Re: [gentoo-user] Verification of audio CD copy?
Mark Knecht markkne...@gmail.com wrote: OK - I'll try that in a minute. Right now I'm stuck. Attempting to run your readcd program it's locked up and I cannot kill it. Ctrl-C didn't work. Trying to kill the process doesn't work. This is on one of the CDs that doesn't play correctly so maybe it's showing the problem? m...@lightning ~/Desktop/cdda2wav $ readcd dev=1000,0,0 -c2scan Read speed: 7056 kB/s (CD 40x, DVD 5x, BD 1x). Write speed: 7056 kB/s (CD 40x, DVD 5x, BD 1x). Capacity: 263932 Blocks = 527864 kBytes = 515 MBytes = 540 prMB Sectorsize: 2048 Bytes readcd: Success. mode select g1: scsi sendcmd: no error CDB: 55 10 00 00 00 00 00 00 14 00 status: 0x2 (CHECK CONDITION) Sense Bytes: 70 00 05 00 00 00 00 0A 00 00 00 00 26 00 00 00 Sense Key: 0x5 Illegal Request, Segment 0 Sense Code: 0x26 Qual 0x00 (invalid field in parameter list) Fru 0x0 Sense flags: Blk 0 (not valid) cmd finished after 0.001s timeout 40s Copy from SCSI (1000,0,0) disk to file '/dev/null' end: 263932 ^C^C^C 1911 cnt: 49 ^C^C^C I cannot get out of this so far. Hate to think I need to reboot! - Mark Humm.rebooted and tried the command above with the CD that plays in the car. Same failure but slightly father along. (I think) It's reading 2009 instead of 1911, but again it seems stuck Bummer... Or is there a command that tells readcd to eventually give up? This SCSI command needs support inside the drive... It will not work for DVD media and some drives do not support it even for CDs. Also note that there is still an unfixed problem on Linux that under unknown conditions modifies the SCSI commands before they are send to the drive. As long as Linux does not allow to retrieve at least the 18 standard bytes of sense data there is no chance to find this problem as the drive writes the information about the location of the unwanted bits into the last 2 bytes that cannot be retrieved on Linux. Jörg -- EMail:jo...@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin j...@cs.tu-berlin.de(uni) joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
Re: [gentoo-user] Verification of audio CD copy?
Mark Knecht markkne...@gmail.com wrote: It seems that the command readcd -c2scan fails on both of my drives for all media types - retail CDs, copies of CDs on CR-R and CD-RW. It always hangs. It hangs at different places if I rerun it. If it fails with different drives from different manufacturers, try to run if from a differnt OS in order to verify whether you are a victim of a long term Linux kernal bug that modifies SCSI commands on their way to the drive. Jörg -- EMail:jo...@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin j...@cs.tu-berlin.de(uni) joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
[gentoo-user] gentoo-sources-2.6.29-r5 ext4
Hi, Yesterday I decided to try gentoo-sources again. I was using vanilla, because of ext4 - when it was declared stable, gentoo-sources still was at the ext4dev phase (a version or so behind the upstream as expected). Yesterday, when migrating from vanilla-2.6.28 to gentoo-sources-2.6.29-5 I haven't noticed any new FS related options except for the btrfs which I don't need, thus have disabled. Kernel compiles w/o errors, but when booted it panics with error msg like unknown mount option ``extents''. Yes, I have extents among the following options in fstab and they work with vannila-2.6.28 /dev/md1 / ext4 stripe=2048,commit=2,extents,orlov,delalloc,\ errors=remount-ro,nodiratime,noatime 0 1 Ideas, suggestions? -- Best regards, Daniel
Re: [gentoo-user] gentoo-sources-2.6.29-r5 ext4
On Donnerstag 04 Juni 2009, Daniel Iliev wrote: Hi, Yesterday I decided to try gentoo-sources again. I was using vanilla, because of ext4 - when it was declared stable, gentoo-sources still was at the ext4dev phase (a version or so behind the upstream as expected). Yesterday, when migrating from vanilla-2.6.28 to gentoo-sources-2.6.29-5 I haven't noticed any new FS related options except for the btrfs which I don't need, thus have disabled. Kernel compiles w/o errors, but when booted it panics with error msg like unknown mount option ``extents''. Yes, I have extents among the following options in fstab and they work with vannila-2.6.28 /dev/md1 / ext4 stripe=2048,commit=2,extents,orlov,delalloc,\ errors=remount-ro,nodiratime,noatime 0 1 Ideas, suggestions? maybe ext4 depends on 'Broken' - as it should - and so it was not shown in menuconfig. As it was not shown, you did not enable it. Without enabling it, it wasn't compiled in
Re: [gentoo-user] gentoo-sources-2.6.29-r5 ext4
2009/6/4 Volker Armin Hemmann volkerar...@googlemail.com On Donnerstag 04 Juni 2009, Daniel Iliev wrote: Hi, Yesterday I decided to try gentoo-sources again. I was using vanilla, because of ext4 - when it was declared stable, gentoo-sources still was at the ext4dev phase (a version or so behind the upstream as expected). Yesterday, when migrating from vanilla-2.6.28 to gentoo-sources-2.6.29-5 I haven't noticed any new FS related options except for the btrfs which I don't need, thus have disabled. Kernel compiles w/o errors, but when booted it panics with error msg like unknown mount option ``extents''. Yes, I have extents among the following options in fstab and they work with vannila-2.6.28 /dev/md1 / ext4 stripe=2048,commit=2,extents,orlov,delalloc,\ errors=remount-ro,nodiratime,noatime 0 1 Ideas, suggestions? maybe ext4 depends on 'Broken' - as it should - and so it was not shown in menuconfig. As it was not shown, you did not enable it. Without enabling it, it wasn't compiled in Using ext4 with gentoo-sources-2.6.29-r4 without any troubles.
