Bug#221234: Reopening 221234
Hi, Maybe I was unclear, this bug does not belong to the locales package. So please tell us which packages are installed on your machine and related to locales configuration, so that this bug can be reassigned to the culprit, otherwise I will close it. Thanks. Denis
Bug#222171: glibc-doc: Floods info-dir with entries.
On Wed, Nov 26, 2003 at 12:04:22AM +0100, Sverker Wiberg wrote:
Package: glibc-doc
Version: 2.3.2.ds1-10
Severity: normal
Tags: sid
Of the 2273 lines in my current info-dir (the Direcory node), 1705 lines
are entry points into the libc info files -- from 'ALTWERASE' to
'{(*__gconv_init_fct)}'. With this many entry points, the directory
isn't nearly as easily browsed as it was with ~570 entries. (When you
know where to go, the 'm' key keeps this from being more of a
pain than it is...except that the 'libc' entry itself has disappeared.)
Hopefully, other maintainers won't find this a good idea, or the
info-dir will most definitely explode.
So, please remove the huge list and add back the 'libc' entry.
The 'libc.info' file does contain a useful index; there's no need to
duplicate it in the directory.
The removal of libc is definitely a problem. I don't know where it
went, we'll investigate.
Let me think about what to do with the others for a bit. It may have
been a bad choice. I'm open to other opinions.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
Bug#222598: thread_info.h:49: global register variable follows a function definition
Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Severity: normal During my work on the alpha boot loader aboot, I stumbled on the following problem: [EMAIL PROTECTED]:~/prog/debian/ABOOT/aboot-0.9b-trisplit/lib$ env LANG=C make gcc -Os -Wall -mno-fp-regs -fno-builtin -D__KERNEL__ -I../include -c -o isolib.o isolib.c In file included from /usr/include/linux/thread_info.h:21, from /usr/include/linux/spinlock.h:12, from /usr/include/linux/time.h:27, from /usr/include/linux/stat.h:57, from isolib.c:15: /usr/include/asm/thread_info.h:49: global register variable follows a function definition /usr/include/asm/thread_info.h:49: warning: call-clobbered register used for global register variable make: *** [isolib.o] Error 1 This looks to me like an error in thread_info.h. If you like, I can provide you with the files, but more easily install the aboot sources, go into the lib/ directory and run make. (On alpha only currently) -- System Information Debian Release: testing/unstable Architecture: alpha Kernel: Linux sixtyfor 2.4.21-pre1 #1 Don Okt 2 18:27:25 CEST 2003 alpha Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] -- Helge Kreutzmann, Dipl.-Phys. [EMAIL PROTECTED] gpg signed mail preferredgpg-key: finger [EMAIL PROTECTED] 64bit GNU powered http://www.itp.uni-hannover.de/~kreutzm Help keep free software libre: http://www.freepatents.org/
Bug#222733: libc6: strchr() broken
Package: libc6
Version: 2.3.2.ds1-10
Severity: normal
I am building a program that at one point uses strchr(). It segfaults in
sysdeps/i386/strchr.S.
I attach a gdb trace. I am not familiar with assembler so I cannot get
any more information without assistance.
gcc is 3.3.2
I am quite prepared to believe that I have made a mistake in my program, but
I can't see anything wrong at the moment.
[EMAIL PROTECTED] LD_LIBRARY_PATH=/usr/lib/debug gdb ./pg_default
GNU gdb 6.0-debian
Copyright 2003 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type show copying to see the conditions.
There is absolutely no warranty for GDB. Type show warranty for details.
This GDB was configured as i386-linux...
(gdb) set args -u olly -c cust1
(gdb) directory
/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2/sysdeps/i386:/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2
Source directories searched:
/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2/sysdeps/i386:/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2:$cdir:$cwd
(gdb) b pg_files.c:397
Breakpoint 1 at 0x804a8a3: file pg_files.c, line 397.
(gdb) r
Starting program: /usr/src/mypackages/postgresql/alioth/common/pg_default -u
olly -c cust1
Breakpoint 1, write_cluster_line (user=0x804d9c8 olly, group=0x804c0c5 *,
cluster=0x804dad8 cust1, db=0x804db60 , forceflag=0) at pg_files.c:397
397 ptr = strchr(curpos, '\n');
(gdb) p curpos
$1 = 0x4014b000 # This file maps users against the database clusters to which
they will\n# connect by default. Any user may use \pg-wrapper -c\ to set a
new\n# default to be stored in ~/.postgresqlrc and the existence...
(gdb) p ptr
$2 = 0x1 Address 0x1 out of bounds
(gdb) s
strchr () at ../sysdeps/i386/strchr.S:37
warning: Source file is more recent than executable.
