Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
on Sat, Nov 08, 2003 at 08:43:07PM -0800, Jeff Bailey wrote: I was hoping to wait until BenC (or another Sparc porter) came up with the solution for the sparc64 borkage. Okay. I've uploaded 2.3.2.ds1-10.0.1 as a binary NMU. It seems to fix the modutils problem. p.
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
Package: libc6
Version: 2.3.2.ds1-10
Severity: grave
Arch: ARM
After upgrading libc6 on an ARM-based handheld running sarge to
sid, all programs in modutils (lsmod, for instance), issue an error
lsmod: QM_MODULES: Function not implemented and exit when run as
root. When run as non-root, lsmod runs expected, the other commands
(modprobe, insmod) fail with the QM_MODULES error message.
Here is the output of strace lsmod when run as root:
execve(/sbin/lsmod, [lsmod], [/* 15 vars */]) = 0
uname({sys=Linux, node=cirrus, ...}) = 0
brk(0) = 0x28494
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or
directory)
open(/etc/ld.so.preload, O_RDONLY)= -1 ENOENT (No such file or
directory)
open(/etc/ld.so.cache, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=11129, ...}) = 0
old_mmap(NULL, 11129, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40015000
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or
directory)
open(/lib/libc.so.6, O_RDONLY)= 3
read(3, \177ELF\1\1\1a\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0(t\1\0004..., 512) =
512
fstat64(3, {st_mode=S_IFREG|0644, st_size=1166132, ...}) = 0
old_mmap(NULL, 1202940, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4001e000
old_mmap(0x40136000, 49152, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x11) = 0x40136000
old_mmap(0x40142000, 6908, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40142000
close(3)= 0
munmap(0x40015000, 11129) = 0
brk(0) = 0x28494
brk(0x49494)= 0x49494
brk(0) = 0x49494
brk(0x4a000)= 0x4a000
open(/proc/sys/kernel/tainted, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0)
= 0x40015000
read(3, 1\n, 4096)= 2
close(3)= 0
munmap(0x40015000, 4096)= 0
brk(0) = 0x4a000
brk(0) = 0x4a000
brk(0x49000)= 0x49000
brk(0) = 0x49000
fstat64(1, {st_mode=S_IFREG|0644, st_size=1891, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0)
= 0x40015000
getuid32() = 0
write(2, lsmod: , 7lsmod: ) = 7
write(2, QM_MODULES: Function not impleme..., 37QM_MODULES: Function not
implemented
) = 37
write(2, \n, 1
) = 1
write(1, Module Size Us..., 54Module
Size Used byTainted: P
) = 54
munmap(0x40015000, 4096)= 0
io_submit(0x1, 0, 0xfbad2088) = -1 ENOSYS (Function not implemented)
_exit(1)= ?
Here is the same for a non-root user (uid 1000):
execve(/sbin/lsmod, [/sbin/lsmod], [/* 14 vars */]) = 0
uname({sys=Linux, node=cirrus, ...}) = 0
brk(0) = 0x28494
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or
directory)
open(/etc/ld.so.preload, O_RDONLY)= -1 ENOENT (No such file or
directory)
open(/etc/ld.so.cache, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=11129, ...}) = 0
old_mmap(NULL, 11129, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40015000
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or
directory)
open(/lib/libc.so.6, O_RDONLY)= 3
read(3, \177ELF\1\1\1a\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0(t\1\0004..., 512) =
512
fstat64(3, {st_mode=S_IFREG|0644, st_size=1166132, ...}) = 0
old_mmap(NULL, 1202940, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4001e000
old_mmap(0x40136000, 49152, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x11) = 0x40136000
old_mmap(0x40142000, 6908, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40142000
close(3)= 0
munmap(0x40015000, 11129) = 0
brk(0) = 0x28494
brk(0x49494)= 0x49494
brk(0) = 0x49494
brk(0x4a000)= 0x4a000
open(/proc/sys/kernel/tainted, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0)
= 0x40015000
read(3, 1\n, 4096)= 2
close(3)= 0
munmap(0x40015000, 4096)= 0
brk(0) = 0x4a000
brk(0) = 0x4a000
brk(0x49000)= 0x49000
brk(0) = 0x49000
fstat64(1,
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sat, 2003-11-08 at 21:14, Major A wrote: After upgrading libc6 on an ARM-based handheld running sarge to sid, all programs in modutils (lsmod, for instance), issue an error lsmod: QM_MODULES: Function not implemented and exit when run as root. When run as non-root, lsmod runs expected, the other commands (modprobe, insmod) fail with the QM_MODULES error message. Argh! It seems that the definition of __NR_query_module has been removed from asm/unistd.h in kernel 2.6. So, glibc is just providing a stub for that function which simply returns ENOSYS. The reason lsmod works as non-root is that it falls back to parsing /proc/modules in that case. p.