Re: [gentoo-user] gentoo-sources-2.6.29-r5 ext4
On Thu, 4 Jun 2009 13:31:15 +0300 Альфар slaughterofshud...@gmail.com wrote: 2009/6/4 Volker Armin Hemmann volkerar...@googlemail.com On Donnerstag 04 Juni 2009, Daniel Iliev wrote: Kernel compiles w/o errors, but when booted it panics with error msg like unknown mount option ``extents''. Yes, I have extents among the following options in fstab and they work with vannila-2.6.28 /dev/md1 / ext4 stripe=2048,commit=2,extents,orlov,delalloc,\ errors=remount-ro,nodiratime,noatime 0 1 Ideas, suggestions? maybe ext4 depends on 'Broken' - as it should - and so it was not shown in menuconfig. As it was not shown, you did not enable it. Without enabling it, it wasn't compiled in Using ext4 with gentoo-sources-2.6.29-r4 without any troubles. Sorry, perhaps I wasn't clear enough. It works if I disable the extents mount option. The problem is that when something is declared stable it should be changed only for bug fixes, not generally and it should be backwards compatible with minor versions at least. Especially when talking about highly important things such as FS. Tonight I'm going to test the latest vanilla and if it works, then I believe it's Gentoo's fault which deserves a report in b.g.o. It would be nice if someone could confirm this issue. -- Best regards, Daniel
Re: [gentoo-user] gentoo-sources-2.6.29-r5 ext4
On 4 Jun 2009, at 11:51, Daniel Iliev wrote: 2009/6/4 Volker Armin Hemmann volkerar...@googlemail.com ... maybe ext4 depends on 'Broken' - as it should - and so it was not shown in menuconfig. As it was not shown, you did not enable it. Without enabling it, it wasn't compiled in ... Sorry, perhaps I wasn't clear enough. It works if I disable the extents mount option. ... it should be changed only for bug fixes, not generally and it should be backwards compatible with minor versions at least. ... I think what Volker is _trying_ to get at is that maybe ext4 isn't compiled into your kernel, because maybe the options have changed in the kernel's .config files between vanilla-2.6.28 and gentoo- sources-2.6.29-5. The error message you see might well be consistent with attempting to mount the volume using the ext3 driver. Suggest you post the .config file against which you built your new kernel. Stroller.
Re: [gentoo-user] OT: buying a keyboard
On Sun, 31 May 2009, Adrian wrote: It's finally getting worn out, keys are sticking pretty bad. I would like to get a new one, but the company I purchased it from (via the internet) seems to not exist any more. Some google action has not resulted in locating any similar keyboards. About once a year, I wash all my keyboards (no, seriously). Dishwasher, quick wash, cold, then leave somewhere warm to dry for a couple days. Afterwards, the tactile response feels like new :-) -- A
Re: [gentoo-user] [OT] Running two apaches and MySQLs on the same server
On Thu, 28 May 2009, Alan McKinnon wrote: It' sin the apache docs, called IP based virtual hosts if memory serves. If you want to run SSL, then IP vhosts is the only way to go. -- A
Re: [gentoo-user] Verification of audio CD copy?
Mark Knecht markkne...@gmail.com wrote: m...@lightning ~/Desktop/cdda2wav $ readcd dev=1000,0,0 -c2scan Read speed: 7056 kB/s (CD 40x, DVD 5x, BD 1x). Write speed: 7056 kB/s (CD 40x, DVD 5x, BD 1x). Capacity: 263932 Blocks = 527864 kBytes = 515 MBytes = 540 prMB Sectorsize: 2048 Bytes readcd: Success. mode select g1: scsi sendcmd: no error CDB: 55 10 00 00 00 00 00 00 14 00 status: 0x2 (CHECK CONDITION) Sense Bytes: 70 00 05 00 00 00 00 0A 00 00 00 00 26 00 00 00 Sense Key: 0x5 Illegal Request, Segment 0 Sense Code: 0x26 Qual 0x00 (invalid field in parameter list) Fru 0x0 Sense flags: Blk 0 (not valid) cmd finished after 0.001s timeout 40s Copy from SCSI (1000,0,0) disk to file '/dev/null' end:263932 ^C^C^C1911 cnt: 49 ^C^C^C I cannot get out of this so far. Hate to think I need to reboot! I already mentioned the reason for the SCSI error.. If you cannot get out of this by typing ^C and you did not disable the signal from your shell, this is a kernel bug. Jörg -- EMail:jo...@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin j...@cs.tu-berlin.de(uni) joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
Re: [gentoo-user] gentoo-sources-2.6.29-r5 ext4
On Thu, 4 Jun 2009 12:27:20 +0100 Stroller strol...@stellar.eclipse.co.uk wrote: On 4 Jun 2009, at 11:51, Daniel Iliev wrote: 2009/6/4 Volker Armin Hemmann volkerar...@googlemail.com ... maybe ext4 depends on 'Broken' - as it should - and so it was not shown in menuconfig. As it was not shown, you did not enable it. Without enabling it, it wasn't compiled in ... Sorry, perhaps I wasn't clear enough. It works if I disable the extents mount option. ... it should be changed only for bug fixes, not generally and it should be backwards compatible with minor versions at least. ... I think what Volker is _trying_ to get at is that maybe ext4 isn't compiled into your kernel, because maybe the options have changed in the kernel's .config files between vanilla-2.6.28 and gentoo- sources-2.6.29-5. The error message you see might well be consistent with attempting to mount the volume using the ext3 driver. Suggest you post the .config file against which you built your new kernel. Stroller. pwd /usr/src/linux-2.6.29-gentoo-r5 grep -i ext4 .config CONFIG_EXT4_FS=y # CONFIG_EXT4DEV_COMPAT is not set CONFIG_EXT4_FS_XATTR=y CONFIG_EXT4_FS_POSIX_ACL=y CONFIG_EXT4_FS_SECURITY=y -- Best regards, Daniel
Re: [gentoo-user] gentoo-sources-2.6.29-r5 ext4
Daniel Iliev daniel.il...@gmail.com writes: pwd /usr/src/linux-2.6.29-gentoo-r5 grep -i ext4 .config CONFIG_EXT4_FS=y # CONFIG_EXT4DEV_COMPAT is not set CONFIG_EXT4_FS_XATTR=y CONFIG_EXT4_FS_POSIX_ACL=y CONFIG_EXT4_FS_SECURITY=y Have you also got ext3 built in, and have you specified the rootfstype kernel boot parameter (in lilo or grub)? Type adding rootfstype=ext4 to your kernel boot parameters.
