- Arch-dev-public - lists.archlinux.org

[arch-dev-public] [signoff] fakeroot-1.14.3-1
by Allan McRae 11 Nov '09

11 Nov '09

Upstream update. Signoff both, Allan

5 8

[arch-dev-public] [PATCH] rc.sysinit: background hwclock calls
by Dan McGee 10 Nov '09

10 Nov '09

hwclock calls appear to block somewhere between 1 and 2 seconds when we have back-to-back calls. My theory (without looking at the code) is that hwclock has to synchronize to the 1 second intervals of the hardware clock, so it can sometimes take up to a second to complete. To get around this unpleasant behavior, we can background the calls at point X in the boot sequence, and then later at point Y in the script (when we absolutely need the clock actions to be complete), we wait on the subprocess. This allows the rest of the boot sequence, after both hwclock code blocks, to continue until the point where we wait on the subprocess. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- For reference, but didn't want to include them in the patch as they will probably disappear: Before on the Eee: http://www.toofishes.net/uploads/bootchart-2.png After: http://www.toofishes.net/uploads/bootchart-3.png Note that there is no longer a huge section of time where the CPU is going unused; this was the second hwclock block before. Also note that rc.multi starts at ~8.5 secs where as before it was ~11 secs. -Dan rc.sysinit | 19 ++++++++++++++++++- 1 files changed, 18 insertions(+), 1 deletions(-) diff --git a/rc.sysinit b/rc.sysinit index 0c934c7..524b2cd 100755 --- a/rc.sysinit +++ b/rc.sysinit @@ -36,6 +36,7 @@ else fi HWCLOCK_PARAMS="--hctosys" +clock_pid="" if [ "$HARDWARECLOCK" = "UTC" ]; then HWCLOCK_PARAMS="$HWCLOCK_PARAMS --utc" elif [ "$HARDWARECLOCK" = "localtime" ]; then @@ -57,7 +58,10 @@ if [ -n "$HWCLOCK_PARAMS" ]; then # Set clock early to fix some bugs with filesystem checks # Clock is set again later to match rc.conf if [ -f /etc/localtime ]; then + ( /sbin/hwclock $HWCLOCK_PARAMS --noadjfile + ) & + clock_pid=$! fi fi @@ -111,6 +115,11 @@ if [ -d /sys/class/net/lo ]; then fi fi +# preliminary hwclock setting needs to be done at this point +if [ -n "$clock_pid" ]; then + wait $clock_pid +fi + # If necessary, find md devices and manually assemble RAID arrays if [ -f /etc/mdadm.conf -a "$(/bin/grep ^ARRAY /etc/mdadm.conf 2>/dev/null)" ]; then status "Activating RAID arrays" /sbin/mdadm --assemble --scan @@ -282,8 +291,11 @@ if [ "$TIMEZONE" != "" -a -e "/usr/share/zoneinfo/$TIMEZONE" ]; then fi if [ -n "$HWCLOCK_PARAMS" ]; then + ( /sbin/hwclock --adjust #Adjust for system drift /sbin/hwclock $HWCLOCK_PARAMS + ) & + clock_pid=$! fi stat_done @@ -396,7 +408,12 @@ fi /bin/dmesg >| /var/log/dmesg.log +# final hwclock setting needs to be done at this point +if [ -n "$clock_pid" ]; then + wait $clock_pid +fi + run_hook sysinit_end # End of file -# vim: set ts=2 noet: +# vim: set ts=2 sw=2 noet: -- 1.6.5.2

1 0

[arch-dev-public] [signoff] glibc-2.11 toolchain rebuild
by Allan McRae 10 Nov '09

10 Nov '09

I have put the glibc-2.11 toolchain rebuild in [testing]. Apart from glibc-2.11, the kernel header patch level got bumped to the latest version and I took a newer snapshot of the binutils-2.20 branch. From a quick look at the changelog there should not be a lot of issues compiling stuff like previous glibc/gcc releases. As always, it has be well tested in Fedora and you should look there for any patches needed. Signoff both, Allan

4 7

[arch-dev-public] Upstream bug closures
by Jan de Groot 10 Nov '09

10 Nov '09

I see a lot of bugs getting closed with "Upstream" lately because they're not packaging bugs. This is not the way to solve bugs. The only bugs that should be closed upstream are the ones in binary modules like flashplugin or nvidia binary drivers. Opensource software can be fixed or debugged, so we should do that instead of using this bogus closure option.

