Re: [gentoo-user]
On Thu, 2007-01-18 at 15:21 -0600, Dan Farrell wrote: [...] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Xgl and direct rendering or 'Would you like Xorg or Xgl, sir?'
On Thu, 2007-01-18 at 23:47 +0100, Jan Stępień wrote: > On 18 Sty, 19:50, "Hemmann, Volker Armin" > <[EMAIL PROTECTED]> wrote: > > nope. It isn't. > > Xgl and direct rendering are exclusive. One or the other. > > Pity. That would be lovely. I'm no expert, but as far as I know, that's the reason why AIGLX succeeded Xgl. So if you ask me, I think AIGLX (if it works with ATI) would be the solution to your problem. Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Gaim upgrade problem
On Thu, 2007-01-18 at 04:08 +0100, Bo Ørsted Andresen wrote: > On Thursday 18 January 2007 04:01, Fredrik Tolf wrote: > > On Thu, 2007-01-18 at 03:48 +0100, Bo Ørsted Andresen wrote: > > > On Thursday 18 January 2007 03:36, Fredrik Tolf wrote: > > > > I've added > > > > `>net-im/gaim-2.0.0_beta5-r3' to /etc/portage/package.mask, but nothing > > > > changes. `emerge gaim' still wishes to emerge gaim-2.0.0_beta5-r3, and > > > > `emerge -Duva world' fails since it cannot build a depgraph including > > > > gaim-2.0.0_beta5-r3. > > > > > > You expect >-r3 to mask -r3? Clearly to mask -r3 you need >=..-r3 or > > > >..-r2. > > > > Sorry, I quoted that wrong. I've tried both >-r2, >=-r3 and =-r3. > > Allright, you're right. package.unmask overrides package.mask. Therefore > you'll have to unmask <-r3. Aha! There was the problem. I had completely forgotten that I had had to p.unmask previously, since it was masked in the tree. > And don't need the mask since it's already masked > in the tree. :) Actually, it seems it isn't any longer, so I could just remove the entry from p.unmask and it worked. Thanks! Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Gaim upgrade problem
On Thu, 2007-01-18 at 03:48 +0100, Bo Ørsted Andresen wrote: > On Thursday 18 January 2007 03:36, Fredrik Tolf wrote: > > I've added > > `>net-im/gaim-2.0.0_beta5-r3' to /etc/portage/package.mask, but nothing > > changes. `emerge gaim' still wishes to emerge gaim-2.0.0_beta5-r3, and > > `emerge -Duva world' fails since it cannot build a depgraph including > > gaim-2.0.0_beta5-r3. > > You expect >-r3 to mask -r3? Clearly to mask -r3 you need >=..-r3 or >..-r2. Sorry, I quoted that wrong. I've tried both >-r2, >=-r3 and =-r3. Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] Gaim upgrade problem
Hi List! I'm having a little problem with package masking that I just can't seem to wrap my head around. See, I emerged the Gaim 2 beta by unmasking it through /etc/portage/package.keywords, and all went fine, until gaim-2.0.0_beta5-r3 came along, which requires a ~x86 dbus, which I don't want (for a variety of obvious reasons). So I thought to continue running with gaim-2.0.0_beta5-r2, which is still in portage. In order to keep `emerge -Duva world' working, I wanted to mask gaim-2.0.0_beta5-r3, though, but there's where my problem is. I've added `>net-im/gaim-2.0.0_beta5-r3' to /etc/portage/package.mask, but nothing changes. `emerge gaim' still wishes to emerge gaim-2.0.0_beta5-r3, and `emerge -Duva world' fails since it cannot build a depgraph including gaim-2.0.0_beta5-r3. I've never had any problems with /etc/portage/package.mask in the past -- even now, it's working to mask >=net-fs/nfs-utils-1.0.8. Does anyone have any idea why it has no effect on gaim? I thought it might be confused by gaim's version numbers or something, but still it seems to be able to see that gaim-2.0.0_beta5-r3 is newer than gaim-2.0.0_beta5-r2, so that can't be it, right? Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Licenses
On Wed, 2007-01-17 at 03:28 +0100, Bo Ørsted Andresen wrote: > On Wednesday 17 January 2007 00:54, Fredrik Tolf wrote: > > I was wondering -- is there a way to find out which of the licenses > > in /usr/portage/licenses can be considered free software licenses > > (without having to read and understand them)? [...] > I don't know of a list which is available now but glep 23 [1] does address > this issue with license groups. There are a number of discussions about this > glep in the archives of the gentoo-dev mailing list and a number of bugs > related to it. I suppose you could try to ask on one of those bugs, irc or > gentoo-dev@ if anyone has a list of what will become the OSI-APPROVED or the > FSF-APPROVED license group. Thanks for the info! I'm in no greater hurry, so I might as well wait until that GLEP becomes implemented. Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] Licenses
