Re: [ARMedslack] libflashplayer.so??
> Greetings: I have armedslack13.27 running except for viewing YouTube > video.. (RasPi-B) > > Adobe site says no such plugin available.. Any suggestions or comments? > TIA.. (I tried a few but they must be for X86, not ARM??) Hi, this has been discussed here: http://lists.armedslack.org/pipermail/armedslack/2012-June/001276.html ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Slackware ARM on Raspberry Pi
> rich@rpi-17:~$ uname -a > Linux rpi-17 3.6.11+ #408 PREEMPT Wed Apr 10 20:33:39 BST 2013 armv6l > ARMv6-compatible processor rev 7 (v6l) BCM2708 GNU/Linux > > also see http://imageshack.us/a/img849/863/screenshot0316201309191.png Looks nice, but it is damn slow when used as desktop, isn't it? -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] A cheap Pogoplug on sale for 10 €
> It probably has a Oxnas 7820 chipset and it can run Linux from a disk > connected to S-ATA: > http://archlinuxarm.org/platforms/armv6/pogoplug-provideov3 > http://archlinuxarm.org/forum/viewtopic.php?f=29&t=2146 > > I've ordered one to install Slackware ARM ;-) I've received mine. Sadly I got a POGO-E02 with Marvell chipset (kirkwood). This actually is the same hardware like my Seagate Dockstar (or other Sheevaplugs), so I'd sell mine for 16 € + plus shipping from Germany. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] A cheap Pogoplug on sale for 10 €
Hi, maybe interesting for some people here: A Pogoplug is on sale for 10 € plus shipping: http://estore.rhiem.com/bestellen/warenkorb.asp?con=DE&artikelnr=POGO-P24&sid=2&wid=1&mid=32&store=&menge=0 It probably has a Oxnas 7820 chipset and it can run Linux from a disk connected to S-ATA: http://archlinuxarm.org/platforms/armv6/pogoplug-provideov3 http://archlinuxarm.org/forum/viewtopic.php?f=29&t=2146 I've ordered one to install Slackware ARM ;-) Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Initrd fails to mount rootfs
> The fact that it worked with 13.37 doesn't really mean a lot, to be > honest. 13.37 had a different version of busybox, a different version of > udev, a different kernel; most likely containted a different set of > modules. Yes, of course. Just wanted to make sure that everything else is ok and I already had it running with 13.37 some time ago. So I gave it a test run. > When I had a similar problem myself, I unpacked the kirkwood's initrd and > put in the classic "print" statements into the init script to find out > where it was broken. I've built my own kernel now (same version like 14.0 and based on your config). I've just added the needed drivers to the kernel (USB, SCSI disk, some filesystems) so I can boot without initrd and everything is ok now. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Initrd fails to mount rootfs
> But while booting the initrd system fails to mount the rootfs. You can find a > complete log here: > http://pastebin.com/XLyCFiGV > > IMHO the relevant part: > > mount: mounting /dev/sda3 on /mnt failed: Invalid argument > > ERROR: No /sbin/init found on rootdev (or not mounted). Trouble ahead. > > You can try to fix it. Type 'exit' when things are done. > > > > /bin/sh: can't access tty; job control turned off Same configuration with kernel and uinitrd from 13.37 (Linux 2.6.38.4-kirkwood) boots fine. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Initrd fails to mount rootfs
> You have four partitions not 3. What is on /dev/sda4? Yes, I've already tested with a second USB stick which has a sda4 (ext4) prepared for home. The first USB stick only has the mentioned 3 partitions. Sorry. > Is your /etc/fstab setup? Should be OK: http://pastebin.com/7n0YmHuz > You have a root shell there - can you share what you have done to > investigate it? It may be obvious what's happened. Tried to mount the rootfs, which works. Tried to run ./init again, which then fails at switch_root But after "exit" it starts booting: http://pastebin.com/Skj1JF4B -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Initrd fails to mount rootfs
Hi, I've reactivated my Seagate Dockstar, that was lying around in my cupboard for some month. I did the following: 1. reinstalled Doozans U-Boot to have a clean system to start 2. prepared an USB stick with sda1=boot, sda2=swap, sda3=rootfs 3. untared the Slackware ARM minirootfs to the correct places on the USB stick (slack-14.0-miniroot_27Sep12.tar.xz) 4. changed some U-Boot settings to match my USB stick (rootfs=sda3, ext4) But while booting the initrd system fails to mount the rootfs. You can find a complete log here: http://pastebin.com/XLyCFiGV IMHO the relevant part: > mount: mounting /dev/sda3 on /mnt failed: Invalid argument > ERROR: No /sbin/init found on rootdev (or not mounted). Trouble ahead. > You can try to fix it. Type 'exit' when things are done. > > /bin/sh: can't access tty; job control turned off You can see in the log that the partitions already appear about 5 seconds before the mount attempt and actually it is possible to mount the partition by hand: > / # mount -o ro -t ext4 /dev/sda3 /mnt > [ 755.989044] EXT4-fs (sda3): mounted filesystem with ordered data mode. > Opts: (null) Any ideas? ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] QEMU support - does anybody care about it?