6 6

Re: [arch-dev-public] Arch Linux: comment about Official Arch Linux Install Guide
by Dieter Plaetinck 09 Nov '09

09 Nov '09

On Wed, 4 Nov 2009 09:04:40 +0900 Hiz Japan <hizdiez(a)gmail.com> wrote: > Dear > > First of all, I am very appreciated and respecting your great > contribute of the Official Arch Linux Install Guide. > I am wornking on your "Official Arch Linux Install Guide" to > translating Japanese language. > and i have question/opinion about .... > > the section Pacman explanation. it says: > > - > > Display specific not installed package info > > # pacman --sync --info foo > > # pacman -Si foo > > Displays information on the not yet installed package foo (size, > install date, build date, dependencies, conflicts, etc.) > > > I feel it is better say like this? because this explanation is not > very clear what actually trying to explain and correctness . > > >> > > - > > Display specific package info from repository database > > # pacman --sync --info foo > > # pacman -Si foo > > Display specific package information details foo from repository > database . (size,* *, build date, dependencies, conflicts, > etc.) > > > *pacman -Si also display package information which already installed.* > *pacman -Si doesnt contain "install date". "install date" info > contained by the command pacman -Qi* . > > > > Best regard > Hiz Good point! Thanks for the idea. http://projects.archlinux.org/aif.git/commit/?h=experimental&id=0b4916b8b07… Dieter

1 0

[arch-dev-public] [signoff] bzip2-1.0.5-5
by Allan McRae 09 Nov '09

09 Nov '09

Added upstream tools to package - FS#14521 Signoff both, Allan

3 2

[arch-dev-public] [signoff] sqlite3 - 3.6.20-1
by Andreas Radke 09 Nov '09

09 Nov '09

next upstream release. please signoff. -Andy SQLite Release 3.6.20 On 2009 Nov 4 (3.6.20) Changes associated with this release include the following: * Optimizer enhancement: prepared statements are automatically re-compiled when a binding on the RHS of a LIKE operator changes or when any range constraint changes under SQLITE_ENABLE_STAT2. * Various minor bug fixes and documentation enhancements.

2 1

[arch-dev-public] The docs/!docs thing again
by Dan McGee 09 Nov '09

09 Nov '09

What are we doing this month with docs? I ask because I was rather shocked to see this: dmcgee@kilkenny /usr/share/doc/gtkmm-2.4 $ pacman -Si gtkmm | grep Size Download Size : 14396.06 K Installed Size : 61660.00 K dmcgee@kilkenny /usr/share/doc/gtkmm-2.4 $ du -sh . 50M . That means ~80% of this package is stuff almost no-one is ever going to look at. That seems a bit overkill to me to include in our packages. -Dan

5 10

[arch-dev-public] [signoff] openssl 0.9.8l-1
by Pierre Schmitz 08 Nov '09

08 Nov '09

Moin, you might have heard from the possible MTM attack against TLS. Openssl has released a new version which disabled the affected renegotiation feature. We should move this to core soon. For more information see http://extendedsubset.com/?p=8 and https://bugzilla.redhat.com/show_bug.cgi?id=CVE-2009-3555 Please note that this is more or less a protocol design flaw which means that every SSL implementation should be affected, not only openssl (e.g. Firefox uses nss and there is also gnutls). So we should have a look at those packages, too. Pierre -- Pierre Schmitz, https://users.archlinux.de/~pierre

6 6

[arch-dev-public] [signoff] mdadm-3.0.3-1
by Tobias Powalowski 07 Nov '09

07 Nov '09

Hi downgrade in testing repository due to potential dataloss problems: http://marc.info/?l=linux-raid&m=125748985908870&w=2 Changelog: http://neil.brown.name/git?p=mdadm;a=blob_plain;f=ANNOUNCE-3.0;hb=7f0066ba7… Please update and recreate your initcpio after upgrading. One thing i observed, when using a custom /etc/mdadm.conf file our rc.sysinit will run mdadm --assemble --scan, which will always give exit code 2, due to devices busy. greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa(a)archlinux.org

1 0