[gentoo-user] How to prevent Firefox from switching VT?
Hello, I'm using mutt as MUA in an xterm on VT1 and calling up, via urlview, FF on VT2 with a web page from a mail URL. Up to some time recently, Firefox only switched VT the first time I did this, now it does it every time, annoyingly. I would like to make FF stick to a specific VT, or just not switch VT when called from another. Possible? Regards, ralf
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tue, Jan 06, 2009 at 07:54:30AM +, Neil Bothwick wrote: That's just FUD. You should tell it to the documentation team. -- Nicolas Sebrecht
[gentoo-user] rsh failed : Connection reset by peer
Hi,
I want to use iozone to test my cluster, and I found that iozone need
rsh. But rsh never worked in my machine.
I have followed this guide
http://www.gentoo.org/doc/en/hpc-howto.xml#doc_chap2, which including
the configuration of rsh.
# eix netkit-rsh
[I] net-misc/netkit-rsh
Available versions: 0.17-r8 ~0.17-r9 {pam}
Installed versions: 0.17-r8(08:51:30 10/20/08)(pam)
Homepage:ftp://ftp.uk.linux.org/pub/linux/Networking/netkit/
Description: Netkit's Remote Shell Suite: rexec{,d}
rlogin{,d} rsh{,d}
# eix xinet
[I] sys-apps/xinetd
Available versions: 2.3.14 {perl tcpd}
Installed versions: 2.3.14(09:02:57 PM 01/05/2009)(perl tcpd)
Homepage:http://www.xinetd.org/
Description: powerful replacement for inetd
# cat /etc/xinetd.d/rsh
service shell
{
socket_type = stream
protocol= tcp
wait= no
user= root
group = tty
server = /usr/sbin/in.rshd
log_type= FILE /var/log/rsh
log_on_success = PID HOST USERID EXIT DURATION
log_on_failure = USERID ATTEMPT
disable = no
}
# cat /etc/xinetd.d/rlogin
service login
{
socket_type = stream
protocol= tcp
wait= no
user= root
group = tty
server = /usr/sbin/in.rlogind
log_type= FILE /var/log/rlogin
log_on_success = PID HOST USERID EXIT DURATION
log_on_failure = USERID ATTEMPT
disable = no
}
I have a node07 whose is is 192.168.1.7 and a master, whose ip is 192.168.1.1.
I tried to rlogin to master from node73 with any user (i mean both
root and other users), I got the error in node07:
node07 # rlogin -l root master
rcmd: master: Connection reset by peer
and I read the log from /var/log/rlogin in master:
# cat /var/log/rsh
[...]
09/1/5...@23:10:36: FAIL: login address
09/1/5...@23:10:36: START: login pid=8961 from=192.168.1.7
09/1/5...@23:10:36: EXIT: login status=0 pid=8961 duration=0(sec)
And the master can't rlogin to itself, too: with root, because t:
master # rlogin -l root localhost
Password:
Password:
Login incorrect
/*rlogin with root, the password is incorrect forever and Of course,
I am sure the password is correct */
master # rlogin -l wcw localhost
Password:
Last login: Mon Jan 5 23:23:06 CST 2009 from localhost on pts/8
rlogin: connection closed.
/*rlogin with other user(wcw), it will succeed, but the connection
will closed immediately */
# cat /var/log/rsh
[...]
09/1/5...@23:20:09: START: login pid=10227 from=127.0.0.1
09/1/5...@23:21:17: EXIT: login status=0 pid=10227 duration=68(sec)
09/1/5...@23:22:13: START: login pid=10954 from=127.0.0.1
09/1/5...@23:22:15: EXIT: login status=0 pid=10954 duration=2(sec)
# cat /etc/hosts.allow
ALL:192.168.1.0/255.255.255.0
# cat /etc/hosts.equiv
master
node07
Anybody can help?
Thanks in advanced!
--
wcw
Re: [gentoo-user] How to prevent Firefox from switching VT?
On Tue, Jan 06, 2009 at 09:21:35AM +0100, Ralf Stephan wrote: I'm using mutt as MUA in an xterm on VT1 and calling up, via urlview, FF on VT2 with a web page from a mail URL. Up to some time recently, Firefox only switched VT the first time I did this, now it does it every time, annoyingly. I would like to make FF stick to a specific VT, or just not switch VT when called from another. Possible? Are you talking about that ? https://bugzilla.mozilla.org/show_bug.cgi?id=425879 I guess that what you call a VT (virtual _terminal_) is a virtual workspace. -- Nicolas Sebrecht
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tue, Jan 6, 2009 at 1:44 AM, Nicolas Sebrecht nicolas.s-...@laposte.net wrote: On Tue, Jan 06, 2009 at 12:59:13AM -0200, Francisco Ares wrote: snip Avoid the HTML format in your mail, please. Also, learn to quote. sorry about that -- Nicolas Sebrecht Francisco
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
Dirk Heinrichs wrote: Am Dienstag, 6. Januar 2009 02:56:28 schrieb Denis: I have Intel network hardware that runs on the E1000 driver Did you try e1000e? Bye... Dirk e1000e had been disabled in 2.6.27 versions, was it put back in by -r7?