Re: [gentoo-user] gentoo-sources-2.6.29-r5 ext4
On Thu, 04 Jun 2009 13:36:41 +0100 Graham Murray gra...@gmurray.org.uk wrote: Daniel Iliev daniel.il...@gmail.com writes: pwd /usr/src/linux-2.6.29-gentoo-r5 grep -i ext4 .config CONFIG_EXT4_FS=y # CONFIG_EXT4DEV_COMPAT is not set CONFIG_EXT4_FS_XATTR=y CONFIG_EXT4_FS_POSIX_ACL=y CONFIG_EXT4_FS_SECURITY=y Have you also got ext3 built in, and have you specified the rootfstype kernel boot parameter (in lilo or grub)? Type adding rootfstype=ext4 to your kernel boot parameters. Yes to both questions: .CONFIG: grep -i ext3 .config CONFIG_EXT3_FS=y CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y CONFIG_EXT3_FS_SECURITY=y GRUB: kernel /boot/vmlinuz real_root=/dev/md1 rootfstype=ext4 -- Best regards, Daniel
[gentoo-user] Re: disable syanptics pad
James Ausmus james.ausmus at gmail.com writes: Sounds like, since you don't have a synaptics driver installed, the synaptics device is being handled like a regular mouse via the evdev driver, and the evdev driver doesn't properly handle the data coming from the touchpad, hence the erratic pointer movement.For disabling, I'm not sure, as I don't have access to a system w/ synaptics on it (until I get home, but that's about 8 hours away yet...), but, after emerging the xf86-input-synaptics driver, and looking at the man page (man synaptics), I see theOption TouchpadOff integeroption - if you set this to 1, then the touchpad is disabled, so I would add in a section for the device in your xorg.conf file, something like: Section InputDevice Identifier touchpad Driver synaptics Option Device /dev/input/whatever-it-is - you can find out by doing cat /proc/bus/input/devices and looking at the Handlers line Option TouchpadOff 1 EndSectionHTH--James Sorry for the delay, I only get access to this laptop, on a sporadic basis. This did not work for me. thanks, James
[gentoo-user] Re: disable syanptics pad
Mick michaelkintzios at gmail.com writes: This is from my /etc/hal/fdi/policy/10-xinput-configuration.fdi: = !-- touchpad -- device match key=info.capabilities contains=input.touchpad match key=info.product contains=SynPS/2 merge key=input.x11_driver type=stringsynaptics/merge merge key=input.x11_options.SHMConfig type=stringtrue/merge merge key=input.x11_options.VertEdgeScroll type=stringtrue/merge merge key=input.x11_options.HorizEdgeScroll type=stringtrue/merge merge key=input.x11_options.TapButton1 type=string1/merge merge key=input.x11_options.ClickButton1 type=string1/merge /match /match /device = thanks mick, Sorry for the delayed response. I only get sporadic access to this laptop. I'm going to research how to remove xorg.conf from a system completely. Once I do this, I'll give your suggestion a whirl and post to a new thread. thanks very much, James
Re: [gentoo-user] OT: buying a keyboard
Ajai Khattri wrote: On Sun, 31 May 2009, Adrian wrote: It's finally getting worn out, keys are sticking pretty bad. I would like to get a new one, but the company I purchased it from (via the internet) seems to not exist any more. Some google action has not resulted in locating any similar keyboards. About once a year, I wash all my keyboards (no, seriously). Dishwasher, quick wash, cold, then leave somewhere warm to dry for a couple days. Afterwards, the tactile response feels like new :-) When I was working on computers, I used to clean them with pure alcohol then lay them on top of the A/C condenser, you know, the hot part. It would dry real good in a couple hours. Heat plus the large volume of air works very well. Be careful that the air doesn't blow the keyboard off tho. Some A/C systems can blow huge amounts of air. That was mostly done on IBM XT/AT and Wyse 50 terminal keyboards by the way. I'm not sure about some of these new keyboards with the little rubber pad thingys. Those old keyboards sure did click loud tho. lol Dale :-) :-)
Re: [gentoo-user] OT: buying a keyboard
On Thursday 04 June 2009 16:39:39 Dale wrote: Ajai Khattri wrote: On Sun, 31 May 2009, Adrian wrote: It's finally getting worn out, keys are sticking pretty bad. I would like to get a new one, but the company I purchased it from (via the internet) seems to not exist any more. Some google action has not resulted in locating any similar keyboards. About once a year, I wash all my keyboards (no, seriously). Dishwasher, quick wash, cold, then leave somewhere warm to dry for a couple days. Afterwards, the tactile response feels like new :-) When I was working on computers, I used to clean them with pure alcohol then lay them on top of the A/C condenser, you know, the hot part. It would dry real good in a couple hours. Heat plus the large volume of air works very well. Be careful that the air doesn't blow the keyboard off tho. Some A/C systems can blow huge amounts of air. That was mostly done on IBM XT/AT and Wyse 50 terminal keyboards by the way. I'm not sure about some of these new keyboards with the little rubber pad thingys. Those old keyboards sure did click loud tho. lol IBM system M keyboard ... best keyboard ever made. They were so loud the switch construction even got given a name : buckling spring :-) -- alan dot mckinnon at gmail dot com
Re: [gentoo-user] OT: buying a keyboard
090604 Alan McKinnon wrote: On Thursday 04 June 2009 16:39:39 Dale wrote: Ajai Khattri wrote: About once a year, I wash all my keyboards (no, seriously). Dishwasher, quick wash, cold, leave somewhere warm to dry couple days. When I was working on computers, I used to clean them with pure alcohol then lay them on top of the A/C condenser, you know, the hot part. IBM system M keyboard ... best keyboard ever made. I'm still using the keyboard which came with my Sanyo XT in 1989 : lives on dust + crumbs, never cleaned, works as well as when I got it. Secret ? -- actually made in Japan ... (grin) -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
Re: [gentoo-user] OT: buying a keyboard
Philip Webb wrote: 090604 Alan McKinnon wrote: On Thursday 04 June 2009 16:39:39 Dale wrote: Ajai Khattri wrote: About once a year, I wash all my keyboards (no, seriously). Dishwasher, quick wash, cold, leave somewhere warm to dry couple days. When I was working on computers, I used to clean them with pure alcohol then lay them on top of the A/C condenser, you know, the hot part. IBM system M keyboard ... best keyboard ever made. I'm still using the keyboard which came with my Sanyo XT in 1989 : lives on dust + crumbs, never cleaned, works as well as when I got it. Secret ? -- actually made in Japan ... (grin) Secret, it actually has a switch instead of some touchy crap. While that switch is not completely sealed, it does keep out most stuff. Of course, Coke will upset it's tummy pretty quick. Jeez, I liked those keyboards. Dale :-) :-)