37 pushl %edi /* Save callee-safe registers used
here. */
Current language: auto; currently asm
(gdb) si
38 movl STR(%esp), %eax
(gdb) si
39 movl CHR(%esp), %edx
(gdb)
46 movb %dl, %dh /* now it is 0|0|c|c */
(gdb)
47 movl %edx, %ecx
(gdb)
48 shll $16, %edx /* now it is c|c|0|0 */
(gdb)
49 movw %cx, %dx /* and finally c|c|c|c */
(gdb)
62 testb $3, %al /* correctly aligned ? */
(gdb) p $cx
$3 = void
(gdb) p $dx
$4 = void
(gdb) p $3
$5 = void
(gdb) p $al
$6 = void
(gdb) si
63 jz L(11)/* yes = begin loop */
(gdb) si
124 L(11): movl (%eax), %ecx /* get word (= 4 bytes) in question */
(gdb) p $eax
$7 = 1075097600
(gdb) p $ecx
$8 = 2570
(gdb) si
Program received signal SIGSEGV, Segmentation fault.
strchr () at ../sysdeps/i386/strchr.S:124
124 L(11): movl (%eax), %ecx /* get word (= 4 bytes) in question */
(gdb)
This is the code of the relevant routine; I can give you the whole lot if
you need it:
/
Write a line in user_clusters
/
void write_cluster_line(const char *user, const char *group,
const char *cluster, const char *db,
const int forceflag) {
FILE *ucf;
char *filename = /etc/postgresql/user_clusters;
char *buf, *curpos;
char *ptr;
char in[BUFSIZ];
char c_user[BUFSIZ], c_group[BUFSIZ], junk[BUFSIZ];
char line[BUFSIZ];
int c, found = 0, set_default = 0;
/* create the line to write */
snprintf(line, BUFSIZ, %s \t%s \t%s \t%s \t%s\n,
user, group, cluster,
(forceflag ? yes : no),
(db[0] == '\0' ? * : db));
line[BUFSIZ - 1] = '\0';
/* start of user_clusters - open read-write */
sprintf(buf, Failed to open %s for writing, filename);
if (!(ucf = fopen(filename, r+))) {
/* try creating it */
if (!(ucf = fopen(filename, w))) {
perror(buf);
exit(LOC_ERR_WRITE_FAIL);
}
/* set permissions and ownership */
if (fchmod(fileno(ucf), S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH)) {
perror (Could not set perms on user_clusters);
exit(LOC_ERR_WRITE_FAIL);
}
fputs(line, ucf);
if (!geteuid()) {
/* we are root - set ownership to postgres.postgres */
struct passwd *pwd;
pwd = getpwnam(postgres);
fchown(fileno(ucf), pwd-pw_uid, pwd-pw_gid);
}
fclose(ucf);
return;
}
/* Read it into memory */
buf = 0;
if ((c = fseek(ucf, 0, SEEK_END)) == -1) {
perror(Error seeking in user_clusters);
exit(LOC_ERR_READ_FAIL);
}
if ((c = ftell(ucf)) == -1) {
perror(Error doing ftell() in user_clusters);
exit(LOC_ERR_READ_FAIL);
}
/* map the file and load an extra page in case the new line expands the
file across the page boundary; adding 2 allows for the truncating
effect of integer division.
Bug#67921: Brand New Marketing
Our New Revolutionary Duplication System can generate you One Hundred Thousand Dollars per month. Starting within 24 hours. And who generates that kind of money? Elite athletes? Top musicians Privileged TV stars? HOW ABOUT YOU? With our Duplication system, you could generate more money than any of them! We've designed an income generating system THE WORLD'S NEVER SEEN ! It's New, It's Never been seen, and it works ! Here it is right below. Presentation: http://www.catchthecurrent.net/?id=20427 OR Call 800-800-8000 Tim generated over twenty seven thousand dollars in just fifty six hours. Would that help? That was last week using the brand new presentation below. Seeing is believing so take a look at this now. If your are on dial up be patient as this will take a few minutes to download. This can change your life now. Presentation: http://www.catchthecurrent.net/?id=20427 OR Call 800-800-8000 With this program everyone can achieve unlimited success and no one is left behind. Meet you at the top. PS: What kind of a car are you driving? That's about to change!
