and it apparently has
no chance of making its way to the upstream.
Another option is to look for an updated version of this patch somewhere.
Signature · 3 lines
Configure bugmail: https://bugs.busybox.net/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.