Hi List! I was wondering -- is there a way to find out which of the licenses in /usr/portage/licenses can be considered free software licenses (without having to read and understand them)? I'm trying to find out what packages I have installed that cannot be considered free, and while it's easy to get a list of the licenses in use, that in itself is pretty useless without knowing anything about the licenses. Btw., shouldn't portage have some kind of flag to at least warn if a proprietary package is being pulled in through dependencies? I just discovered that I have realplayer installed, which was pulled in by mplayer. Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Gnome 2.16 and FAM
On Wed, 2006-12-13 at 00:09 +0100, Thomas Rösner wrote: > Fredrik Tolf wrote: > > Does anyone know if it's really necessary to use Gamin instead of SGI > > FAM with Gnome 2.16, and, if not, how to make emerge not do that? > > gnome-base/gnome/gnome-2.16.1.ebuild doesnt dep on either of them, and > contains: > [...] Indeed, the gnome ebuild itself does not depend on gamin, but it seems that with the ldap USE flag, it depends on sabayon, and sabayon depends on gamin. Why Sabayon would be related to LDAP, I have no idea whatsoever. I guess this means that I can just remove the ldap USE flag from gnome and it will work, though. Furthermore, it seems that Ekiga is only dep'd with USE=ldap as well. Has anyone an idea why? Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] almost completely OT: mouses
On Tue, 2006-12-12 at 13:05 -0500, Ryan Sims wrote: > Might this also be related to the use of "mouse" as a verb? I.e. > "mouse over the image to see it change," > > I mouse > You mouse > He mouses? > We all.mice? Well, that had me laughing quite a bit. However, are you really sure that the verb would be "to mouse over", rather than "to mouseover"? I would suspect that the usage of that term as a verb has been grabbed from maybe JavaScript or some GUI builder tool, and thus should be verbified as the entire word that it is in its original usage. Of course, one might consider it normal to seperate prepositions from verb stems. Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] Gnome 2.16 and FAM
I just noticed that Gnome 2.16 wants Gamin instead of SGI FAM. I don't, though, since I use NFS and, last I looked, Gamin doesn't support monitoring NFS exports, unlike SGI FAM. Does anyone know if it's really necessary to use Gamin instead of SGI FAM with Gnome 2.16, and, if not, how to make emerge not do that? Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] Emerging world selectively
Here's a portage question that I guess is quite elementary, but I haven't been able to find an answer to it on my own: If, when I emerge -Duva world, it calculates a package that I, for any reason, would like to delay/not emerge at all, how would I do to emerge all the packages except that one. This far, I've been running emerge -1va on all the others in the list (manually, that is), but I'm getting quite tired of that method, and I'm guessing that portage has some way of avoiding it. Thanks for any answers! Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] Tomcat logging problems
Hi list!
I just tried upgrading from Tomcat 5.0 to 5.5 (since the former has been
p.masked since some time), and while it is working, I am experiencing
logging problems.
First of all, it refused to log anything at all, but I fixed that. It
appears to have been a typo in the init script -- it defined
java.util.logging.config.file as ${CATALINA_BASE}/logging.properties,
while it should have been ${CATALINA_BASE}/conf/logging.properties.
However, I fixed that, but found that it won't log the JVM's
std{out,err} anywhere. According to all documentation about Tomcat,
there should be two logfiles called stdout.log and stderr.log placed in
the ${CATALINA_BASE}/logs directory, but there aren't, and I haven't
been able to find out why. Has anyone else encountered this problem and
managed to solve it?
Thanks for your time!