I've never used QEMU and I think nowadays ARM hardware is really cheap to get. So even for a first try beginners may rather buy an ARM board than using QEMU. 2012/8/31 Stuart Winter : > > Hi > > I am thinking about removing the QEMU packages from -current. > I only added these years ago because my real ARM hardware died and I > needed a stop gap. > > The thing is that QEMU is *so slow* that I cannot imagine it being useful > at all, apart from a 10 minute novelty for x86 users. > It takes time to test whether the installation works, and takes time to > build the Versatile kernel. > > Is anybody going to miss it if I drop the packages and install docs? > > -- > Stuart Winter > www.slackware.com/~mozes > Slackware for ARM: www.armedslack.org > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] libflashplayer
Am Wed, 4 Jul 2012 14:51:29 +0100 (BST) schrieb Davide : > why the hell should one use flash for authentication on a web site > (on some sites if you don't have flash you cannot even see the box > for typing the authentication) ! I've never seen that and actually it's kind of funny nowadays. I guess the webmaster does not want to have many users? :-D The mobile phone/tablet market is growing every day and neither iOS nor recent Android is meant to run flash. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] libflashplayer
Am Wed, 4 Jul 2012 09:03:22 -0600 schrieb Old Man : > Chrome on ARM is just an android package: > https://www.google.com/intl/en/chrome/browser/mobile/android.html#utm_campaign=en&utm_source=en-ha-na-us-bk&utm_medium=ha No, it should be possible to build "real" chromium on ARMed Slack and run it in X. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] libflashplayer
2012/7/4 Ottavio Caruso : > On 2 July 2012 01:55, stanley garvey wrote: > >>> It is a can of worms. It is closed source and lets hope it's finished. >>> Adobe don't fully support it and have publicly said so. > > Yes, Adobe has officially end-of-lifed Flash Player on Android: > > http://blogs.adobe.com/flashplayer/2012/06/flash-player-and-android-update.html Not only on Android, but on Linux in general: http://tech.slashdot.org/story/12/02/22/1323204/adobe-makes-flash-on-gnulinux-chrome-only Adobe Flash will only be available as part of Google Chrome/Chromium. Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] libflashplayer
Am Wed, 27 Jun 2012 19:10:00 + schrieb Ottavio Caruso : > On 27 June 2012 18:56, Thorsten Mühlfelder wrote: > > Various Android Browser are shipping a Flash plugin. AFAIK the > > latest Firefox for Android (version 14.x) has one, too. Maybe you > > can find something in their source/app/ML/chat... :-/ > > > I strongly doubt that at the moment you can take random pieces from > the Android userspace and just transplant it into any Linux distro, > but I'd be pleased to be proven wrong. > Well, that's right. Precompiled probably won't work. No idea what C library Android uses. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] libflashplayer
Am Wed, 27 Jun 2012 19:10:53 +0100 schrieb Ottavio Caruso : > On 27 June 2012 15:34, Davide wrote: > > Has anyone got a working version of libflashplayer.so for > > armedslack ? > > There's no official flash player for Linux on ARM. You can try to > google around but you would probably find either pirated of obsolete > versions. Various Android Browser are shipping a Flash plugin. AFAIK the latest Firefox for Android (version 14.x) has one, too. Maybe you can find something in their source/app/ML/chat... :-/ -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Slackware ARM current is being re-based to armv5te