Bug#222620: Upgrade to 2.3.2.ds1-10 segfaults on kernel 2.6.0-test9-1-386
Version: 2.3.2.ds1-10 Package: libc6 I was running the 2.6.0-test7-1-386 kernel package when I tried to ungrade; bash, ldd and many other programs segfaulted. 'Ash' did not. After checking with bugs.debian.org I switched back to a 2.4.21 kernel which has completed the upgrade successfully. On attempted reboot to 2.6, lots of segfaults and: INIT:PANIC: segmentation violation at 0x8049f52 (code)! sleeping for 30 seconds. Hard reboot required. I then tried debian kernel image 2.6.0-test9-1-386 and this gives the same segfaults. BUT! After all this I've found that this is only when running inside vmware. Therefor I think the severity for this is at most 'normal' and it is likely that this is a vmware bug. For vmware: Host windows 2k-sp4, version workstation 4.0.5.6030 contact email [EMAIL PROTECTED] -- Rob. (Robert de Bath robert$ @ debath.co.uk) http://www.cix.co.uk/~mayday Google Homepage: http://www.google.com/search?btnIq=Robert+de+Bath
Bug#222709: */time.h: Some one is not using _STRUCT_TIMEVAL.
Package: libc6-dev Version: 2.3.2.ds1-10 Severity: normal This is what I get from... #include linux/time.h #include sys/time.h In file included from /usr/include/sys/time.h:28, from tmp.c:2: /usr/include/bits/time.h:70: error: redefinition of struct timeval' In file included from /usr/include/sys/select.h:44, from /usr/include/sys/time.h:30, from tmp.c:2: /usr/include/time.h:119: error: redefinition of struct timespec' In file included from /usr/include/sys/time.h:30, from tmp.c:2: /usr/include/sys/select.h:78: error: conflicting types for fd_set' /usr/include/linux/types.h:22: error: previous declaration of fd_set' In file included from tmp.c:2: /usr/include/sys/time.h:57: error: redefinition of struct timezone' /usr/include/sys/time.h:94: error: parse error before numeric constant /usr/include/sys/time.h:108: error: redefinition of struct itimerval' This is scary, how did this get distributed? -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux overrun 2.6.0-test9-1-386 #1 Sun Oct 26 22:32:52 EST 2003 i686 Locale: LANG=C, LC_CTYPE=C Versions of packages libc6-dev depends on: ii libc6 2.3.2.ds1-10 GNU C Library: Shared libraries an ii linux-kernel-headers 2.5.999-test7-bk-9 Linux Kernel Headers for developme -- no debconf information
Bug#221401: libc6: Not able to load libglx.so (nvidia)
Package: libc6 Version: 2.3.2.ds1-10 Severity: normal Followup-For: Bug #221401 -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Same, with www.minion.de patches. But xfree server doesn't load the GLX driver: - /var/log/XFree86.0.log- [...] (II) LoadModule: glx (II) Loading /usr/X11R6/lib/modules/extensions/libglx.so dlopen: /usr/lib/libGLcore.so.1: undefined symbol: __gl_tls_var0 (EE) Failed to load /usr/X11R6/lib/modules/extensions/libglx.so (II) UnloadModule: glx (EE) Failed to load module glx (loader failed, 135874371) [...continues and xfree runs, without glx...]] - --- - -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux 2.6.0-test11 #2 Wed Dec 3 17:53:03 CET 2003 i686 Locale: LANG=C, LC_CTYPE=C (ignored: LC_ALL set to [EMAIL PROTECTED]) Versions of packages libc6 depends on: ii libdb1-compat 2.1.3-7The Berkeley database routines [gl - -- no debconf information -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.3 (GNU/Linux) iD8DBQE/zlgjAzbsrEhTuckRAl8AAKCAYGzpZ+rNGMfks/3MNBA88l5qsgCeIjkU ACm3eAvk3FXUEvNwkMGZL5I= =kn08 -END PGP SIGNATURE-
Bug#222777: /usr/include/asm/byteorder.h:56: error: parse error before ':' token
Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Severity: normal Compiling KDE from CVS, i got the following message: In file included from /usr/include/linux/cdrom.h:14 from plat_linux_cdda.c:43: /usr/include/asm/byteorder.h: In function '___arch__swab64': /usr/include/asm/byteorder.h:56: error: parse error before ':' token The relevant part from plat_linux_cdda.c is: #ifndef __GNUC__ #define __GNUC__ 1 #endif #undef __STRICT_ANSI__ #include asm/types.h /* ugly workaround for broken glibc shipped in SuSE 9.0 */ #define inline __inline__ #include linux/cdrom.h #undef inline And the referenced (buggy?) line from byteorder.h: asm(xchgl %0,%1 : =r (v.s.a), =r (v.s.b) : 0 (v.s.a),1 (v.s.b)); -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux fenix 2.6.0-test9leo #1 Thu Nov 13 14:06:27 CLST 2003 i686 Locale: LANG=es_CL, LC_CTYPE=es_CL (ignored: LC_ALL set to C) -- no debconf information