Re: [gentoo-user] Re: weird cron mail problem
090105 Harry Putnam wrote: Philip Webb purs...@ca.inter.net writes: The problem originated 090104 c0520 , when I edited ~/.fetchmailrc to delete the reference to a logfile. However, attempts to restore the STATVS QVO ANTE have failed: I've restored the previous version of .fetchmailrc without success I've remerged Fetchmail, rebooted then run fetchmailconf , but the crazy mails continue to appear every 5 min in my inbox. I think that looks like normal cron mail... Looks like you may have once had a redirect to `/dev/null' in the crontab line and inadvertently removed it. No, I have never touched the cron part of it since installing 0710xx . It looks to me as if Fetchmail reacts badly if you change anything by hand in ~/.fetchmailrc : I did exactly that without any problem when I changed ISPs 0809xx, but I also find it odd that Fetchmail is sending the mail to Uniserve, but not to my U Toronto e-ddress, which is also in .fetchmailrc . I have a few ideas of things to try today will try yours too (thanks). Any suggestions from others are very welcome. My fetchmail line in crontab: */15 * * * * /usr/bin/fetchmail -f /home/reader/.fetchmailrc /dev/null 21 Note the difference with yours: */5 * * * * /usr/bin/fetchmail Not redirect in yours but note that I dump any output to /dev/null You may first want to just say `cmd /dev/null' , so that any errors are still send to you but once its working smoothly you can add 21 like `cmd /dev/null 21', so that both stderr and stdout go to dev/null I think you can also do the same thing like this `cmd /dev/null -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
Am Dienstag, 6. Januar 2009 11:14:08 schrieb Steven Susbauer: e1000e had been disabled in 2.6.27 versions, was it put back in by -r7? Don't know. In vanilla-sources, it has always been there. Bye... Dirk signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On 06.01.2009 12:14, Steven Susbauer wrote: Dirk Heinrichs wrote: Am Dienstag, 6. Januar 2009 02:56:28 schrieb Denis: I have Intel network hardware that runs on the E1000 driver Did you try e1000e? Bye... Dirk e1000e had been disabled in 2.6.27 versions, was it put back in by -r7? Supposed to be back in 2.6.27-r1. Can't check at the moment tho. http://bugs.gentoo.org/show_bug.cgi?id=238489 -- Eray
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tue, 6 Jan 2009 09:34:31 +0100, Nicolas Sebrecht wrote: That's just FUD. You should tell it to the documentation team. The last time I looked at the docs for this, which was a while ago, they recommended against using it for major version updates, which makes perfect sense. -- Neil Bothwick Artificial Intelligence usually beats real stupidity. signature.asc Description: PGP signature
Re: [gentoo-user] init script freezes when xdm (kdm) starts
On Tuesday 06 January 2009 02:02:16 Francisco Ares wrote: This is a snippet of the end of the boot screen: * Setting framebuffer console images ... [ ok ] * Starting gpm ... [ ok ] * Setting up kdm ... [ ok ] * Loading ALSA modules ... * Loading: snd-seq-oss ... [ ok ] * Loading: snd-pcm-oss ... [ ok ] * Restoring Mixer Levels ... [ ok ] * Starting * no interface module has been loaded That looks suspicious to me. Init seems to be trying to run an (in effect) empty file in /etc/init.d/. What else have you got in there that doesn't belong? This is mine: $ ls -w 76 /etc/init.d acpiddbus lm_sensors numlock spamd alsasounddepscan.shlocal pwcheck sshd bootmisc dnsextd localmount pydoc-2.4 syslog-ng checkfs fancontrolmdnsd pydoc-2.5 udev-postmount checkrootfunctions.sh mDNSResponderPosix reboot.sh urandom chronyd gkrellmd modules rmnologin vixie-cron clockgpm net.eth0rsyncdxdm consolefont halt.sh net.lo runscript.sh crypto-loop hostname netmountsaslauthd cupsdkeymaps nscdshutdown.sh Four of those are soft links: depscan.sh, functions.sh, net.eth0 and runscript.sh, thus: $ (cd /etc/init.d ls -l depscan.sh functions.sh net.eth0 runscript.sh) lrwxrwxrwx 1 [...] depscan.sh - ../../sbin/depscan.sh lrwxrwxrwx 1 [...] functions.sh - ../../sbin/functions.sh lrwxrwxrwx 1 [...] net.eth0 - net.lo lrwxrwxrwx 1 [...] runscript.sh - ../../sbin/runscript.sh Your script execution order doesn't look right to me. This is what happens on my system: INIT: Entering runlevel: 3 * Starting eth0 * Bringing up eth0 * 192.168.2.6 [ ok ] * Adding routes * default via 192.168.2.1 ... [ ok ] * Mounting network filesystems ... [ ok ] * Setting up kdm ... [ ok ] * Loading ALSA modules ... * Loading: snd-pcm-oss ... [ ok ] * Loading: snd-seq-oss ... [ ok ] * Restoring Mixer Levels ... [ ok ] * Starting chronyd ... [ ok ] * Starting D-BUS system messagebus ... [ ok ] * Starting cupsd ... [ ok ] * Loading lm_sensors modules... * Loading i2c-nforce2 ...[ ok ] * Loading adm1026 ...[ ok ] * Loading w83627hf ... [ ok ] * Loading k8temp ... [ ok ] * Initializing sensors ... [ ok ] * Starting sshd ...[ ok ] * Starting vixie-cron ... [ ok ] * Starting local ... [ ok ] (I don't run gpm if I'm starting X; I have a separate soft-level for console operation, and it does run there.) Have you sync'd recently? Which profile did you install from? This is mine: $ ls -l /etc/make.profile lrwxrwxrwx 1 root root 50 2008-12-10 11:24 /etc/make.profile - ../usr/portage/profiles/default/linux/amd64/2008.0 You might like to look at these again, as they also affect the execution order: $ grep PLUG /etc/conf.d/rc | grep -v \# RC_HOTPLUG=yes RC_COLDPLUG=yes RC_PLUG_SERVICES=!* -- Rgds Peter
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tuesday 06 January 2009 11:42:58 Neil Bothwick wrote: On Tue, 6 Jan 2009 09:34:31 +0100, Nicolas Sebrecht wrote: That's just FUD. You should tell it to the documentation team. The last time I looked at the docs for this, which was a while ago, they recommended against using it for major version updates, which makes perfect sense. I've seen a dev complain that most of the problems he's had to deal with have been due to oldconfig, so I don't use it at all. I copy the config in from the previous tree, then I run menuconfig and search for lines ending in [NEW] or [DEPRECATED]. This is far quicker and less error-prone than going through the entire set of options. -- Rgds Peter