Am Fri, 16 Dec 2011 05:16:40 -0500 schrieb John O'Donnell : > Well you obviously wont be processing anything like that on these > plugs without an FPU, Well, AFAIK mp3, flac and tremor is all integer processing. Only vorbis requires FPU. And a dockstar is a nice mpd server ;) Basically this was just meant as example, but I'm quite sure changing from armv4te to armv5te will be far is far away from changing from march=i486 to march=k8 on my x86 system. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Slackware ARM current is being re-based to armv5te
Am Fri, 16 Dec 2011 00:10:31 +0100 schrieb "John O'Donnell" : > My web site is down at the moment but I recompiled the core libraries > and a few of the packages and ran cpu tests on sheevas, guruplug, and > a seagate dockstar and posted the results with stock armedslack and > the arm5te compiled version with no differences. What kind of tests? On x86 platform I could measure slight differences on audio/video decoding/encoding. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Which packages to install for gcc/make/etc ?
Am Sat, 8 Oct 2011 15:06:49 +0100 (BST) schrieb Stuart Winter : > ARMedslack's installpkg differs from x86's because ARMedslack's is > patched for several speed improvements on ARM. Take a look at spkg. It's so much faster than the original package tools ;) -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Which packages to install for gcc/make/etc ?
Am Wed, 5 Oct 2011 00:40:58 +0100 schrieb Ottavio : > On 5 October 2011 00:34, Rich wrote: > > Ah... yes it does if you use the Salix repo, which is a mirror of > > the ARMedSlack repository with dependency checking. The Salix ARM > > slapt-get package available already has the repo setup in the > > slapt-getrc config file. > > Dependency checking! Horror! Sacrilege! > Geek :-P -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Flash Plugin for ARM
Please note the info that TI provides, especially point 1+2: IMPORTANT: * This Evaluation/Demo code is provided ONLY for users developing products with Texas Instruments Cortex(tm)-A8 based microprocessors (i.e.: Sitara(tm) and DaVinci(tm) platforms). * Request for access is approved on a case-by-case basis to end equipment manufacturers/customers. * This code is NOT intended for production use. If you intend to ship this software with your product, please contact Adobe or Adobe partners directly for any necessary qualification and licensing. * For more details, please read carefully the TI license terms and conditions provided in the installer. 2011/9/6 Davide : > While looking for Flash plugin for ARM linux I found this: > http://www.ti.com/tool/adobeflash-a8 > > Not sure if there is an easier way to get Flash plugin on ARM linux. > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack > ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Logitech Squeezebox Touch
Hi, I wonder if there is any information how to change the OS on a Logitech Squeezebox Touch? According to this blog it's an ARM based device running Linux: http://blog.sarine.nl/2010/09/30/squishybox-a-better-squeezebox-touch-experience/ https://github.com/DaveDavenport/squishybox/wiki/SqueezeBox-Touch Greetings Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Got a toshiba AC100
Take a look here: http://people.canonical.com/~ogra/tegra/2.6.37/ and https://launchpad.net/~ac100/+archive/ppa Perhaps you'll find a working kernel there. Otherwise you should find newer kernel sources there and the config of the 2.6.29 kernel here: http://ac100.gudinna.com/kernel/config.txt You could use Codesourcery's toolchain to cross compile this kernel then. Make sure you are using the EABI toolchain. You also could compile the kernel on a Dockstar or some other ARMed Slack system. Am Friday 15 July 2011 11:23:06 schrieb Davide: > To reduce possible