Re: [gentoo-user] Re: [gentoo-dev] No XFree86 w/ new license
ALL the distro's are having this same problem though. The current version of X works and is still available under the license it was released under. There's nothing stopping anyone from continuing to work with it for the time being, it just means new versions aren't necessarily going to find their way in to gentoo or several other distros for that matter. I don't think this is a case of saying "oh well, no more GUI" but we may be staring at the same version of X for a while. Bryn On Feb 26, 2004, at 3:15 PM, William Kenworthy wrote: I must agree that a gui is essential. Without this, most of gentoo's user base will disappear overnight, as it will be useless for any desktop application. Less users=less support, less development and eventually oblivion. BillK On Fri, 2004-02-27 at 00:05, Donnie Berkholz wrote: On Thu, 2004-02-26 at 10:57, Stewart wrote: Yes, it is a major component. Nay, it is a critical component. Without a GUI, we may as well write our source code on toilet paper and distribute it to the developers to be used appropriately. Graphical User Environment = Market Acceptance. Period. You make the assumption that part of what makes Gentoo what it is, is having X. This is not the case. It doesn't matter what you say about market acceptance, if Gentoo is not provided by default with X (meaning X is part of "system" in my interpretation), it is not a "major component" of the "operating system." D -- [EMAIL PROTECTED] mailing list -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] boot partition
On Feb 26, 2004, at 4:00 PM, Jorge Almeida wrote: On Thu, 26 Feb 2004, Sami Samhuri wrote: On Thu, 2004-02-26 at 10:41, Jorge Almeida wrote: I don't think you can boot from an extended partition directly without a bootloader, such as GRUB. I don't understand what you mean. I intend to install GRUB on the MBR. Can /boot be a logical partition then? (Or do you mean a bootloader on the boot sector of the logical partition?) AFAIK grub/lilo cannot find the kernel if it is on a logical or extended partition, they specifically require the kernel to be on a primary regular partition. On the ix86 platform using DOS disklabels you can have up to 4 primary partitions. One of those 4 can be an extended partition which can then have many more partitions created within it. grub/lilo can't go through that many levels of partitioning so they need /boot to be on one of the 4 primary partitions. As for the size, I think 32M and 100M are reasonable seeing as most drives are >=10G these days but if you want a smaller one, it's your box. Even with 5 kernels on /boot you could probably get away with 16M. I think it's more of a "why not" situation where if it is ever needed for whatever reason, it will be there. I see. So there is no special reason, except perhaps keeping several kernels? For example, is /boot/grub expected to get fat? Exactly... You don't necessarily need much space there but with disk being so abundant these days it's safer to give more than you need just in case. I personally wouldn't expect to need a boot partition much over 15 megs ever but who knows what the future may hold... a new boot loader could come around for instance which needs more space. Bryn -- [EMAIL PROTECTED] mailing list
[gentoo-user] xinetd won't restart service after connect
Hi everyone,
I have xinetd-2.3.13 installed on my gentoo-ppc system. I'm trying to
use amanda on this server (which I have used successfully on many other
servers) but I seem to be having some trouble specifically with the
amanda service. The first time a backup process connects to the server
the amanda service is started properly by xinetd. From running xinetd
-d (debug mode) I am able to see that 2 connections are made, the first
one exits and xinetd resumes the service properly. After the second
connection exits however xinetd does NOT resume the amanda service.
This means any future connections fail. If you look at the log below
you can see the first connection detach and xinetd resume the service.
I can wait for any amount of time and xinetd never seems to resume
after the second connection. After the second connection xinetd starts
using CPU time like crazy too even though it isn't doing anything.
Below is an example run of xinetd -d:
04/2/[EMAIL PROTECTED]:47:45: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/amanda [file=/etc/xinetd.conf]
[line=13]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/chargen-tcp
[file=/etc/xinetd.d/chargen-tcp] [line=37]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/chargen-udp
[file=/etc/xinetd.d/chargen-udp] [line=13]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/daytime-tcp
[file=/etc/xinetd.d/daytime-tcp] [line=14]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/daytime-udp
[file=/etc/xinetd.d/daytime-udp] [line=13]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/echo-tcp
[file=/etc/xinetd.d/echo-tcp] [line=14]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/echo-udp
[file=/etc/xinetd.d/echo-udp] [line=13]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/time-tcp
[file=/etc/xinetd.d/time-tcp] [line=14]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {handle_includedir} Reading included
configuration file: /etc/xinetd.d/time-udp
[file=/etc/xinetd.d/time-udp] [line=13]
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing
chargen
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing
chargen
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing
daytime
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing
daytime
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing echo
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing echo
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing time
04/2/[EMAIL PROTECTED]:47:46: DEBUG: 23802 {remove_disabled_services} removing time
Service defaults
Instances = 15
Bind = All addresses.