Bug#222777: /usr/include/asm/byteorder.h:56: error: parse error before ':' token
On Wed, Dec 03, 2003 at 12:31:51AM -0300, Leonardo Soto wrote: Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Severity: normal Compiling KDE from CVS, i got the following message: In file included from /usr/include/linux/cdrom.h:14 from plat_linux_cdda.c:43: /usr/include/asm/byteorder.h: In function '___arch__swab64': /usr/include/asm/byteorder.h:56: error: parse error before ':' token The relevant part from plat_linux_cdda.c is: #ifndef __GNUC__ #define __GNUC__ 1 #endif #undef __STRICT_ANSI__ #include asm/types.h /* ugly workaround for broken glibc shipped in SuSE 9.0 */ #define inline __inline__ #include linux/cdrom.h #undef inline And the referenced (buggy?) line from byteorder.h: asm(xchgl %0,%1 : =r (v.s.a), =r (v.s.b) : 0 (v.s.a),1 (v.s.b)); This is an absolutely, wildly, crazily dumb way to work around a broken header. Remove the #undef, and either linux-kernel-headers will work or the next version uploaded will when uploads are re-enabled. I believe Debian kde packages have already removed the #undef. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer
bug report : error: variable or field `__user' declared void
Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Hello, I look the development of driver eagle for my adsl modem. And we notice an error during compilation on Debian. I think I found the problem. /usr/include/linux/usbdevice_fs.h use __user but the definition of __user is not include automaticaly (there is in linux/compiler.h). Is it a bug? Or must we include linux/compiler.h before linux/usbdevice_fs.h ? Here an example if my explanation is too bad :-). [EMAIL PROTECTED]:/tmp/ -su2.05b$ echo #include linux/usbdevice_fs.h test.c [EMAIL PROTECTED]:/tmp/ -su2.05b$ gcc -D__KERNEL__ test.c In file included from test.c:1: /usr/include/linux/usbdevice_fs.h:49: error: variable or field `__user' declared void ... Excuse my poor english. mcoolive. _ Trouvez l'âme soeur sur MSN Rencontres http://g.msn.fr/FR1000/9551
Bug#222777: /usr/include/asm/byteorder.h:56: error: parse error before ':' token
El mié, 03-12-2003 a las 20:52, Daniel Jacobowitz escribió: [...] The relevant part from plat_linux_cdda.c is: #ifndef __GNUC__ #define __GNUC__ 1 #endif #undef __STRICT_ANSI__ #include asm/types.h /* ugly workaround for broken glibc shipped in SuSE 9.0 */ #define inline __inline__ #include linux/cdrom.h #undef inline And the referenced (buggy?) line from byteorder.h: asm(xchgl %0,%1 : =r (v.s.a), =r (v.s.b) : 0 (v.s.a),1 (v.s.b)); This is an absolutely, wildly, crazily dumb way to work around a broken header. Remove the #undef, and either linux-kernel-headers will work or the next version uploaded will when uploads are re-enabled. OK, undef removed and kscd compiles. I didn't remove it before submit the bug because I fear collateral troubles... Thanks you. -- Leonardo Soto M. [EMAIL PROTECTED] Jabber: [EMAIL PROTECTED]
Bug#221234: Reopening 221234
Hi, Maybe I was unclear, this bug does not belong to the locales package. So please tell us which packages are installed on your machine and related to locales configuration, so that this bug can be reassigned to the culprit, otherwise I will close it. Thanks. Denis -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222171: glibc-doc: Floods info-dir with entries.
On Wed, Nov 26, 2003 at 12:04:22AM +0100, Sverker Wiberg wrote:
Package: glibc-doc
Version: 2.3.2.ds1-10
Severity: normal
Tags: sid
Of the 2273 lines in my current info-dir (the Direcory node), 1705 lines
are entry points into the libc info files -- from 'ALTWERASE' to
'{(*__gconv_init_fct)}'. With this many entry points, the directory
isn't nearly as easily browsed as it was with ~570 entries. (When you
know where to go, the 'm' key keeps this from being more of a
pain than it is...except that the 'libc' entry itself has disappeared.)
Hopefully, other maintainers won't find this a good idea, or the
info-dir will most definitely explode.
So, please remove the huge list and add back the 'libc' entry.
The 'libc.info' file does contain a useful index; there's no need to
duplicate it in the directory.
The removal of libc is definitely a problem. I don't know where it
went, we'll investigate.