faliures to a minimum I did the same with clashNG. > Maybe I did not look well enough the first time but this time I got for > sure another intresting message before init killing itself: Kernel too old > > Don't think a picture is necessary for that. > > Now I'm not 100% sure that I also had this on slackware 13.1 miniroot but > since I use slackware 13.1 to build clashNG I'm strongly suspecting that I > had the same message but did not notice it since it wan not on the last > line. > > Damn > > Anyone know what boot loader is on the AC100 ... maybe it's possible to > have more then one kernel lingering around ? Anyone know how to build > kernel for the ac100 that will run armedslack ? > > Regards > David > > --- Ven 15/7/11, Thorsten Mühlfelder ha scritto: > > Da: Thorsten Mühlfelder > > Oggetto: Re: [ARMedslack] Got a toshiba AC100 > > A: "Slackware ARM port" > > Data: Venerdì 15 luglio 2011, 10:14 > > Am Friday 15 July 2011 09:21:53 > > > > schrieb Davide: > > > Ok I followed instructions from http://ac100.gudinna.com/README/ to get > > > ubuntoo 10.10 working off the SD. > > > > > > I then mixed up an armedslack miniroot 13.1 (any > > > > higher would hit the udev > > > > > kernel version requirement issue) with the kernel > > > > modules from ubuntu 10.10 > > > > > for ac100 in an attempt to get a base system booting > > > > off the SD. > > > > > All looked right while starting to boot until I hit a > > > > kernel panic: > > > attempting to kill init > > > > Could you make some photos of the interesting error > > messages? If you do, don't > > attach them to the mail. Instead upload them to an image > > hoster. > > > > > > -- > > Thorsten Mühlfelder > > Salix OS: www.salixos.org > > _______ > > ARMedslack mailing list > > ARMedslack@lists.armedslack.org > > http://lists.armedslack.org/mailman/listinfo/armedslack > > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Got a toshiba AC100
Am Friday 15 July 2011 09:21:53 schrieb Davide: > Ok I followed instructions from http://ac100.gudinna.com/README/ to get > ubuntoo 10.10 working off the SD. > > I then mixed up an armedslack miniroot 13.1 (any higher would hit the udev > kernel version requirement issue) with the kernel modules from ubuntu 10.10 > for ac100 in an attempt to get a base system booting off the SD. > > All looked right while starting to boot until I hit a kernel panic: > attempting to kill init Could you make some photos of the interesting error messages? If you do, don't attach them to the mail. Instead upload them to an image hoster. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Got a toshiba AC100
Am Thu, 14 Jul 2011 19:07:31 +0100 (BST) schrieb Davide : > I'm worried about binary compatibility as "armedslack > binaries will run on almost any arm platform" but that does not mean > they will run on every arm platform. They are built for armv4 and the Tegra should be at least armv6 so they should run without problems. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Got a toshiba AC100
Am Thursday 14 July 2011 17:29:59 schrieb Dave Dowell: > Here's a page which covers the bootloader > http://tosh-ac100.wetpaint.com/page/Ubuntu And there also is a kernel available and a rootfs (which probably can be exchanged with the ARMed Slack one): http://ac100.gudinna.com/ > > On 14/07/2011 16:15, Davide wrote: > > Hum well I can do with the ordinary fluxbox as long as the touchpad > > works. I could even accept to run on it a framebuffer kdrive as xserver > > as long as I can have an ordinary linux distribution on it (possibly > > slackware or slackware based). > > > > The real questions are: > > do I haveto stick with the original bootloader whatever it is or has > > someone ported uboot or something