Please do not discuss spam on list (was: Re: K12 Broward Virtual [...])
One time to everyone on the list: Please do not discuss spam on the mailinglist, but see below. Should it happen again I will inform people privately. Am Sonntag 30 August 2009 schrieben Sie: > On Wed, 26 Aug 2009, Cavan Mejias wrote: > > WTH? > > Both the threads thus far, relating to K12, appear to me to be spam. > > I do not know whether the mailing list administrator can block further > messages originating from the same domain name, or can get that domain > name blacklisted (notified to the blacklists that companies use in > anti-spam software) as a spammer, but, they are options. Please never ever reply to spam on a mailinglist in such a way that your answers goes to the mailing list as well. I did not see that spam due to the good work of policyd-weight and crm114. But I do see your answers. And I even trained one of them as no spam in crm114. I am tempted to train discussions about spam as spam as well however. Please contact the list administrators directly. AFAIR there is a way to report spam to them. It might even be accessible from the web interface of the mailing list archives. Yes, just see there: http://lists.debian.org/debian-laptop/2009/08/threads.html Search the mail. And click "Report as spam". Please do not reply on list unless you have anything significant to add. Thanks. Ciao, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 signature.asc Description: This is a digitally signed message part.
ThinkPad T520: uefi/gpt based debian install
Hi! I am about to receive a ThinkPad T520 with Intel SSD 320 real soon hopefully. I wondered whether it might make sense to install with UEFI and GPT. Informations from Arch Wiki hint at that for better alignment and less space waste for SSD: https://wiki.archlinux.org/index.php/SSD#Using_GPT_-_RECOMMENDED_METHOD https://wiki.archlinux.org/index.php/Unified_Extensible_Firmware_Interface Did anyone here try this with any laptop with uefi BIOS? Did it work? Were there problems? I have seen some Debian bug reports that were closed. As far as I understand, I it should work when I Debian installer to select UEFI/GPT somehow and then add an about 1 MiB sized 0xEF02 Bios Boot Partition as GPT doesn't guerantee free space aber (legagy) MBR. I intend to use BTRFS. I hope that it will align automatically, but I am not sure about this. I will use debian unstable, possible even some parts from experimental with kernel 2.6.38 or better 2.6.39 for best Sandybridge graphics support. If interested I report back what I find when I do the Linux installation. Ciao, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to debian-laptop-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201105131833.53986.mar...@lichtvoll.de
Re: ThinkPad T520: uefi/gpt based debian install
Am Freitag, 13. Mai 2011 schrieb Martin Steigerwald: > Hi! > > I am about to receive a ThinkPad T520 with Intel SSD 320 real soon > hopefully. > > I wondered whether it might make sense to install with UEFI and GPT. > Informations from Arch Wiki hint at that for better alignment and less > space waste for SSD: > > https://wiki.archlinux.org/index.php/SSD#Using_GPT_-_RECOMMENDED_METHOD > > https://wiki.archlinux.org/index.php/Unified_Extensible_Firmware_Interf > ace > > > Did anyone here try this with any laptop with uefi BIOS? Did it work? > Were there problems? I have seen some Debian bug reports that were > closed. Just for the record. The ThinkPad BIOS Version: 8AET42WW (1.22 ), didn´t like to boot from a GPT partitioned SSD with BIOS Boot Partition and EFI System Partition. I made the latter just for the case I want to switch to EFI later on, but let grub-pc install to the BIOS Boot Partition. Even after switching to MBR it didn´t work. Only after I zerod sector 2 to 2047, the BIOS Boot Partition and the EFI System Partition and set the partition /boot bootable the BIOS booted from the SSD again. I do not think it was the bootable flag. Maybe it was the EFI System Partition, but I also tried after removing it in gdisk. A parted mklabel msdos which should wipe the pre-existing GPT headers did not suffice. So the exact circumstances why it didn´t work are unknown to me, but I now stick to MBR for the while being. Doesn´t matter anyway, as fdisk -cu also aligns at 1 MiB and displays partitions offsets and sizes in sectors. Well I saw an report that it doesn´t work with a ThinkPad W520 as well. -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 signature.asc Description: This is a digitally signed message part.
Re: tp_smapi 0.40-9 incompatible with thinkpad w520?
Am Montag, 23. Mai 2011 schrieb CHERN Yen-Chieh: > Hi, Hi, > Before the bios update made available on 19th May, my thinkpad w520 > couldn't boot into debian wheezy/sid even after a successful > installation. I assume that was the BIOS with a fix to also boot Linux when no partition is marked bootable? Been there with my T520. But didn´t install the BIOS Update yet, just marked partition with /boot as bootable even when Linux itself does not need it. > However the latest bios update solved that issue. But > still the tp_smapi can not be loaded. > > usr@debian-thinkpadw520:~$ sudo modprobe tp_smapi > [sudo] password for usr: > FATAL: Error inserting tp_smapi > (/lib/modules/2.6.39-1-amd64/updates/dkms/tp_smapi.ko): No such device Same here. I think the current tp-smapi doesn´t properly support the new models. I didn´t get anything back regarding my post on that topic on linux thinkpad mailing list. -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 signature.asc Description: This is a digitally signed message part.
Re: Lenovo T520i X11 Intel i810 "static display" issue
Am Montag, 11. Juli 2011 schrieb Dr. Ed Morbius:
> on 01:46 Sun 10 Jul, Bill Brelsford (w...@k2di.net) wrote:
> > In article ,
> >
> > Dr. Ed Morbius wrote:
> > > I've been running into an issue with a Lenovo Thinkpad T520i
> > > running squeeze where the X11 display goes into a sort of "static"
> > > state. It's usually after a pm-suspend or pm-hibernate (but not
> > > after all suspend/hibernates), and sometimes simply when awakening
> > > the screensaver. Restarting X generally (always?) clears the
> > > issue, but doesn't do much for preserving session state.
> > >
> > > The symptoms are that the display doesn't update -- while it's
> > > viewable. There's no visible response to keyboard or mouse inputs.
> > > The pointer itself /does/ move around the display, but windows
> > > don't take focus, won't resize or move, etc.
> > >
> > > However if I toggle away () then back to
> > > ( or ) the display -- the changes /are/
> > > reflected. I've taken advantage of this to shut down windows/apps
> > > cleanly.
> > >
> > > If i restart X ("sudo /etc/init.d/kdm stop; sudo /etc/init.d/kdm
> > > start" from another virtual console), X works properly, usually.
> > >
> > > I've noticed this under several desktops / window managers (GNOME,
> > > KDE, XFCE, twm, wmaker). It doesn't appear to be specific to the
> > > window manager.
> > >
> > > Restarting the window manager doesn't clear the behavior.
> > >
> > > Occasionally when xscreensaver is activated, moving the mouse won't
> > > bring up the password dialog, but the mouse is limited to moving
> > > within the bounds of where that dialog would be.
> > >
> > > Graphics devicde appears to be Intel Integrated Graphics Chipsets:
> > > i810 famiily.
> > >
> > > 1: Any idea what the bug is? Or a fix?
> > > 2: What package(s) should I file it against?
> >
> > It appears to be a bug in xserver-xorg-video-intel that occurs
> > under kernels 2.6.38 and 2.6.39. The latter seems to help some --
> > in my case anyway -- but the problem is still there. See bug
> > 622308 (which has been closed); bugs 626337 and 627249 are similar.
>
> I'm seeing the same issue with a 2.6.39 kernel. I'll try a 3.0 rc as
> well, but so far no joy.
>
> Linux altaira 2.6.39-2-amd64 #1 SMP Tue Jul 5 02:51:22 UTC 2011
> x86_64 GNU/Linux
I strongly recommend 3.0 kernel for Intel Sandybridge for various reasons
including a strange lockup bug that was discussed on linux thinkpad
mailing list apparently has been fixed since then.
And then the new Sandybridge Acceleration Architecture which easily
doubles FreeDroid RPG framerate with Full HD display from 30 to 60 fps is
only in 3.0.