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tue, 6 Jan 2009 12:13:05 +, Peter Humphrey wrote: I've seen a dev complain that most of the problems he's had to deal with have been due to oldconfig, so I don't use it at all. I copy the config in from the previous tree, then I run menuconfig and search for lines ending in [NEW] or [DEPRECATED]. How is this different from using make oldconfig, apart from the UI? -- Neil Bothwick Approx. 1 in 36000 people will break a leg within 3 weeks of reading this post signature.asc Description: PGP signature
[gentoo-user] Re: rsh failed : Connection reset by peer
I think I have fixed part of the problem, I found this option only_from = localhost in file /etc/xinetd.conf, and mask it. And now I can rlogin into localhost or remotehost: node07 # rlogin master Password: Last login: Tue Jan 6 20:40:36 CST 2009 from node07 on pts/4 rlogin: connection closed. master # rlogin localhost Password: Last login: Tue Jan 6 20:56:42 CST 2009 from node07 on pts/8 rlogin: connection closed. But as you can see, after login successfully, the connection is closed immediately. I dont' know whether it's normal or not, as I never used rsh before. And now, although rlogin can be use, but rsh still don't work: -- wcw
[gentoo-user] Re: rsh failed : Connection reset by peer
I am very sorry for the last e-mail, as I intented to save the email but clicked the send button. I will continue the last letter below. On Tue, Jan 6, 2009 at 9:08 PM, Chuanwen Wu wcw8...@gmail.com wrote: I think I have fixed part of the problem, I found this option only_from = localhost in file /etc/xinetd.conf, and mask it. And now I can rlogin into localhost or remotehost: node07 # rlogin master Password: Last login: Tue Jan 6 20:40:36 CST 2009 from node07 on pts/4 rlogin: connection closed. master # rlogin localhost Password: Last login: Tue Jan 6 20:56:42 CST 2009 from node07 on pts/8 rlogin: connection closed. But as you can see, after login successfully, the connection is closed immediately. I dont' know whether it's normal or not, as I never used rsh before. And now, although rlogin can be use, but rsh still don't work: /*rsh into remotehost*/ node07 # rsh -l wcw master pwd assword: ^C /* rsh into localhost */ master # rsh localhost pwd assword: a ^C It's very weird as you can see above, after I entered the rsh command, the assword but not password appleared. The is the characters I typed. Any help will be very appreciated! -- wcw
[gentoo-user] wpa_supplicant with madwifi partially stopped working after remerge
Hello. A few hours ago I tried to upgrade from 2.6.26-gentoo-r4 to 2.6.27-gentoo-r7. As expected, WLAN stopped working due to missing modules. I remerged madwifi-hal and madwifi-hal-tools contained in the overlay from allenjb to get the needed kernel modules. Afterwards remerged wpa_supplicant. But WLAN still did not work. So I booted to my prevoius kernel and found WLAN non functional there too. I switched the link /usr/src/linux back to the previous kernel and tried to remerge madwifi-hal, the tools and wpa_supplicant again. Now the modules do correctly (I suppose) create the needed devices and the init scripts pull the interface up. But the boot messages complain about not having set WEP-Keys, which are set in the config file residing at /etc/wpa_supplicant/wpa_supplicant.conf. That results in my WLAN is associated to an unsecured access point in my area. Since I did not touch the config files, I wondered if the standard location of config files regarding wpa_supplicant moved, but could not find anything related. I tried wpa_cli to get some information, but it can't connect to wpa_supplicant. So I tried to run wpa_supplicant manually: wpa_supplicant -Dmadwifi -iwifi0 -c/etc/wpa_supplicant/wpa_supplicant.conf ioctl[SIOCSIWMODE]: Operation not supported Could not configure driver to use managed mode ioctl[SIOCGIWRANGE]: Operation not supported ioctl[IEEE80211_IOCTL_SETPARAM]: Operation not supported ioctl[SIOCSIWAP]: Operation not supported Failed to initialize driver interface Speicherzugriffsfehler Now I am stuck. Any suggestions how to get WLAN back working? Jan
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tuesday 06 January 2009 12:45:15 Neil Bothwick wrote: On Tue, 6 Jan 2009 12:13:05 +, Peter Humphrey wrote: I've seen a dev complain that most of the problems he's had to deal with have been due to oldconfig, so I don't use it at all. I copy the config in from the previous tree, then I run menuconfig and search for lines ending in [NEW] or [DEPRECATED]. How is this different from using make oldconfig, apart from the UI? I don't know; I just pick up my clues where I can. Perhaps there's a difference in handling of unchanged or default values. On the other hand, maybe the dev was complaining about the way people use the two methods rather than the way they operate, in which case I can be a bit less paranoid. As I don't have the time to investigate how Make works, I propose to continue to take the path of least resistance. -- Rgds Peter
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tue, 6 Jan 2009 15:06:28 +, Peter Humphrey wrote: I've seen a dev complain that most of the problems he's had to deal with have been due to oldconfig, so I don't use it at all. I copy the config in from the previous tree, then I run menuconfig and search for lines ending in [NEW] or [DEPRECATED]. How is this different from using make oldconfig, apart from the UI? I don't know; I just pick up my clues where I can. Perhaps there's a difference in handling of unchanged or default values. Unchanged values are just that. When a config option is new, make oldconfig prompts for a choice,with a default option (the same default that menuconfig uses). The only real difference is that with menuconfig you have to go through the options, looking for those marked NEW (and risking missing an important one) while oldconfig presents them to you in sequence,asking for your choice on each one. I have boxes running 2.6.28 on which I have used oldconfig on every change since switching them from2.4 to 2.6. -- Neil Bothwick You know it's going to be a bad day when you forget your new password. signature.asc Description: PGP signature
[gentoo-user] Re: kernel linux-2.6.27-gentoo-r7 won't load network!