Only from: localhost(HOST) localhost.localdomain(HOST)
bluemanie.nashira.ca(HOST) bluemeanie-prv(HOST)
No access: No blocked sites
Logging to syslog. Facility = authpriv, level = info
Log_on_success flags = HOST PID
Log_on_failure flags = HOST
Service configuration: amanda
id = amanda
flags = IPv4
socket_type = dgram
Protocol (name,number) = (udp,17)
port = 10080
wait = yes
user = 87
group = 6
Groups = yes
PER_SOURCE = -1
Bind = All addresses.
Server = /usr/libexec/amandad
Server argv = amandad
Only from: localhost(HOST) localhost.localdomain(HOST)
bluemanie.nashira.ca(HOST) bluemeanie-prv(HOST)
No access: No blocked sites
Logging to syslog. Facility = authpriv, level = info
Log_on_success flags = HOST PID
Log_on_failure flags = HOST
Service configuration: amandaidx
id = amandaidx
flags = IPv4
socket_type = stream
Protocol (name,number) = (tcp,6)
port = 10082
wait = no
user = 87
Groups = yes
PER_SOURCE = -1
Bind = All addresses.
Server = /usr/libexec/amindexd
Server argv = amindexd
Only from: localhost(HOST) localhost.localdomain(HOST)
bluemanie.nashira.ca(HOST) bluemeanie-prv(HOST)
No access: No blocked sites
Logging to syslog. Facility = authpriv, level = info
Log_on_success flags = HOST PID
Log_on_failure flags = HOST
Service configuration: amidxtape
id = amidxtape
flags = IPv4
socket_type = stream
Protocol (name,number) = (tcp,6)
port = 10083
wait = no
Re: [gentoo-user] portage tree on USB stick (was: )
There's no problem copying portage trees around on removable media, but the portage tree on my gentoo box here is about 2.2GB, so you'd need to make a couple of trips with your 128MB stick. Is it possible for you to burn to disc (CD/DVD)? as this would probably be a lot less effort for you. By the way - a relevant subject line on all your emails will help you to get more answers, as this is a high-traffic list a lot of folks won't bother to read messages without a subject line, the same also goes for HTML email in general - a lot of people's mail readers won't format it, so HTML email just goes straight to /dev/null... cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] emerge win4lin doesn't work
Try 'emerge sync' first - that will update your local portage database to match that of the server. Bryn On Jan 16, 2004, at 3:12 PM, Klaus Neumann wrote: Hi, trying to emerge win4lin, I got this: - Connecting to ftp.gtlib.cc.gatech.edu[130.207.108.134]:21... connected. Logging in as anonymous ... Logged in! ==> SYST ... done.==> PWD ... done. ==> TYPE I ... done. ==> CWD /pub/gentoo/distfiles ... done. ==> PASV ... done.==> RETR Win4Lin-5.5.8e-d.i386.rpm ... No such file `Win4Lin-5.5.8e-d.i386.rpm'. !!! Couldn't download Win4Lin-5.5.8e-d.i386.rpm. Aborting. - I'm lost now. Could somebody please tell me what I can do about it? TIA! -- Best regards, Klaus -- [EMAIL PROTECTED] mailing list -- [EMAIL PROTECTED] mailing list
[gentoo-user] xinetd 2.3.11 - can't find protocols
I've got xinetd 2.3.11 installed on my gentoo-ppc 1.4 server. All I'm planning on using xinetd for is amanda but it seems I can't get any services whatsoever to work. xinetd appears to not be able to read (or process) the contents of /etc/protocols. Here's what I get in my logs: [...] Jan 16 13:17:45 bluemeanie xinetd[10119]: Reading included configuration file: /etc/xinetd.d/amanda [line=13] Jan 16 13:17:45 bluemeanie xinetd[10119]: Protocol udp not in /etc/protocols [line=8] Jan 16 13:17:45 bluemeanie xinetd[10119]: Error parsing attribute protocol - DISABLING SERVICE [line=8] Jan 16 13:17:46 bluemeanie xinetd[10119]: Protocol tcp not in /etc/protocols [line=22] Jan 16 13:17:46 bluemeanie xinetd[10119]: Error parsing attribute protocol - DISABLING SERVICE [line=22] Jan 16 13:17:46 bluemeanie xinetd[10119]: Protocol tcp not in /etc/protocols [line=34] Jan 16 13:17:46 bluemeanie xinetd[10119]: Error parsing attribute protocol - DISABLING SERVICE [line=34] Jan 16 13:17:46 bluemeanie xinetd[10119]: Reading included configuration file: /etc/xinetd.d/chargen-tcp [line=42] Jan 16 13:17:47 bluemeanie xinetd[10119]: Protocol tcp not in /etc/protocols [line=9] Jan 16 13:17:47 bluemeanie xinetd[10119]: Error parsing attribute protocol - DISABLING SERVICE [line=9] Obviously I've checked /etc/protocols for tcp and udp: [...] tcp 6 TCP # transmission control protocol udp 17 UDP # user datagram protocol And verified that /etc/protocols is readable: -rw-r--r--1 root root 1.8K Jan 2 09:44 /etc/protocols My past experience with xinetd has generally been pretty good - it usually just works baring any silly config file errors. Anyone have any ideas? Bryn Bryn Hughes Network Specialist Linux/OS X Support Vancouver Community College -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] equivelant of Nero's Image Writer