--
Martin 'Helios' Steigerwald - http://www.Lichtvoll.de
GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7
--
To UNSUBSCRIBE, email to debian-laptop-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/201108121955.34136.mar...@lichtvoll.de
Re: Lenovo T520i X11 Intel i810 "static display" issue
Am Dienstag, 16. August 2011 schrieb kei...@strucktower.com: > In response to a previous thread: > > Dr. Mobius, Bill Brelsford, et all: > > Did anyone find a solution or even a cause for this "static display" > problem upon resume from suspend? > > I see that Martin Steigerwald claims that upgrading to 3.0 solves the > problem, but it hasn't solved the problem for me- I am running > 3.1.0-rc1 with all upgraded packages on my Thinkpad T520 and the > problem persists. In looking around on the Internet I see that others > have the same problem, and _sometimes_ upgrading mysteriously helps, > but no one seems to know a definitive answer. > > I would appreciate any help anyone can offer. Please try exactly 3.0.0 from the debian package. This is what I am using with Debian Sid. Might well be that 3.1.0-rc1 has some bug. Its a rc1. I am using KDE 4.6.5 with Powerdevil via pm-utils hibernation or suspend to ram and both work out of the box. -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to debian-laptop-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201108251842.33284.mar...@lichtvoll.de
Re: Lenovo T520i X11 Intel i810 "static display" issue
Am Donnerstag, 25. August 2011 schrieb kei...@strucktower.com: > Hi Martin- Hi Keith, > Thanks for your reply. > > Can you tell me _how_ to "try exactly 3.0.0 from the debian package"? > If I donwload the source for 3.0.3 (the most recent stable kernel on > kernel.org) I will then need to build a .config file- and obviously my > attempts so far have been wrong. I'm new to Linux, and guessing at the > thousands of configuration lines in make menuconfig isn't working for > me... > > The most recent debian distribution only uses 2.6.39 (or so), so I am > not sure how you mean to try 3.0.0 from "the debian package". > > Can you explain further? Is there some way to get a 3.0.0 kernel with > "default" or generic .config file that will work with debian on my > T520? Get linux-image-3.0.0-1-amd64 3.0.0-2 from Debian Sid. No need to build a kernel, just use the binary package. Thanks, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to debian-laptop-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201108260002.51057.mar...@lichtvoll.de
Re: Lenovo T520i X11 Intel i810 "static display" issue
Hi Keith! Please use proper quoting. Thanks. TOFU repaired. Am Freitag, 26. August 2011 schrieb kei...@strucktower.com: > > Am Donnerstag, 25. August 2011 schrieb kei...@strucktower.com: > >> Hi Martin- > > > > Hi Keith, > > > >> Thanks for your reply. > >> > >> Can you tell me _how_ to "try exactly 3.0.0 from the debian > >> package"? If I donwload the source for 3.0.3 (the most recent > >> stable kernel on kernel.org) I will then need to build a .config > >> file- and obviously my attempts so far have been wrong. I'm new to > >> Linux, and guessing at the thousands of configuration lines in make > >> menuconfig isn't working for me... > >> > >> The most recent debian distribution only uses 2.6.39 (or so), so I > >> am not sure how you mean to try 3.0.0 from "the debian package". > >> > >> Can you explain further? Is there some way to get a 3.0.0 kernel > >> with "default" or generic .config file that will work with debian > >> on my T520? > > > > Get linux-image-3.0.0-1-amd64 3.0.0-2 from Debian Sid. No need to > > build a kernel, just use the binary package. > > > > Thanks, > > -- > > Martin 'Helios' Steigerwald - http://www.Lichtvoll.de > > GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 > > Hi Martin- > > Thanks once again, I really appreciate your help. > > I downloaded the linux-image-3.0.0-1-amd64 deb file from Debian Sid as > you suggested, but when I tried to install it I got this error message > from the package instaler: > > "Error: Dependency is not satisfiable: linux-base (<=3~) > > Hmm... I thought at first that referred to the fact that I had booted > into my 3.1.0-rc1 kernel, but I also tried it after booting into my > old 2.6.32-5-amd64 kernel and still I get the same error message. No, this is related some to a package dependency. > Can you explain how to work around this? I bet you need linux-base 3.3 from unstable as well. > When I search inside Synaptic for linux-base I only see the one that's > installed- 2.6.32-35, even though I am running the 3.1.0-rc1 kernel... > so I am confused. Do I need to add a new/different repository to my > sources.list file in order to install a later version? You could do something like this: deb http://ftp.de.debian.org/debian/ squeeze main non-free contrib deb-src http://ftp.de.debian.org/debian/ squeeze main non-free contrib deb http://ftp.de.debian.org/debian/ sid main non-free contrib deb-src http://ftp.de.debian.org/debian/ sid main non-free contrib and then in /etc/apt/preferences to avoid updating everything to Sid: Explanation: lower priority than stable (500), 26.8.2011 Package: * Pin: release a=unstable Pin-Priority: 400 BTW when you used 2.6.32 before, are you using Squeeze completely? Then I suggest upgrading to Wheezy. Squeeze is not recent enough for Sandybridge. So some of your problems might just be related to using Squeeze. I just run everything at Sid at the moment and it works marvellous. A bit more conservative approach would be using Wheezy which should be recent enough by now. At least 3.0 kernel already made it to Wheezy: martin@merkaba:~> rmadison linux-image-3.0.0-1-amd64 linux-image-3.0.0-1-amd64 | 3.0.0-1 | wheezy | amd64, i386 linux-image-3.0.0-1-amd64 | 3.0.0-2 | sid| amd64, i386 Thanks, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to debian-laptop-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201108261517.18848.mar...@lichtvoll.de
Re: Lenovo T520i X11 Intel i810 "static display" issue- SOLVED
Am Mittwoch, 31. August 2011 schrieben Sie: > I wanted to get back to you and tell you that I was finally able to > install a 3.0 kernel (Wheezy- not Sid) and you are correct- now my > Thinkpad T520 will resume from sleep properly. Thanks so much! > > For some background for others: I had been using Squeeze (kernel > 2.6.32-5) and my T520 would not resume from sleep properly. Looking > around on the web I saw that a number of other people were having this > problem. > > I first tried to compile my own 3.0 kernel within Squeeze- but alas, > choosing the proper kernel compile options for this beginner proved to > be too difficult. [...] > Now my T520 does resume from sleep properly. So there you are after quite an odyssey ;). Often its easier to just use the standard package. I used to compile my own kernels for my older ThinkPads, but on the T520 the packaged 3.0 works so great, that I didn´t feel like compiling my own one, although it should compile pretty fast on that T520 with Intel SSD 320. A good starting point for an own kernel is the config from the debian one and the package kernel-package. But that for sure it out of the scope of debian-laptop. Maybe sometime I subscribe to debian-user, but for now I am just on debian-user-german. -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to debian-laptop-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201108312042.25413.mar...@lichtvoll.de
Re: Laptop
Am Samstag, 11. April 2015, 16:57:27 schrieb Tobias: > Hi, Hi Tobias, > can somebody recommend me a Laptop that will be able to run Jessie and > meets business-related requirements (long battery life, low noise > emission, SSD optional) and is not too expansive? Thanks in advance! I recommend getting a used business line ThinkPad. Either T series or X series depending on the display size and optical drive requirements. Intel only graphics. For SSD you need at least T43 or similar recent X series model. But I recommend at least T500 for T series. For X series you already got recommendations. Ciao, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to debian-laptop-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/1886751.HvkMUmB9b6@merkaba
Re: Home Directory in SSD
Am Samstag, 6. Februar 2016, 16:41:59 CET schrieb Aleksandar Atanasov: > You have to determine where exactly is speed required with your files. > As it was already mentioned (twice) speed is required only for certain > things. Putting a video file that is opened once every couple of days is > most certainly not crutial to the system's performance. I haven't heard > anyone ever suggesting that you store your video collection on an SSD... I still store larger files like audio files on SSD as well. Why? Laptop is so much more silent with only SSD equipped. That said, I believe that if you buy a good SSD they are not less reliable than harddisk. The Intel SSD 320 is more than 4 years old and according to smart data still considers it self to be new (smartctl -a excerpt): ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE 233 Media_Wearout_Indicator 0x0032 098 098 000Old_age Always - 0 Also the mSATA SSD still seems fine: 173 Wear_Leveling_Count 0x0032 092 092 000Old_age Always - 267 202 Percent_Lifetime_Used 0x0031 092 092 000Pre-fail Offline - 8 Granted, chips can fail. But a harddisk also has chips that can fail. And mechanics that can fail. Anyway, I had very good experiences with harddisk as well as SSDs so far. In my private use I have not one harddisk or SSD hardware failure with data loss in the last ten years. I suggest tough that you really take the time to read reviews and if in doubt rather go for the more reliable device. I did with Intel SSD 320 and then later with Crucial m500. The Crucial device is said to be a bit slower in reviews I found, but more reliable than other brands like Samsung at that time. Its some years since that already, and I didn´t do any research recently. So don´t take my findings for current state. Another suggestion: Many laptops can hold two SSDs. I do have Intel SSD 320 300 GB SATA SSD + Crucial M500 480 GB mSATA in the same ThinkPad T520. And the most critical data in my home directory is on BTRFS RAID 1. So first I can scrub checksums and second I can repair the BTRFS if only one SSD returns crap. I had some checksums errors with Crucial M500 SSD once with I attribute to a sudden power loss. The Intel SSD 320 does have condensators that should allow the SSD to still write out all the data in case of power loss, but I didn´t see any place where Crucial engineers could have fitted condensators on the really tiny mSATA SSD. BTRFS scrubbing repaired this just fine, there were no SMART errors and it never happened again, so… I left it at that assumption. That said only critical stuff like OS and /home is on BTRFS RAID 1 filesystems. The music collection and photo collection I update more rarely is only on the bigger Crucial mSATA SSD. The photo import directory is on RAID 1 again as I backup photos from SD card more often than I actually sort them into Digikam. So in any case if you want to be on the safe side with SSDs I suggest you use two of them. I´d also still leave some space of them unused if possible, although others think current SSDs have enough spare area, I still think that will extend lifetime and long time performance of SSDs due to the way SSDs work internally. Thanks, -- Martin signature.asc Description: This is a digitally signed message part.
Re: Home Directory in SSD
Am Sonntag, 7. Februar 2016, 00:11:30 CET schrieb Aleksandar Atanasov: > As for the quality of SSDs - yes, it has improved greatly (example for > some test of longevity: > http://techreport.com/review/27436/the-ssd-endurance-experiment-two-freaking > -petabytes). I have read however that SSDs are much more susceptible to > power failure compared to HDDs leading to a higher data corruption in case > a poor electrical grid is used to power the device up. However in terms of > protecting your data from shocks (dropping the notebook onto the floor) > SSDs are definitely much better. Because of the mechanics that are involved > in HDDs, such drives are less suitable for the task of storing data on > mobile devices. Depends on whether the SSD uses condensators to provide some energy in case of sudden power loss or not. And on whether there are bugs with that like in the Intel SSD 320. Luckily I never encountered the bug (that basically trashes the device), but I also keep my laptop battery inside the laptop at all time just for safety. -- Martin signature.asc Description: This is a digitally signed message part.