Let me think about what to do with the others for a bit. It may have
been a bad choice. I'm open to other opinions.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222620: Upgrade to 2.3.2.ds1-10 segfaults on kernel 2.6.0-test9-1-386
Version: 2.3.2.ds1-10 Package: libc6 I was running the 2.6.0-test7-1-386 kernel package when I tried to ungrade; bash, ldd and many other programs segfaulted. 'Ash' did not. After checking with bugs.debian.org I switched back to a 2.4.21 kernel which has completed the upgrade successfully. On attempted reboot to 2.6, lots of segfaults and: INIT:PANIC: segmentation violation at 0x8049f52 (code)! sleeping for 30 seconds. Hard reboot required. I then tried debian kernel image 2.6.0-test9-1-386 and this gives the same segfaults. BUT! After all this I've found that this is only when running inside vmware. Therefor I think the severity for this is at most 'normal' and it is likely that this is a vmware bug. For vmware: Host windows 2k-sp4, version workstation 4.0.5.6030 contact email [EMAIL PROTECTED] -- Rob. (Robert de Bath robert$ @ debath.co.uk) http://www.cix.co.uk/~mayday Google Homepage: http://www.google.com/search?btnIq=Robert+de+Bath -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222709: */time.h: Some one is not using _STRUCT_TIMEVAL.
Package: libc6-dev Version: 2.3.2.ds1-10 Severity: normal This is what I get from... #include linux/time.h #include sys/time.h In file included from /usr/include/sys/time.h:28, from tmp.c:2: /usr/include/bits/time.h:70: error: redefinition of struct timeval' In file included from /usr/include/sys/select.h:44, from /usr/include/sys/time.h:30, from tmp.c:2: /usr/include/time.h:119: error: redefinition of struct timespec' In file included from /usr/include/sys/time.h:30, from tmp.c:2: /usr/include/sys/select.h:78: error: conflicting types for fd_set' /usr/include/linux/types.h:22: error: previous declaration of fd_set' In file included from tmp.c:2: /usr/include/sys/time.h:57: error: redefinition of struct timezone' /usr/include/sys/time.h:94: error: parse error before numeric constant /usr/include/sys/time.h:108: error: redefinition of struct itimerval' This is scary, how did this get distributed? -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux overrun 2.6.0-test9-1-386 #1 Sun Oct 26 22:32:52 EST 2003 i686 Locale: LANG=C, LC_CTYPE=C Versions of packages libc6-dev depends on: ii libc6 2.3.2.ds1-10 GNU C Library: Shared libraries an ii linux-kernel-headers 2.5.999-test7-bk-9 Linux Kernel Headers for developme -- no debconf information -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222733: libc6: strchr() broken
Package: libc6
Version: 2.3.2.ds1-10
Severity: normal
I am building a program that at one point uses strchr(). It segfaults in
sysdeps/i386/strchr.S.
I attach a gdb trace. I am not familiar with assembler so I cannot get
any more information without assistance.
gcc is 3.3.2
I am quite prepared to believe that I have made a mistake in my program, but
I can't see anything wrong at the moment.
[EMAIL PROTECTED] LD_LIBRARY_PATH=/usr/lib/debug gdb ./pg_default
GNU gdb 6.0-debian
Copyright 2003 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type show copying to see the conditions.
There is absolutely no warranty for GDB. Type show warranty for details.
This GDB was configured as i386-linux...
(gdb) set args -u olly -c cust1
(gdb) directory
/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2/sysdeps/i386:/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2
Source directories searched:
/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2/sysdeps/i386:/usr/src/glibc-2.3.2.ds1/build-tree/glibc-2.3.2:$cdir:$cwd
(gdb) b pg_files.c:397
Breakpoint 1 at 0x804a8a3: file pg_files.c, line 397.
(gdb) r
Starting program: /usr/src/mypackages/postgresql/alioth/common/pg_default -u olly -c
cust1
Breakpoint 1, write_cluster_line (user=0x804d9c8 olly, group=0x804c0c5 *,
cluster=0x804dad8 cust1, db=0x804db60 , forceflag=0) at pg_files.c:397
397 ptr = strchr(curpos, '\n');
(gdb) p curpos
$1 = 0x4014b000 # This file maps users against the database clusters to which they
will\n# connect by default. Any user may use \pg-wrapper -c\ to set a new\n#
default to be stored in ~/.postgresqlrc and the existence...
(gdb) p ptr
$2 = 0x1 Address 0x1 out of bounds
(gdb) s
strchr () at ../sysdeps/i386/strchr.S:37
warning: Source file is more recent than executable.