Denis denis.che at gmail.com writes: I was just updating from kernel linux-2.6.24-gentoo-r7 to linux-2.6.27-gentoo-r7 using the oldconfig method, and I cannot get the new kernel to load my network. I have Intel network hardware that runs on the E1000 driver, which I generally compile right into the kernel - never had any problem. For some reason, the linux-2.6.27-gentoo-r7 kernel leaves me netless. We seem to be getting this a lot lately. I upgraded a sony laptop (intel cpu, eth etc) and had no problems with 2.6.27-gentoo-r7, right out of the box. However, I do not use 'make oldconfing'. What I do is use 'make menuconfig' and most of the meaningful options are carried forward (automatically) into the new kernel configurations. I also save both the old and new configs for comparison and trouble shooting. for Example on an amd 64 after the kernel is built: cd /usr/src make menuconfig (make choices in gui-menu) make make modules_install cp System.map /boot/System.map-2.6.27-gentoo-r7 cp arch/x86_64/boot/bzImage /boot/kernel-2.6.27-gentoo-r7 cp .config /boot/config-2.6.27-gentoo-r7 edit grub.conf The .config file saving is really worth keeping around for a few revs in case you experience problems. I know lost of folks use 'oldconfig' but I think it is problematic. (Surely, I'll get a few flames here). Particularly when your options do need to change as the kernel is routinely reorganized (sometimes) when new version numbers change I.E. (2.6.26 -- 2.6.27) for example. Using the .config files top parse line by line (when needed) is a very excellent methodology, methinks, to troubleshoot. YMMV, James
Re: [gentoo-user] Re: kernel linux-2.6.27-gentoo-r7 won't load network!
Looks like there are other bugs filed elsewhere on the net about E1000 not loading with the 2.6.27 kernel. Here's a curious note from https://bugs.launchpad.net/launchpad-foundations/+bug/275611 = If I remove the line of the card in /etc/udev/rules.d/70-persistent-net.rules and boot with kernel 2.6.27-4, nothing happens, it's like the card does not exist. If I modprobe e1000, and try to bring up manually eth0: eth0: error fetching interface information: Device not found If I boot back in kernel 2.6.27-3, everything works fine, /etc/udev/rules.d/70-persistent-net.rules is correctly configured, and so on. = Sounds like something's up. I will try to configure my 2.6.27 using menuconfig, but why do I get the sense that I'll just waste time doing that and end up at square one?
[gentoo-user] Re: rsh failed : Connection reset by peer
I guessed it is the pam problem, and I re-installed netkit-rsh with USE=-pam, now I can use rsh with common users, but root still not work: # rsh -l wcw master date Tue Jan 6 23:33:35 CST 2009 # rsh -l root master date Permission denied. Everytime when I tried to use root to login, I got the error Permission denied. master # cat /root/.rhosts node07 root node07 wcw master # ls -l /root/.rhosts -rw--- 1 root root 33 Jan 6 23:31 /root/.rhosts I guess maybe rsh does not allow to login as root. So anyway to solved this problem? Thanks! -- wcw
Re: [gentoo-user] How to install libselinux if it is masked?