else to it or do I haveto work with > > kexecboot ? > > > > do I haveto stick with the origina kernel to get all the onboard devices > > working or can a vanilla kernel be configured to handle all the devices > > on this thing ? > > > > will the armedslack binaries run on the ac100 or do they need to be > > compiled specially for this specific arm processor ? > > > > Regards > > David > > > > --- Gio 14/7/11, Manfred Müller ha scritto: > >> Da: Manfred Müller > >> Oggetto: Re: [ARMedslack] Got a toshiba AC100 > >> A: "Slackware ARM port" > >> Data: Giovedì 14 luglio 2011, 16:42 > >> Hi list, > >> > >>> You can search the ML archive for this message and ask > >> > >> Manfred: > >> > >> although I still own this littele thing I gave up on > >> porting ARMedslack to it > >> because I have to finish the tablet version of FluxFlux > >> first... > >> > >> @Thorsten: > >> > >> ot: was it really needed to reinvent the wheel with Salt > >> instead of improving the legacy LLS? > >> > >> Regards, Manfred > >> ___ > >> ARMedslack mailing list > >> ARMedslack@lists.armedslack.org > >> http://lists.armedslack.org/mailman/listinfo/armedslack > > > > ___ > > ARMedslack mailing list > > ARMedslack@lists.armedslack.org > > http://lists.armedslack.org/mailman/listinfo/armedslack > > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Got a toshiba AC100
Am Thursday 14 July 2011 16:42:16 schrieb Manfred Müller: > @Thorsten: > > ot: was it really needed to reinvent the wheel with Salt instead of > improving the legacy LLS? I'm not a Linux Live user, so you have to ask this question on our ML ;-) -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Got a toshiba AC100
Am Thursday 14 July 2011 15:06:01 schrieb Davide: > My last day at the old work the work mate gave me a Toshiba AC100 as a > present (one of those nvidia tegra dual core arm based netbooks with > android): how do I put armedslack in this thing ? You can search the ML archive for this message and ask Manfred: > > I wonder If anyone has tried ARMed Slack on a Toshiba AC100 netbook? > > Hi Thorsten, > > at the very moment I've started porting FluxFlux to ARMed Slack using the > ac100 as dev machine. > > I'm at a very early stage, but I think this netbook is a very good start > to be prepared for the release of tegra 4. > > Regards, > > Manfred aka Quax -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] File system on my Freeagent Dockstar got corrupted
Am Fri, 6 May 2011 07:59:14 +0100 (BST) schrieb Davide : > This is probably not the cause of your problems but it's something > you want to keep in mind anyway: Running non flash oriented journaled > file-systems on flash based devices can ware them out very quickly No that's not the problem because this is a Seagate Freeagent Go hard disk. > especially if you are using MLB flash based device with a typical > 1 cycle rewrite capability. To minimize damage on flash devices > at least make sure you mount the file-system with noatime option to > avoid rewriting an entire erase block just because something accessed > a file. The error message indicates that the USB part of the hard disk is broken. Unfortunately I cannot open the USB case without damaging a warranty seal, so I cannot save my data. I have to find a Windows PC now to run the Seagate test tools. This is mandatory for a warranty request. > --- Ven 6/5/11, Thorsten Mühlfelder ha scritto: > > > Da: Thorsten Mühlfelder > > Oggetto: Re: [ARMedslack] File system on my Freeagent Dockstar got > > corrupted A: armedslack@lists.armedslack.org > > Data: Venerdì 6 maggio 2011, 01:29 > > Am Fri, 18 Feb 2011 09:20:44 + > > (GMT) > > schrieb Stuart Winter : > > > > > The problem could be anywhere -- perhaps take out the > > disc