Re: Home Directory in SSD
Am Sonntag, 7. Februar 2016, 00:57:24 CET schrieb Matus UHLAR - fantomas: > On 06.02.16 19:30, Jos Collin wrote: > >I have Debian/testing installed completely in my 120GB SSD. I have > >learned that if an SSD fails, it is difficult to recover data from > >them. An SSD often does not give much warning before it fails. > >Electronic components don’t begin to grind or buzz as they grow > >older. They work – and then they don’t. > > > >So do I have to consider this risk and move the /home and /root > >directories to an HDD as they contain the Personal Data of each user, > >and only keep the Operating System files in the SSD ? How do you > >people keep the /home and /root directories, when you install the OS > >in an SSD ? (I have an Ultrabay Caddy, in which I can connect the HDD > >also in my Thinkpad T61). > > this is debian-laptop list, are you really talking about a laptop? > > I consider two possibilities for my home PC > 1. mirror SSD and HDD (I currently have two old 120 HDDs in mirror) I think this can be done with regular rsync or btrfs send/receive. I probably wouldn´t try to use BTRFS RAID 1 or SoftRAID 1, cause I think the slower disk will slow down write accesses. Also I am not sure whether the RAID implementations are aware of disk speed and for example shuffle most of the reads to the SSD in that case. I bet they aren´t (yet). > 2. using bcache, dm-cache or flashcache (cache HDD on SSD) If you use bcache with write caching you basically double your risk of data loss. > maybe someone could share experience with some of these... bcache worked in a test more than a year ago I think. But I don´t use it, so… no idea about long term reliability. And alternative is also using a hybrid harddisk that uses a little amount of flash as caching. You have one device and its all transparent. But you have lots of logic in firmware. Seagate does these, I am not sure whether other vendors do these meanwhile as well. Personally I only stuff flash into laptops anymore cause laptop is just much more silent this way. I didn´t think back then that at one time I´d consider even a 2,5 inch harddisk as loud, but compared to the SSD – which can make some noices on heavy write accesses, at least my Intel SSD 320 does – they are. Thanks, -- Martin
Re: Home Directory in SSD
Am Sonntag, 7. Februar 2016, 08:48:55 CET schrieb Jos Collin: > Hi, Hi Jos, > First of all, I apologize for starting an off-topic discussion. > > i'm using an Intel 535 Series 120GB SSD. So I believe that it is > reliable and good quality product. I am very pleased with my Intel SSD 320 and back then only Intel published a statistic on failure rates. It was about 0,6% of devices being sent back (I think X-25M still back then) and about 0,4% of devices being actually really faulty. I didn´t have exact figures of other vendors but I think at least for some of them back then it has been more like 2-3% of device failures. Now 0,4 is still one device out of 1000 failing. I am not sure about the current state of affairs, but I do think Intel generally does reliable SSDs. Maybe not always the fastest ones, but honestly I don´t care that much. If its more reliable I tolerate a bit lower performance. Of course it shouldn´t crawl down on write access. > I have installed the entire OS in the SSD now. But considering the > failure rate of all SSDs in common, I'm thinking of creating a partition > for /home in the HDD and mount it during boot. But as I have already > created /home in the SSD, I have to move that to the new partition in > HDD now. I can leave /root in the SSD itself, as suggested by Matus > UHLAR in the other email thread. This method will secure the user data > on the fly, even though there is a compromise in performance (not a big > deal). If you have both an SSD and a HD in your machine, how about rsync or BTRFS send and receive to regularily backup /home from SSD to HD? With BTRFS on the HD, you can even snapshot your backup states and if you accidentally delete a file get it back from a backup. I think I´d use it if that laptop would still have an SSD. Right now I use BTRFS RAID 1 on two SSDs for important data like the data in / home: merkaba:~#1> btrfs scrub status -d /home scrub status for […] scrub device /dev/dm-0 (id 1) history scrub started at Sat Feb 6 17:56:55 2016 and finished after 00:10:08 total bytes scrubbed: 152.88GiB with 0 errors scrub device /dev/mapper/sata-home (id 2) history scrub started at Sat Feb 6 17:56:55 2016 and finished after 00:11:09 total bytes scrubbed: 152.88GiB with 0 errors merkaba:~> btrfs device stats /home [/dev/mapper/msata-home].write_io_errs 0 [/dev/mapper/msata-home].read_io_errs0 [/dev/mapper/msata-home].flush_io_errs 0 [/dev/mapper/msata-home].corruption_errs 0 [/dev/mapper/msata-home].generation_errs 0 [/dev/mapper/sata-home].write_io_errs 0 [/dev/mapper/sata-home].read_io_errs0 [/dev/mapper/sata-home].flush_io_errs 0 [/dev/mapper/sata-home].corruption_errs 0 [/dev/mapper/sata-home].generation_errs 0 I really like it that BTRFS can basically prove to me that the data is okay. So I also use it on my 2 TB backup harddisk, where I use its snapshot functionality to keep old backup states as well. I still use rsync for backup, but I consider switching to btrfs send/receive after testing it maybe with the smaller / (in addition to rsync) for a while. Remembers me to scrub / and /daten (for larger files only on the larger SSD) as well again. :) Thanks, -- Martin
Re: Home Directory in SSD
Am Sonntag, 7. Februar 2016, 22:06:16 CET schrieb Jos Collin: > On 02/07/2016 03:50 PM, Martin Steigerwald wrote: > > If you have both an SSD and a HD in your machine, how about rsync or BTRFS > > send and receive to regularily backup /home from SSD to HD? > > > > With BTRFS on the HD, you can even snapshot your backup states and if you > > accidentally delete a file get it back from a backup. I think I´d use it > > if > > that laptop would still have an SSD. > > > > Right now I use BTRFS RAID 1 on two SSDs for important data like the data > > in / > > home: > Hi Martin, > > I'm new to btrfs. Does this sync the /home in the SSD to HDD on the fly > (like a periodic rsync), without me entering commands ? I need something > that happens automatically and regularly. > > For the time being, I am mounting a partition in HDD to /home until I > find a better method. btrfs send/receive are implemented in the btrfs command. As with rsync you would have to setup some script / cron job to them to happen regularily. -- Martin
Re: Home Directory in SSD
Hi! I see no issues against uses btrfs send/receive or rsync as backup from SSD to HDD. With btrfs send/receive I think after the initial sync it can even work quickly enough to have it run every 10 minutes or so. But if RAID allows write-mostly, why not. I think BTRFS RAID does not (yet). Am Sonntag, 7. Februar 2016, 17:49:14 CET schrieb Matus UHLAR - fantomas: > >> 2. using bcache, dm-cache or flashcache (cache HDD on SSD) > > > >If you use bcache with write caching you basically double your risk of data > >loss. > > double? explain, please. If one of two disk fails => boom. Unless you use bcache not for caching writes, but even then I am not sure as it creates a new block device on top of the ssd and harddisk – I think it can at least lead to some filesystem inconsistencies. In case you use RAID 1 or regular rsync / btrfs send/receive one disk can fail and you are still good. So comparing these two approaches its even more than double the risk. Its about double the risk comparing with storing different data on SSD and HDD. You loose the data that the broken device stores, but with bcache you loose the data if any of the devices break. > >And alternative is also using a hybrid harddisk that uses a little amount > >of flash as caching. You have one device and its all transparent. But you > >have lots of logic in firmware. Seagate does these, I am not sure whether > >other vendors do these meanwhile as well. > > hybrid HDD is very nice alternative, and in fact it is what I would > like to "emulate" using either flash cache mechanism mentioned above. > The main difference is that current hybrids only have 8GB of cache, so the > speedup is not that noticeable. Even though you have two parts in such a hybrid HDD that can fail, it is at least one device. -- Martin
Re: Home Directory in SSD
Am Donnerstag, 11. Februar 2016, 21:40:20 CET schrieb herve: > On 11/02/2016 15:23, Stefan Monnier wrote > > I have it all on my SSD, and I use daily backups. > > Before that, I used an HDD, with the same daily backups. > > > > If you don't perform regular backups, then clearly you don't care about > > your > > data, so why bother trying to distinguish if the HDD is ever so slightly > > less untrustworthy than the SSD? > > > > And if you do perform regular backups, then you only need your storage > > media to be "reliable enough", so again the minute differences in > > failure scenarios for HDDs and SSDs don't matter because both of them > > are "reliable enough". > > > > The differences are important, tho: > > - SSDs are silent. > > - HDDs are not silent. > > - Some operations are much faster with the SSD (tho those don't > >affect me very much, in practice, so it's not the main selling point). > > - Oh, did I mention how little noise SSDs emit? > > > > > > Stefan > That is not untruth, but if you think about ecology, a media that lives > 5-10 years is not the same thing that one living 3-5 years. > > You trade lifetime for comfort , maybe another one will make a different > choice ? To take en example : you can decide to drive a Ferrari at 160 > Mph, and you'll gain a lot of time in transfer, but you'll have the risk > of an accident and a shorter life. It's your choice, not THE only choice. I don´t think it is proven that SSDs fail earlier than HDDs. So far none of the SSDs I use have failed and one is almost 5 years, still thinking about itself that it is actually almost new according to SMART data. And the only reason it isn´t older is that it is the first SSD I got. I expect it to live on for years to come. So do you have any factual data to prove your claim? So far I didn´t see any proof that SSDs fail more often or earlier than harddisks. Thanks, -- Martin
Re: Home Directory in SSD
Am Freitag, 12. Februar 2016, 11:48:10 CET schrieb herve: > On 11/02/2016 22:25, Martin Steigerwald wrote: > > I don´t think it is proven that SSDs fail earlier than HDDs. So far none of > > the SSDs I use have failed and one is almost 5 years, still thinking about > > itself that it is actually almost new according to SMART data. And the only > > reason it isn´t older is that it is the first SSD I got. I expect it to live > > on for years to come. > > > > So do you have any factual data to prove your claim? > > > > So far I didn´t see any proof that SSDs fail more often or earlier than > > harddisks. > > > > Thanks, > > There can be any proof or factual data yet because this is a very young > technology. But what we know is that there is a limited number of write > cycle. Of course hdd can fail with mechanical problems ssd don't have. > From what i read, ssd are much much faster, but, at the time i read, > hdd can be more reliable. And the cost of reliable ssd are no comparable > with hdd. Heise and others tried to destroy SSDs by writing to it. And they found out that it is not easy to do so. The Intel SSD 320 is rated for 20 GiB of writes a *day* for a minimum of 5 years usable life-time. With usable they mean: It remains fast. Now I ask you: Do you have 20 GiB each day of writes? That is 20 * 365 = 7300 GiB or about 7,12 TiB. Now this Intel SSD is almost 5 years old. That will make almost 7,12*5 = 35,6 TiB. I do have: 241 Host_Writes_32MiB 0x0032 100 100 000Old_age Always - 1000827 Thats 1000827 * 32 MiB = about 30 TiB. I admit thats pretty close and sometimes I wonder about Plasma 5 desktop with KDEPIM + Akonadi + Baloo and before that Nepomuk just writing a tiny bit much onto the SSD. But I also compile KDE Frameworks + KDEPIM + Linux Kernel from source and that adds more I/O. But my main point is: The SSD wrote about 30 TiB. Yet, according to smartctl -a it thinks: 233 Media_Wearout_Indicator 0x0032 098 098 000Old_age Always - 0 this media wearout indicator which according to Intel is related to write cycle consumption start at the value 100. Now it is 098. So the SSD considers itself to be almost new. Intel writes in a PDF about the media wearout indicator[1]: > The value of the E9 SMART attribute is defined by Intel as the media wearout > indicator of an Intel SSD. When the value reads 1, this