Am Dienstag, 6. Januar 2009 17:54:29 schrieb Frank Schwidom: Hi, i want to setup selinux, but it seems to failure because of no installable tools and libs. All necessary packages are masked by /usr/portage/profiles/default/linux/package.mask and no alternative libs and tools are within reach. How can i solve this problem? Hmm, maybe by switching to a selinux profile? HTH... Dirk signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] weird cron mail problem
On Mon, Jan 05, 2009 at 05:28:41PM -0500, Penguin Lover Philip Webb squawked: 090105 Willie Wong wrote: On Mon, Jan 05, 2009 at 01:04:45PM -0500, Penguin Lover Philip Webb squawked: have you checked your outgoing mail logs to see if cron attempts to send mail when fetchmail is called? Thanks for the suggestion: the last few lines of /var/log/syslog are : Jan 5 17:10:01 localhost cron[4753]: (root) CMD (test -x /usr/sbin/run-crons /usr/sbin/run-crons ) Jan 5 17:10:01 localhost cron[4754]: (purslow) CMD (/usr/bin/fetchmail) Jan 5 17:10:11 localhost sSMTP[4765]: Sent mail for r...@ca.inter.net (221 smtp-relay1.uniserve.ca closing connection) uid=1000 username=purslow outbytes=1049 Jan 5 17:15:01 localhost cron[4908]: (purslow) CMD (/usr/bin/fetchmail) Jan 5 17:15:09 localhost sSMTP[4909]: Sent mail for r...@ca.inter.net (221 smtp-relay2.uniserve.ca closing connection) uid=1000 username=purslow outbytes=952 Okay, this suggests that either cron or your mailer is misconfigured. From the logs, you are running ssmtp. Check the config files for it to see why mail intended for r...@localhost is delivered to r...@ca.inter.net. And check /etc/crontab to see whether the MAILTO field is set properly to where you want the mail to be delivered. Also, I don't know which implementation of cron you are using, but if your crontab is under userid, then usually cron will demand notification e-mails sent to userid... unless, hum, do a ls -l /var/spool/cron/crontabs/ Is your user's crontab owned by root for some reason? Actually, what is your cron recipe anyway? /var/log/spool/cron/crontabs/userid is : # DO NOT EDIT THIS FILE - edit the master and reinstall. # (/tmp/crontab.wsshN7 installed on Sun Nov 4 23:46:22 2007) # (Cron version V5.0 -- $Id: crontab.c,v 1.12 2004/01/23 18:56:42 vixie Exp $) */5 * * * * /usr/bin/fetchmail I considered trying 'fetchmail -s', but I'm not sure how to add a flag in a cron file (perhaps by using a script). for what it's worth, I have my fetchmail line as /usr/bin/fetchmail /dev/null so I only get e-mails when fetchmail writes to stderr (so when there's an error). Writing fetchmail -s instead of what you have really shouldn't be a problem. Why do you think it needs a script? Also, have you updated either cron or fetchmail recently? The problem originated 090104 c0520 , when I edited ~/.fetchmailrc to delete the reference to a logfile. However, attempts to restore the STATVS QVO ANTE have failed: I've restored the previous version of .fetchmailrc without success I've remerged Fetchmail, rebooted then run fetchmailconf , but the crazy mails continue to appear every 5 min in my inbox. Okay, running fetchmail -s (or redirecting fetchmail output to /dev/null) will probably cure the problem of the deluge. But I still think you MDA is misconfigured for local mail. W -- Q: What's grey and proves the nondenumerability of the Reals? A: Cantor's Diagonal Elephant Sortir en Pantoufles: up 760 days, 15:25
Re: [gentoo-user] weird cron mail problem
On Tue, Jan 06, 2009 at 11:56:55AM -0500, Penguin Lover Willie Wong squawked: Okay, running fetchmail -s (or redirecting fetchmail output to /dev/null) will probably cure the problem of the deluge. But I still think you MDA is misconfigured for local mail. Oops, my mistake. I meant MTA, not MDA in the above. Sorry, W -- Pintsize: Nooo! I'm lactose intolerant! Sortir en Pantoufles: up 760 days, 15:48
Re: [gentoo-user] Network printing
On Tue, Jan 6, 2009 at 7:39 AM, Peter Humphrey pe...@humphrey.ukfsn.org wrote: Hello, As it's the New Year, perhaps it's time to resume banging my head on this wall again. I have a server with two printers connected, set up using the cups Web page and operating properly. Now I want to send print jobs to them from my workstation, which is on the same network, and with the same version of cups: 1.3.9-r1. One of the printers is an HP Deskjet D4260, so I also have hplip version 2.8.6b installed on both machines. I can connect either printer to either machine and print locally without any problems. However, I cannot get anything to print over the network. If, on the workstation, I declare the network laser printer and connect to it, all appears to work until I send a print job to it; the job sits in the lp queue locally, and when I next look at the status of the printer it says Destination printer does not exist! If I try to set up the Deskjet as a remote printer in the local cups server, I get Filter foomatic-rip for printer HP_Deskjet_D4260 not available: No such file or directory. I don't know what to try next. Anyone any idea here? -- Rgds Peter Hi Peter, For the sake of conversation how about emerge flags? My server: Sector9 ~ # emerge -pv cups hplip These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild R ] net-print/cups-1.3.9-r1 USE=X acl dbus jpeg ldap pam perl png ppds python samba ssl tiff -avahi -gnutls -java -kerberos -php -slp -static -xinetd -zeroconf LINGUAS=en -de -es -et -fr -he -id -it -ja -pl -sv -zh_TW 0 kB [ebuild R ] net-print/hplip-2.8.6b USE=dbus ppds qt3 qt4 -cupsddk -doc -fax -minimal -parport -scanner -snmp 0 kB One of my clients: dragonfly ~ # emerge -pv cups hplip These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild R ] net-print/cups-1.3.9-r1 USE=X acl dbus jpeg ldap pam perl png ppds python ssl tiff -avahi -gnutls -java -kerberos -php -samba -slp -static -xinetd -zeroconf LINGUAS=en -de -es -et -fr -he -id -it -ja -pl -sv -zh_TW 0 kB [ebuild R ] net-print/hplip-2.8.6b USE=dbus ppds qt3 qt4 -cupsddk -doc -fax -minimal -parport -scanner -snmp 0 kB I can send config files again if you need them but they haven't changed in the few weeks since you looked at this. Cheers, Mark
Re: [gentoo-user] How to install libselinux if it is masked?