[gentoo-user] emerge --resume finds an old resume rather than the previous one
I have a shell script which loops on failure from emerge to use --skipfirst --resume to continue as much as possible. Every once in a while, it gets forgetful and picks up an emerge from several days previous to try to resume instead of the one that just failed. This is pretty annoying. It started several months ago, sporadically, and seems to be happening more and more often, but not always, and not in any pattern I can discern. This is sys-apps/portage-2.1.6.13 on ~amd64. -- ... _._. ._ ._. . _._. ._. ___ .__ ._. . .__. ._ .. ._. Felix Finch: scarecrow repairman rocket surgeon / fe...@crowfix.com GPG = E987 4493 C860 246C 3B1E 6477 7838 76E9 182E 8151 ITAR license #4933 I've found a solution to Fermat's Last Theorem but I see I've run out of room o
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
On Thu, Jun 04, 2009 at 09:02:21AM -0700, Penguin Lover fe...@crowfix.com squawked: I have a shell script which loops on failure from emerge to use --skipfirst --resume to continue as much as possible. May I suggest using the --keep-going option for emerge to do what you want to do instead of a shell script? W -- All of this is on the web, so other people know it too. ~DeathMech, S. Sondhi. P-town PHY 205 Sortir en Pantoufles: up 909 days, 14:55
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
On Thu, Jun 4, 2009 at 11:02 AM, fe...@crowfix.com wrote: I have a shell script which loops on failure from emerge to use --skipfirst --resume to continue as much as possible. Every once in a while, it gets forgetful and picks up an emerge from several days previous to try to resume instead of the one that just failed. This is pretty annoying. It started several months ago, sporadically, and seems to be happening more and more often, but not always, and not in any pattern I can discern. This is sys-apps/portage-2.1.6.13 on ~amd64. It may be easier to use the --keep-going switch instead. It allows emerge to continue emerging everything else (when possible) and skipping unresolved dependencies caused by the failed merges.
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
On Thu, Jun 04, 2009 at 11:08:26AM -0500, Paul Hartman wrote: On Thu, Jun 4, 2009 at 11:02 AM, fe...@crowfix.com wrote: I have a shell script which loops on failure from emerge to use --skipfirst --resume to continue as much as possible. Every once in a while, it gets forgetful and picks up an emerge from several days previous to try to resume instead of the one that just failed. This is pretty annoying. It started several months ago, sporadically, and seems to be happening more and more often, but not always, and not in any pattern I can discern. This is sys-apps/portage-2.1.6.13 on ~amd64. It may be easier to use the --keep-going switch instead. It allows emerge to continue emerging everything else (when possible) and skipping unresolved dependencies caused by the failed merges. Good grief. That's *zackly* what I have wanted for a long time. When did it get added? I will certainly try that next merge. Heck, I might try the most recent one again anyway just to see miracles in action :-) -- ... _._. ._ ._. . _._. ._. ___ .__ ._. . .__. ._ .. ._. Felix Finch: scarecrow repairman rocket surgeon / fe...@crowfix.com GPG = E987 4493 C860 246C 3B1E 6477 7838 76E9 182E 8151 ITAR license #4933 I've found a solution to Fermat's Last Theorem but I see I've run out of room o
[gentoo-user] Re: f-secure linux security 7.03 on gentoo?
Jarry wrote: Stefan G. Weichinger wrote: at a customers site they have some company-license for f-secure-products. I run a mail-gateway there, using gentoo, it runs amavisd which utilizes clamav and fsav ... (the customer *wants* me to use both as he pays for the f-secure-licenses ...) What mail-server are your running there, may I ask? I'm trying to get amavisd-new working with sendmail, but it is rather difficult. There is only brief documentation with amavisd-new, I do not know how to modify sendmail start-up script. Any help from someone having experience with sendmail + amavisd-new would be appreciated. (sorry for stealing topic) Jarry By some accountings ( http://www.av-comparatives.org/comparativesreviews/main-tests http://www.virusbtn.com/news/2008/09_02 ), Avira/Antivir is one of the better if not best virus/Trojan signature scanners out there. 1. It provides transparent on-access scanning. You stipulate which directories should be considered, and it monitors them. Much easier, IMHO, than fooling with agents and servers. 2. In addition to Windows signatures and heuristics, it includes hundreds of Linux/Unix Trojan, rootkit, and virus signatures - so I also scan user directories where browsers and mail clients work, and work directories where stuff is downloaded and compiled. 3. It is remarkably easy to install - a script both installs the scanner, and optionally builds the kernel module (dazuko) required to do the on access scanning. http://www.free-av.com/en/download/download_servers.php HTH
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
On Thu, Jun 4, 2009 at 11:18 AM, fe...@crowfix.com wrote: Good grief. That's *zackly* what I have wanted for a long time. When did it get added? From bug 12768 it looks like it was added by Zac Medico on 2008-06-27 :)
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
On Thu, Jun 04, 2009 at 09:18:14AM -0700, fe...@crowfix.com wrote: On Thu, Jun 04, 2009 at 11:08:26AM -0500, Paul Hartman wrote: On Thu, Jun 4, 2009 at 11:02 AM, fe...@crowfix.com wrote: I have a shell script which loops on failure from emerge to use --skipfirst --resume to continue as much as possible. Every once in a while, it gets forgetful and picks up an emerge from several days previous to try to resume instead of the one that just failed. This is pretty annoying. It started several months ago, sporadically, and seems to be happening more and more often, but not always, and not in any pattern I can discern. This is sys-apps/portage-2.1.6.13 on ~amd64. It may be easier to use the --keep-going switch instead. It allows emerge to continue emerging everything else (when possible) and skipping unresolved dependencies caused by the failed merges. Good grief. That's *zackly* what I have wanted for a long time. When did it get added? I will certainly try that next merge. Heck, I might try the most recent one again anyway just to see miracles in action :-) Unfortunately, it does exactly the same thing. *** Resuming merge... Calculating dependencies ..^H^H... done! * One or more packages are either masked or have missing dependencies: and it picks up the same several day old failed emerge instead of the one it was running. -- ... _._. ._ ._. . _._. ._. ___ .__ ._. . .__. ._ .. ._. Felix Finch: scarecrow repairman rocket surgeon / fe...@crowfix.com GPG = E987 4493 C860 246C 3B1E 6477 7838 76E9 182E 8151 ITAR license #4933 I've found a solution to Fermat's Last Theorem but I see I've run out of room o