indicates that the > SSD is reaching the predetermined maximum media wearout limit. Intel > recommends that you replace the SSD at this point or back up the SSD to > help prevent the loss of data. And in addition this: > E9 SMART Attribute > The E9 SMART attribute reports a > normalized value of 100 (when the SSD > is brand new out of the factory) and > declines to a minimum value of 1. > The normalized value decreases as the > NAND erase cycles increase from 0 to the > maximum-rated cycles. Once the > normalized value reaches 1, the number > will not decrease, although it is likely that > additional wear can be put on the device. > Figure 1 shows how the value of the E9 > attribute decreases over time in a sample > usage model with a consistent workload. and the figure shows a linear scale. So here is facts. Please stop spreading FUD about SSDs in general. This Intel SSD 320 did 30 TiB of host writes and it is as fast as ever. There are a tiny bit of erase failures: 172 Erase_Fail_Count0x0032 100 100 000Old_age Always - 169 it raised to 169 I think within the first or second year of the SSD and didn´t raise then anymore (I can look it up if desired. So I bet thats just some cells that weren´t up to par with the rest. Still the normalized value was 100 and is 100, so I don´t care either. So a *good* SSD is very reliable. So just don´t buy cheap crap and read some reviews. Good SSDs are built to last. I am not sure about the current generation, but from what I read here even SSDs with TLC chips are designed to last long if not longer than the ones with MLC chips I have here. I wondered a bit about the Crucial m500 480 GiB mSATA SSD. As it reports: 202 Percent_Lifetime_Used 0x0031 091 091 000Pre-fail Offline - 9 But with 91% of lifetime left and it almost 2 years old as well, I can extrapolate the lifetime to nine times * 2 years that 18 years. Quite good if you ask me! Even it is just lasts for 8 more years thats fine. So while I still recommend to leave some free space (I did) and use noatime and so, I do think except for crazy server workloads you can only destroy SSDs by writing too much if you actually do this with full intention. It won´t happen on normal use, it won´t even happen on heavy use like the crazy things I do on this machine (like storing a maildir with 1,7 million single mai
Thank you (was: Re: [spam subject elided] (fwd))
Dear Alex, dear Dave, dear everyone. Alexander Wirt - 04.12.18, 07:47: > On Mon, 03 Dec 2018, Marvin Renich wrote: > > * Jim Popovitch [181203 15:21]: > > > How long should Dave wait out the problem before insulting folks? > > > 1 > > > month, 2 months, 6 months, 1 year, 5 years, 10 years, 25 years? > > > > The correct answer is "much longer than 25 years". > > He should at least wait until he gets its facts right. Afair/afaik > there never was any offer to help or constructive help we received > from him. First off: Alex, thank you very much for your persistent and continued dedication to serve the project and all of us discussing on Debian mailinglists as a listmaster. I imagine that it can be challenging to deal with negativity like this. Now some random thoughts about this: Regarding spam there for me there is one most important rule: Never *ever* reply to spam, and never *ever* do so publicly. Why? In case of answering privately you tell the spam sender that your mail address exists or in case of forged from address to tell that to some other random account. In the case of responding publicly you *highlight* the spam. I am highly successful with a Postscreen and rspamd based Spam-filtering setup and I did not see the original spam at all. My mail server just refused to receive it. Replying to the spam publicly just caused me to see it as well and so was in dis-service for me. Recently I did not even need any additional rules after having blocked whole domains of a mail provider with dubious domain names after they insulted me for sending them abuse complaints. But its mostly just my mail, I do not have the responsibility for anyone else. So I can afford to make it more aggressive and take a greater risk regarding false positives. While I see rspamd as a very viable alternative to my previous Spam- filtering setup, I am currently focused on other things than to offer help to the listmasters. But then I accept and am grateful for all I receive. Also I see the challenge to change the an existing setup, especially when it is as large-scale as I'd expect the existing setup for these mailing lists. That written: I also complained to listmasters about a certain repeated spam mail that appeared on almost every Debian mailing list I was subscribed to and was not acted upon. However I learned that I am much more at peace of mind if I just block it first with my mail server… and then notify listmasters in a friendly and constructive way. Cause then I do not give my power away and make my own peace of mind dependent on what listmasters may or may not do about it. Listmasters volunteer. They owe me *nothing*. They have a life outside of volunteering as a listmaster. All they give, they give freely without asking anything in return. The most appropriate reaction to this in my eyes is gratefulness. So again: Thank you, Alex, thank you, all listmasters for your time, your dedication and even that you put up with all the negativity that people who complain (including me as I did it) send towards you. I commit to stop complaining and to start helping where I can, and even if its just by reporting the spam mail via the web interface. Of course I know not everyone has their own mail server, but Dave, if you are proficient with anti spam measures and run your own mail server… why are you even affected? My anti spam measures, and I am certainly not an expert in this, blocked the original mail just fine. So if you do not like to spend the time to help listmasters, how about you just fix your own mail server not to accept mails like this in you are so proficient with anti spam measures? Thanks, -- Martin
Hibernation with ThinkPad T14 AMD Gen 1 + 2
Hi! Back then when I bought the first ThinkPad T14 AMD Gen 1 I had memory corruption issues after resuming from hibernation several times. Which led to general protection faults and BTRFS filesystem errors. This also happened with sleep mode set to Linux. It was one with AMD Ryzen 7 PRO 4750U and 32 GiB RAM. Since that time I do not hibernate any of the ThinkPad T14 AMD laptops anymore. Including a ThinkPad T14 AMD Gen 2 one. But I'd like to revisit that decision since there have been a lot of major kernel updates in between. But before risking the integrity of any of the filesystems on one of those machines again during testing, I'd like to gather feedback: Anyone here hibernating a ThinkPad T14 AMD 1 or 2 regularly and can tell whether it works reliably? I am interested in experiences regarding both models, but for now mostly model 1. Of course if someone also uses BTRFS that would be even better. On my research on the web I concluded that hibernation may work okay meanwhile. But as I do like to avoid any filesystem corruption issues, I'd like to read about more experiences before I make a decision. Best, -- Martin
Fwd: Working 3D acceleration on IBM ThinkPad T23 ;-)
Hello, well maybe this is of interest here to: Its an easy description on how to setup 3D acceleration for any IBM ThinkPad T23 user using Etch/Sid who has not yet setup 3D. According to a posting of Nate Bargmann on the linux thinkpad mailing list this works also with a T42 when using the radeon snapshot instead. Nate also observed: " I notice that with udev and the stock Debian 2.6.14 kernel (Sid) starting Xorg 6.9 automatically loads the savage and drm modules on the T23 and the radeon and drm modules on the T42." Date: Sat, 31 Dec 2005 22:42:13 -0600 Subject: Re: [ltp] Fwd: Working 3D acceleration on IBM ThinkPad T23 ;-) List-Id: This list for users of Linux on IBM Thinkpads. On the T23 it needed a complete reboot not only a X server reload till the driver worked. Before that glxgears and glxinfo both gave a segmentation fault. Have a Happy New Year!! Regards, Martin -- Weitergeleitete Nachricht -- Subject: Working 3D acceleration on IBM ThinkPad T23 ;-) Date: Samstag 31 Dezember 2005 18:08 From: Martin Steigerwald <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Hello, finally I got it working. I had a try compiling the needed parts on my own but I had no luck with that and no time to dig deeper. Now I didn't need to compile a single bit. I just used X.org 6.9.0.dfsg.1-1 from Debian unstable from today ;-) with the savage drm module from this self compiled kernel (with also sports software suspend 2 which is working nicely ;): Linux version 2.6.14.5-tp23-sws2-2.2-rc15 ([EMAIL PROTECTED]) (gcc-Version 4.0.3 20051201 (prerelease) (Debian 4.0.2-5)) #1 PREEMPT Fri Dec 30 11:53:37 CET 2005 Together with the latest binary DRI driver snapshot: http://dri.freedesktop.org/snapshots/savage-20051220-linux.i386.tar.bz2 Using a mixture of Debian Etch/Sid. I just copied savage-20051220-linux.i386/savage/savage_dri.so to /usr/X11R6/lib/modules/dri/ and restarted the X server. You also have to make sure that the savage module is loaded, I put "savage" into /etc/modules in case its not automatically loaded by udev's hotplug functionality. I got this: [EMAIL PROTECTED]:~/Computer/deepdance/3D-Treiber -> glxgears 1182 frames in 5.0 seconds = 236.227 FPS 1159 frames in 5.0 seconds = 231.676 FPS 1175 frames in 5.0 seconds = 234.958 FPS Compared to this: [EMAIL PROTECTED]:~ -> glxgears 907 frames in 5.0 seconds = 181.400 FPS 690 frames in 5.0 seconds = 138.000 FPS 853 frames in 6.0 seconds = 142.167 FPS 711 frames in 5.0 seconds = 142.200 FPS And tested with supertux, chromium, powermanga and they all worked well and smooth compared to the jerkyness I had before. I also tested with celestia. It also worked well, but showed some rendering bugs in some cases. Sometimes it seemed the driver got the rendering area wrong. Well, thanks to everyone who made this possible! Could well be that Debian Etch will come with Savage 3D acceleration out of the box then, I guess. Special thanks to Alex for hanging out on this mailing list and helping people getting things set up. I hope my little description above will help other TP23 users who are not scared away from using Debian Etch/Sid to get 3D acceleration the easy and lazy way. ;) Have a happy new year! Now I just need to find out the name of the large free software elite alike space flight game again I could never play cause it was way to slow on my hardware. ;) Regards, -- [... sig ...] --- -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [ltp] Re: Fwd: Working 3D acceleration on IBM ThinkPad T23 ;-)
Am Dienstag 03 Januar 2006 05:24 schrieb Nate Bargmann: > > The xlibmesa-dri package in Sid provides the necessary radeon_dri.so > > file and places it in the right place. It supports several other > > cards, but not the savage (yet), unfortunately. > > Let me add that as of today, Jan 2, 2006, that installing the > libgl1-mesa-dri package will remove the xlibmesa-gl package, but won't > break Xorg like a couple of days ago. The savage_dri.so file is > provided by libgl1-mesa-dri and performance is unchanged from the > manual installation method on my T23. > > Painless 3D on the T23! Hello, does anyone know the reason for those different DRI driver packages and the reason, why savage is not included in xlibmesa-dri? Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [ltp] Re: Fwd: Working 3D acceleration on IBM ThinkPad T23 ;-)