Hi, You probably wish to read the Gentoo SELinux Handbook [1] and the Gentoo Security Handbook [2] ;) [1] http://www.gentoo.org/proj/en/hardened/selinux/selinux-handbook.xml [2] http://www.gentoo.org/doc/en/security/security-handbook.xml Sincerely, Jil Frank Schwidom a écrit : Hi, i want to setup selinux, but it seems to failure because of no installable tools and libs. All necessary packages are masked by /usr/portage/profiles/default/linux/package.mask and no alternative libs and tools are within reach. How can i solve this problem? Regards
Re: [gentoo-user] Microphone not working
The microphone on my laptop isn't working and I'm not sure how to troubleshoot it. I've tried using it in twinkle and arecord but it doesn't work in either. The headset works fine on a different system. Can anyone help with this? - Grant * Do you get an error? I don't get an error. * Have you unmuted the mic and set it as the input device (in alsamixer)? It actually doesn't appear in alsamixer although it used to when it was working. I have Master, PCM, Line In, Mic Boost, and Internal. It shows View: [Playback] at the top. * My laptop has both an internal mic and an external mic port. Have you selected the correct one for the input device? xfce's volume control does allow me to select between the input sources Mic and Front Mic but I've tried both. Is the mic input addressed as the same device hw(0,0) as the output? - Grant
Re: [gentoo-user] weird cron mail problem
090106 Willie Wong wrote: this suggests that either cron or your mailer is misconfigured. From the logs, you are running ssmtp. Check the config files for it to see why mail intended for r...@localhost is delivered to r...@ca.inter.net. /etc/ssmtp/ssmtp.conf has the lines (the last refers to my ISP): ... # The person who gets all mail for userids 1000 # Make this empty to disable rewriting. root=postmaster ... # The full hostname hostname=ca.inter.net ... It looks as if this combination causes it to send to r...@ca.inter.net . I can't find a man file for ssmtp.conf , only for ssmtp , tho' the latter says the former should exist, so I can't check in any more detail just how the lines above work, eg what rewriting refers to or how the hostname is used. And check /etc/crontab to see whether the MAILTO field is set properly to where you want the mail to be delivered. MAILTO=root this is for system cron jobs, which are run by for root. Also, I don't know which implementation of cron you are using, but if your crontab is under userid, ... Yes, with mail that's necessary for security. ... then usually cron will demand notification e-mails sent to userid: Try 'ls -l /var/spool/cron/crontabs/' : is your user's crontab owned by root for some reason? No : root:550 crontabs ls -l -rw--- 1 purslow crontab 234 2007-11-04 23:46 purslow I have my fetchmail line as /usr/bin/fetchmail /dev/null so I only get e-mails when it writes to stderr (when there's an error). That looks like the quick fix. Writing fetchmail -s instead shouldn't be a problem. Why do you think it needs a script? I didn't know if cron will ignore stuff after the space otherwise, but obviously it does, if your line above works. I still think you MTA is misconfigured for local mail. Like you, I'ld like to understand the real cause of the problem: Gentoo is for people like us ... (smile) Rewriting .fetchmailrc using fetchmailconf thro'out didn't help, so that's one set of possibilities ruled out I will try the suggestions above (thanks) next. Your or others' further comments are very welcome. -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
Re: [gentoo-user] Network printing
I think you are very close to getting it to work but just need to get LP configured correctly. It sounds like you have both systems configured as the Server, which is not correct. For your workstation, follow the Client directions below: http://www.gentoo.org/doc/en/quick-samba-howto.xml#doc_chap5 Which is basically two steps: 1) Modify '/etc/cups/client.conf' and tell it where the server is. 2) Configure LP: - use lpstat to see the available printers - use lpoptions to set the default printer The beauty of CUPS is that any new printer you install on the server automatically becomes available to all clients. On Windows, you need additional drivers; but for other CUPS clients, it takes care of it for you internally (to my understanding). I believe you only need HPLIP on the server side, not the client side. But having it there shouldn't do any harm. HTH, Ben - Original Message From: Peter Humphrey pe...@humphrey.ukfsn.org To: gentoo-user@lists.gentoo.org Sent: Tuesday, January 6, 2009 10:39:48 AM Subject: [gentoo-user] Network printing Hello, As it's the New Year, perhaps it's time to resume banging my head on this wall again. I have a server with two printers connected, set up using the cups Web page and operating properly. Now I want to send print jobs to them from my workstation, which is on the same network, and with the same version of cups: 1.3.9-r1. One of the printers is an HP Deskjet D4260, so I also have hplip version 2.8.6b installed on both machines. I can connect either printer to either machine and print locally without any problems. However, I cannot get anything to print over the network. If, on the workstation, I declare the network laser printer and connect to it, all appears to work until I send a print job to it; the job sits in the lp queue locally, and when I next look at the status of the printer it says Destination printer does not exist! If I try to set up the Deskjet as a remote printer in the local cups server, I get Filter foomatic-rip for printer HP_Deskjet_D4260 not available: No such file or directory. I don't know what to try next. Anyone any idea here? -- Rgds Peter
Re: [gentoo-user] [Solved] Bash completion problem: can anyone confirm?