Fredrik Tolf
--
gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Init scripts acting weird with baselayout 1.12
On Thu, 2006-09-07 at 22:46 -0700, Richard Fish wrote: > On 9/7/06, Fredrik Tolf <[EMAIL PROTECTED]> wrote: > > start-stop-daemon -S -p /var/run/cidd.pid -qx /usr/local/sbin/cidd > > Is cidd the actual daemon, or a starter shell script? IIRC, > start-stop-daemon has a problem with starting things that end up with > a different exe name, and thus the --startas option. It is indeed the actual daemon. > You also might try fiddling with the RC_WAIT_ON_START setting in > /etc/conf.d/rc. If the service can't be detected as running 100ms > after starting, then you might have a problem. What does this actually mean? How does it "detect" if it is running? The case with these daemons (as usual) is that they do some initialization and then daemonize using the standard C daemon() call (fork and exit in the parent). What is it that is detected? Thanks for responding! Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] Init scripts acting weird with baselayout 1.12
Ever since I got baselayout upgraded to 1.12, the init scripts that I
have written for a couple of program are acting really weird. What
happens is that they start the program normally, but then, when the
program daemonizes and detaches, it immediately kills it (with a
SIGTERM).
These scripts are also really simple, so I don't understand at all what
I could be doing wrong. Here's a sample:
#!/sbin/runscript
opts="start stop restart"
depend() {
need net
}
start() {
ebegin "Starting cidd"
start-stop-daemon -S -p /var/run/cidd.pid -qx /usr/local/sbin/cidd
eend $?
}
stop() {
ebegin "Stopping cidd"
start-stop-daemon -K -p /var/run/cidd.pid -qx /usr/local/sbin/cidd
eend $?
}
Is there anything wrong with this script that should cause it do act as
I described?
Fredrik Tolf
--
gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Too many X modules pulled in
On Mon, 2006-08-21 at 21:18 +0200, Fredrik Tolf wrote: > On Sun, 2006-08-20 at 11:56 -0700, Donnie Berkholz wrote: > > Fredrik Tolf wrote: > > > I'm not an expert with portage, but the fact that xorg-server is > > > indented one space from gtk+ means that gtk+ depends on it directly, > > > doesn't it? If so, whence does that dependency come from (it's obviously > > > not on the DEPEND or RDEPEND variables)? > > > It's because it inherits the virtualx.eclass -- gtk+ requires an X > > server (Xvfb) to build. What you can do is set USE="minimal -dmx -xorg > > -xprint" in your package.use for xorg-server to minimize this. Also try > > VIDEO_CARDS="" and INPUT_DEVICES="" in make.conf. > > > > However, this only works well on xorg-server-1.1, so you will also need > > to add it to package.keywords. > > I see... how troublesome. Well, I guess there's no choice, but would > anyone happen to know the reason why gtk+ requires Xvfb? Actually, it seems that it doesn't require Xvfb, or any other X server, after all. I just tried emerge gtk+ with USE=-X, and it works perfectly. It didn't emerge any X server, and it works just fine with all the programs I've tried so far. Does that mean that the virtualx eclass is unnecessary, and that I should file a bug report to have it removed? Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Too many X modules pulled in
On Sun, 2006-08-20 at 11:56 -0700, Donnie Berkholz wrote: > Fredrik Tolf wrote: > > I'm not an expert with portage, but the fact that xorg-server is > > indented one space from gtk+ means that gtk+ depends on it directly, > > doesn't it? If so, whence does that dependency come from (it's obviously > > not on the DEPEND or RDEPEND variables)? > It's because it inherits the virtualx.eclass -- gtk+ requires an X > server (Xvfb) to build. What you can do is set USE="minimal -dmx -xorg > -xprint" in your package.use for xorg-server to minimize this. Also try > VIDEO_CARDS="" and INPUT_DEVICES="" in make.conf. > > However, this only works well on xorg-server-1.1, so you will also need > to add it to package.keywords. I see... how troublesome. Well, I guess there's no choice, but would anyone happen to know the reason why gtk+ requires Xvfb? Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] Too many X modules pulled in
Dear List,
I'm trying to install Gentoo on a headless machine. I want programs to
use the X libraries, so that I can SSH in from another machine and use X
over TCP, but I'm not too fond of installing the entire X package
(especially the X server) on it, since it won't ever be used.
