search for: sternwelten

----- Forwarded message from General Stone <generalstone at gmx.net> ----- X-Original-To: maks at sternwelten.at Date: Mon, 2 Jan 2006 14:59:03 +0100 From: General Stone <generalstone at gmx.net> To: maximilian attems <maks at sternwelten.at> Subject: Re: [Logcheck-devel] Bug#344832: correct subject header On Mon, Jan 02, 2006 at 02:09:48PM +0100, maximilian attems wrote: > please provide t...

A non-text attachment was scrubbed... Name: parse_logcheck Type: application/octet-stream Size: 3860 bytes Desc: not available Url : http://lists.alioth.debian.org/pipermail/logcheck-devel/attachments/20040901/b1727d89/attachment.obj

From: Steve Langasek <vorlon@debian.org> Fix the definition of struct statfs64, required for run-init to work on alpha. verified to have no regressions on amd64. Signed-off-by: maximilian attems <janitor@sternwelten.at> Signed-off-by: Frederik Sch?ler <fschueler-guest@costa.debian.org> --- klibc-1.1.1.orig/include/sys/vfs.h +++ klibc-1.1.1/include/sys/vfs.h @@ -32,17 +32,17 @@ #else /* _BITSIZE == 64 */ struct statfs { - uint64_t f_type; - uint64_t f_bsize; + uint32_t f_type;...

sparc still needs gcc-3.3 for a working run-init. belows makes use of the exported CC var. Signed-off-by: maximilian attems <maks@sternwelten.at> diff --git a/scripts/Kbuild.klibc b/scripts/Kbuild.klibc index 66e16e7..70e4071 100644 --- a/scripts/Kbuild.klibc +++ b/scripts/Kbuild.klibc @@ -72,7 +72,7 @@ KLIBCROSS := $(CROSS_COMPILE) # binutils KLIBCLD := $(KLIBCROSS)ld -KLIBCCC := $(KLIBCROSS)gcc +KLIBCCC...

From: Sjoerd Simons <sjoerd@spring.luon.net> It seems that klibc uses __sparc64__ to determine if it's compiled in 64 bit mode on sparc, unfortunatly gcc doesn't define __sparc64__ :) It does define __arch64__ though. Signed-off-by: maximilian attems <maks@sternwelten.at> --- klibc-1.2.2.orig/include/sys/vfs.h 2006-02-15 18:32:10.000000000 +0100 +++ klibc-1.2.2/include/sys/vfs.h 2006-02-15 18:32:35.000000000 +0100 @@ -13,7 +13,7 @@ /* struct statfs64 -- there seems to be two standards - one for 32 and one for 64 bits, and they're incompatible... */...

patch from Hannes Reinecke <hare at suse.de> reduces noise on s390 build logs. while we are it remove check for early 2.X gcc versions. Signed-off-by: maximilian attems <maks at sternwelten.at> diff --git a/usr/include/klibc/compiler.h b/usr/include/klibc/compiler.h index 893f8a9..b7d9b50 100644 --- a/usr/include/klibc/compiler.h +++ b/usr/include/klibc/compiler.h @@ -91,12 +91,12 @@ # define __mallocfunc #endif /* likely/unlikely */ -#if defined(__GNUC__) && (__GNUC__...

please pull git://charm.itp.tuwien.ac.at/mattems/klibc/.git maks commit b0c5c2b8ee60d91cb96541339c6be8108d0e2131 Author: maximilian attems <maks at sternwelten.at> Date: Thu Aug 24 22:40:23 2006 +0200 [klibc] mkdir add usage when printing errors about cli switches, point out the expected usage. patches adds usage function and calls to it. Signed-off-by: maximilian attems <maks at sternwelten.at> diff --git a/usr/utils/mkd...

add object rules so that the division, remainder and friends get really build on sparc, patch from Fabio M. Di Nitto <fabbione@ubuntu.com>. reworked to apply on latest git tree. Signed-off-by: maximilian attems <maks@sternwelten.at> --- Has been since long in the Debian and Ubuntu klibc. diff --git a/klibc/arch/sparc/Makefile.inc b/klibc/arch/sparc/Makefile.inc index 41c23c5..d77030d 100644 --- a/klibc/arch/sparc/Makefile.inc +++ b/klibc/arch/sparc/Makefile.inc @@ -23,25 +23,33 @@ KLIBCARCHOBJS = \ libgcc/__umoddi3....