Am Donnerstag 05 Januar 2006 13:31 schrieb Nate Bargmann: > > Hello, > > > > does anyone know the reason for those different DRI driver packages > > and the reason, why savage is not included in xlibmesa-dri? > > There is an existing Debian bug report on just this issue: > > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=338047 > > The last comment seems to state authoratively, "libgl1-mesa-dri comes > from the mesa source package, not xorg-x11." > > I'm not willing to download the entirety of Xorg 6.9 just to verfiy > this claim. At least a package exists and one does not have to resort > to manually locating and installing a file not compiled with the rest > of the system. Hello Nate, I considered installing it, but actually I do not like this too much, as I would like to be able to compile a KDE application spontaneously ;): [EMAIL PROTECTED]:/sys -> LANG=en_GB apt-get install libgl1-mesa-dri/unstable Reading package lists... Done Building dependency tree... Done Selected version 6.3.2-2 (Debian:unstable) for libgl1-mesa-dri The following extra packages will be installed: libdrm1 libgl1-mesa-dri The following packages will be REMOVED: kdelibs4-dev kspy libarts1-dev libglu1-xorg-dev libopenexr-dev libqt3-mt-dev x-window-system x-window-system-core xlibmesa-dri xlibmesa-gl xlibmesa-gl-dev xlibosmesa-dev xlibosmesa4 The following NEW packages will be installed: libdrm1 libgl1-mesa-dri 0 upgraded, 2 newly installed, 13 to remove and 0 not upgraded. Need to get 11.6MB of archives. After unpacking 10.4MB disk space will be freed. Do you want to continue [Y/n]? n Abort. This one looks better however: [EMAIL PROTECTED]:/sys -> LANG=en_GB apt-get install libgl1-mesa-dri/unstable libgl1-mesa-dri-dev/unstable mesa-common-dev/unstable Reading package lists... Done Building dependency tree... Done Selected version 6.3.2-2 (Debian:unstable) for libgl1-mesa-dri Selected version 6.3.2-2 (Debian:unstable) for libgl1-mesa-dri-dev Selected version 6.3.2-2 (Debian:unstable) for mesa-common-dev The following extra packages will be installed: libdrm1 libgl1-mesa-dri libgl1-mesa-dri-dev mesa-common-dev The following packages will be REMOVED: x-window-system x-window-system-core xlibmesa-dri xlibmesa-gl xlibmesa-gl-dev xlibosmesa-dev xlibosmesa4 The following NEW packages will be installed: libdrm1 libgl1-mesa-dri libgl1-mesa-dri-dev mesa-common-dev 0 upgraded, 4 newly installed, 7 to remove and 0 not upgraded. Need to get 11.8MB of archives. After unpacking 8380kB of additional disk space will be used. Do you want to continue [Y/n]? I think I stick with the manual installation approach unless I know more about the background of those packaging decisions ;-) or unless xlibmesa-dri contains the savage driver. If libgl1-mesa-dri is to be used in future as Alex comment suggests the x-window-system and x-window-system-core dependencies should be adjusted. Maybe this whole affair is worth a bug report. Well I will add this to the existing one you mentioned Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
getting correct time on a laptop (IBM ThinkPad T23)
Hello, I have an IBM ThinkPad T23 with highly inaccurate clock. It misses accurate time by several minutes a day. Long time I used chrony. It worked most of the time, but not always. Right now the problems I have with chrony - see below - increased and none of the other approaches worked like I want. I want the following: 1) When the laptop has internet access it should synchronise with NTP time 2) Otherwise a program should correct the clock by its average inaccuracy (like chrony is supposed to do). That is be important cause sometimes the laptop is without internet access for several days. 3) Suspend to disk should be handled nicely. 4) Ideally on resume and boot the time is corrected due to NTP time or the inaccuracy factor via one big step. Right now I am fighting with chrony and different other approaches and none of them worked like I want it to. My problems with chrony: 1) I had chrony claiming that system time had 0 seconds difference to NTP time while there was a difference of 10 or more minutes. This is with a standard chrony installation (after aptitude purge chrony). 2) Chrony apparantly cannot set the hardware clock. I get a input/output error on modprobe rtc, and it seems that the module genrtc doesn't do the trick. I reverted on not letting chrony do that but the usually debian hwclock scripts which seems to work. 3) Sometimes net access is not detected properly. I have copied /etc/ppp/ip-up.d/chrony and /etc/ppp/ip-down.d/chrony to /etc/network/if-up.d and /etc/network/if-down.d respectively in order to have chrony switched online and offline accordingly. I am using ifplugd and guessnet for automatic network configuration on demand which is worked quite well. 4) After suspend to disk chrony does not seem to have any NTP sources at all... I tried different approaches, right now I am using: #!/bin/sh /etc/init.d/ifplugd stop ifdown eth0 /etc/init.d/chrony stop # Noch nicht gespeicherte Daten sichern sync # Einschlafen hibernate /etc/init.d/chrony start /etc/init.d/ifplugd start Rationale for this: chrony and network are stopped before suspend since they do not work anyway while the notebook is powered off. After resume chrony is started and should be offline. Then ifplugd is started. Now whenever a network link is availble network is configured and chrony should be switched to online mode due to /etc/network/if-up.d/chrony that I copied there. But apparently that doesn't work. I also tried openntpd, which didn't even start on my system as well as the full blown ntp server which sort of worked, but then it does not provide . Once I even was desperate enough to just start ntpdate every hour with a cron job. On other approach would be to fix the hardware clock such that it would be that much too slow. Maybe replacing the battery could help? Any other ideas? Right now its extremely frustrating cause none of the approaches work to properly and its that I didnt try hard to make it work. Ideally I would have something like chrony but with the enhancement that it would simply work even for suspend and that it just finds out itself whether internet access is there or not. Switching it to online and offline mode via scripts is just extremely error prone IMHO. Any tricks I can play with the chrony configuration to make it work? chrony seems to be quite old, is upstream still alive? Any sensible replacement available? Any ideas? I greatly appreciate it when I could be able to fix up this time stuff without investing another dozen of hours. If I cannot fix chrony easily I probably revert to ntp-simple for now and accept wrong time when the laptop is not with internet access. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: getting correct time on a laptop (IBM ThinkPad T23)
Am Mittwoch 21 Juni 2006 13:05 schrieb teefour: > hello martin, > > > I also tried openntpd, which didn't even start on my system as well > > as the full blown ntp server which sort of worked, but then it does > > not provide . Once I even was desperate enough to just start ntpdate > > every hour with a cron job. > > openntpd works fine here. not just on my thinkpad, but on most servers > and other machines as well. > after realizing that chrony did not do what i want i ditched it for > openntpd and my timekeeping problems were gone. Hello, I had it running for several days now... it tells valid sources in syslog and that it adjusts the time, but still its 2 minute off... I am now installing ntp-simple for now. If nothing else helps I go with a cron job that uses ntpdate to the hardware clock with -set and uses the adjtime feature of it. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: getting correct time on a laptop (IBM ThinkPad T23)
Am Mittwoch 21 Juni 2006 17:02 schrieb Paul Kimoto: > On Wed, Jun 21, 2006 at 08:34:54AM +0200, Martin Steigerwald wrote: > > 1) I had chrony claiming that system time had 0 seconds difference to > > NTP time while there was a difference of 10 or more minutes. This is > > with a standard chrony installation (after aptitude purge chrony). > > > > 2) Chrony apparantly cannot set the hardware clock. I get a > > input/output error on modprobe rtc, and it seems that the module > > genrtc doesn't do the trick. I reverted on not letting chrony do that > > but the usually debian hwclock scripts which seems to work. > > Do you invoke the (chronyc) "trimrtc" command? (If this works, you > probably want to "writertc" as well.) Hello Paul, okay I missed that. I just used the default configuration, but it complained about missing rtc support anyway until I modprobed genrtc (rtc does not work). > Does your computer have a way to get IP addresses of external NTP > servers when chronyd starts (even if the network is down)? I believe > that otherwise chronyd forgets about those servers and never tries to > contact them. No... it uses automatic network configuration via guessnet and ifplugd. Network is down when no network is available. So I may be better off using the IP addresses of some NTP servers? I just don't get it how chrony can tell me my system clock is 0.0 seconds ahead of NTP time, while "sources" show that there are three sources available and just to make completely sure I manually put it online by issuing "password blabla" and "online" and the system clock is definately several minutes behind NTP time... well sure its not fast of NTP time then but shouldn't chrony adjust the clock anyhow, like it sometimes does? I just install ntpdate right now, run it once and sure the clock is right up to the second with my radio controlled solar alarm clock. Why do chrony and openntpd fail to do that for me? Well probably worth a bug report and further investigation, but that will have to wait until I manage to take the time for it. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: getting correct time on a laptop (IBM ThinkPad T23)
Am Mittwoch 21 Juni 2006 17:02 schrieb Paul Kimoto: > > 2) Chrony apparantly cannot set the hardware clock. I get a > > input/output error on modprobe rtc, and it seems that the module > > genrtc doesn't do the trick. I reverted on not letting chrony do that > > but the usually debian hwclock scripts which seems to work. > > Do you invoke the (chronyc) "trimrtc" command? (If this works, you > probably want to "writertc" as well.) > > Does your computer have a way to get IP addresses of external NTP > servers when chronyd starts (even if the network is down)? I believe > that otherwise chronyd forgets about those servers and never tries to > contact them. Hello Paul, I think I found a setup that should do. 1) realtime clock support should be working. at least trimrtc does work. rtcfile is enabled in the config file. so maybe I was seeing a different problem which I interpreted as chrony could not set the hardware clock. 2) according to chrony faq hwclock should not update the hardwareclock on suspend or shutdown cause chrony wouldnt know that and cannot adjust the hardware clock accordingly on the next boot or resume... on bootup hwclock --hctosys may be used before starting chrony... so I edited out /etc/init.d/hwclock.sh stop with a "return 0". /etc/hibernate/common.conf seems to be configured correctly already: ### clock SaveClock restore-only 3) I use this hibernate script: #!/bin/sh /etc/init.d/ifplugd stop ifdown eth0 /etc/init.d/chrony stop sync hibernate /etc/init.d/chrony start /etc/init.d/ifplugd start This way chrony should notice that the network goes down, before it is stopped, and when the network goes up (as soon as a link is detected) after it was started... I had the wrong order before and thus chrony may have missed that the network went up. Let's see how that goes. If that doesn't work I probably try replacing pool.ntp.org with some IP addresses... I wonder whether chrony retries the DNS lookups when an online command is issued, but the FAQ suggests that it does not... ;(. "Q: I have problems if I put the names of my NTP servers in the chrony.conf file. If you have no connection to the Internet at boot time, chrony won't be able to turn the names into IP addresses when it starts. There seem to be 2 solutions: 1. Put the numeric IP addresses in the chrony.conf file or 2. Put the server->IP address mappings in your /etc/hosts file and ensure that /etc/host.conf reads 'order hosts,bind'. The problem is that chronyd (currently) isn't designed in a way that allows hostname->IP address lookups during normal operation. I hope to work on this problem very soon." Hmmm, seems that I need IP addresses. Chrony should be more robust there IMHO and at least retry DNS lookups when an online command is issued and not only on /etc/chrony restart. Workaround could be to restart chrony whenever the network is brought up and put it online directly after that. instead of only putting it online. Maybe thats the best solution. I will try that. I am not completely sure whether it makes really sense to use hwclock --hctosys on boot or resume - since from what I understand chrony should be able handle the RTC clock in such cases too. But it probably won't harm either. I found a hint why chrony sometimes shows system clock is 0.0 seconds ahead of NTP time even when they differ. It does not seem to have done any measurement in that case. When I manually start one via burst, it shows the correct inaccuracy. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: getting correct time on a laptop (IBM ThinkPad T23)
Am Mittwoch 21 Juni 2006 17:02 schrieb Paul Kimoto:
> Do you invoke the (chronyc) "trimrtc" command? (If this works, you
> probably want to "writertc" as well.)
