Bug#410819: Please add VIA C7 processor definition
Package: base-installer Version: 1.72 Severity: wishlist Tags: patch Hi, A popular French datacenter called Dedibox [1] is using a specifically built motherboard with a VIA C7 processor. Via a secure web interface, you can install Debian Sarge among other distros (with their untrusty closed source system). I am currently setting up a documentation to use the official Debian Installer with the official Debian kernels on that machine. Once done, I will put a link on the Debian Wiki. By default, only the -486 subarch is available at most, because /proc/cpuinfo is different from other x86 machines, i.e. neither AMD nor Intel... Therefore, I did a patch to get the -686 subarch available on that distinct hardware with all the flavours, e.g. xen, vserver, xen+vserver. Please consider the attached patch against the current SVN. Cheers, [1] http://www.dedibox.fr/ -- .''`. Aurélien GÉRÔME : :' : `. `'` Free Software Developer `- Unix Sys Net Admin Index: kernel/tests/i386/via-c7.test === --- kernel/tests/i386/via-c7.test (révision 0) +++ kernel/tests/i386/via-c7.test (révision 0) @@ -0,0 +1,22 @@ +cpuinfo via-c7.cpuinfo +machine i686 +majors 2.6 +flavour 686 +env SPEAKUP /nonexistent +kernel-2.6 \ + linux-image-2.6-686 \ + linux-image-2.6-486 \ + kernel-image-2.6-686 \ + kernel-image-2.6-386 +usable \ + linux-image-2.6-486 \ + linux-image-2.6-686-smp \ + linux-image-2.6-686 \ + linux-image-2.6.18-1-486 \ + linux-image-2.6.18-1-686-smp \ + linux-image-2.6.18-1-686 +unusable \ + linux-image-2.6-k7-smp \ + linux-image-2.6-k7 \ + linux-image-2.6.18-1-k7-smp \ + linux-image-2.6.18-1-k7 Index: kernel/tests/i386/via-c7.cpuinfo === --- kernel/tests/i386/via-c7.cpuinfo(révision 0) +++ kernel/tests/i386/via-c7.cpuinfo(révision 0) @@ -0,0 +1,18 @@ +processor : 0 +vendor_id : CentaurHauls +cpu family : 6 +model : 10 +model name : VIA Esther processor 2000MHz +stepping: 9 +cpu MHz : 1995.084 +cache size : 128 KB +fdiv_bug: no +hlt_bug : no +f00f_bug: no +coma_bug: no +fpu : yes +fpu_exception : yes +cpuid level : 1 +wp : yes +flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge cmov pat clflush acpi mmx fxsr sse sse2 tm nx pni est tm2 rng rng_en ace ace_en ace2 ace2_en phe phe_en pmm pmm_en +bogomips: 3994.32 Index: kernel/i386.sh === --- kernel/i386.sh (révision 45107) +++ kernel/i386.sh (copie de travail) @@ -9,7 +9,7 @@ *) echo 386 ;; esac ;; -GenuineIntel| GenuineTMx86*) +GenuineIntel| GenuineTMx86| CentaurHauls*) case $FAMILY in 6| 15) echo 686 ;; 5) echo 586tsc ;; Index: debian/changelog === --- debian/changelog(révision 45107) +++ debian/changelog(copie de travail) @@ -1,3 +1,9 @@ +base-installer (1.73) unstable; urgency=low + + * Add VIA C7 processor definition. + + -- Aurélien GÉRÔME [EMAIL PROTECTED] Tue, 13 Feb 2007 17:58:27 +0100 + base-installer (1.72) unstable; urgency=low [ Colin Watson ] signature.asc Description: Digital signature
Bug#410819: Please add VIA C7 processor definition
tags 410819 + pending thanks On Tuesday 13 February 2007 18:10, Aurélien GÉRÔME wrote: I am currently setting up a documentation to use the official Debian Installer with the official Debian kernels on that machine. Once done, I will put a link on the Debian Wiki. OK, so that is targeted at supporting Etch? By default, only the -486 subarch is available at most, because /proc/cpuinfo is different from other x86 machines, i.e. neither AMD nor Intel... Therefore, I did a patch to get the -686 subarch available on that distinct hardware with all the flavours, e.g. xen, vserver, xen+vserver. I have asked myself a little bit if it really makes sense to support such a very specific use case. As the patch is minimal and maintenance should be negligible, I guess it does not do any harm. Please consider the attached patch against the current SVN. Thanks for providing the test case too. Next time, please make sure your name is UTF-8 encoded in the changelog. Bonus points for marking the new changelog entry as UNRELEASED. Cheers, FJP pgpZqXomK1S8L.pgp Description: PGP signature
Bug#410819: Please add VIA C7 processor definition