However, I just tried to install a couple of program linking against GTK
(such avahi), and GTK, for some reason, tries to pull in lots of stuff
that I don't really want (among others, the X server). Looking in the
GTK ebuild, I found this:
RDEPEND="|| ( (
x11-libs/libXrender
x11-libs/libX11
x11-libs/libXi
x11-libs/libXt
x11-libs/libXext
x11-libs/libXcursor
x11-libs/libXrandr
x11-libs/libXfixes
xinerama? ( x11-libs/libXinerama ) )
virtual/x11 )
>=dev-libs/glib-2.10.1
>=x11-libs/pango-1.9
>=dev-libs/atk-1.10.1
>=x11-libs/cairo-0.9.2
media-libs/fontconfig
x11-misc/shared-mime-info
>=media-libs/libpng-1.2.1
jpeg? ( >=media-libs/jpeg-6b-r2 )
tiff? ( >=media-libs/tiff-3.5.7 )"
DEPEND="${RDEPEND}
sys-devel/autoconf
>=dev-util/pkgconfig-0.9
=sys-devel/automake-1.7*
|| ( (
x11-proto/xextproto
x11-proto/xproto
x11-proto/inputproto
x11-proto/xineramaproto )
virtual/x11 )
doc? (
>=dev-util/gtk-doc-1.4
~app-text/docbook-xml-dtd-4.1.2 )"
I even emerged all those X11 dependencies myself, but even then, it
wants to emerge xorg-server. The first few lines of the output of
"emerge -tvp gtk+" look like this:
[ebuild N] x11-libs/gtk+-2.8.19 USE="X jpeg -debug -doc -tiff
-xinerama" 12,015 kB
[ebuild N] x11-misc/shared-mime-info-0.17-r2 558 kB
[ebuild N] dev-libs/libxml2-2.6.26 USE="ipv6 python readline
-debug -doc -test" 3,338 kB
[ebuild N] dev-util/intltool-0.35.0 126 kB
[ebuild N] x11-libs/pango-1.12.3 USE="-debug -doc" 1,197 kB
[ebuild N] x11-base/xorg-server-1.0.2-r7 USE="dri ipv6 -debug
-minimal -xprint" 8,716 kB
I'm not an expert with portage, but the fact that xorg-server is
indented one space from gtk+ means that gtk+ depends on it directly,
doesn't it? If so, whence does that dependency come from (it's obviously
not on the DEPEND or RDEPEND variables)?
Please help me -- what should I do to make gtk+ only pull in the minimal
amount of deps?
Thanks for your time!
Fredrik Tolf
--
gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Network connection repeatedly failing!
On Wed, 2006-08-16 at 11:47 -0400, Samuel Baldwin wrote: > Also, does anyone know of a tool that allows you to play audio from a > terminal? Basically, sending audio to standard audio out? I'd like to > be able to listen to music (all .ogg and .mp3, if it makes a > difference), while in runlevel 3. This is a side comment, and do not > waste time answering this/looking it up if you don't know off hand. My personal favorite is mpg321 (or ogg123 for OGG). Very efficient and very customizable. Normally, it plays directly to your default libao plugin (probably OSS or ALSA), but with the -s option, you can send raw (headerless) 44100 kHz, 8 bit signed short audio to stdout for further processing by e.g. sox. Call me a simpleton, but I often have lots of fun with changing the playback sampling rate :) mpg123 -vs somefile.mp3 | sox -t .raw -r 48000 -sw -c 2 - -t \ ossdsp /dev/dsp Have fun, Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] YP domain name in baselayout 1.12
On Mon, 2006-08-14 at 14:40 -0500, Paul Varner wrote: > On Mon, 2006-08-14 at 19:05 +0200, Fredrik Tolf wrote: > > Hi list! > > > > I just noticed that /etc/init.d/domainname has disappeared in baselayout > > 1.12, but I can't find what is supposed to replace it. > > > > So, in short: How am I supposed to set my YP domain name with the new > > baselayout? > > > > It is now set in /etc/conf.d/net Oh, I see. That explains why I hadn't found it, as well (I had been grepping through /etc/init.d after domainname and hostname, but not /lib/rcscripts/net). Thank you very much for clearing that up for me! That does lead me to another question, though -- the interface-boundedness of these settings makes me a bit suspicious. Will /etc/init.d/net.* start rewriting /etc/resolv.conf and/or /etc/yp.conf for me or something like that, and will the NIS domain name be kept unset if net.eth0 fails to start for some or another reason? Fredrik Tolf -- gentoo-user@gentoo.org mailing list
[gentoo-user] YP domain name in baselayout 1.12
Hi list! I just noticed that /etc/init.d/domainname has disappeared in baselayout 1.12, but I can't find what is supposed to replace it. So, in short: How am I supposed to set my YP domain name with the new baselayout? Thanks for your time! Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Font problems with xorg-x11 7.0
On Fri, 2006-07-14 at 22:07 +0200, Fredrik Tolf wrote:
> On Thu, 2006-07-13 at 23:31 -0700, Donnie Berkholz wrote:
> > Fredrik Tolf wrote:
> > > My main problem is that xfs (the X font server, not the filesystem) hogs
> > > the CPU for about 2-3 seconds whenever I start a new xterm, blocking the
> > > new xterm during that time. I haven't found the exact extent of programs
> > > affected by the problem -- right now, I only know that xterm and twm are
> > > affected.