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
fe...@crowfix.com wrote: On Thu, Jun 04, 2009 at 11:08:26AM -0500, Paul Hartman wrote: On Thu, Jun 4, 2009 at 11:02 AM, fe...@crowfix.com wrote: I have a shell script which loops on failure from emerge to use --skipfirst --resume to continue as much as possible. Every once in a while, it gets forgetful and picks up an emerge from several days previous to try to resume instead of the one that just failed. This is pretty annoying. It started several months ago, sporadically, and seems to be happening more and more often, but not always, and not in any pattern I can discern. This is sys-apps/portage-2.1.6.13 on ~amd64. It may be easier to use the --keep-going switch instead. It allows emerge to continue emerging everything else (when possible) and skipping unresolved dependencies caused by the failed merges. Good grief. That's *zackly* what I have wanted for a long time. When did it get added? I will certainly try that next merge. Heck, I might try the most recent one again anyway just to see miracles in action :-) You can also add that option to make.conf so it will do it even if you forget to type it in. I did that a while back when I found out about it. Dale :-) :-)
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
On Thu, 4 Jun 2009 09:55:59 -0700, fe...@crowfix.com wrote: and it picks up the same several day old failed emerge instead of the one it was running. Try emaint -f cleanresume -- Neil Bothwick A computer without Microsoft is like a chocolate cake without mustard. signature.asc Description: PGP signature
Re: [gentoo-user] emerge --resume finds an old resume rather than the previous one
On Thu, Jun 04, 2009 at 06:30:37PM +0100, Neil Bothwick wrote: On Thu, 4 Jun 2009 09:55:59 -0700, fe...@crowfix.com wrote: and it picks up the same several day old failed emerge instead of the one it was running. Try emaint -f cleanresume Same result. emain did something, but the emerge still picks up the old emerge. -- ... _._. ._ ._. . _._. ._. ___ .__ ._. . .__. ._ .. ._. Felix Finch: scarecrow repairman rocket surgeon / fe...@crowfix.com GPG = E987 4493 C860 246C 3B1E 6477 7838 76E9 182E 8151 ITAR license #4933 I've found a solution to Fermat's Last Theorem but I see I've run out of room o
Re: [gentoo-user] OT: buying a keyboard
On Thu, Jun 4, 2009 at 12:49, Dale rdalek1...@gmail.com wrote: Philip Webb wrote: 090604 Alan McKinnon wrote: On Thursday 04 June 2009 16:39:39 Dale wrote: Ajai Khattri wrote: About once a year, I wash all my keyboards (no, seriously). Dishwasher, quick wash, cold, leave somewhere warm to dry couple days. When I was working on computers, I used to clean them with pure alcohol then lay them on top of the A/C condenser, you know, the hot part. IBM system M keyboard ... best keyboard ever made. I'm still using the keyboard which came with my Sanyo XT in 1989 : lives on dust + crumbs, never cleaned, works as well as when I got it. Secret ? -- actually made in Japan ... (grin) Secret, it actually has a switch instead of some touchy crap. While that switch is not completely sealed, it does keep out most stuff. Of course, Coke will upset it's tummy pretty quick. Jeez, I liked those keyboards. I have an ancient Dell QuietKey keyboard. I cleaned it (yeah, removed all keys) and its the best for my work. I is kinda heavy to press, but that avoid lots of problems, like accidental key presses. I also have a Logitech Dinovo Edge at home, connected to my HTPC, and for that purpose, its the best. I guess it all depends on how you'll use it. What I hate most about some keyboards are: 1) The damn layout changes (like FN key at the left side of the BACKSPACE, and backspace not a double size key, or right shift also not double size, ENTER is another one sometimes manufacturers screw around). 2) Useless keys inserted in the middle of the layout (like sleep, standby and power down above INSERT, HOME and PAGE UP). 3) Absence of the pads used to lift the back of the keyboard (I can't believe some people don't use them). -- Daniel da Veiga
Re: [gentoo-user] OT: buying a keyboard
On Thu, Jun 4, 2009 at 1:19 PM, Daniel da Veiga danieldave...@gmail.com wrote: What I hate most about some keyboards are: 1) The damn layout changes (like FN key at the left side of the BACKSPACE, and backspace not a double size key, or right shift also not double size, ENTER is another one sometimes manufacturers screw around). 2) Useless keys inserted in the middle of the layout (like sleep, standby and power down above INSERT, HOME and PAGE UP). 3) Absence of the pads used to lift the back of the keyboard (I can't believe some people don't use them). 4) Cable so short it can't even reach the computer. I have a keyboard with an 18-inch cord. It is barely long enough to reach from one side of the keyboard to the other! Absolutely useless.
[gentoo-user] lvm2 questions
Hi group, Creating LVM partitions on SSD and SD card using systemrescuecd-1.2.0 while following doc, 'Gentoo LVM2 installation'. In the doc it says to edit the 'filter =' statement in lvm.conf in order to scan the correct devices. But just below it says to use #pvcreate with the appropriate devices. But if you use pvcreate to specify the device you're planning to partition what's the point in editing lvm.conf? Edited it anyways. Now have: ... filter= [ r|/dev/ndb.*|, a|/dev/sd[ab]| ] # only changed 'accept' line ... /root % vgscan Reading all physical volumes. This may take a while... Well, no, it was more or less instant. There was no further output and the prompt reappeared. Does this mean everything is cool or that vgscan found nothing? Maxim
Re: [gentoo-user] Verification of audio CD copy?