On 08:15 Fri 16 Jan , Andrew Gaffney wrote: > Barry Marler wrote: > >On Fri, 16 Jan 2004 07:57:38 -0600 > >Andrew Gaffney <[EMAIL PROTECTED]> wrote: > > > > > >>Is there a program/kernel module for Linux that emulates a CD burner > >>for the purpose of generating ISOs? > > > > > >mkisofs? > > Not what I'm looking for. Basically, it could be useful for converting one > rip format (BIN/CUE or CCD/CUE/IMG/SUB) to another (ISO). > You'll need to be a bit more specific about what you're trying to do. 'generating ISOs' (from a directory tree) is handled by mkisofs, as suggested by the other posters. I guess a lot of people (m'self included) on here never use Nero, so some more info on how you make use of it's 'Image Writer' function would help.. cdrdao will allow you to burn CUE/BIN format to a disc, but I'm not sure about redirecting output to an image file as apposed to a physical device, so you could later recover a regular ISO form it.. is that yr goal? otoh, the formats are probably not terribly complicated - isos can be prepared from a physical disk like so with dd: # dd if=/dev/cdrom of=~/mydisc.iso I'd guess the BIN part of CUE/BIN is very simillar, afaik BIN is just images of consecutive tracks concattenated and CUE is a plaintext index of htese tracks. Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Proxy server
On 11:02 Tue 06 Jan , Nuckerl Stefan wrote: > If you have a fast connection like cable or dsl you won't speed up your > connection but rather slow it down. > Not necessarily so... It all depends how contended the ISP's border routers are, and how much repeat content the guy views. -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Running Linux with no file systems mounted...
On 16:45 Mon 05 Jan , Andrew Gaffney wrote: > opportunity to install Gentoo, but even with my dual Athlon MP 2200+ > processors and 768MB DDR, it still took 9 hours to go from bootstrap to > ready to go. I don't want to have to do that again ;) > > -- Why not back up all your built packages separately aswell as the actual system? That way you can rebuild with your own optimisations and no building - should take a lot less than 9 hours. Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] from arch-linux back to gentoo
On 18:09 Tue 13 Jan , Fabian Braennstroem wrote: > Hello, > > does anybody tried arch-linux before and can tell me now, why gentoo is > better. I need some good arguments for a friend, who doesn't know which > distribution (arch/gentoo) he should use. > Is it just the speed which is very important? > > Greetings! > Fabian > > > -- > [EMAIL PROTECTED] mailing list > Just looking at the front page of http://www.archlinux.org tells me that arch-linux is optimised for i686. This is a big difference from gentoo, for one thing I assume it is x86 only. Gentoo will optimize packages however you ask it too, based on your /etc/make.conf settings - it can do this because it builds everything (if you use a stage1 tarball) on the local machine as you install it. Gentoo has portage as a package manager, arch-linux has pacman, again the homepage tells me: "Pacman package manager, which couples a simple binary package format with an easy-to-use build system" So there you are. arch-linux has a (primarily) binary package manager, gentoo (primarily) source-based. I don't think it is a question of which is 'better' in absolute terms, it all depends what you want to do with it, and what your personal tastes and goals for a distro dictate. Your best option is probably to read the About section of their website and see how well it fits your needs. Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] What happened today?
