Michal Schmidt [Thu, 26 May 2011 23:29:34 +0000 (01:29 +0200)]
pam-module: add debug= parameter
It is customary that pam modules do not log debugging information by default.
Usually they offer a 'debug' option.
Add a boolean debug= option to pam_systemd.so.
This will solve bug
https://bugzilla.redhat.com/show_bug.cgi?id=705427
Commit 53d5582fa006b0eb528f5dc3f4ba978abd8ac5a3 was not sufficient to fix it,
because in Fedora rsyslog is configured to write even LOG_DEBUG messages to
/var/log/secure by default.
Tom Gundersen [Sat, 11 Jun 2011 13:55:11 +0000 (15:55 +0200)]
remote-fs.target: do not order after network.target
remote-fs.target is ordered after the {auto,}mount units. In case of automount
we do not want to wait for the network to come up before proceeding. In case
of a regular mount unit, the unit will be ordered after network.target
so the behavior is unchanged.
This speeds up boot quite a bit for me when having some services needing
NetworkManager-wait-online.service, and having my home partition on nfs
under an automountpoint.
Michal Schmidt [Fri, 10 Jun 2011 12:37:21 +0000 (14:37 +0200)]
readahead-common: fix total memory size detection
sysinfo returns the total memory size in multiples of mem_unit bytes.
As long as the size in bytes fits into unsigned long, the kernel uses
mem_unit = 1, but this is not true on i386 with more than 4 GB RAM.
Michal Schmidt [Mon, 6 Jun 2011 20:59:19 +0000 (22:59 +0200)]
cryptsetup-generator: fix /etc/cryptsetup options
cryptsetup-generator parses the options in /etc/cryptsetup incorrectly.
It fails to find the 'swap' option in
swap,foo
and instead it matches on
swaplalala,foo
The condition for the comma separator is reversed.
Kay Sievers [Wed, 25 May 2011 16:21:36 +0000 (18:21 +0200)]
drop hwclock-save.service
We don't want to fiddle around changing the RTC, not on bootup, not
on shutdown.
If we don't run NTP, we have absolutely no clue what's the current
time to store in the RTC. If we run NTP, the kernel syncs the system
time every 11 minutes to the RTC.
Especially in multi-boot environents we must not call hwclock(8)
which tries to be smart with calculating/storing/applying drifts
and such.
Live-CDs must never touch the RTC, because we don't know if it is
running in UTC or locatime.
Michael Olbrich [Wed, 4 May 2011 09:01:54 +0000 (11:01 +0200)]
build-sys: create dbussystemservicedir
otherwise building fails if it doesn't exist:
( cd <DESTDIR>/usr/share/dbus-1/services && \
rm -f org.freedesktop.systemd1.service && \
ln -s ../system-services/org.freedesktop.systemd1.service org.freedesktop.systemd1.service )
/bin/sh: line 0: cd: <DESTDIR>/usr/share/dbus-1/services: No such file or directory
Paul Menzel [Mon, 2 May 2011 12:13:08 +0000 (14:13 +0200)]
util: include `stdarg.h`
Otherwise using uClibc compilation (OpenEmbedded [1] with `minimal-uclibc` for `MACHINE = "beagleboard"`) fails with the following error message.
| make[1]: Entering directory `/oe/build-minimal-uclibc/minimal-uclibc-dev/work/armv7a-oe-linux-uclibceabi/systemd-v26-r0/git'
| CC src/libsystemd_basic_la-util.lo
| src/util.c: In function 'acquire_terminal':
| src/util.c:2484:29: warning: cast increases required alignment of target type [-Wcast-align]
| src/util.c:2497:37: warning: cast increases required alignment of target type [-Wcast-align]
| src/util.c: In function 'rename_process':
| src/util.c:2905:17: warning: passing argument 1 of 'strncpy' discards qualifiers from pointer target type
| /oe/build-minimal-uclibc/minimal-uclibc-dev/sysroots/armv7a-oe-linux-uclibceabi/usr/include/string.h:87:14: note: expected 'char * restrict' but argument is of type 'const char *'
| src/util.c: In function 'cpu_set_malloc':
| src/util.c:3229:17: warning: implicit declaration of function 'CPU_ALLOC' [-Wimplicit-function-declaration]
| src/util.c:3229:25: warning: assignment makes pointer from integer without a cast
| src/util.c:3232:17: warning: implicit declaration of function 'CPU_ALLOC_SIZE' [-Wimplicit-function-declaration]
| src/util.c:3233:25: warning: implicit declaration of function 'CPU_ZERO_S' [-Wimplicit-function-declaration]
| src/util.c:3241:17: warning: implicit declaration of function 'CPU_FREE' [-Wimplicit-function-declaration]
| CC src/libsystemd_basic_la-label.lo
| In file included from src/label.c:27:0:
| src/util.h:358:41: error: expected declaration specifiers or '...' before 'va_list'
| make[1]: *** [src/libsystemd_basic_la-label.lo] Error 1
| make[1]: Leaving directory `/oe/build-minimal-uclibc/minimal-uclibc-dev/work/armv7a-oe-linux-uclibceabi/systemd-v26-r0/git'
| make: *** [all] Error 2
| + die 'oe_runmake failed'
| + oefatal 'oe_runmake failed'
| + echo FATAL: 'oe_runmake failed'
| FATAL: oe_runmake failed
| + exit 1
NOTE: package systemd-v26-r0: task do_compile: Failed
ERROR: Function 'do_compile' failed (see /oe/build-minimal-uclibc/minimal-uclibc-dev/work/armv7a-oe-linux-uclibceabi/systemd-v26-r0/temp/log.do_compile.4020 for further information)
ERROR: Task 13 (/oe/openembedded/recipes/systemd/systemd_git.bb, do_compile) failed with exit code '1'
ERROR: '/oe/openembedded/recipes/systemd/systemd_git.bb' failed
We simply keep open copies of the dbus connections across the
reexecution and close them as last step of it. A client can thus simply
wait until its connection is dropped to know when the reexecution is
finished.
missing: Fix broken syscall(__NR_fanotify_mark... on o32 mips.
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=%23618562
MIPS o32 ABI enforces that 64bit arguments should be 64bit-aligned, and
the third argument of syscall(__NR_fanotify_mark, ...) is 64bit and not
64bit-aligned on o32 mips platform, thus extra padding is inserted
before it. The syscall function doesn't know the prototype of
fanotify_mark, so the padding will be passed into kernel as valid
argument.