from the > > > usb caddy and put it onto a bare IDE interface and try > > SMART? > > > The disc or USB caddy could be faulty. > > > > I'm quite sure now that the USB disk is faulty. After > > running for some > > month now a similiar error occured: > > Dockstar was still running, the DSL connection was up and > > it did route > > to the internet. But all daemons apparently have crashed: > > no ssh login > > anymore, no dhcpd service and http server crashed, too. > > > > When I plug the disk to my PC or notebook I'm getting these > > messages > > in /var/log/messages in an endless loop: > > ay 5 20:54:33 pinkfloyd kernel: [ 440.520511] > > sd 15:0:0:0: [sdd] Attached SCSI disk > > May 5 20:54:36 pinkfloyd kernel: [ 442.989369] > > sd 15:0:0:0: [sdd] Unhandled sense code > > May 5 20:54:36 pinkfloyd kernel: [ 442.989373] > > sd 15:0:0:0: [sdd] Result: hostbyte=0x00 driverbyte=0x08 > > May 5 20:54:36 pinkfloyd kernel: [ 442.989376] > > sd 15:0:0:0: [sdd] Sense Key : 0x3 [current] > > May 5 20:54:36 pinkfloyd kernel: [ 442.989379] > > sd 15:0:0:0: [sdd] ASC=0x11 ASCQ=0x0 > > May 5 20:54:36 pinkfloyd kernel: [ 442.989381] > > sd 15:0:0:0: [sdd] CDB: cdb[0]=0x28: 28 00 00 41 73 4c 00 00 > > 07 00 > > May 5 20:54:39 pinkfloyd kernel: [ 445.794887] > > sd 15:0:0:0: [sdd] Unhandled sense code > > May 5 20:54:39 pinkfloyd kernel: [ 445.794890] > > sd 15:0:0:0: [sdd] Result: hostbyte=0x00 driverbyte=0x08 > > May 5 20:54:39 pinkfloyd kernel: [ 445.794893] > > sd 15:0:0:0: [sdd] Sense Key : 0x3 [current] > > May 5 20:54:39 pinkfloyd kernel: [ 445.794896] > > sd 15:0:0:0: [sdd] ASC=0x11 ASCQ=0x0 > > May 5 20:54:39 pinkfloyd kernel: [ 445.794898] > > sd 15:0:0:0: [sdd] CDB: cdb[0]=0x28: 28 00 00 41 73 4c 00 00 > > 07 00 > > > > > > -- > > Thorsten Mühlfelder > > Salix OS: www.salixos.org > > ___ > > ARMedslack mailing list > > ARMedslack@lists.armedslack.org > > http://lists.armedslack.org/mailman/listinfo/armedslack > > > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] File system on my Freeagent Dockstar got corrupted
Am Fri, 18 Feb 2011 09:20:44 + (GMT) schrieb Stuart Winter : > The problem could be anywhere -- perhaps take out the disc from the > usb caddy and put it onto a bare IDE interface and try SMART? > The disc or USB caddy could be faulty. I'm quite sure now that the USB disk is faulty. After running for some month now a similiar error occured: Dockstar was still running, the DSL connection was up and it did route to the internet. But all daemons apparently have crashed: no ssh login anymore, no dhcpd service and http server crashed, too. When I plug the disk to my PC or notebook I'm getting these messages in /var/log/messages in an endless loop: ay 5 20:54:33 pinkfloyd kernel: [ 440.520511] sd 15:0:0:0: [sdd] Attached SCSI disk May 5 20:54:36 pinkfloyd kernel: [ 442.989369] sd 15:0:0:0: [sdd] Unhandled sense code May 5 20:54:36 pinkfloyd kernel: [ 442.989373] sd 15:0:0:0: [sdd] Result: hostbyte=0x00 driverbyte=0x08 May 5 20:54:36 pinkfloyd kernel: [ 442.989376] sd 15:0:0:0: [sdd] Sense Key : 0x3 [current] May 5 20:54:36 pinkfloyd kernel: [ 442.989379] sd 15:0:0:0: [sdd] ASC=0x11 ASCQ=0x0 May 5 20:54:36 pinkfloyd kernel: [ 442.989381] sd 15:0:0:0: [sdd] CDB: cdb[0]=0x28: 28 00 00 41 73 4c 00 00 07 00 May 5 20:54:39 pinkfloyd kernel: [ 445.794887] sd 15:0:0:0: [sdd] Unhandled sense code May 5 20:54:39 pinkfloyd kernel: [ 445.794890] sd 15:0:0:0: [sdd] Result: hostbyte=0x00 driverbyte=0x08 May 5 20:54:39 pinkfloyd kernel: [ 445.794893] sd 15:0:0:0: [sdd] Sense Key : 0x3 [current] May 5 20:54:39 pinkfloyd kernel: [ 445.794896] sd 15:0:0:0: [sdd] ASC=0x11 ASCQ=0x0 May 5 20:54:39 pinkfloyd kernel: [ 445.794898] sd 15:0:0:0: [sdd] CDB: cdb[0]=0x28: 28 00 00 41 73 4c 00 00 07 00 -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Improving USB audio sound quality on the Dockstar