On 11:41 Tue 13 Jan , Thomas T. Veldhouse wrote: > Can anybody explain what happened to the email lists and the rsync server > today? I notice that there is still a stale posting on the website about > the rsync problem (doesn't mention that the lists were down yet again), but > no update. > > Anybody? > > > -- > [EMAIL PROTECTED] mailing list > well, I'm somebody, but not anybody 'triffically special :) rsync.gentoo.org had it's dns record changed to 127.0.0.1, I guess to reduce the flood of traffic while the chaps at gentoo.org were working on the box... I couldn't find anything bar the note on the website either, it was only posted today I think, but atleast it gave you some clue as to what was going on... B. -- [EMAIL PROTECTED] mailing list
[gentoo-user] gnome 2.4 logout strange behaviour
Hi all, I installed gnome-2.4 last week with no real problems. Up until now, all has been fine. This morning, the panel started playing up; clicking actions -> logout would freeze the panel and leave the 'actions' menu button highlighted. Killing the panel caused it to reload fine but trying to logout had the same effect. Restarting the whole of X/gdm/gnome also didn't help. After a reboot, the hanging problem is gone, but gnome is taking a very long time to startup when logging in, and the logout confirm dialog does not appear until several minutes after seleting logout from the actions menu. I've had strange behaviour with the gnome panel in earlier versions, but normally it's fine after a restart. Any suggestions? Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] background process
rh wrote: When I start a process running in the background, as in, "xplanet -projection rectangular 2>/dev/null &" , why when I exit the terminal window, does that process die? Kind of defeats the purpose of shoving it to the background if you still have to keep the eterm open. rh -- [EMAIL PROTECTED] mailing list It does that because that's what it's meant to do ;-) and look for stuff on job control (for bash, "man bash" then "/^JOB CONTROL" will take you to the right bit). When an interactive shell exits, it sends a SIGHUP to all the processes in it's job table, you can remove jobs from the table using the bash builtin disown or by running them under the nohup program. Some programs exhibit different behaviour (staying alive after the shell exits) but that's usually because they explicitly catch the HUP signal (daemons esp.) hope that helps Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] portage group issues
Bryce wrote: Thanks for the responce Tom; however this is a long standing problem. Since i put in a new harddrive and rebuilt gentoo. I didn't have this problem before( on the old harddrive) and i find it quite annoying now. Could it maybe be that my account is in a group that somehow writes over being in the portage group? Or maybe i just belong to too many groups. Could some people please send me their output for "groups" so i have an idea of where mine should be??? I realize its an odd request, but private emails are welcomed ;). bryce On Wednesday 30 July 2003 04:26 am, Tom Knight-Markiegi wrote: Hi Bryce On Wed, Jul 30, 2003 at 09:56:36PM -0700, Bryce wrote: Hey all, my normal user account is in the portage group( i even checked /etc/group), however, whenever i use my account to do some emergeing ( short of compiling) i always get that "not part of portage group" error. Has anyone else seen or dealt with this problem?? Have you logged out and back in again? I've noticed that you have to do this before any changes to /etc/groups take effect. If you type: groups this will tell you what groups you are in, if this doesn't match up to /etc/groups try logging in again. Tom thanks, bryce -- [EMAIL PROTECTED] mailing list -- [EMAIL PROTECTED] mailing list -- [EMAIL PROTECTED] mailing list Here's the output of groups on my box, and also the portage line from my group file: [EMAIL PROTECTED] bryn $ groups users wheel slocate portage [EMAIL PROTECTED] bryn $ grep portage /etc/group portage::250:portage,bryn [EMAIL PROTECTED] bryn $ What do you get? Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Listing all init scripts at any runlevel
Dhruba Bandopadhyay wrote: Hello How can I list all init scripts that are currently added to the default run level and to the boot level? A solution other than looking at bootup screen would be helpful. With regards. ls /etc/runlevels/$RUNLEVEL Where $RUNLEVEL is the name of the runlevel you're interested in. More info here: http://www.gentoo.org/doc/en/rc-scripts.xml hth Bryn -- "A computer programmer is someone who, when told to "Go to Hell!", sees the "Go to", rather than the destination, as harmful." -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Which ADSL devices work on Gentoo?
Jannu Hatinen wrote: Cheers, I'm planning to get an ADSL connection, and was wondering what kind of modem I should buy. It's quite obvious that the external stand-alone modem/routers should work just fine, but what about the other options? Do internal PCI modems and external USB modems work on Gentoo? I'm planning to use one Gentoo system as a router/firewall for the rest of the network, so I have no need for the router/firewall capacities of the stand-alone modems. It's not so much a question of which will work with Gentoo as it is 'Which work with Linux?'. The alcatel (frog) USB modems are pretty well supported, as are most modems based on the same chipset (fuji do one iirc). Having said that, I just bought a router for the home setup. hth Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] extend a running shell command
Christopher Fisk wrote: use ^Z to pause the task at hand, then use 'fg ; next-command' to put the paused command back into the FG and add the new command to the end. The problem with that approach is that once you've entered 'fg ; command' you loose job control on that shell. Check the bugs section of the man page - builtins like fg aren't supported elegantly by job control and you cant properly suspend the command. If you try, the next job in sequence is executed immediately even though it should wait on fg's termination. While this won't effect the example given, it does criple the shell for future use until the sequence completes. It's totally a personal thing but really bugs me so I avoid doing it (do fg voodoo, forget I'm using a builtin, hit -z... oops!) I think this is a very long standing bash issue and probably won't get fixed anytime soon. Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] extend a running shell command
Ohad Lutzky wrote:
You guys aren't getting it... the command is already running. He's
thinking along the lines of this: "Okay, I'm downloading this file, it's
not resumable and already at 60%. It'll take a few hours more to
download and quite a while to untar it. I'm going to sleep now... I wish
I could tell it to untar once it's done downloading. Too bad I can't
start over without losing all the progress it already made."