37 pushl %edi /* Save callee-safe registers used here. */
Current language: auto; currently asm
(gdb) si
38 movl STR(%esp), %eax
(gdb) si
39 movl CHR(%esp), %edx
(gdb)
46 movb %dl, %dh /* now it is 0|0|c|c */
(gdb)
47 movl %edx, %ecx
(gdb)
48 shll $16, %edx /* now it is c|c|0|0 */
(gdb)
49 movw %cx, %dx /* and finally c|c|c|c */
(gdb)
62 testb $3, %al /* correctly aligned ? */
(gdb) p $cx
$3 = void
(gdb) p $dx
$4 = void
(gdb) p $3
$5 = void
(gdb) p $al
$6 = void
(gdb) si
63 jz L(11)/* yes = begin loop */
(gdb) si
124 L(11): movl (%eax), %ecx /* get word (= 4 bytes) in question */
(gdb) p $eax
$7 = 1075097600
(gdb) p $ecx
$8 = 2570
(gdb) si
Program received signal SIGSEGV, Segmentation fault.
strchr () at ../sysdeps/i386/strchr.S:124
124 L(11): movl (%eax), %ecx /* get word (= 4 bytes) in question */
(gdb)
This is the code of the relevant routine; I can give you the whole lot if
you need it:
/
Write a line in user_clusters
/
void write_cluster_line(const char *user, const char *group,
const char *cluster, const char *db,
const int forceflag) {
FILE *ucf;
char *filename = /etc/postgresql/user_clusters;
char *buf, *curpos;
char *ptr;
char in[BUFSIZ];
char c_user[BUFSIZ], c_group[BUFSIZ], junk[BUFSIZ];
char line[BUFSIZ];
int c, found = 0, set_default = 0;
/* create the line to write */
snprintf(line, BUFSIZ, %s \t%s \t%s \t%s \t%s\n,
user, group, cluster,
(forceflag ? yes : no),
(db[0] == '\0' ? * : db));
line[BUFSIZ - 1] = '\0';
/* start of user_clusters - open read-write */
sprintf(buf, Failed to open %s for writing, filename);
if (!(ucf = fopen(filename, r+))) {
/* try creating it */
if (!(ucf = fopen(filename, w))) {
perror(buf);
exit(LOC_ERR_WRITE_FAIL);
}
/* set permissions and ownership */
if (fchmod(fileno(ucf), S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH)) {
perror (Could not set perms on user_clusters);
exit(LOC_ERR_WRITE_FAIL);
}
fputs(line, ucf);
if (!geteuid()) {
/* we are root - set ownership to postgres.postgres */
struct passwd *pwd;
pwd = getpwnam(postgres);
fchown(fileno(ucf), pwd-pw_uid, pwd-pw_gid);
}
fclose(ucf);
return;
}
/* Read it into memory */
buf = 0;
if ((c = fseek(ucf, 0, SEEK_END)) == -1) {
perror(Error seeking in user_clusters);
exit(LOC_ERR_READ_FAIL);
}
if ((c = ftell(ucf)) == -1) {
perror(Error doing ftell() in user_clusters);
exit(LOC_ERR_READ_FAIL);
}
/* map the file and load an extra page in case the new line expands the
file across the page boundary; adding 2 allows for the truncating
effect of integer division.
Bug#222598: thread_info.h:49: global register variable follows a function definition
Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Severity: normal During my work on the alpha boot loader aboot, I stumbled on the following problem: [EMAIL PROTECTED]:~/prog/debian/ABOOT/aboot-0.9b-trisplit/lib$ env LANG=C make gcc -Os -Wall -mno-fp-regs -fno-builtin -D__KERNEL__ -I../include -c -o isolib.o isolib.c In file included from /usr/include/linux/thread_info.h:21, from /usr/include/linux/spinlock.h:12, from /usr/include/linux/time.h:27, from /usr/include/linux/stat.h:57, from isolib.c:15: /usr/include/asm/thread_info.h:49: global register variable follows a function definition /usr/include/asm/thread_info.h:49: warning: call-clobbered register used for global register variable make: *** [isolib.o] Error 1 This looks to me like an error in thread_info.h. If you like, I can provide you with the files, but more easily install the aboot sources, go into the lib/ directory and run make. (On alpha only currently) -- System Information Debian Release: testing/unstable Architecture: alpha Kernel: Linux sixtyfor 2.4.21-pre1 #1 Don Okt 2 18:27:25 CEST 2003 alpha Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] -- Helge Kreutzmann, Dipl.-Phys. [EMAIL PROTECTED] gpg signed mail preferredgpg-key: finger [EMAIL PROTECTED] 64bit GNU powered http://www.itp.uni-hannover.de/~kreutzm Help keep free software libre: http://www.freepatents.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#67921: Brand New Marketing
Our New Revolutionary Duplication System can generate you One Hundred Thousand Dollars per month. Starting within 24 hours. And who generates that kind of money? Elite athletes? Top musicians Privileged TV stars? HOW ABOUT YOU? With our Duplication system, you could generate more money than any of them! We've designed an income generating system THE WORLD'S NEVER SEEN ! It's New, It's Never been seen, and it works ! Here it is right below. Presentation: http://www.catchthecurrent.net/?id=20427 OR Call 800-800-8000 Tim generated over twenty seven thousand dollars in just fifty six hours. Would that help? That was last week using the brand new presentation below. Seeing is believing so take a look at this now. If your are on dial up be patient as this will take a few minutes to download. This can change your life now. Presentation: http://www.catchthecurrent.net/?id=20427 OR Call 800-800-8000 With this program everyone can achieve unlimited success and no one is left behind. Meet you at the top. PS: What kind of a car are you driving? That's about to change! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#219940: Please fix this trivial bug, it cause packages to FTBFS.