Am Thursday 05 May 2011 10:32:52 schrieb Thorsten Mühlfelder: > Hi, > > I'm using a high quality USB audio device (Burr-Brown Japan PCM2702) with > my Dockstar. I've configured mpd to use the hardware interface without dmix > because dmix seems to be broken on the ARM platform. > > But there is one problem left: sometimes clicks and pops can be heard. They > appear randomly up to 2 times a minute. My impression is that hard disk > data transfer causes them, because using "hdparm -t" makes the situation > worse. On the Dockstar all USB devices are connected via hub to the same > USB controller: > # lsusb > Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub > Bus 001 Device 002: ID 05e3:0608 Genesys Logic, Inc. USB-2.0 4-Port HUB > Bus 001 Device 003: ID 0bc2:2120 Seagate RSS LLC > Bus 001 Device 005: ID 08bb:2702 Texas Instruments Japan Speakers > > Is it possible to give the USB audio data transfers a higher priority? Another idea is that network traffic may cause this because eth0 has higher priority IRQ than USB: thorsten@dreamtheater:~$ cat /proc/interrupts CPU0 1: 68571661 orion_irq orion_tick 5: 2 orion_irq mv_xor.0 6: 2 orion_irq mv_xor.1 7: 2 orion_irq mv_xor.2 8: 2 orion_irq mv_xor.3 11: 94038944 orion_irq eth0 19: 37108202 orion_irq ehci_hcd:usb1 22: 0 orion_irq mv_crypto 33:290 orion_irq serial 46: 25 orion_irq mv643xx_eth Err: 0 -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Improving USB audio sound quality on the Dockstar
Hi, I'm using a high quality USB audio device (Burr-Brown Japan PCM2702) with my Dockstar. I've configured mpd to use the hardware interface without dmix because dmix seems to be broken on the ARM platform. But there is one problem left: sometimes clicks and pops can be heard. They appear randomly up to 2 times a minute. My impression is that hard disk data transfer causes them, because using "hdparm -t" makes the situation worse. On the Dockstar all USB devices are connected via hub to the same USB controller: # lsusb Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 001 Device 002: ID 05e3:0608 Genesys Logic, Inc. USB-2.0 4-Port HUB Bus 001 Device 003: ID 0bc2:2120 Seagate RSS LLC Bus 001 Device 005: ID 08bb:2702 Texas Instruments Japan Speakers Is it possible to give the USB audio data transfers a higher priority? -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Slackware ARM on Small NAS (NS-K330)
Am Fri, 22 Apr 2011 09:32:33 +0100 (BST) schrieb Davide : > > I was told that reply made a mess by some other user. What I've meant was: Don't use the reply button for new threads and change the subject only because this generates a mess in the thread view ;-) > I'll fix it manually just tell me what fields I should leave and > where. This is what my client would normaly do as a reply: > "Re: [ARMedslack] Slackware ARM on Small NAS (NS-K330)" > > What's good in that and what's bad in that ? That's fine. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] jeff Doozan's uboot
Am Thu, 21 Apr 2011 11:12:44 +0100 (BST) schrieb Davide : > This is a mix ov a few I built myself and some gotten from current. > This is waht I'll be working with and should fit in a compressed > jffs2 image 64Mb big. root@slackware:/usr/src/surap_packages# du -ms > * | sort -n 1 busybox-1.18.4-arm-1.tgz > 1 dropbear-0.53.1-arm-1.tgz > 1 hostapd-0.7.3-arm-1.tgz > 1 iptables-1.4.10-arm-1.tgz > 1 iw-0.9.20-arm-1.tgz > 1 ppp-2.4.5-arm-1.tgz > 1 udev-165-arm-2.tgz > 1 usb_modeswitch-1.1.6-arm-1.tgz > 1 wireless-tools-29-arm-2.tgz > 2 httpd-2.2.17-arm-2.tgz > 2 kernel-firmware-2.6.38.3-noarch-1.tgz > 5 glibc-solibs-2.13-arm-1.tgz > 8 kernel_kirkwood-2.6.38.3-arm-1.tgz > 10 php-5.3.5-arm-1.tgz > 15 kernel-modules-kirkwood-2.6.38.3_kirkwood-arm-1.tgz > root@slackware:/usr/src/surap_packages# du -ms . > 43 . > root@slackware:/usr/src/surap_packages# > > > Since booting from jffs2 image does not require initrd ... and maybe > one can do without documentation I'll see if I can fit that in a > 32Mb image. Build a custom kernel with few modules ;) -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Kernel kirkwood-2.6.38.2: CONFIG_N_HDLC