Okay, like always, this is an I-D-ten-T error. On Mon, Jan 05, 2009 at 02:20:19PM -0500, Penguin Lover Willie Wong squawked: So this suggests that, though I have not checked on my laptop, that on my aterms bash-completion is actually NOT activated, since my .bashrc almost certainly included the line to source bash-completion and not bash-completion.sh, which explains the different behaviour between an aterm and a login console. This is confirmed. My .bashrc didn't have the right file for bash-completion. Now it does and the behaviour on the console agrees with that in the aterm. The next postulate is that since Graham noticed the behaviour only manifesting itself for vim and turning off the vim completion (I remember vaguely that on my laptop 'eselect bashcomp list --global' shows vim completion as enabled) that this maybe a bug with the module relating to bash completion for vim. This is related to the other part of the ELOG which I didn't carefully read. The _filedir problem was that the _filedir function is defined in the base module for bash completion, which is a new addition in the current version. By reading the configs for the previous setup, the base module is not enabled by default. The vim module calls _filedir, which is undefined, and hence the error. The solution: eselect bashcomp enable --global 2 (on my box 2 corresponds to the base module). Best, W -- Proof that Suki is good: according to spaceballs: [now we know evil will always triumph, because] good is dumb according to S: ok, now I feel really dumb the proof follows. ~~~ (on W's attempt to prove the phrase suki is good by syllogism) S: it's okay, you don't need to proove [sic] it: it's a definition W: nice S: i spelled prove wrong S: geez W: don't worry W: that doesn't make you evil. Sortir en Pantoufles: up 760 days, 17:38
Re: [gentoo-user] weird cron mail problem
On Tue, Jan 06, 2009 at 01:22:28PM -0500, Philip Webb wrote: 090106 Willie Wong wrote: this suggests that either cron or your mailer is misconfigured. From the logs, you are running ssmtp. Check the config files for it to see why mail intended for r...@localhost is delivered to r...@ca.inter.net. /etc/ssmtp/ssmtp.conf has the lines (the last refers to my ISP): ... # The person who gets all mail for userids 1000 # Make this empty to disable rewriting. root=postmaster ... # The full hostname hostname=ca.inter.net ... It looks as if this combination causes it to send to r...@ca.inter.net . I can't find a man file for ssmtp.conf , only for ssmtp , tho' the latter says the former should exist, so I can't check in any more detail just how the lines above work, eg what rewriting refers to or how the hostname is used. Okay, I am now a bit confused. What is ca.inter.net? A whois shows no match. Does it refer to your ISP? or to your personal box? In any case, you may want to change the root line to root=purslow so the mail gets sent to purslow instead of postmaster (which according to /etc/mail/aliases becomes root again). Also, just a personal preference: but on systems that are on for long enough of periods to have daemon process that will be emailing stuff to sys admins (processes like cron and such), I'd usually install an actual MTA instead of something like ssmtp. W -- Willie W. Wong ww...@math.princeton.edu 408 Fine Hall, Department of Mathematics, Princeton University, Princeton A mathematician's reputation rests on the number of bad proofs he has given.
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
Neil Bothwick wrote: On Tue, 6 Jan 2009 15:06:28 +, Peter Humphrey wrote: I've seen a dev complain that most of the problems he's had to deal with have been due to oldconfig, so I don't use it at all. I copy the config in from the previous tree, then I run menuconfig and search for lines ending in [NEW] or [DEPRECATED]. How is this different from using make oldconfig, apart from the UI? I don't know; I just pick up my clues where I can. Perhaps there's a difference in handling of unchanged or default values. Unchanged values are just that. When a config option is new, make oldconfig prompts for a choice,with a default option (the same default that menuconfig uses). The only real difference is that with menuconfig you have to go through the options, looking for those marked NEW (and risking missing an important one) while oldconfig presents them to you in sequence,asking for your choice on each one. I have boxes running 2.6.28 on which I have used oldconfig on every change since switching them from2.4 to 2.6. I been using oldconfig a long time too and never had trouble with it. It basically will make the same kernel but with bug/security fixes. I even copied the config over then forgot to run oldconfig and it compiled and booted just fine. I think FUD was the right term for this one. Major versions, make it by hand; incremental upgrades works fine with oldconfig. Dale :-) :-)
Re: [gentoo-user] weird cron mail problem
090106 Willie Wong wrote: you may want to change the root line to root=purslow, so the mail gets sent to purslow instead of postmaster (which according to /etc/mail/aliases becomes root again). That doesn't work, but adding ' /dev/null' or '-s' in crontab does. The latter seems simpler, so that's what I've done. It doesn't explain why the problem suddenly arose last Sunday after I made a simple editing change in .fetchmailrc nothing like this had happened before with the same crontabssmtp.conf : perhaps there's an obscure bug, but the irritating problem has been resolved I have other jobs today. Thanks for the helpful advice (smile). -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
[gentoo-user] gentoo notebook
Hi group, My ISP just disappeared. *Poof*, gonzo. So I going to have to forage for wi-fi hotspots. Been looking at netbooks. I like the idea of the SS drive but I've heard that they're so compact they tend to deteriorate after lots of read/write cycles. What's been the group's experience? Must have a matte screen and hopefully a keyboard that's not too cramped. And gentoo-worthy, of course. Maxim
Re: [gentoo-user] gentoo notebook
On Tue, 6 Jan 2009 13:01:13 -0800 (PST), maxim wexler wrote: Been looking at netbooks. I like the idea of the SS drive but I've heard that they're so compact they tend to deteriorate after lots of read/write cycles. SSD drives have wear levelling, unlike flash memory cards and sticks, so they should last a lot longer. The most used directory on a Gentoo system is probably $PORTAGE_TMPDIR, which I have on an SDHC card. What's been the group's experience? Must have a matte screen and hopefully a keyboard that's not too cramped. And gentoo-worthy, of course. I've been using this Eee PC 900 for around eight months, now that the 2.6.28 kernel has support for both wired and wireless interfaces, just about everything works fine without and special software. -- Neil Bothwick Sex is better than logic. You can't prove it, but it is. signature.asc Description: PGP signature