Package: logcheck wanted to work on #195935, but found a less than funny issue, easy to reproduce: * remove some lines in front of your logfile * invoke logcheck you'll get a big email with all not matching lines from that log. not setting that to high priority because you are getting also the newer loglines. don't know if i find time that weekend. wanted to document it anyways. a++

hello, the debconf configuration of logcheck was removed since woody release: as logcheck's user base is targeting server admins, we don't see the need of a debconf based question regarding the frequency of logcheck. beside once per day is a good default. a note was added to documentation README.Debian that this value may be changed in /etc/cron.d/logcheck. so these bugreports will be

...l i got: Setting up logcheck (1.2.32+1.2.33rc1) ... /var/lib/dpkg/info/logcheck.postinst: 87: touch: not found i guess it's harmless, just sending out as note, reinstall of coreutils + logcheck worked. ;-) -- maks kernel janitor http://janitor.kernelnewbies.org/ personal weblog http://www.sternwelten.at/weblog.shtml

...nt from other syscalls on alpha, which didn't make sense; then I realized that none of the other syscalls I was looking at were *returning* a non-zero errno, just a return value. verified by indepent test boots that both patches fixes alpha klibc. Signed-off-by: maximilian attems <janitor@sternwelten.at> --- libc-1.1.1.orig/klibc/arch/alpha/syscall.S 2005-09-06 13:49:34.000000000 -0700 +++ klibc-1.1.1/klibc/arch/alpha/syscall.S 2005-12-15 23:42:02.000000000 -0800 @@ -17,8 +17,8 @@ 2: ldgp gp, 0(pv) lda a1, errno + stl v0, 0(a1) lda v0, -1(zero) - stl a3, 0(a1) 1: ret z...

...e/arch/sparc64/archstat.h. That required changing the storage size for some fields, rearranging them, and adding some padding fields where neccessary. Attached patch implements these changes. With this change I was able to successfully boot my ultra5 box. Signed-off-by: maximilian attems <maks@sternwelten.at> --- a/include/arch/sparc64/klibc/archstat.h 2005-09-06 13:49:34.000000000 -0700 +++ b/include/arch/sparc64/klibc/archstat.h 2005-10-29 10:31:26.000000000 -0700 @@ -3,20 +3,26 @@ /* No nsec fields?! */ struct stat { - unsigned st_dev; + unsigned long st_dev; + unsigned short __pad1;...

klibc doesn't currently build on s390 since commit 1e8e338325fa0f2bd5c1c800086e900cf188d2cd belows patch adds some kernel ifdef for the newly added section. Signed-off-by: maximilian attems <maks@sternwelten.at> --- linux-2.6.15/include/asm-s390/setup.h.orig 2006-02-01 11:27:00.000000000 +0100 +++ linux-2.6.15/include/asm-s390/setup.h 2006-02-01 11:31:54.000000000 +0100 @@ -67,6 +67,8 @@ #define SET_CONSOLE_3215 do { console_mode = 2; } while (0) #define SET_CONSOLE_3270 do { console_mode = 3; }...

klibc/arch/i386/openat.o: In function `openat':klibc/arch/i386/openat.S:19: undefined reference to `__NR_openat' make[3]: *** [klibc/libc.so] Error 1 belows allows to build against a tree based on 2.6.15 Signed-off-by: maximilian attems <maks@sternwelten.at> diff --git a/klibc/arch/i386/openat.S b/klibc/arch/i386/openat.S index 3955eae..40ab299 100644 --- a/klibc/arch/i386/openat.S +++ b/klibc/arch/i386/openat.S @@ -6,6 +6,7 @@ * and the need for O_LARGEFILE. */ +#ifdef __NR_openat #include <asm/unistd.h> /* <asm/fcntl.h>,...