> > >
> > > I've tried turning off xfs and moving the FontPaths into the X server
> > > itself, but that only made the problem worse (taking 6+ seconds instead).
> >
> > Does a symlink /usr/share/fonts/fonts exist?
>
> Nope, and I never had it. Should it exist? In that case, whence should
> it point?
To follow up on this, the problem turned out to be the lack of CJK
fonts. When I installed font-{jis,isas,daewoo}-misc, it started working
just fine.
Fredrik Tolf
--
gentoo-user@gentoo.org mailing list
[gentoo-user] Font problems with xorg-x11 7.0
Hi List! I recently upgraded xorg-x11 to 7.0, as it was unmasked in stable. However, I was given some font problems along with the upgrade. I haven't found anything on either Google or on bugs.gentoo.org, so I'm turning to the list now. My main problem is that xfs (the X font server, not the filesystem) hogs the CPU for about 2-3 seconds whenever I start a new xterm, blocking the new xterm during that time. I haven't found the exact extent of programs affected by the problem -- right now, I only know that xterm and twm are affected. I've tried turning off xfs and moving the FontPaths into the X server itself, but that only made the problem worse (taking 6+ seconds instead). Also, when I'm starting xterms from other machines, they don't seem to care at all, so I'm guessing that the xterm on my own machine is looking for some fonts for which the parameters have somehow changed in 7.0 (I'm not very familiar with the X font system, so I don't really know what could have changed). It might be related that I am able to view CJK characters in the local xterms, but not the remote ones (they display `??' instead of one double-width character). There is also another, unrelated and rather minor, problem: The 8x16 fixed font seems to have disappeared in 7.0. Does anyone know where it went? Thanks for reading! Fredrik Tolf -- gentoo-user@gentoo.org mailing list
OT: Re: [gentoo-user] Update xorg-x11
On Thu, 13 Jul 2006, Richard Fish wrote: Your 11pt font looks absolutely ridiculous on my 133dpi screen. I'm not a typography expert, so correct me if I'm wrong, but should 11pt (being 11/72" by the DTP system) not be the same on any display or other device, regardless of resolution? Are you sure that it isn't your display settings that are wrong? Fredrik -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] LVM boot problem
On Thu, 2006-04-06 at 15:50 -0500, Boyd Stephen Smith Jr. wrote: > On Thursday 06 April 2006 11:44, Fredrik Tolf <[EMAIL PROTECTED]> wrote > about '[gentoo-user] LVM boot problem': > > I'm having a bit of a problem with LVM2 on Gentoo. See, I have a > > computer with a couple of hard drives in an LVM, and when it boots, I > > can see it loading the driver modules for the controller cards properly > > (I've added them to /etc/modules.autoload.d), but then when it comes to > > initializing the LVM, it complains that it cannot find all the PVs to > > initialize the VG. In fact, it complains about _all_ the PVs, as though > > it found none. > > Is / on LVM? Are you using an initrd / initramfs? Nope. Here's my fstab (minus pseudo-FS:s): /dev/hda2 / ext3 defaults,acl,user_xattr 0 1 /dev/hda1 noneswap sw0 0 /dev/vg1/newhome /home reiserfs defaults,acl,user_xattr 0 2 /dev/vg1/dlwsroot /srv/dlwsroot xfs defaults0 3 /dev/vg1/site /usr/site xfs defaults0 4 > Does your early userspace start a /dev daemon? If you mean udev, yes. > Does your early userspace have a different lvm.conf then your final userspace? Nope. > > Later, when the system is booted, I can log in