As for the solution - I got nothing concrete, but here's my idea: Figure
out the PID of the process you want to be finished. Then create a script
that loops while the PID is existant, and once it's not - does whatever
it is that you want to do. Actually, that would be a one-liner. Here's
my idea:
Great minds think alike ;-)
First, get the PID using pidof. I'll call it $THE_PID
Then what you need is this:
Not sure about pidof for this tho - it'll return the pid of every
process running the command you supply. Try pidof mozilla-bin with moz
running. jobs lets you see just the jobs running in the current shell
- an incantation like the following might be a better way to get the
pid for a script:
[EMAIL PROTECTED] bryn $ jobs -l
[1] 1530 Stopped updatedb
[2]- 1534 Stopped updatedb
[3]+ 1535 Stopped ls -lR /
[EMAIL PROTECTED] bryn $ jobs -l | grep ls | awk '{print $2}'
1535
[EMAIL PROTECTED] bryn $
To be honest tho this is all a bit hardcore for me! Maybe someone on a
bash-specific mailing list would be able to give better advice?
Cheers
Bryn
--
[EMAIL PROTECTED] mailing list
Re: [gentoo-user] Re: extend a running shell command
CrPy wrote: Hi ng, sorry, but I think you don't get the problem. Type sleep 100h in your shell. And now don't kill it or stop it. Now imagine that you actually forgot what you really wanted: To halt the machine after the program (that already runs and may be it will need to run for some hours because it is a compile process or similar). And imagine that you don't want to sit all the time behind your computer and wait until the program finishes to be able to halt the computer. So, what I want is, to append to the aready running command line a new command without stopping the old (because then it would have to start it from the beginning). I think, it would be ok to suspend the running program temporarily and to resume it after a short time where I append the additional command to it. After all it should look like i never executed sleep 100h but sleep 100h; halt One solution I have is to suspend the command with and type fg; halt But what if I have a longer queue of commands and I like to edit the command line. e.g. make clean && make dep && make bzImage && make module && make modules_install And now I see that I have a typo in the line (the make module -> make modules). How can I change the entered command line while excuting this command line? Sorry, it was not easy for me to explain it in a way that somebody else could understand it. Maybe it is more clear now. THX /CrPy -- [EMAIL PROTECTED] mailing list I got what you meant from the first post but what you're asking for is not trivial!! First stop is the job control section of the bash manual (If you use less as your pager, do a man bash and hit /^J ). You should also check out the bugs section and the comments regarding compound commands & sequences. You will definately want to suspend the running job with -z (no needed!) or start it in the background with & You can get the process id of a running job from jobs -p, you may also want to explore the -l option. A hackish solution might be a script which greps/awks these values and awaits the disappearance of the pid from /proc or ps's output before executing the command but I can't see a way you'd be able to retrospectively capture the exit code of the job in order to make a decision based on it. I don't know when bash parses the command string you enter but I guess it's probably all done by the time the process is started - allowing you to go back and change it sounds tricky and would need work on bash itself. You could try making a feature request to the maintainers of bash - their emails are in the man page but if it conflicts with any of the sh compatibility I doubt they'd go for it.. Alternatively check out the other shells available like ksh, csh etc. They all have slightly different job control features - maybe there's one to suit you better than bash out there. Having said all that I just read Robert's post - if all you want to do is tack on extra commands like the halt example you gave then that's surely the easiest way to go. Good luck Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] initrd - how to build?
brett holcomb wrote:
I don't know where it is either . I used that on RH 7.3. If it's in
portage it's probably in a package with other tools.
On Tue, 17 Jun 2003 09:56:59 -0700
"Mark Knecht" <[EMAIL PROTECTED]> wrote:
-Original Message-
From: brett holcomb [mailto:[EMAIL PROTECTED]
Sent: Tuesday, June 17, 2003 9:54 AM
To: [EMAIL PROTECTED]
Subject: Re: [gentoo-user] initrd - how to build?
Is there a man mkinitrd? I've seen that on other distros.
It makes an initrd for the kernel you give it.
No, but I haven't got mkinitrd on my system either. That was the first
likely clue I got from Google.
Maybe if I emerged something I'd get it and a man page? (What to emerge?
emerge mkinitrd didn't do anything.)
qpkg mkinitrd?