...__sparc64__ to determine if it's compiled in 64 bit mode on sparc, gcc doesn't define __sparc64__ use __sparc_v9__ instead as proposed by Kyle McMartin <kyle@parisc-linux.org>, based on a patch by Sjoerd Simons <sjoerd@spring.luon.net> Signed-off-by: maximilian attems <maks@sternwelten.at> --- resend --- klibc-1.2.2.orig/include/sys/vfs.h 2006-02-15 18:32:10.000000000 +0100 +++ klibc-1.2.2/include/sys/vfs.h 2006-02-15 18:32:35.000000000 +0100 @@ -13,7 +13,7 @@ /* struct statfs64 -- there seems to be two standards - one for 32 and one for 64 bits, and they're incompa...

...has not been cleared with pvremove, fstype would recognise the ext3 partition as lvm2. workaround that by detection lvm2 after ext3. Bonuspoint remove long fs list as this one just generates conflicts. Signed-off-by: David H?rdeman <david@2gen.com> Signed-off-by: maximilian attems <maks@sternwelten.at> diff --git a/usr/kinit/fstype/fstype.c b/usr/kinit/fstype/fstype.c index c892c6b..17a89ce 100644 --- a/usr/kinit/fstype/fstype.c +++ b/usr/kinit/fstype/fstype.c @@ -6,8 +6,8 @@ * FSTYPE - filesystem type * FSSIZE - filesystem size (if known) * - * We currently detect (in order): -...

found by Christian T. Steigies <cts@debian.org> usr/klibc/arch/m68k/crt0.o: In function `_start': usr/klibc/arch/m68k/crt0.S:(.text+0xe): undefined reference to `___libc_init' too many '_'. never seen jbsr, converted that to jsr. Signed-off-by: maximilian attems <maks@sternwelten.at> diff --git a/usr/klibc/arch/m68k/crt0.S b/usr/klibc/arch/m68k/crt0.S index 6eb3b51..780d7a5 100644 --- a/usr/klibc/arch/m68k/crt0.S +++ b/usr/klibc/arch/m68k/crt0.S @@ -22,6 +22,6 @@ _start: # Push ELF argument pointer on the stack move.l %d0, -(%a7) - jbsr ___libc_init + jsr __libc...

found by Christian T. Steigies <cts@debian.org> usr/include/arch/m68k/klibc/archstat.h:12: error: syntax error before '__dev64' better use __stdev64 defined in klibc/stathelp.h Signed-off-by: maximilian attems <maks@sternwelten.at> diff --git a/usr/include/arch/m68k/klibc/archstat.h b/usr/include/arch/m68k/klibc/archstat.h index 89c0341..dce25f9 100644 --- a/usr/include/arch/m68k/klibc/archstat.h +++ b/usr/include/arch/m68k/klibc/archstat.h @@ -9,7 +9,7 @@ * insane padding around dev_t's. */ struct stat { -...

...d7 removed err declaration usr/klibc/arch/s390/syscall.c: In function '__syscall_common': usr/klibc/arch/s390/syscall.c:10: error: parameter name omitted usr/klibc/arch/s390/syscall.c:12: error: 'err' undeclared (first use in this function) Signed-off-by: maximilian attems <maks@sternwelten.at> diff --git a/usr/klibc/arch/s390/syscall.c b/usr/klibc/arch/s390/syscall.c index 60fab13..e1d201d 100644 --- a/usr/klibc/arch/s390/syscall.c +++ b/usr/klibc/arch/s390/syscall.c @@ -7,7 +7,7 @@ */ #include <errno.h> -unsigned long __syscall_common(unsigned long) +unsigned long __s...