On Thu, Jun 4, 2009 at 1:02 AM, Joerg Schilling joerg.schill...@fokus.fraunhofer.de wrote: Mark Knecht markkne...@gmail.com wrote: It seems that the command readcd -c2scan fails on both of my drives for all media types - retail CDs, copies of CDs on CR-R and CD-RW. It always hangs. It hangs at different places if I rerun it. If it fails with different drives from different manufacturers, try to run if from a differnt OS in order to verify whether you are a victim of a long term Linux kernal bug that modifies SCSI commands on their way to the drive. Jörg Expand please. Run it on a different OS? The machine is dual-boot Gentoo/XP but your site doesn't seem to have an updated Windows version. (Unless it's not in the win32 directory.) I presume you did not mean 'different kernel' as your assertion is that the bug is long-standing and therefore will be in other kernels. How do I run in another OS? Build recent source under Windows and somehow run it there? I'm up for trying that but I'm not a developer, don't have a Win 32 compiler, and wouldn't know where to start. Does gcc run under Windows? Could I compile in Linux for Windows and test the drive under Windows that way? Thanks! - Mark
Re: [gentoo-user] Verification of audio CD copy?
On Thursday 04 June 2009 21:13:39 Mark Knecht wrote: On Thu, Jun 4, 2009 at 1:02 AM, Joerg Schilling joerg.schill...@fokus.fraunhofer.de wrote: Mark Knecht markkne...@gmail.com wrote: It seems that the command readcd -c2scan fails on both of my drives for all media types - retail CDs, copies of CDs on CR-R and CD-RW. It always hangs. It hangs at different places if I rerun it. If it fails with different drives from different manufacturers, try to run if from a differnt OS in order to verify whether you are a victim of a long term Linux kernal bug that modifies SCSI commands on their way to the drive. Jörg Expand please. Run it on a different OS? The machine is dual-boot Gentoo/XP but your site doesn't seem to have an updated Windows version. (Unless it's not in the win32 directory.) He likely means Use a Unix that is not Linux. One of the *BSDs perhaps Unlike most folk here, Linux is to Joerg just another OS. For best response, do the test on Solaris. -- alan dot mckinnon at gmail dot com
[gentoo-user] Re: gentoo-sources-2.6.29-r5 ext4
Daniel Iliev wrote: Hi, Yesterday I decided to try gentoo-sources again. I was using vanilla, because of ext4 - when it was declared stable, gentoo-sources still was at the ext4dev phase (a version or so behind the upstream as expected). Yesterday, when migrating from vanilla-2.6.28 to gentoo-sources-2.6.29-5 I haven't noticed any new FS related options except for the btrfs which I don't need, thus have disabled. Kernel compiles w/o errors, but when booted it panics with error msg like unknown mount option ``extents''. Yes, I have extents among the following options in fstab and they work with vannila-2.6.28 /dev/md1 / ext4 stripe=2048,commit=2,extents,orlov,delalloc,\ errors=remount-ro,nodiratime,noatime 0 1 Ideas, suggestions? The extents mount option has been removed from 2.6.29: http://kerneltrap.org/mailarchive/linux-ext4/2009/1/6/4613614
[gentoo-user] KDE4 halts - how to catch a reason?
Last week or two (I mostly use Gnome and can not say more precisely) after KDE4 starting session apps starts also, as well as bg and panels. But then I get black screen (during this KDE4 start process), and Ctrl-Alt-Backspace does nothing. The only way I have in hand to continue is to use SysRq kernel hacking (s, u, b). Have you an idea how to catch the halt reason?
Re: [gentoo-user] OT: buying a keyboard
On Thu, 4 Jun 2009 15:19:50 -0300, Daniel da Veiga wrote: 3) Absence of the pads used to lift the back of the keyboard (I can't believe some people don't use them). Some clam that raising the back of the keyboard increases the risk of RSI, and that you should raise the front if anything, to reduce the amount you bend your wrists. -- Neil Bothwick Time is an illusion but never so much as when you're using a modem. signature.asc Description: PGP signature
[gentoo-user] Re: KDE4 halts - how to catch a reason?
Andrew Gaydenko wrote: Last week or two (I mostly use Gnome and can not say more precisely) after KDE4 starting session apps starts also, as well as bg and panels. But then I get black screen (during this KDE4 start process), and Ctrl-Alt-Backspace does nothing. The only way I have in hand to continue is to use SysRq kernel hacking (s, u, b). Have you an idea how to catch the halt reason? Try to ssh into the box from another machine and inspect the logs (~/.xsession-errors and /var/log/Xorg.0.log).
Re: [gentoo-user] OT: buying a keyboard
On Thu, Jun 4, 2009 at 3:04 PM, Neil Bothwick n...@digimed.co.uk wrote: On Thu, 4 Jun 2009 15:19:50 -0300, Daniel da Veiga wrote: 3) Absence of the pads used to lift the back of the keyboard (I can't believe some people don't use them). Some clam that raising the back of the keyboard increases the risk of RSI, and that you should raise the front if anything, to reduce the amount you bend your wrists. I think this is an area where people with good posture have worse typing conditions in general. A person who slouches in their chair (like me, for example) has a very natural typing angle even when the rear of the keyboard is raised. :) My elbows are well below the plane of the desk... maybe my spine will look like a Lacrosse stick someday, but I am comfortable for now.
[gentoo-user] Re: gentoo-sources-2.6.29-r5 ext4
Daniel Iliev wrote: Hi, Yesterday I decided to try gentoo-sources again. I was using vanilla, because of ext4 - when it was declared stable, gentoo-sources still was at the ext4dev phase (a version or so behind the upstream as expected). Yesterday, when migrating from vanilla-2.6.28 to gentoo-sources-2.6.29-5 I haven't noticed any new FS related options except for the btrfs which I don't need, thus have disabled. Kernel compiles w/o errors, but when booted it panics with error msg like unknown mount option ``extents''. Yes, I have extents among the following options in fstab and they work with vannila-2.6.28 /dev/md1 / ext4 stripe=2048,commit=2,extents,orlov,delalloc,\ errors=remount-ro,nodiratime,noatime 0 1 Ideas, suggestions? The extents mount option has been removed from 2.6.29: http://kerneltrap.org/mailarchive/linux-ext4/2009/1/6/4613614
Re: [gentoo-user] Verification of audio CD copy?