- Mark
--
[EMAIL PROTECTED] mailing list
--
[EMAIL PROTECTED] mailing list
The original mkinitrd was a bash script from RedHat which used
modules.conf to put all the necessary boot modules (originally iirc by
just including ALL scsi modules) into a ram fs file.
You can do the same by making an ext2 file system as a file (use the
loopback device) containing the modules you need.
I've attached mkinitrd from a RedHat 7.2 system - it prolly won't work
out of the box for gentoo but should give you a good idea of what's
going on.
Basically I think any fs the kernel can mount without any mods is ok
for an initrd - the loopback device is general purpose that way.
I just tried this out but can't seem to find a way to associate a file
with a loopback device in gentoo - I used losetup under redhat, anyone
know the way?
Cheers
Bryn
#!/bin/bash
# mkinitrd
#
# Written by Erik Troan <[EMAIL PROTECTED]>
#
# Contributors:
# Elliot Lee <[EMAIL PROTECTED]>
# Miguel de Icaza <[EMAIL PROTECTED]>
# Christian 'Dr. Disk' Hechelmann <[EMAIL PROTECTED]>
# Michael K. Johnson <[EMAIL PROTECTED]>
# Pierre Habraken <[EMAIL PROTECTED]>
# Jakub Jelinek <[EMAIL PROTECTED]>
# Carlo Arenas Belon ([EMAIL PROTECTED]>
# Keith Owens <[EMAIL PROTECTED]>
# Bernhard Rosenkraenzer <[EMAIL PROTECTED]>
# Matt Wilson <[EMAIL PROTECTED]>
PATH=/sbin:$PATH
export PATH
VERSION=3.2.6
compress=1
target=""
kernel=""
force=""
verbose=""
MODULES=""
img_vers=""
modulefile=/etc/modules.conf
if [ `uname -m` = "ia64" ]; then
IMAGESIZE=4000
else
IMAGESIZE=3000
fi
PRESCSIMODS="scsi_mod sd_mod unknown"
fstab="/etc/fstab"
usage () {
echo "usage: `basename $0` [--version] [-v] [-f] [--ifneeded] [--preload
]" >&2
echo " [--omit-scsi-modules] [--omit-raid-modules] [--with=]" >&2
echo " [--image-version] [--fstab=] [--nocompress] " >&2
echo " " >&2
echo " (ex: `basename $0` /boot/initrd-2.2.5-15.img 2.2.5-15)" >&2
exit 1
}
findmodule() {
skiperrors=""
modName=$1
if [ $(echo $modName | cut -b1) = "-" ]; then
skiperrors=1
modName=$(echo $modName | cut -b2-)
fi
if [ "$modName" = "i2o_block" ]; then
findmodule i2o_pci
findmodule i2o_core
modName="i2o_block"
fi
if [ "$modName" = "pluto" ]; then
findmodule fc4
findmodule soc
modName="pluto"
fi
if [ "$modName" = "fcal" ]; then
findmodule fc4
findmodule socal
modName="fcal"
fi
if [ "$modName" = "ext3" ]; then
if [ -n "$skiperrors" ]; then
findmodule -jbd
else
findmodule jbd
fi
modName="ext3"
fi
fmPath=`(cd /lib/modules/$kernel; find -type f -name $modName.o)`
if [ ! -f /lib/modules/$kernel/$fmPath ]; then
if [ -n "$skiperrors" ]; then
return
fi
# ignore the absence of the scsi modules
for n in $PRESCSIMODS; do
if [ "$n" = "$modName" ]; then
return;
fi
done;
echo "No module $modName found for kernel $kernel" >&2
exit 1
fi
# only need to add each module once
if echo $MODULES | grep $fmPath >/dev/null 2>&1 ; then : ; else
MODULES="$MODULES $fmPath"
fi
}
inst() {
if [ "$#" != "2" ];then
echo "usage: inst "
return
fi
[ -n "$verbose" ] && echo "$1 -> $2"
cp $1 $2
}
while [ $# -gt 0 ]; do
case $1 in
--fstab*)
if echo $1 | grep '=' >/dev/null ; then
Re: [gentoo-user] ssh keeps stopping!
MAL wrote: bryn wrote: The [2]+ Stopped occurs as a side effect of issuing ls - no kb interrupts sent. No! the first ls has completed, thus the job has finished, thus it has stopped. It hadn't completed, I was just pointing out that (completed|finished)!=stopped. Sorry if you didn't get what I meant. try: $(ssh [EMAIL PROTECTED] command) & Sorry, no I don't get what you mean. What has running ssh inside a bash command substitution have to do with the problem I was asking about? It behaves oddly because it gets run in a sub-shell. When I say this 'happens as a side effect of issuing ls', like I said, that was an example. It indicated the point in the process when it happened, I wasn't trying to say that that's the cause of the problem. I've also just checked and the problem is independent of the shell - csh shows the same behaviour. Anyway, this is all kinda academic as I just tried ssh'ing an account which needs a password while using the -f switch. According to the man page this won't work but it appears to go just fine. Problem Solved. Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] ssh keeps stopping!