On Tuesday 13 February 2007 18:10, Aurélien GÉRÔME wrote: Please consider the attached patch against the current SVN. I've made one correction relative to the patch: -GenuineIntel| GenuineTMx86| CentaurHauls*) +GenuineIntel| GenuineTMx86*| CentaurHauls) Keeping the * at the end changes the test for GenuineTMx86, which would be a regression. pgpLGqmQ9NK3I.pgp Description: PGP signature
Processed: Re: Bug#410819: Please add VIA C7 processor definition
Processing commands for [EMAIL PROTECTED]: tags 410819 + pending Bug#410819: Please add VIA C7 processor definition Tags were: patch Tags added: pending thanks Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#410819: Please add VIA C7 processor definition
On Tue, Feb 13, 2007 at 07:00:55PM +0100, Frans Pop wrote: Keeping the * at the end changes the test for GenuineTMx86, which would be a regression. Damn me, nice you spotted that one. I should have tested it back into qemu afterwards... Cheers, -- .''`. Aurélien GÉRÔME : :' : `. `'` Free Software Developer `- Unix Sys Net Admin signature.asc Description: Digital signature
Bug#410819: Please add VIA C7 processor definition
On Tue, Feb 13, 2007 at 07:20:22PM +0100, Aurélien GÉRÔME wrote: On Tue, Feb 13, 2007 at 07:00:55PM +0100, Frans Pop wrote: Keeping the * at the end changes the test for GenuineTMx86, which would be a regression. Damn me, nice you spotted that one. I should have tested it back into qemu afterwards... Actually, I remember that I tested it back, but qemu emulates a GenuineIntel, period. :) Cheers, -- .''`. Aurélien GÉRÔME : :' : `. `'` Free Software Developer `- Unix Sys Net Admin signature.asc Description: Digital signature
Bug#410819: Please add VIA C7 processor definition
On Tue, Feb 13, 2007 at 06:52:38PM +0100, Frans Pop wrote: On Tuesday 13 February 2007 18:10, Aurélien GÉRÔME wrote: I am currently setting up a documentation to use the official Debian Installer with the official Debian kernels on that machine. Once done, I will put a link on the Debian Wiki. OK, so that is targeted at supporting Etch? Sure, it works well. At least, better than in -486 flavour. There is a kernel optimisation for VIA C3/C7 processors which is disabled by default in the Debian kernel package. I will have a look whether it is fine or not with other 686 machines to activate it... By default, only the -486 subarch is available at most, because /proc/cpuinfo is different from other x86 machines, i.e. neither AMD nor Intel... Therefore, I did a patch to get the -686 subarch available on that distinct hardware with all the flavours, e.g. xen, vserver, xen+vserver. I have asked myself a little bit if it really makes sense to support such a very specific use case. As the patch is minimal and maintenance should be negligible, I guess it does not do any harm. Well, it does no harm. The machine is a 686, it is a bit unsatisfying not to run the kernel as such. Please consider the attached patch against the current SVN. Thanks for providing the test case too. Next time, please make sure your name is UTF-8 encoded in the changelog. $ wget http://bugs.debian.org/cgi-bin/bugreport.cgi/via-c7.patch?bug=410819;msg=5;att=1; file via* via-c7.patch?bug=410819;msg=5;att=1: UTF-8 Unicode text I am always unicode-enabled as you can see... Maybe an issue on your end? Bonus points for marking the new changelog entry as UNRELEASED. Indeed. :) Cheers, -- .''`. Aurélien GÉRÔME : :' : `. `'` Free Software Developer `- Unix Sys Net Admin signature.asc Description: Digital signature
Bug#410819: Please add VIA C7 processor definition
On Tue, Feb 13, 2007 at 07:20:22PM +0100, Aurélien GÉRÔME wrote: On Tue, Feb 13, 2007 at 07:00:55PM +0100, Frans Pop wrote: Keeping the * at the end changes the test for GenuineTMx86, which would be a regression. Damn me, nice you spotted that one. I should have tested it back into qemu afterwards... ... which leads me to think that the testsuite should have barfed. $ grep GenuineTMx86 kernel/tests/i386/* kernel/tests/i386/oqo1.cpuinfo:vendor_id: GenuineTMx86 There is no test case for a GenuineTMx86* string... Cheers, -- .''`. Aurélien GÉRÔME : :' : `. `'` Free Software Developer `- Unix Sys Net Admin signature.asc Description: Digital signature