as root and just run > > "vgchange -ay" to get the VG up. Note that I don't have to run pvscan > > first, which I'm assuming has to mean that it has somehow managed to > > find the PVs somewhere in the meantime. > > Are all of your pvs standard disks or disk partitions? Partitions. Here's the output of pvs: PV VG Fmt Attr PSize PFree /dev/hde1 vg1 lvm2 a- 111.79G 0 /dev/hdf1 vg1 lvm2 a- 111.79G 0 /dev/hdg1 vg1 lvm2 a- 111.79G 0 /dev/hdh1 vg1 lvm2 a- 111.79G 60.68G /dev/sda1 vg1 lvm2 a- 186.30G 186.30G /dev/sdb1 vg1 lvm2 a- 279.46G 0 /dev/sdc1 vg1 lvm2 a- 279.46G 0 > What modules are being loaded by the gentoo /etc/modules.autoload system? pcspkr pdc202xx_new sata_sil sd_mod rpcsec_gss_krb5 > What is your RC_VOLUME_ORDER in /etc/conf.d/rc? RC_VOLUME_ORDER="raid evms lvm dm" > Does your /lib/rcscripts/addons/lvm-start.sh > contain "/sbin/vgscan --mknodes --ignorelockingfailure >/dev/null"? Yep. > > This problem is quite annoying, since it means that I have to restart > > NFS and boot the system without the network cable plugged in (if it was, > > the NFS clients would try to reconnect and get handles to the wrong > > filesystem, which makes them mark the mount points as stale). > > You might investigate the mpt= /etc/exports option to only export > filesystems that have been successfully mounted, at least temporarily. Interesting. I'll definitely use that. For completeness, though, I should mention that looking at the manpage, it seems to be named "mp" rather than "mpt". Thanks! Fredrik Tolf -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] LVM boot problem
On Thu, 2006-04-06 at 10:09 -0700, Richard Fish wrote:
> On 4/6/06, Fredrik Tolf <[EMAIL PROTECTED]> wrote:
> > Hi list!
> >
> > I'm having a bit of a problem with LVM2 on Gentoo.
Actually, it worked now that I just rebooted it! I noticed that the time
it took to load the modules was longer than usual (I just added another
controller card, so I'm guessing that's why), so I'm beginning to
suspect that it's just a race condition. Could it be that the kernel is
scanning for partitions in a background thread, so that the disks simply
haven't finished scanning by the time the LVM scans for PVs?
I'll give you the data you requested anyway, though. Please tell me if
something is strange.
Fredrik Tolf
--
(dmesg output moved to the end of the mail)
Bonus data:
# pvs
PV VG Fmt Attr PSize PFree
/dev/hde1 vg1 lvm2 a- 111.79G 0
/dev/hdf1 vg1 lvm2 a- 111.79G 0
/dev/hdg1 vg1 lvm2 a- 111.79G 0
/dev/hdh1 vg1 lvm2 a- 111.79G 60.68G
/dev/sdb1 vg1 lvm2 a- 279.46G 0
/dev/sdc1 vg1 lvm2 a- 279.46G 0
> 2. grep -v -e "^ *#.*$" -e "^ *$" /etc/lvm/lvm.conf
devices {
dir = "/dev"
scan = [ "/dev" ]
filter = [ "a/.*/" ]
cache = "/etc/lvm/.cache"
write_cache_state = 1
sysfs_scan = 1
md_component_detection = 1
}
log {
verbose = 0
syslog = 1
overwrite = 0
level = 0
indent = 1
command_names = 0
prefix = " "
}
backup {
backup = 1
backup_dir = "/etc/lvm/backup"
archive = 1
archive_dir = "/etc/lvm/archive"
retain_min = 10
retain_days = 30
}
shell {
history_size = 100
}
global {
umask = 077
test = 0
activation = 1
proc = "/proc"
locking_type = 1
locking_dir = "/var/lock/lvm"
}
activation {
missing_stripe_filler = "/dev/ioerror"
mirror_region_size = 512
reserved_stack = 256
reserved_memory = 8192
process_priority = -18
}
> 3. grep -v -e "^ *#.*$" -e "^ *$" /etc/modules.autoload.d/kerrnel-2.6
pcspkr
pdc202xx_new <-- This is for my P-ATA controller card
sata_sil <-- This is for my two S-ATA controller cards
sd_mod
rpcsec_gss_krb5
> 4. lsmod
Module Size Used by
ehci_hcd 30344 0