The [2]+ Stopped occurs as a side effect of issuing ls - no kb interrupts sent. No! the first ls has completed, thus the job has finished, thus it has stopped. It hadn't completed, I was just pointing out that (completed|finished)!=stopped. I never disputed that it was stopping as opposed to completing, I was merely mentioning explaining how the shell printed it's stopped response, and that you could run it _not_ in the background to possibly find out the reason why it was stopping.. does it only happen when run in the background? No, this is an issue which only affects background ssh processes. If I run the same app locally, or in the foreground, no problem. Thing is I often want to log into a machine, open a few X apps in the background and check things out, close them down and log out. I just noticed tho that this effects another machine here running redhat & a different version of ssh so maybe this is a 'feature'. The -f switch redirects stdin from /dev/null, which is fine for the boxes here which don't need a password but less good for remote ones where I don't control the login method. I've not long had bandwidth here to make it worth running X apps remotely so this has only just become an issue so I guess I'll go check in the ssh archives see if it's been a problem for anyone else. Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] ssh keeps stopping!
MAL wrote:
No! the first ls has completed, thus the job has finished, thus it has
stopped. The shell waits for the next prompt before it alerts you of
this fact.
No, 'fraid not.
Compare:
[EMAIL PROTECTED] bryn $ ssh hex ls -lR > /dev/null 2>&1 &
[2] 1889
[EMAIL PROTECTED] bryn $ ls
FrozenPenguins.png foo.javashare shot0003.jpg
Screenshot.png questframe shot0002.jpg
[2]+ Stopped ssh hex ls -lR >/dev/null 2>&1
[EMAIL PROTECTED] bryn $
with:
[EMAIL PROTECTED] bryn $ ls / > /dev/null &
[3] 1048
[EMAIL PROTECTED] bryn $
[3] Donels / >/dev/null
A stopped job is one who's execution is suspended by the shell. It
would return 'Done' if the job had completed. The use of ls was merely
an example to demonstrate; ls -lR takes quite a while on my home
directory on hex!
The problem I was refering to was background ssh jobs spontaneously
stopping just because I ran another command or hit enter.
This is simply because the background job has already finished.
Try running it _not_ in the background, and see how long it takes to
complete.
[EMAIL PROTECTED] bryn $ fc 465
time ssh hex ls -lR > /dev/null 2>&1 &
[3] 3439
[EMAIL PROTECTED] bryn $
real0m3.342s
user0m0.290s
sys 0m0.115s
Long enough to see the problem I'm referring to - maybe this example
with gimp makes it clearer:
[EMAIL PROTECTED] bryn $ ssh hex gimp&
[3] 3443
(1) [EMAIL PROTECTED] bryn $
[3]+ Stopped ssh hex gimp
[EMAIL PROTECTED] bryn $ bg
[3]+ ssh hex gimp &
(2) [EMAIL PROTECTED] bryn $
[3]+ Stopped ssh hex gimp
[EMAIL PROTECTED] bryn $
All I did at (1) & (2) was hit enter. I'm using a normal UK keyboard &
keymap so I don't think the KB is spitting out funny chars...
I can work around this locally by using the -f switch because I use
public key logins, but this will break keyboard logins for remote sites.
Cheers
Bryn
--
[EMAIL PROTECTED] mailing list
Re: [gentoo-user] Re: Regarding your posts on the GUML
Stroller wrote: On 16/6/03 1:01 pm, "Zack Gilburd" <[EMAIL PROTECTED]> wrote: Can we have a bit more context please? I have no idea where this thread came from! I've noticed that on your posts to the GUML, you advertise your want to be employed; Gizza job, mister..! >... However, I do not think it would be wise to advertise yourself as a sysadmin, then post to a ML with your root address. Pardon..? [snip] Erm... but, of course, I do. I rather assumed that would be obvious. Not if the account appears to be named 'root' - there's so many CVs in the pile these days it's just too easy to bin it on little things like that. I know it's a bad assumption but it's probably one I'd make just because of how it *looks*! It's not really THAT seemingly, Sir: [silva:~] stroller% host stellar.eclipse.co.uk stellar.eclipse.co.uk mail is handled (pri=10) by mx1.ex.eclipse.net.uk stellar.eclipse.co.uk mail is handled (pri=20) by mx2.ex.eclipse.net.uk [silva:~] stroller% nslookup stellar.eclipse.co.uk Server: gentoo.lan Address: 192.168.1.43 *** No address (A) records available for stellar.eclipse.co.uk [silva:~] stroller% As you can see, stellar.eclipse is virtually hosted. Is it really a risk for me to post using this address..? If I worked in Human Resources I'd never have got past 'Sir'. ;-) Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] ssh keeps stopping!