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, Nov 09, 2003 at 12:18:23AM +, Phil Blundell wrote: On Sat, 2003-11-08 at 21:14, Major A wrote: After upgrading libc6 on an ARM-based handheld running sarge to sid, all programs in modutils (lsmod, for instance), issue an error lsmod: QM_MODULES: Function not implemented and exit when run as root. When run as non-root, lsmod runs expected, the other commands (modprobe, insmod) fail with the QM_MODULES error message. Argh! It seems that the definition of __NR_query_module has been removed from asm/unistd.h in kernel 2.6. So, glibc is just providing a stub for that function which simply returns ENOSYS. The reason lsmod works as non-root is that it falls back to parsing /proc/modules in that case. I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, Nov 09, 2003 at 12:36:50AM +, Phil Blundell wrote: On Sun, 2003-11-09 at 00:33, Daniel Jacobowitz wrote: I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? Sure. In fact, I just checked in a patch for that, and was on the point of preparing an upload. But I guess I'll hang back and let you upload once you've taken care of ia64. Oh, I missed that :) Go right ahead. I meant to say I've already fixed this for ia64 in the last release, so your analysis is definitely right. Note that you need to bump glibc's build dependencies for this. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, 2003-11-09 at 00:33, Daniel Jacobowitz wrote: I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? Sure. In fact, I just checked in a patch for that, and was on the point of preparing an upload. But I guess I'll hang back and let you upload once you've taken care of ia64. p.
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, 2003-11-09 at 00:33, Daniel Jacobowitz wrote: I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? Ah, maybe I read too much into the present tense there. I guess I'll do my upload after all. p.
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, 2003-11-09 at 00:38, Daniel Jacobowitz wrote: Go right ahead. I meant to say I've already fixed this for ia64 in the last release, so your analysis is definitely right. Note that you need to bump glibc's build dependencies for this. Right, done. Are we planning another source upload of glibc in the near future? If not, I can make a binary NMU for arm with the fixed headers. p.
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, Nov 09, 2003 at 12:54:18AM +, Philip Blundell wrote: Go right ahead. I meant to say I've already fixed this for ia64 in the last release, so your analysis is definitely right. Note that you need to bump glibc's build dependencies for this. Right, done. Are we planning another source upload of glibc in the near future? If not, I can make a binary NMU for arm with the fixed headers. I was hoping to wait until BenC (or another Sparc porter) came up with the solution for the sparc64 borkage. Besides, it would probably be a good test to do a binary NMU with the new packaging to make sure all the dependancies do the right thing. Tks, Jeff Bailey
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
Package: libc6
Version: 2.3.2.ds1-10
Severity: grave
Arch: ARM
After upgrading libc6 on an ARM-based handheld running sarge to
sid, all programs in modutils (lsmod, for instance), issue an error
lsmod: QM_MODULES: Function not implemented and exit when run as
root. When run as non-root, lsmod runs expected, the other commands
(modprobe, insmod) fail with the QM_MODULES error message.
Here is the output of strace lsmod when run as root:
execve(/sbin/lsmod, [lsmod], [/* 15 vars */]) = 0
uname({sys=Linux, node=cirrus, ...}) = 0
brk(0) = 0x28494
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or directory)
open(/etc/ld.so.preload, O_RDONLY)= -1 ENOENT (No such file or directory)
open(/etc/ld.so.cache, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=11129, ...}) = 0
old_mmap(NULL, 11129, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40015000
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or directory)
open(/lib/libc.so.6, O_RDONLY)= 3
read(3, \177ELF\1\1\1a\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0(t\1\0004..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=1166132, ...}) = 0
old_mmap(NULL, 1202940, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4001e000
old_mmap(0x40136000, 49152, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x11) = 0x40136000
old_mmap(0x40142000, 6908, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40142000
close(3)= 0
munmap(0x40015000, 11129) = 0
brk(0) = 0x28494
brk(0x49494)= 0x49494
brk(0) = 0x49494
brk(0x4a000)= 0x4a000
open(/proc/sys/kernel/tainted, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
0x40015000
read(3, 1\n, 4096)= 2
close(3)= 0
munmap(0x40015000, 4096)= 0
brk(0) = 0x4a000
brk(0) = 0x4a000
brk(0x49000)= 0x49000
brk(0) = 0x49000
fstat64(1, {st_mode=S_IFREG|0644, st_size=1891, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
0x40015000
getuid32() = 0
write(2, lsmod: , 7lsmod: ) = 7
write(2, QM_MODULES: Function not impleme..., 37QM_MODULES: Function not
implemented
) = 37
write(2, \n, 1
) = 1
write(1, Module Size Us..., 54Module Size Used
byTainted: P
) = 54
munmap(0x40015000, 4096)= 0
io_submit(0x1, 0, 0xfbad2088) = -1 ENOSYS (Function not implemented)
_exit(1)= ?