psmouse35972 0
parport_pc 25060 0
parport32360 1 parport_pc
floppy 57956 0
tulip 47584 0
crc32 3808 1 tulip
i2c_piix4 7568 0
i2c_core 17168 1 i2c_piix4
uhci_hcd 30928 0
intel_agp 20412 1
agpgart29160 1 intel_agp
usbcore 117188 3 ehci_hcd,uhci_hcd
xfs 606840 2
reiserfs 259600 1
nfsd 233700 17
exportfs5088 2 xfs,nfsd
lockd 62280 2 nfsd
nfs_acl 2816 1 nfsd
dm_mod 52472 4
rpcsec_gss_krb5 8072 0
auth_rpcgss38496 3 rpcsec_gss_krb5
sunrpc136892 15
nfsd,lockd,nfs_acl,rpcsec_gss_krb5,auth_rpcgss
sd_mod 15056 4
sata_sil6468 2
libata 55084 1 sata_sil
scsi_mod 87240 2 sd_mod,libata
pdc202xx_new7872 0 [permanent]
pcspkr 1476 0
> 1. dmesg
Linux version 2.6.15-gentoo-r1 ([EMAIL PROTECTED]) (gcc version 3.4.4 (Gentoo
3.4.4-r1, ssp-3.4.4-1.0, pie-8.7.8)) #1 PREEMPT Mon Feb 27 02:00:44 CET
2006
BIOS-provided physical RAM map:
BIOS-e820: - 0009e000 (usable)
BIOS-e820: 0009e000 - 000a (reserved)
BIOS-e820: 000e6c00 - 0010 (reserved)
BIOS-e820: 0010 - 040fd800 (usable)
BIOS-e820: 040fd800 - 040ff800 (ACPI data)
BIOS-e820: 040ff800 - 040ffc00 (ACPI NVS)
BIOS-e820: 040ffc00 - 1000 (usable)
BIOS-e820: fff8 - 0001 (reserved)
256MB LOWMEM available.
On node 0 totalpages: 65536
DMA zone: 4096 pages, LIFO batch:0
DMA32 zone: 0 pages, LIFO batch:0
Normal zone: 61440 pages, LIFO batch:15
HighMem zone: 0 pages, LIFO batch:0
DMI 2.1 present.
ACPI: RSDP (v000 PTLTD ) @ 0x000f6bb0
ACPI: RSDT (v001 PTLTDRSDT 0x0001 PTL 0x0100) @
0x040fdbcb
ACPI: FADT (v001 DELL KUBLAI 0x20001013 PTL 0x000f4240) @
0x040ff78c
ACPI: DSDT (v001 Intel S2440BX 0x0001 MSFT 0x0104) @
0x
Allocating PCI resources starting at 2000 (gap: 1000:eff8)
Built 1 zonelists
Kernel command line: ro root=/dev/hda2
Local APIC disabled by BIOS -- you can enable it with "lapic"
mapped APIC to d000 (01201000)
Initializing CPU#0
PID hash ta
Re: [gentoo-user] LVM boot problem
On Thu, 2006-04-06 at 11:51 -0500, John Jolet wrote:
> sounds like you're using device mapper as a module...compile it into
> the kernel. module loading happens too late
Thanks for your reply, but I seriously doubt that to be the problem,
since /etc/rcscripts/addons/lvm-start explicitly contains the following
code:
if [[ -e /proc/modules ]] && \
! grep -qs 'device-mapper' /proc/{devices,misc}
then
modprobe dm-mod &>/dev/null
fi
Also, if the dm module wasn't loaded, I think that is what the LVM tools
would complain at, not at not finding the PVs, right?
Fredrik Tolf
--
gentoo-user@gentoo.org mailing list
[gentoo-user] LVM boot problem
Hi list! I'm having a bit of a problem with LVM2 on Gentoo. See, I have a computer with a couple of hard drives in an LVM, and when it boots, I can see it loading the driver modules for the controller cards properly (I've added them to /etc/modules.autoload.d), but then when it comes to initializing the LVM, it complains that it cannot find all the PVs to initialize the VG. In fact, it complains about _all_ the PVs, as though it found none. Later, when the system is booted, I can log in as root and just run "vgchange -ay" to get the VG up. Note that I don't have to run pvscan first, which I'm assuming has to mean that it has somehow managed to find the PVs somewhere in the meantime. This problem is quite annoying, since it means that I have to restart NFS and boot the system without the network cable plugged in (if it was, the NFS clients would try to reconnect and get handles to the wrong filesystem, which makes them mark the mount points as stale). Does anyone have any idea what could cause it? Thanks for your time! Fredrik Tolf -- gentoo-user@gentoo.org mailing list