severity 219940 serious thanks Raising this bug report, since it seems to be the only way to get attention to this. There are already two packages which seem to FTBFS because of this smallest of changes, and the proposed fix has also been discussed on LKML, altough i am not sure if positive action was taken or not about this, downloading -test11 to see if it was fixed there or not. That said, even if the proposed fix is not the right way of solving this, it would be nice to have at least feedback from the glibc maintainers on this issue. And parted FTBFS because of this, and i am thus sitting on a NMU upload of parted since almost a month. And parted is used by the debian-installer, which in turn is needed to make the sarge release a reality. So, please, as aj said, this is a small thing, so let's get it out of the way and fix it, so we can all go and work on other things, or if there is a problem with it, temm it, so i can go and investigate more, and maybe find a better solution. Friendly, Sven Luther -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222514: libc6-dev: restrict_arr should be #define'd empty if __STRICT_ANSI__ is defined
Package: libc6-dev Version: 2.3.2.ds1-10 Severity: minor ANSI/ISO C90 does not know the restrict keyword for arrays. In sys/cdefs.h it should be checked for the __STRICT_ANSI__ macro. Line 245: #if __GNUC_PREREQ (3,1) !defined __GNUG__ Should be: #if __GNUC_PREREQ (3,1) !defined __GNUG__ !defined __STRICT_ANSI__ -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux wwid 2.4.21-3-k7 #1 Sun Jul 20 19:23:36 EST 2003 i686 Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] (ignored: LC_ALL set to [EMAIL PROTECTED]) Versions of packages libc6-dev depends on: ii libc6 2.3.2.ds1-10 GNU C Library: Shared libraries an ii linux-kernel-headers 2.5.999-test7-bk-9 Linux Kernel Headers for developme -- no debconf information -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222777: /usr/include/asm/byteorder.h:56: error: parse error before ':' token
Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Severity: normal Compiling KDE from CVS, i got the following message: In file included from /usr/include/linux/cdrom.h:14 from plat_linux_cdda.c:43: /usr/include/asm/byteorder.h: In function '___arch__swab64': /usr/include/asm/byteorder.h:56: error: parse error before ':' token The relevant part from plat_linux_cdda.c is: #ifndef __GNUC__ #define __GNUC__ 1 #endif #undef __STRICT_ANSI__ #include asm/types.h /* ugly workaround for broken glibc shipped in SuSE 9.0 */ #define inline __inline__ #include linux/cdrom.h #undef inline And the referenced (buggy?) line from byteorder.h: asm(xchgl %0,%1 : =r (v.s.a), =r (v.s.b) : 0 (v.s.a),1 (v.s.b)); -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux fenix 2.6.0-test9leo #1 Thu Nov 13 14:06:27 CLST 2003 i686 Locale: LANG=es_CL, LC_CTYPE=es_CL (ignored: LC_ALL set to C) -- no debconf information -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#221401: libc6: Not able to load libglx.so (nvidia)
Package: libc6 Version: 2.3.2.ds1-10 Severity: normal Followup-For: Bug #221401 -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Same, with www.minion.de patches. But xfree server doesn't load the GLX driver: - /var/log/XFree86.0.log- [...] (II) LoadModule: glx (II) Loading /usr/X11R6/lib/modules/extensions/libglx.so dlopen: /usr/lib/libGLcore.so.1: undefined symbol: __gl_tls_var0 (EE) Failed to load /usr/X11R6/lib/modules/extensions/libglx.so (II) UnloadModule: glx (EE) Failed to load module glx (loader failed, 135874371) [...continues and xfree runs, without glx...]] - --- - -- System Information: Debian Release: testing/unstable Architecture: i386 Kernel: Linux 2.6.0-test11 #2 Wed Dec 3 17:53:03 CET 2003 i686 Locale: LANG=C, LC_CTYPE=C (ignored: LC_ALL set to [EMAIL PROTECTED]) Versions of packages libc6 depends on: ii libdb1-compat 2.1.3-7The Berkeley database routines [gl - -- no debconf information -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.3 (GNU/Linux) iD8DBQE/zlgjAzbsrEhTuckRAl8AAKCAYGzpZ+rNGMfks/3MNBA88l5qsgCeIjkU ACm3eAvk3FXUEvNwkMGZL5I= =kn08 -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222777: /usr/include/asm/byteorder.h:56: error: parse error before ':' token