On Thu, Jun 4, 2009 at 12:38 PM, Alan McKinnon alan.mckin...@gmail.com wrote: On Thursday 04 June 2009 21:13:39 Mark Knecht wrote: On Thu, Jun 4, 2009 at 1:02 AM, Joerg Schilling joerg.schill...@fokus.fraunhofer.de wrote: Mark Knecht markkne...@gmail.com wrote: It seems that the command readcd -c2scan fails on both of my drives for all media types - retail CDs, copies of CDs on CR-R and CD-RW. It always hangs. It hangs at different places if I rerun it. If it fails with different drives from different manufacturers, try to run if from a differnt OS in order to verify whether you are a victim of a long term Linux kernal bug that modifies SCSI commands on their way to the drive. Jörg Expand please. Run it on a different OS? The machine is dual-boot Gentoo/XP but your site doesn't seem to have an updated Windows version. (Unless it's not in the win32 directory.) He likely means Use a Unix that is not Linux. One of the *BSDs perhaps Unlike most folk here, Linux is to Joerg just another OS. For best response, do the test on Solaris. Thanks for the explanation. No option for me to do any of the above ideas. Available: Gentoo 64-bit Gentoo 32-bit Win XP Win Vista FreeDOS No idea how to build for the last 3. Cheers, Mark
Re: [gentoo-user] Touchpad wont click with xorg-server-1.5.3-r6
Hi Adam, On Wednesday 03 June 2009, Adam Carter wrote: I can no longer tap the touchpad to cause a mouse click. It worked in 1.5.3-r5, but i think r5 was -hal, whereas i have built -r6 +hal. I cant see any touchpad related patches in r6. Should i try -hal, reverting to -r5 or something else? I have no input devices defined in my xorg.conf. Dell laptop running amd64 2.6.28-gentoo-r5. This is probably related to the latest xorg update. Have you had a look at this: http://www.gentoo.org/proj/en/desktop/x/x11/xorg-server-1.5-upgrade-guide.xml and the link it mentions about synaptics (in case evdev does not work). Alternatively, google for synaptics to find what entries you should make in your xorg.conf to affect the mouse click function. There have also been a number of threads in this list on the new xorg and associated issues over the last month or so. Check in Gmane for older posts. HTH. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Re: Verification of audio CD copy?
On Wed, June 3, 2009 22:23, Joerg Schilling wrote: Nikos Chantziaras rea...@arcor.de wrote: BTW: this is where you see that cdparranoia(1) is now 12 year old and unmaintained since ~ 10 years. cdda2wav -paranoia gives much better results that cdparanoia. Pardon me for getting in the middle but I am really curious about this. I know that paranoia is old but according to their web page the last update was September 11, 2008. However maybe it was just a maintenance update to fix compilation errors or something like that. I have researched about cdda2wav vs. cdparanoia vs. cdrdao in the past and came to the conclussion that cdparanioa is probably the worst nowadays, however I am not sure it's dead. -- Jesús Guerrero
Re: [gentoo-user] [solved] Re: gentoo-sources-2.6.29-r5 ext4
On Thu, 04 Jun 2009 23:13:47 +0300 Nikos Chantziaras rea...@arcor.de wrote: Daniel Iliev wrote: [--snip--] Kernel compiles w/o errors, but when booted it panics with error msg like unknown mount option ``extents''. Yes, I have extents among the following options in fstab and they work with vannila-2.6.28 /dev/md1 / ext4 stripe=2048,commit=2,extents,orlov,delalloc,\ errors=remount-ro,nodiratime,noatime 0 1 Ideas, suggestions? The extents mount option has been removed from 2.6.29: http://kerneltrap.org/mailarchive/linux-ext4/2009/1/6/4613614 Thanks. I'm back to gentoo-sources w/o this mount option. I believe the ext4 team deserves a few bad words for this (and many good words for the whole work), but I'll spare the rant. :) -- Best regards, Daniel
Re: [gentoo-user] lvm2 questions
On Thu, 2009-06-04 at 13:01 -0600, Maxim Wexler wrote: Hi group, Creating LVM partitions on SSD and SD card using systemrescuecd-1.2.0 while following doc, 'Gentoo LVM2 installation'. In the doc it says to edit the 'filter =' statement in lvm.conf in order to scan the correct devices. But just below it says to use #pvcreate with the appropriate devices. But if you use pvcreate to specify the device you're planning to partition what's the point in editing lvm.conf? Edited it anyways. Now have: ... filter= [ r|/dev/ndb.*|, a|/dev/sd[ab]| ] # only changed 'accept' line ... /root % vgscan Reading all physical volumes. This may take a while... Well, no, it was more or less instant. There was no further output and the prompt reappeared. Does this mean everything is cool or that vgscan found nothing? You still need pvcreate to create PVs. What the filter parameter is for is for when LVM scans for PVs. Basically pvcreate puts a header on the device so that vgscan and friends knows that it's a PV. It may be undesirable to scan some devices (e.g. optical drives) for PVs. Having said that, usually the defaults for filter and exclude are fine and you don't need to play with them (as is the case for many defaults). You'll know if LVM has detected your PVs by running pvdisplay. If your PV doesn't show up either a: you didn't pvcreate it or b: it didn't get scanned due to your filter or exclude params. -a
[gentoo-user] Preventing Midnight Commander from putting junk in .bash_history
My .bash_history contains lots of stuff like: cd `printf %b '\0057home\0057realnc'` They seem to come from Midnight Commander and are very annoying. Any way to prevent this stuff from ending up in the bash history?