>
> Does your computer have a way to get IP addresses of external NTP
> servers when chronyd starts (even if the network is down)? I believe
> that otherwise chronyd forgets about those servers and never tries to
> contact them.
Hello Paul,
my setup didn't work yet... I had to add further foo magic to make it
work... this is getting ridiculous. ;(
When network is up, I add a at job to handle chrony at a later time.
Background: My DSL router my take some time to initialize the DSL
connection and chrony simply would not work correctly when connection is
not available when it starts. I have to do this asynchronously, as
otherwise ifplugd would block for at least one minute:
[EMAIL PROTECTED]:~ -> cat /etc/network/if-up.d/chrony
#!/bin/sh
# Start chrony stuff in at least one minute, lets see whether that works,
26.6.2006
at -f /etc/chrony/network-up.sh now + 2 minutes
So I restart chrony at least 1 minute after network has been brought up,
then I wait a while and then I tell it to go online... if I do not wait
chronyc will hang:
[EMAIL PROTECTED]:~ -> cat /etc/chrony/network-up.sh
#!/bin/sh
# This script tells chronyd that the connection is up so that it can
# contact the server. John Hasler <[EMAIL PROTECTED]> 1998-2003
# Any possessor of a copy of this program may treat it as if it
# were in the public domain. I waive all rights.
# Debugging?
#set -x
# Put a hint in the log file that we are running
logger "/etc/chrony/network-up.sh: restarting chrony and put it online."
# Make sure that chrony does DNS lookups, 25.6.2006, martin
# Should better wait till pool.ntp.org can be accessed!
# but for now I tell at to defer it about a minute, 25.6.2006, martin
/etc/init.d/chrony restart
# Otherwise chronyc will hang ;(, oh man this is sucking so extremely,
26.6.2006, martin
sleep 5
# Copied from /etc/ppp/ip-up.d/chrony, just modified the /var/run
filename, 25.6.2006, martin
/bin/pidof chronyd > /dev/null || exit 0
KEY=$(awk '$1 ~ /^commandkey$/ { print $2; exit}' /etc/chrony/chrony.conf)
PASSWORD=`awk '$1 ~ /^'$KEY'$/ {print $2; exit}' /etc/chrony/chrony.keys`
/usr/bin/chronyc << EOF
password $PASSWORD
online
EOF
touch /var/run/chrony-eth0-up
exit 0
I had a "burst 5/5" after the "online" command, but right now I am testing
whether it works without it.
Regards,
--
Martin 'Helios' Steigerwald - http://www.Lichtvoll.de
GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
getting correct time on a laptop *solved*
Hello,
just wanted to inform you, that I solved the getting correct time on a
laptop problem since quite some time (see older thread with that
subject). It works in a ifplugd / guessnet controlled automatic network
adaption setup and it works nicely together with suspend to disk
(currently sws2, but that shouldn't matter).
I still use chrony, but I made some adaptions for it in order to work.
1) on suspend chrony is stopped after the network is brought down
2) on resume chrony is started before the network is brought up again (and
thus the mechanism to switch chrony into online mode when network it
brought up works)
3) as a work-around for my DSL environment and chrony's feature not to try
again on DNS failure chrony is restarted again 1-2 minutes after the
network has been brought up. The DSL modem needs some time to setup the
internet connection.
4) hwclock is disabled:
[EMAIL PROTECTED]:~ -> head -1 /etc/init.d/hwclock.sh
exit 0
[EMAIL PROTECTED]:~ -> head -1 /etc/init.d/hwclockfirst.sh
exit 0
5) chrony runs with following configuration options:
server de.pool.ntp.org offline minpoll 8
server de.pool.ntp.org offline minpoll 8
server de.pool.ntp.org offline minpoll 8
Regulate realtime clock
rtcfile /var/lib/chrony/chrony.rtc
This gives a fast adaption if clock differs by more than 30 seconds...:
initstepslew 30 de.pool.ntp.org de.pool.ntp.org
6) For real time clock I added "genrtc" module to /etc/modules
7) Chrony is brought online when network is brought up and offline when it
is brought down: I copied the scripts /etc/ppp/ip-up.d/chrony
and /etc/ppp/ip-down.d/chrony to /etc/network/if-up.d/chrony
and /etc/network/if-down.d/chrony. I adapted up and down scripts to use a
different /var/run filename. I adapted the if-up.d-Script to trigger the
chrony restarting after to minutes
/etc/acpi/hibernate.sh:
#!/bin/sh
# Detect change of network environment and handle chrony
/etc/init.d/ifplugd stop
ifdown eth0
/etc/init.d/chrony stop
sync
hibernate
# Detect change of network enviroment and handle chrony
/etc/init.d/chrony start
/etc/init.d/ifplugd start
/etc/network/if-up.d/chrony:
#!/bin/sh
# Start chrony stuff in at least one minute, lets see whether that works,
26.6.2006
#nohup /etc/chrony/network-up.sh
at -f /etc/chrony/network-up.sh now
#/etc/chrony/network-up.sh &1 >/dev/null &
#screen /etc/chrony/network-up.sh
The nohup, /dev/null variante blocked ifplugd and a screen can not be
started from a script...
/etc/network/if-down.d/chrony:
#!/bin/sh
# This script tells chronyd that the connection is down
# so that it won't try to contact the server.
# John Hasler <[EMAIL PROTECTED]> 1998-2003
# Any possessor of a copy of this program may treat it as if it
# were in the public domain. I waive all rights.
# Put a hint in the logfile that we are running
logger "/etc/network/if-down.d/chrony: Putting chrony offline..."
/bin/pidof chronyd > /dev/null || exit 0
# Don't mark the connection offline unless we know ppp brought it up.
test -e /var/run/chrony-eth0-up || exit 0
KEY=$(awk '$1 ~ /^commandkey$/ { print $2; exit}' /etc/chrony/chrony.conf)
PASSWORD=`awk '$1 ~ /^'$KEY'$/ {print $2; exit}' /etc/chrony/chrony.keys`
/usr/bin/chronyc >/dev/null << EOF
password $PASSWORD
offline
EOF
rm -f /var/run/chrony-eth0-up
exit 0
/etc/chrony/network-up.sh:
#!/bin/sh
# This script tells chronyd that the connection is up so that it can
# contact the server. John Hasler <[EMAIL PROTECTED]> 1998-2003
# Any possessor of a copy of this program may treat it as if it
# were in the public domain. I waive all rights.
# Debugging?
#set -x
# Wait a bit till the DSL router likely built up an internet connection
sleep 60
# Put a hint in the log file that we are running
logger "/etc/chrony/network-up.sh: restarting chrony, put it online and do
some rapid measurements..."
# Make sure that chrony does DNS lookups, 25.6.2006, martin
# Should better wait till pool.ntp.org can be accessed!
# but for now I tell at to defer it about a minute, 25.6.2006, martin
/etc/init.d/chrony restart >/dev/null
# Otherwise chronyc will hang ;(, oh man this is sucking so extremely,
26.6.2006, martin
sleep 5
# Copied from /etc/ppp/ip-up.d/chrony, just modified the /var/run
filename, 25.6.2006, martin
/bin/pidof chronyd > /dev/null || exit 0
KEY=$(awk '$1 ~ /^commandkey$/ { print $2; exit}' /etc/chrony/chrony.conf)
PASSWORD=`awk '$1 ~ /^'$KEY'$/ {print $2; exit}' /etc/chrony/chrony.keys`
/usr/bin/chronyc > /dev/null << EOF
password $PASSWORD
online
EOF
touch /var/run/chrony-eth0-up
exit 0
I know my solution is a hack and it would be way better if chrony handled
network outages automatically and would retry DNS lookups from time to
time, but at least this works for my highly inaccurate IBM ThinkPad T23
clock.
And no right now I didn't yet replace the CMOS battery ;).