Mark Knecht wrote:
Bryn,
I have not seen this problem. I'm on a slightly newer gentoo release
(1.4rc4)
I'm running using
shh -X -C -c blowfish remote_IP
I'm Gentoo on both ends. I run Evolution multiple times every day with no
big problems. I did write this list a week or two ago about the
disconnection of ssh hanging on the local end, and needing to do a Ctrl-C to
get around that. I got no responses, and the problem seems to have
disappeared a few days ago.
Cheers,
Mark
Mark,
Thanks for the reply - I've done an emerge -u world recently, so we
should be at or around the same versions, apart from the very basic
stuff but this seems very ssh-specific.
I've just confirmed it's not just X apps - even ls will do it:
[EMAIL PROTECTED] bryn $ ssh hex ls -lR > /dev/null 2>&1 &
[2] 1889
[EMAIL PROTECTED] bryn $ ls
FrozenPenguins.png foo.javashare shot0003.jpg
Screenshot.png questframe shot0002.jpg
[2]+ Stopped ssh hex ls -lR >/dev/null 2>&1
[EMAIL PROTECTED] bryn $
The [2]+ Stopped occurs as a side effect of issuing ls - no kb
interrupts sent.
If you do anything on a terminal with a background ssh job, it stops.
This just feels like a silly config issue to me but I can't see what
I've done wrong!
Also checked and it happens on a vanilla xterm or vt too.
Cheers
Bryn
--
[EMAIL PROTECTED] mailing list
[gentoo-user] ssh keeps stopping!
I've a problem using ssh's X11 forwarding under gentoo-1.4rc2. The service works, but if you start an ssh session say with: # ssh hex /usr/local/mozilla/mozilla & The app opens fine, but ocassionally (and annoyingly!) the job becomes stopped and the app will hang. Sometimes hitting bg will sort the problem, others it's necessarry to do a fg and hit a couple of carriage-returns to unblock the app. This happens with all the gui apps I've tried. I'm using gnome-terminal 2.2.1, bash 2.05b.0(1), OpenSSH_3.6.1p2 all on a pretty much stock 2.4.20-gentoo kernel. Am I doing something dumb here? Cheers Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] play [amsn]
MooktaKiNG wrote: You'll need to start a sound server. When using KDE arts (which is a sound server) is started. arts puts all thos programs that want to use sound together. So one program doesn't have to wait for the other to finish. You can start arts without using kde. just do "artsd &". Then use the play command "artsplay" for amsn. This will make amsn using arts instead of play, which i think is used by OSS. [EMAIL PROTECTED] wrote: Hi, I´m not sure how many of you use amsn, but when you go to tools>options>preferences...>sounds there it says ¨play¨. I think I´m missing that package because I don´t hear any sound when my pals go online, or anything else, and I think that you´re supposed to hear sound by default in amsn. How do I get that package? Thank you,fg ZiM I think the problem's more basic than that - if play cannot be found, SoX probably hasn't been emerged. I just checked in the amsn ebuild and it doesn't depend on sox. imho, it's more likely that's missing than a sound daemon. Try 'emerge sox' and see if that sorts the problem. I don't use amsn anymore (too nasty! isn't there a better im with msn integration for *nix?) but maybe someone should tell the ebuild maintainer about the missing dependancy. hope that helps Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Which java-sdk?
Actually, those are native threads, unless you have an old JVM. Linux lists all threads in the output of ps aux. The only way I know to tell when they are actually threads in the same process is that a lot of the stats reported (memory size, resident memory, etc.) are identical. I haven't looked carefully at the ps man page in a while, though. - richard ps -ax --forest gives a tree for each process group - Bryn -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Performing Global Updates
Aaron Stout wrote: Just curious what this message is all about. I have been receiving this message lately in when emerge syncing. Performing Global Updates: /usr/portage/profiles/updates/2Q-2003 (Could take a couple minutes if you have a lot of binary packages.) .='update pass' *='binary update' @='/var/db move' [EMAIL PROTECTED] ** Skipping packages. Run 'fixpackages' or set it in FEATURES to fix the tbz2's in the packages directory. Note: This can take a very long time. The "Run 'fixpackeges'" part. is that part of the gentoolkit or emerge? It updates binary packages to reflect changes in the location of packages in the portage tree. afaik it's not needed often and runs quick unless you have lots of binary packages. You'll find it in /usr/lib/portage/bin there's also a question about it in the faq section of the forums Bryn. -- [EMAIL PROTECTED] mailing list