Here is the same for a non-root user (uid 1000):
execve(/sbin/lsmod, [/sbin/lsmod], [/* 14 vars */]) = 0
uname({sys=Linux, node=cirrus, ...}) = 0
brk(0) = 0x28494
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or directory)
open(/etc/ld.so.preload, O_RDONLY)= -1 ENOENT (No such file or directory)
open(/etc/ld.so.cache, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=11129, ...}) = 0
old_mmap(NULL, 11129, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40015000
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK) = -1 ENOENT (No such file or directory)
open(/lib/libc.so.6, O_RDONLY)= 3
read(3, \177ELF\1\1\1a\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0(t\1\0004..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=1166132, ...}) = 0
old_mmap(NULL, 1202940, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4001e000
old_mmap(0x40136000, 49152, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x11) = 0x40136000
old_mmap(0x40142000, 6908, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40142000
close(3)= 0
munmap(0x40015000, 11129) = 0
brk(0) = 0x28494
brk(0x49494)= 0x49494
brk(0) = 0x49494
brk(0x4a000)= 0x4a000
open(/proc/sys/kernel/tainted, O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) =
0x40015000
read(3, 1\n, 4096)= 2
close(3)= 0
munmap(0x40015000, 4096)= 0
brk(0) = 0x4a000
brk(0) = 0x4a000
brk(0x49000)= 0x49000
brk(0) = 0x49000
fstat64(1, {st_mode=S_IFREG|0644,
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sat, 2003-11-08 at 21:14, Major A wrote: After upgrading libc6 on an ARM-based handheld running sarge to sid, all programs in modutils (lsmod, for instance), issue an error lsmod: QM_MODULES: Function not implemented and exit when run as root. When run as non-root, lsmod runs expected, the other commands (modprobe, insmod) fail with the QM_MODULES error message. Argh! It seems that the definition of __NR_query_module has been removed from asm/unistd.h in kernel 2.6. So, glibc is just providing a stub for that function which simply returns ENOSYS. The reason lsmod works as non-root is that it falls back to parsing /proc/modules in that case. p. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, Nov 09, 2003 at 12:18:23AM +, Phil Blundell wrote: On Sat, 2003-11-08 at 21:14, Major A wrote: After upgrading libc6 on an ARM-based handheld running sarge to sid, all programs in modutils (lsmod, for instance), issue an error lsmod: QM_MODULES: Function not implemented and exit when run as root. When run as non-root, lsmod runs expected, the other commands (modprobe, insmod) fail with the QM_MODULES error message. Argh! It seems that the definition of __NR_query_module has been removed from asm/unistd.h in kernel 2.6. So, glibc is just providing a stub for that function which simply returns ENOSYS. The reason lsmod works as non-root is that it falls back to parsing /proc/modules in that case. I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, Nov 09, 2003 at 12:36:50AM +, Phil Blundell wrote: On Sun, 2003-11-09 at 00:33, Daniel Jacobowitz wrote: I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? Sure. In fact, I just checked in a patch for that, and was on the point of preparing an upload. But I guess I'll hang back and let you upload once you've taken care of ia64. Oh, I missed that :) Go right ahead. I meant to say I've already fixed this for ia64 in the last release, so your analysis is definitely right. Note that you need to bump glibc's build dependencies for this. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, 2003-11-09 at 00:33, Daniel Jacobowitz wrote: I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? Sure. In fact, I just checked in a patch for that, and was on the point of preparing an upload. But I guess I'll hang back and let you upload once you've taken care of ia64. p. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, 2003-11-09 at 00:33, Daniel Jacobowitz wrote: I'm already fixing this for ia64 in linux-kernel-headers. Mind doing the same for ARM? Ah, maybe I read too much into the present tense there. I guess I'll do my upload after all. p. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#219769: libc6-2.3.2.ds1-10 on ARM breaks modutils
On Sun, 2003-11-09 at 00:38, Daniel Jacobowitz wrote: Go right ahead. I meant to say I've already fixed this for ia64 in the last release, so your analysis is definitely right. Note that you need to bump glibc's build dependencies for this. Right, done. Are we planning another source upload of glibc in the near future? If not, I can make a binary NMU for arm with the fixed headers. p. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