My solution produces undesirable results when you use more than one
network interface for internet access at once. If you switch network
Re: iTunes & Linux (Debian)
Am Mittwoch 27 Dezember 2006 04:28 schrieb Baz: > So, it appears the only way to play music purchased from Apple (mp4 > audio files) on Linux is to burn a CD, then rip them - or, use > CrossOver Office. Yes? Hello, I recommend having a look at magnatune.com (http://www.magnatune.com) for DRM free music. Amarok (http://amarok.kde.org) provides support for this music shop. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: file recovery - urgent
Am Mittwoch 14 Februar 2007 schrieb Micha: > Hello, > > Please excuse this is not strictly debian or laptop related, > but i'm in urgent trouble. > I accidently deleted a folder with 3500 files on my laptop, > just when i wanted to do a really overdue backup. > It's about 5 GB, nearly 3 months of work, and the last > backup is a month ago where most of the stuff was added > just recently, including hundreds of hours of the working time. > > When i realized what happened i shutdown the laptop. > The files are in the user's home on the root filesystem, > a journaling ext3. debian sid. > I was so shocked i didn't look for the exact time, but i > think i could boil it down to a 10 minute time frame. > I've no experience with file recovery. > I would be able to boot into a cd or dvd which i could > download and burn on another machine. > I'm aware a time based recovery could mean i'd have to reinstall > the KDE session and do more cleanup afterward, but that's > absolutely unimportant, if i only could recover the files. > It would be worth a complete reinstall. > > Any good idea what i can do ? Hallo Micha! BIG DISCLAIMER first: Try at your own risk. I have no practical experience with these tools... ... but they might just do what you want so it might be worth a try: [EMAIL PROTECTED]:~> apt-cache search ext3 undelete e2undel - Undelete utility for the ext2 file system recover - Undelete files on ext2 partitions (search http://packages.debian.org) Cauion: They do not work with ext3 only with ext2 as their package descriptions claim! I do not know whether it might be enough to remove the journal via tune2fs, but AFAIK then ext3 should behave just like ext2. Important: I recommend to make a block based copy of the partition with via ddrescue or dd, for example FIRST! So you have a second chance if anything goes wrong at your first attempt. ddrescue /dev/hda2 hda2.img (store hda2.img on some external harddrive, replace /dev/hda2 with your partition device file) You could then work with the image file only not altering whats on the harddisk until you are confident that it works! Use somehting like "mkdir /mnt/recover ; mount -o loop -t ext2 hda2.img /mnt/recover". (or even better make another copy of the file first;-) You should do everything with a Live CD. I recommend one of these (I am not sure whether they have e2undel or recover pre-installed, but if anything else fails you could install it yourself, I think on recent Knoppix this even works via aptitude otherwise you might have to compile the tool): - GRML: http://grml.org (text mode based, as limited GUI support) - Knoppix: http://www.knoppix.org (has full GUI support) Maybe someone who has experience recovering deleted files from ext2/ext3 filesystem can give you better hints. This is all untried and untested! Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ACPI with IBM ThinkPad T23
Hello, has anyone with an IBM ThinkPad T23 a Debian Linux setup with working ACPI? Last time I looked at the list somewhere in the Wiki at http://acpi.sourceforge.net the IBM ThinkPad T23 is listed as working. Well I tried with standard Debian 2.6.8, 2.6.7, 2.6.6 kernels and has no full success. Kernel 2.6.9 does work reliable on my notebook. It halts on some ACPI message (like some earlier versions of Kernel 2.6.8, see bug #282905) At least it went to standby mode, but after awakening ALSA sound did not work anymore and its not possible to switch from X11 to tty console anymore. Well I can switch but the X11 display will still remain on the screen. Today I tried it with the IBM ACPI stuff at: http://ibm-acpi.sourceforge.net/ For simplicity I used the Debian package at: http://debian.isg.ee.ethz.ch/public/ (This is not the most recent version of IBM ACPI) I compiled the ibm-acpi module and it loads fine. Now when I tell the laptop to go to standbye mode it goes to sleep but when awakening it shows a lot of ACPI messages and then halts completey (had no time to write these down and cut and paste obviously didn't work at that time anymore). I am not sure whether ibm-acpi makes that difference in behaviour, or wheter its just a newer version of the debian kernel 2.6.8 package. Anyone with an IBM ThinkPad who got farer than me (had time to compile an own kernel or whatever)? Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ACPI with IBM ThinkPad T23
Am Mittwoch, 19. Januar 2005 00:17 schrieb Michael Perry: > On Tue, 18 Jan 2005 23:00:10 +0100, Martin Steigerwald > > <[EMAIL PROTECTED]> wrote: Hello Michael, > I've been very successful with a T23 and ACPI. My recipe for success > is to use the 2.6.9 kernel with no additional acpi patches added and > upgrade the bios to the most recent revision. This sounds rather nice. ;) So you didn't use the standard Debian kernel (kernel.org + Debian patches) at all. I already thought about that but hesitated to take the time compiling my own kernel without knowing whether that helps. Too good that you tried. I think sooner or later I would have tried it too, but its good to know that this approach works. > I also have added the > ibm-acpi package for additional functionality. The ways I can suspend > now using acpi are lid close events and Fn-F4 key combinations. I've > written scripts for acpi controls and suspend scripts to suspend the > laptop with a lid close event and I found an issue on the T23 with the > earlier eepro100 driver and replaced it with the e100 in my kernel. An ACPI related issue or are there other problems with this driver? > I also seem to have issues with pcmcia and only the orinoco card so I > make pcmcia stop at a suspend event and then restart it when the > system comes back. Okay, that's fair. So it would stop an ISDN connection with my AVM Fritzcard before suspending, which would be a good idea anyway. Only thing would be that USB 2.0 PCMCIA card (which works marvellous under Linux BTW). I would have to unmount things plugged into it, before suspending. > My success rate with the T23 has been very good of late and I've > shared the same approaches with a few others on the ibm thinkpad > mailing list and they have been successful. Can you give me a hint where I can find that mailinglist? Is it an official thing from IBM? > If you want my scripts and my kernel config for the 2.6.9 kernel.org > kernel, let me know. As an aside with a kernel before 2.6.8.1 I > always had to patch the kernel with the acpi patchs from acpi.sf.net. > With 2.6.9 I dont any longer. Sure! Thanks a lot Michael. I didn't think about the 2.6.9 kernel, since the Debian packaged one didn't work at all on my ThinkPad. Seems that the Debian patches have negative impact on running kernels on IBM ThinkPad with full functionality. I really should have tried compiling my own kernel! When I get around taking the time I will write bug reports on that. I will also try with Debian kernel package for 2.6.10 which should be out now. Maybe it works better. > For me and a few others the primary success thing has been upgrading > the bios to the most recent version on IBM's website. I did so shortly after I bought that used IBM ThinkPad T23 laptop. It was fairly recent (well what you can call recent for T23) anyway. I upgraded from 1.15 to 1AET61WW (1.17). Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ACPI on IBM ThinkPad T23: suspend works with 2.6.10
Hello, today I installed 2.6.10 debian kernel package from unstable. Now suspend seems to work properly. ;) Michael, I still would like to see your acpi event setup and scripts, since 2.6.10 has IBM ACPI included already ;-). Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ACPI on IBM ThinkPad T23: suspend works with 2.6.10
Am Donnerstag 20 Januar 2005 10:01 schrieb Martin Steigerwald: > Hello, > > today I installed 2.6.10 debian kernel package from unstable. > > Now suspend seems to work properly. ;) Hello, suspend with an IBM ThinkPad R51 also works with this one. ;) Regards, -- Martin SteigerwaldMail: [EMAIL PROTECTED] Systemadministrator Fon: +49 (0)911 30999-0 team(ix) GmbH Fax: +49 (0)911 30999-99 Süd-West-Park 35, 90449 Nürnberg, Deutschland
ALSA sound not working after suspend
Hello, as I told suspend now works with IBM ThinkPad T23 and R51 when using Debian kernel 2.6.10. But I still have one problem: ALSA sound does not work anymore after a suspend through ARTS. On the ThinkPad R51 it is enough to restart the ARTS soundsystem via the KDE control panel. On the ThinkPad T23 I have to reboot to get ALSA working. Well so next step will be to compile my own kernel out of kernel.org sources and try it with that. Any other hints? Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ALSA sound not working after suspend
Am Donnerstag, 20. Januar 2005 17:05 schrieb Jeremy Brooks: > I had the same problem with a Dell X300. I was able to get sound > working after suspend by using kernel.org 2.6.10 sources, and patching > them with the software suspend 2 stuff. Now I can suspend to disk, and > when I resume sound works. The only thing that doesn't work (yet -- > I'm still playing with the config options) is DRM. After coming back > from suspend, the GL screensavers no longer animate, and glxgears just > sits, without spinning. Hello Jeremy, I am currently reading software suspend 2 HOWTO... did you try the option: "UseDummyXServer This option may also help if you experience strange things such as 3D not working upon resume. It launches a fake xserver that should cause proper initialization of the chipset." to get 3D working? http://softwaresuspend.berlios.de/HOWTO-4.html#ss4.1 Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
whole lot of ACPI problems with self compiled kernel
Hello, I now compiled an own kernel. I used unmodified kernel.org sources for 2.6.10 together with the config file from the Debian 2.6.10 kernel. I just added SOFTWARE_SUSPEND=y to also get suspend-to-disk. I made it as a initrd kernel just like the debian packaged kernel. I added "xfs" to /etc/mkinitrd/modules so that it will mount my mount filesystem. I also made the initrd and re-run update-grub. So well I can boot the thing, but 1) ALSA sound still does not work after suspend 2) suspend-to-disk doesn't work at all. I get "swsusp: FATAL: cannot find swap device, try swapon -a" I added "resume=/dev/hda7" to kopt in GRUBs menu.list and re-run update-modules. I checked that the swap partition is active as well. "free" showed swap memory. I even did "swapoff /dev/hda7", "mkswap /dev/hda7" (which created a new style V1 swap partition) and "swapon /dev/hda7" but this didnt help. One option will be to add /dev/hda7 to the kernel configuration. I thought a resume= parameter would be enough, but maybe it doesn't work in my case. 3) my AVM PCMCIA Fritzcard ISDN doesn't seem to work after suspend anymore. I cannot go online after suspending. Even when I do "/etc/init.d/pcmcia stop" and "/etc/init.d/pcmcia start" it won't work but that doesn't surprise me a lot, since I never was able to unload the ISDN driver modules once they were loaded. I am currently using isdn4linux emulation layer since Knoppix Red Hat ISDN config works with it properly. Anyone experience whether the CAPI driver stuff works with suspend? 4) During suspend-to-ram time goes considerably too fast. This was reported for the Debian kernel, but it seems to be true for the generic kernel as well. Next thing will be trying with software suspend 2 patches. A guy from this mailing list told me that this solved his ALSA not working after suspend issues on a Dell laptop. Maybe it also has to do with the default Debian kernel configuration or initrd. Well I don't need initrd so if it causes problems I go without it. I hope some day this will work as described on the ACPI homepage: transparently (without a need for end user support). Regards -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: whole lot of ACPI problems with self compiled kernel