On Wed, Dec 03, 2003 at 12:31:51AM -0300, Leonardo Soto wrote: Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Severity: normal Compiling KDE from CVS, i got the following message: In file included from /usr/include/linux/cdrom.h:14 from plat_linux_cdda.c:43: /usr/include/asm/byteorder.h: In function '___arch__swab64': /usr/include/asm/byteorder.h:56: error: parse error before ':' token The relevant part from plat_linux_cdda.c is: #ifndef __GNUC__ #define __GNUC__ 1 #endif #undef __STRICT_ANSI__ #include asm/types.h /* ugly workaround for broken glibc shipped in SuSE 9.0 */ #define inline __inline__ #include linux/cdrom.h #undef inline And the referenced (buggy?) line from byteorder.h: asm(xchgl %0,%1 : =r (v.s.a), =r (v.s.b) : 0 (v.s.a),1 (v.s.b)); This is an absolutely, wildly, crazily dumb way to work around a broken header. Remove the #undef, and either linux-kernel-headers will work or the next version uploaded will when uploads are re-enabled. I believe Debian kde packages have already removed the #undef. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
bug report : error: variable or field `__user' declared void
Package: linux-kernel-headers Version: 2.5.999-test7-bk-9 Hello, I look the development of driver eagle for my adsl modem. And we notice an error during compilation on Debian. I think I found the problem. /usr/include/linux/usbdevice_fs.h use __user but the definition of __user is not include automaticaly (there is in linux/compiler.h). Is it a bug? Or must we include linux/compiler.h before linux/usbdevice_fs.h ? Here an example if my explanation is too bad :-). [EMAIL PROTECTED]:/tmp/ -su2.05b$ echo #include linux/usbdevice_fs.h test.c [EMAIL PROTECTED]:/tmp/ -su2.05b$ gcc -D__KERNEL__ test.c In file included from test.c:1: /usr/include/linux/usbdevice_fs.h:49: error: variable or field `__user' declared void ... Excuse my poor english. mcoolive. _ Trouvez l'âme soeur sur MSN Rencontres http://g.msn.fr/FR1000/9551 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222777: /usr/include/asm/byteorder.h:56: error: parse error before ':' token
El mié, 03-12-2003 a las 20:52, Daniel Jacobowitz escribió: [...] The relevant part from plat_linux_cdda.c is: #ifndef __GNUC__ #define __GNUC__ 1 #endif #undef __STRICT_ANSI__ #include asm/types.h /* ugly workaround for broken glibc shipped in SuSE 9.0 */ #define inline __inline__ #include linux/cdrom.h #undef inline And the referenced (buggy?) line from byteorder.h: asm(xchgl %0,%1 : =r (v.s.a), =r (v.s.b) : 0 (v.s.a),1 (v.s.b)); This is an absolutely, wildly, crazily dumb way to work around a broken header. Remove the #undef, and either linux-kernel-headers will work or the next version uploaded will when uploads are re-enabled. OK, undef removed and kscd compiles. I didn't remove it before submit the bug because I fear collateral troubles... Thanks you. -- Leonardo Soto M. [EMAIL PROTECTED] Jabber: [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#222777: /usr/include/asm/byteorder.h:56: error: parse error before ':' token
El mié, 03-12-2003 a las 20:52, Daniel Jacobowitz escribió: [...] The relevant part from plat_linux_cdda.c is: #ifndef __GNUC__ #define __GNUC__ 1 #endif #undef __STRICT_ANSI__ #include asm/types.h /* ugly workaround for broken glibc shipped in SuSE 9.0 */ #define inline __inline__ #include linux/cdrom.h #undef inline And the referenced (buggy?) line from byteorder.h: asm(xchgl %0,%1 : =r (v.s.a), =r (v.s.b) : 0 (v.s.a),1 (v.s.b)); This is an absolutely, wildly, crazily dumb way to work around a broken header. Remove the #undef, and either linux-kernel-headers will work or the next version uploaded will when uploads are re-enabled. OK, undef removed and kscd compiles. I didn't remove it before submit the bug because I fear collateral troubles... Thanks you. -- Leonardo Soto M. [EMAIL PROTECTED] Jabber: [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