RE: [gentoo-user] Touchpad wont click with xorg-server-1.5.3-r6
I can no longer tap the touchpad to cause a mouse click. It worked in 1.5.3-r5, but i think r5 was -hal, whereas i have built -r6 +hal. I cant see any touchpad related patches in r6. Should i try -hal, reverting to -r5 or something else? I have no input devices defined in my xorg.conf. Dell laptop running amd64 2.6.28-gentoo-r5. This is probably related to the latest xorg update. Have you had a look at this: http://www.gentoo.org/proj/en/desktop/x/x11/xorg-server-1.5-up grade-guide.xml A rebuild with -hal sorted it out.
Re: [gentoo-user] Preventing Midnight Commander from putting junk in .bash_history
On Thu, Jun 4, 2009 at 8:33 PM, Nikos Chantziarasrea...@arcor.de wrote: My .bash_history contains lots of stuff like: cd `printf %b '\0057home\0057realnc'` They seem to come from Midnight Commander and are very annoying. Any way to prevent this stuff from ending up in the bash history? I use MC constantly and don't have anything at all like that (or anything unusual) in my .bash_history
[gentoo-user] Re: Preventing Midnight Commander from putting junk in .bash_history
Paul Hartman wrote: On Thu, Jun 4, 2009 at 8:33 PM, Nikos Chantziarasrea...@arcor.de wrote: My .bash_history contains lots of stuff like: cd `printf %b '\0057home\0057realnc'` They seem to come from Midnight Commander and are very annoying. Any way to prevent this stuff from ending up in the bash history? I use MC constantly and don't have anything at all like that (or anything unusual) in my .bash_history It is a problem occurring with mc but I think I've found the solution now: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=529590
[gentoo-user] External USB HDD: device descriptor read/64, error -62 (solved?)
I thought I would post this here for the archives and google in case someone else has the same problem. Tonight I rebooted for the first time in 60 days. In that time I've changed from gentoo-sources-2.6.29-r1 to 2.6.29-r5 and countless updates to my ~amd64 world. I have an external LaCie 2TB USB hard drive which has been in use those whole 60 days without problems. Much to my dismay, after rebooting and plugging it in I got this in my dmesg: [ 144.963008] usb 2-7: new full speed USB device using ohci_hcd and address 9 [ 145.127007] usb 2-7: device descriptor read/64, error -62 [ 145.796006] usb 2-7: device not accepting address 9, error -62 [ 145.953007] usb 2-7: new full speed USB device using ohci_hcd and address 10 [ 146.117007] usb 2-7: device descriptor read/64, error -62 [ 146.382007] usb 2-7: device descriptor read/64, error -62 [ 146.640007] usb 2-7: new full speed USB device using ohci_hcd and address 11 [ 147.044914] usb 2-7: device not accepting address 11, error -62 [ 147.201007] usb 2-7: new full speed USB device using ohci_hcd and address 12 [ 147.605006] usb 2-7: device not accepting address 12, error -62 [ 147.605010] hub 2-0:1.0: unable to enumerate USB device on port 7 I tried unplugging and replugging both the USB and power cables several times but still had the same error. After a bit of Googling I saw one old post on another distro's site that suggested it might be a power issue -- not with the power to the drive, but power to the USB hub on the motherboard. In fact, I had at the time plugged in these USB devices: keyboard, mouse, joystick, hard drive, monitor (which has a two-port hub), webcam, and scanner. I also charge my cell phone via USB, though it was not connected at the time. All 4 USB ports on the back of the motherboard were in use, which is where the HDD was plugged in, as well as 2 more of the external (case) ports which also come from the motherboard. So I unplugged everything except for the keyboard and mouse, then plugged the HDD back in to the same USB port and magic! It works! fsck found no errors, my data is safe... and I am /yet/ /again/ reminded to make backups of everything. I cannot be sure that my USB power was being over-utilized, but I have unplugged and re-plugged it a few times and it has kept working so far. Perhaps I have a bad port or cable that only sometimes works and happened to work by coincidence after I unplugged the other devices. I thought I'd post it here in case anyone else like me has a mild heart-attack thinking their drive has died. There is hope! :)
Re: [gentoo-user] Re: Preventing Midnight Commander from putting junk in .bash_history
On Thu, Jun 4, 2009 at 10:15 PM, Nikos Chantziarasrea...@arcor.de wrote: Paul Hartman wrote: On Thu, Jun 4, 2009 at 8:33 PM, Nikos Chantziarasrea...@arcor.de wrote: My .bash_history contains lots of stuff like: cd `printf %b '\0057home\0057realnc'` They seem to come from Midnight Commander and are very annoying. Any way to prevent this stuff from ending up in the bash history? I use MC constantly and don't have anything at all like that (or anything unusual) in my .bash_history It is a problem occurring with mc but I think I've found the solution now: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=529590 It is strange, I do not have that environment variable set and checked my /etc/bashrc and ~/.bashrc and neither has anything like that either. I'm using mc-4.6.2_pre1 and bash-4.0_p24 in case it makes a difference. I hope that solution works for you! :)
Re: [gentoo-user] lvm2 questions
On 6/4/09, Albert Hopkins mar...@letterboxes.org wrote: You'll know if LVM has detected your PVs by running pvdisplay. Check! Thanks Albert Another please, from the doc: Note: As Terje Kvernes commented, it is easier to increase the size of a partition then to shrink it. You might therefore want to start with smaller partitions and increase their size as needed. Code Listing 2.5: Creating and extending logical volumes Then why give the logical volumes any size at all? If they can be expanded at will, why not just the let the files fill them up as much as they need? This point has me stymied. Maxim
[gentoo-user] Endless stream of valgrind errors with glibc-2.10.1
I'm getting a big load of errors when running valgrind on every executable on the system. Even a valgrind ls / results in tons of this: ==16779== Use of uninitialised value of size 8 ==16779==at 0x3100C3: ??? (in /lib64/ld-2.10.1.so) ==16779==by 0x30397B: ??? (in /lib64/ld-2.10.1.so) ==16779==by 0x31401B: ??? (in /lib64/ld-2.10.1.so) ==16779==by 0x301450: ??? (in /lib64/ld-2.10.1.so) ==16779==by 0x300BA7: ??? (in /lib64/ld-2.10.1.so) Can anyone confirm? I'm on AMD64, GCC 4.3.3-r2, gentoo-sources 2.6.28-r6, linux-headers 2.6.28-r1. I've rebuild system and valgrind, but it doesn't help.