Am Samstag, 22. Januar 2005 09:01 schrieb Martin Steigerwald: > 4) During suspend-to-ram time goes considerably too fast. This was > reported for the Debian kernel, but it seems to be true for the generic > kernel as well. > > Next thing will be trying with software suspend 2 patches. A guy from > this mailing list told me that this solved his ALSA not working after > suspend issues on a Dell laptop. Hello, I tried this (kernel 2.6.9 from kernel.org with stable software suspend patch for it) at all it gives me is: BIG FAT WARNING! Failed to translate the device name into a device id. The hibernate script won't work. It says software suspend is disabled due to missing resume2= parameter. Well but I am sure I have it: [EMAIL PROTECTED]:/boot -> grep "SUSPEND" config-2.6.9sws2-1 # CONFIG_SOFTWARE_SUSPEND is not set CONFIG_SOFTWARE_SUSPEND2_CORE=y CONFIG_SOFTWARE_SUSPEND2=y CONFIG_SOFTWARE_SUSPEND2_WRITER=y CONFIG_SOFTWARE_SUSPEND_SWAPWRITER=y CONFIG_SOFTWARE_SUSPEND_LZF_COMPRESSION=y CONFIG_SOFTWARE_SUSPEND_GZIP_COMPRESSION=y CONFIG_SOFTWARE_SUSPEND_DEVICE_MAPPER=m CONFIG_SOFTWARE_SUSPEND_TEXT_MODE=y CONFIG_SOFTWARE_SUSPEND_DEFAULT_RESUME2="swap:/dev/hda7" # CONFIG_SOFTWARE_SUSPEND_KEEP_IMAGE is not set # CONFIG_SOFTWARE_SUSPEND_DEBUG is not set # CONFIG_SOFTWARE_SUSPEND_CHECKSUMS is not set # CONFIG_APM_IGNORE_USER_SUSPEND is not set # CONFIG_USB_SUSPEND is not set And I even set it in "kopt" in "/boot/grub/menu.lst" as well (including running update-grub). Like software suspend 1 it doesn't seem to like swap partition. But to me it just seems perfectly fine. Well I put enough hours into this for now. I might try something else at a later time. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: whole lot of ACPI problems with self compiled kernel
Am Samstag, 22. Januar 2005 16:12 schrieb Lukas Ruf: > > Martin Steigerwald <[EMAIL PROTECTED]> [2005-01-22 13:27]: [...] > > I tried this (kernel 2.6.9 from kernel.org with stable software > > suspend patch for it) at all it gives me is: > > I am running 2.6.10-ac9 with swuswp2-2.1.5.14-for-2.6.10 with 1GB of > RAM and a swap partition of 1.6GB on an IBM T40p. > > Initially, I have had similar problems. They went away when I edited > /etc/sound/ Hello Lukas, what in there? I have only an "events" directory in there with configuration files for GNOME sound events (like what sample is to be played on a certian event). I cannot see how that relates to my problem. > > CONFIG_SOFTWARE_SUSPEND_DEFAULT_RESUME2="swap:/dev/hda7" > >^^ > is this your swap device? "/dev/hda7" is my swap device, but I just used the notation that is shown in the HOWTO at the software suspend 2 webpage at berlios. > I wrote in mine '/dev/hda5' simply. Ah, I got another hint, that it might work this way and the original software suspend implementation which made it into the 2.6 kernel uses "swap:/dev/hda5" instead. But this would oppose the documentation I found. Well I try it nevertheless as documentation might be wrong. > > Like software suspend 1 it doesn't seem to like swap partition. But > > to me it just seems perfectly fine. > > what is the size of your swap partition? Is there room for current > swap usage + memory image? 1003MB. My ThinkPad T23 has 384 MB RAM... usually swap space isnt used at all or not a lot at least: total used free sharedbuffers cached Mem:386124 306684 79440 0 4948 134548 -/+ buffers/cache: 167188 218936 Swap: 979924 2764 977160 > > Well I put enough hours into this for now. I might try something else > > at a later time. > > Don't give up! Well I won't, not yet. But I took a break. When something does not work I often want to run with my head through a wall (well I don't know whether you have that idiom in english but I hope you get what I mean) and then I realize that it is better to give it a break, ask for hints and such instead of doing it the hard way ;-) Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: whole lot of ACPI problems with self compiled kernel
Am Sonntag, 23. Januar 2005 22:53 schrieb Lukas Ruf: > > > Initially, I have had similar problems. They went away when I > > > edited /etc/sound/ > > /etc/hibernate > /etc/mkinitrd/scripts > > were thos I meant sorry, seems I had sound in my brain > flying'roundd Hello Lukas, yes, I found /etc/hibernate. But /etc/mkinitrd/scripts is an empty directory here. What should I add there? > > > Ah, I got another hint, that it might work this way and the original > > software suspend implementation which made it into the 2.6 kernel > > uses "swap:/dev/hda5" instead. But this would oppose the > > documentation I found. Well I try it nevertheless as documentation > > might be wrong. > > has it helped? I am currently compiling a 2.6.10 kernel with latest ACPI patches and software suspend 2 patches and I will try it with that one after it is ready (probably tomorrow). Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Suspend on T23: a small step forward
A small note: I am not sure whether I should continue this topic on this mailing list or just use the linux-thinkpad-ml in future. Well actually I am using a IBM ThinkPad T23 on Debian Sarge, so it might belong into both mailinglists. What do you think? Hello, I now used kernel 2.6.10, newest ACPI patch (acpi-20050114-2.6.10.diff) and a kernel configuration based on that from Michael Perry. I just added some stuff I needed or thought I may need (ISDN). And I made an initrd. I also applied the clock drift fix posted by Tino Keitel on thinkpad-ml (CONFIG_X86_PM_TIMER=y). Well I have one full and two partial successes. The full one: 1) Clock does not seems to drift anymore. I am not sure whether its by the clock drift fix posted by Tino Keitel or by using newest ACPI patches, but at least it works. I guess its due to Tino Keitel's solution cause I did not read anything regarding a clock drift fix in the ACPI changelog. The partial ones: 1) Suspend-to-disk works, well sort off: I again use "resume=/dev/hda7" parameter as "swap:/dev/hda7" does not work. This time it actually suspends to disk. Well and it even awakes from suspending and goes back into my KDE desktop. But mouse (and probably other USB device connected to the USB 1.1 on the ThinkPad") do not work anymore. This may be indicated by this message I get on wakeup: uhci_hcd :00:1d.1: Unloink after no-IRQ? Different ACPI or APIC settings may help usb 2-2: klaptop_acpi_he timed out on ep0in 2) When trying the first time, ALSA sound worked after resume. In this case I did not use ALSA sound before, but tried it after the resume. It worked then, which is more than I had before. Motivated this way I became adventorous and tried suspend to ram during music playback by JuK (KDE's jukebox;). This time ALSA didn't work after suspend. Fair enough, I thought I can stop the music playback before suspending and then tried the following: I played a MP3 by JuK, pressed stop, left the program open, did suspend-to-ram, and pressed play again. Well also in this case ALSA did not work after suspending. One thing that still doesn't work: 1) My AVM Fritzcard ISDN adapter is not operational after suspending. So in the whole I am not really that much farer. Now I am a bit stuck on what to try next. I think I try compiling a kernel without initrd. I hope someday that this all will become easier. In an ideal world, hardware vendors would provide all the information or even contributions to the Linux Kernel to make the whole ACPI stuff work like it is described on the ACPI stuff: Transparently without requiring that end users fix issues. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Unloading hisax PCMCIA driver fails
Hello, I tried unloading the ISDN drivers to make them survive a suspend. This is what I get: [EMAIL PROTECTED]:/home/martin -> /etc/init.d/isdn stop Shutting down isdnlog Shutting down ipppd Unloading ISDN modules FATAL: Module hisax is in use. [EMAIL PROTECTED]:/home/martin -> rmmod avma1_cs ERROR: Module avma1_cs is in use [EMAIL PROTECTED]:/home/martin -> rmmod hisax ERROR: Module hisax is in use by avma1_cs [EMAIL PROTECTED]:/home/martin -> rmmod isdn ERROR: Module isdn is in use by hisax (I am using RedHat's isdn config as Klaus Knopper provides it ported to Debian on the Knoppix CD as I wrote in my Linux User article in 6/2004) Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: compiling kernels newer than 2.6.8
Am Dienstag, 1. Februar 2005 21:56 schrieb Martin Hauser: > > > I don't think it's even true with 2.4. I'm sure I've used vanilla > > > kernels and, despite the warning, got a working initrd. "make-kpkg > > > --initrd..."-- > > > > Hard to believe because cramfs only works with debian patched > > kernelsources. Even the vanilla-kernel 2.6.10 does not support cramfs > > on initrd. > > Hmm... > I'm using Cramfs initrd on a vanilla 2.6.10 kernel to boot my debian > desktop PC... and it's using lvm which it required to have in the > initrd cause the root is on the lvm... (don't ask me why, i have not > yet figured ;) ) > > I'm pretty sure the initrd works.. cause without the kernel won't boot. Hallo Martin, I am also pretty sure, since I have SGI's XFS on my root filesystem and compiled XFS as kernel module for my initrd image with kernel 2.6.9 and 2.6.10 from kernel.org sources. Well and the machine booted from XFS with those kernels. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: /dev/thinkpad keeps disappearing
Am Montag, 31. Januar 2005 00:07 schrieb Pollywog: > This is what I had tried: > > #mkdir --mode=755 /dev/thinkpad > #mknod --mode=664 /dev/thinkpad/thinkpad c 10 170 > #chown root:thinkpad /dev/thinkpad/thinkpad > > This worked but the device would disappear after a reboot. Hello, its probably not recommended but I solved similar problems with udev by adding some stuff to the non existant file /etc/udev/links.conf For example for support of my AVM Fritzcard via ISDN 4 Linux: # ISDN M isdn0 c 45 0 M isdnctrl0 c 45 64 L isdnctrl /dev/isdnctrl0 M ippp0 c 45 128 M isdn1 c 45 1 M isdnctrl1 c 45 65 M ippp1 c 45 129 M isdn2 c 45 2 M isdnctrl2 c 45 66 M ippp2 c 45 130 M isdn3 c 45 3 M isdnctrl3 c 45 67 M ippp3 c 45 131 M isdn4 c 45 4 M isdnctrl4 c 45 68 M ippp4 c 45 132 M isdn5 c 45 5 M isdnctrl5 c 45 69 M ippp5 c 45 133 M isdn6 c 45 6 M isdnctrl6 c 45 70 M ippp6 c 45 134 M isdn7 c 45 7 M isdnctrl7 c 45 71 M ippp7 c 45 135 M isdninfo c 45 255 > > Alternatively you can add 'thinkpad' to /etc/modules. > > I had not tried this, I had added nvram to that file and solved that > problem. I will try adding thinkpad to that file also. If that works okay. I tried that with "hisax" and it didn't work. Also udev didn't remember manually added device nodes. So I used the approach I described above and this works fairly well. At work I used this approach to make udev VMware aware. Still the cleaner approach would be some nice udev rules I guess. But I already invested lots of time into fixing those issues with udev. So I will use that work-around for now. Regards, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: whole lot of ACPI problems with self compiled kernel
Am Dienstag, 8. Februar 2005 23:25 schrieb martin f krafft: > also sprach Martin Steigerwald <[EMAIL PROTECTED]> [2005.01.22.0901 +0100]: > > 2) suspend-to-disk doesn't work at all. I get > > > > "swsusp: FATAL: cannot find swap device, try swapon -a" > > FYI: > > http://marc.theaimsgroup.com/?l=linux-kernel&m=110789536921510&w=2 Hallo Martin, Danke für die Info. Hmmm, ich bilde mir ein, das auch mit einem Vanilla-Kernel probiert zu haben? Geht es bei Dir mit einem Vanilla-Kernel? Bist Du mittlerweile mit dem Thema weitergekommen? Das ist echt noch eine große Baustelle bei Linux. Grüße, -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de