[gentoo-user] Re: weird cron mail problem
Philip Webb purs...@ca.inter.net writes: 090106 Willie Wong wrote: you may want to change the root line to root=purslow, so the mail gets sent to purslow instead of postmaster (which according to /etc/mail/aliases becomes root again). That doesn't work, but adding ' /dev/null' or '-s' in crontab does. The latter seems simpler, so that's what I've done. It doesn't explain why the problem suddenly arose last Sunday after I made a simple editing change in .fetchmailrc nothing like this had happened before with the same crontabssmtp.conf : perhaps there's an obscure bug, but the irritating problem has been resolved I have other jobs today. Thanks for the helpful advice (smile). Looking back thru the thread I don't see the actual change made to fetchmailrc. Maybe just blind. From my experience fetchmail is a very robust (non buggy) and easily configured tool. At least for my simple usage. It's the only config I can think of that is edited much like you might talk to friend while walking along. Something that happens from time to time is introducing an unprintable CHAR into a *.rc file and not being able to see it. I'm not sure if fetchmail would respond poorly to that. If there is any chance of that; you might want to use vim to check each line. You can hit the el (l) lowercase, on each line to expose most kinds of unprintable char. It takes 3 key strokes to show the line. 1) : 2) l 3) enter Then the line appears in the command area along with any unprintable chars, As Willie mentioned the mail mta is capable of rewriting stuff in its configurations. Do you control this machine? Sorry if you've already covered that. Another unlikely thing that can catch you ... happened to me on a remote account I didn't control. The machine underwent some kind of mishap that required serious backup effort replacing all us users files from backups. Turned out the backups were pretty old and further I missed the notification about the mishap... the next thing I knew lots of strange things began to happen. Some of my older configurations were re-introduced in place of things I had changed due to new circumstances.
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
I'm doubtful that using oldconfig would make a driver not work because I have done it this way for a while, going between various 2.6.x versions, like 2.6.21 to 2.6.24... And there was never an issue. Maybe the E1000 driver somehow got messed up in this particular version of the kernel. Now, that e1000e driver someone mentioned - how is it different from the e1000 driver, and does the kernel float two versions, or it depends on which kernel it is? Thanks, Denis
Re: [gentoo-user] Re: kernel linux-2.6.27-gentoo-r7 won't load network!
Denis wrote: Looks like there are other bugs filed elsewhere on the net about E1000 not loading with the 2.6.27 kernel. Here's a curious note from https://bugs.launchpad.net/launchpad-foundations/+bug/275611 = If I remove the line of the card in /etc/udev/rules.d/70-persistent-net.rules and boot with kernel 2.6.27-4, nothing happens, it's like the card does not exist. If I modprobe e1000, and try to bring up manually eth0: eth0: error fetching interface information: Device not found If I boot back in kernel 2.6.27-3, everything works fine, /etc/udev/rules.d/70-persistent-net.rules is correctly configured, and so on. = Sounds like something's up. I will try to configure my 2.6.27 using menuconfig, but why do I get the sense that I'll just waste time doing that and end up at square one? Because if it is in the kernel, it is in the kernel. It doesn't matter if it is put there by oldconfig or menuconfig or some other config. It's either there built in, as a module or it is not there. Really simple. Dale :-) :-)
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Tuesday January 6 2009 20:37:31 Denis wrote: I'm doubtful that using oldconfig would make a driver not work because Unless the driver (module) changed name ;) (ata-pata/sata anyone?) Anyway, I've always used oldconfig except when moving from 2.4 to 2.6 Regards, Norberto
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
On Wednesday 07 January 2009 00:37:31 Denis wrote: I'm doubtful that using oldconfig would make a driver not work because I have done it this way for a while, going between various 2.6.x versions, like 2.6.21 to 2.6.24... And there was never an issue. Maybe the E1000 driver somehow got messed up in this particular version of the kernel. Now, that e1000e driver someone mentioned - how is it different from the e1000 driver, and does the kernel float two versions, or it depends on which kernel it is? e1000e breaks the hardware e1000 does not break the hardware Or maybe it's the other way round Anyway, two different drivers for the same hardware, the broken one was removed from the sources till it gets fixed. You can always build both drivers in a case like this but obviously you can only use one of them at a time -- alan dot mckinnon at gmail dot com
Re: [gentoo-user] kernel linux-2.6.27-gentoo-r7 won't load network!
Alan McKinnon wrote: On Wednesday 07 January 2009 00:37:31 Denis wrote: I'm doubtful that using oldconfig would make a driver not work because I have done it this way for a while, going between various 2.6.x versions, like 2.6.21 to 2.6.24... And there was never an issue. Maybe the E1000 driver somehow got messed up in this particular version of the kernel. Now, that e1000e driver someone mentioned - how is it different from the e1000 driver, and does the kernel float two versions, or it depends on which kernel it is? e1000e breaks the hardware e1000 does not break the hardware Or maybe it's the other way round Anyway, two different drivers for the same hardware, the broken one was removed from the sources till it gets fixed. You can always build both drivers in a case like this but obviously you can only use one of them at a time Actually, there was a bug with Ftrace overwriting the flash ROM of a NIC (which shouldn't have been mapped RW anyway) with e1000e in certain cases. IIRC this was only present in the 2.6.27 release candidates and got fixed by mapping it read-only for the final release. 2.6.28 then had the fix for the Ftrace bug. Regards, Andi