Am Sat, 9 Apr 2011 15:33:08 +0200 schrieb Thorsten Mühlfelder : > Am Thu, 7 Apr 2011 19:39:16 +0100 (BST) > schrieb Stuart Winter : > > > I'll add the kernel option before the 13.37 release - either as a > > kernel rebuild, or a kernel update - which ever comes first. > > Thanks! > Just noticed: pppoe uses about 11 % CPU with downstram fully loaded > (16 MBit/s). > With synchronous mode pppoe only uses 5 % CPU on my dockstar. Sadly the connection is not really stable. The download rate fluctuates and opening web pages in the browser is slow. In the syslog are some errors: Apr 21 10:57:22 dreamtheater pppd[2911]: Protocol-Reject for unsupported protocol 0xfc7a Apr 21 10:57:27 dreamtheater pppd[2911]: Protocol-Reject for unsupported protocol 0xdb Apr 21 10:58:43 dreamtheater pppoe[2912]: read (syncReadFromPPP): Session 649: Input/output error So I'm back to asynchronous mode now. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] jeff Doozan's uboot
Am Wednesday 20 April 2011 12:01:10 schrieb Davide: > Now that I've a developement environment I'll be replicating what I did > with miniroot to get ap/3G/nas/router on a minimal busybox based > environment. So you have a Slackware system with busybox instead of the GNU tools? How much space saves this? > As an alternatibe place to have this would it be a crazy > thing to have a microroot armedslack ? The bigges NAND partition on the Dockstar is about 220 MB and could be used by a rescue/fallback root partition if no USB drive is connected. UBIFS would be a good choice for it. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] jeff Doozan's uboot
Am Tuesday 19 April 2011 18:37:20 schrieb richard.lapoi...@gmail.com: > Yeah, I got fooled by this. Would had replied but you had it solve before > I finished reading the thread. BTW, I have three Dockstars all with their > own Seagate Freeagent Go drives running ARMedSlack 13.37RC1 and booting > from Doozan's uboot. I also use his Rescue System. What rescue system did you use? It seems my original pogoplug system (that one the dockstar is deliverd with) does not boot anymore. So some other fallback system would be nice. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] jeff Doozan's uboot
Am Tuesday 19 April 2011 15:18:26 schrieb Davide: > The last cpuinfo i sent was produced by pogoplug system ... from slackware > it looks different but still is not what one would expect from a dockstar: > root@slackware:~# cat /proc/cpuinfo > Processor : Feroceon 88FR131 rev 1 (v5l) > BogoMIPS: 1192.75 > Features: swp half thumb fastmult edsp > CPU implementer : 0x56 > CPU architecture: 5TE > CPU variant : 0x2 > CPU part: 0x131 > CPU revision: 1 > > Hardware: Marvell SheevaPlug Reference Board > Revision: > Serial : > root@slackware:~# > > The board actually looks the same as my other dockstar ... I guess it's > just a different SOC that make sthe difference. As I've said: you are using a wrong "arcNumber". Doozan's u-boot is patched so that the arcNumber can be set as u-boot environment variable. Basically arcNumber is just some term used by Doozan. In the Linux kernel it is called machine type. Have a look at /usr/src/linux/include/generated/mach-types.h for a complete list. There you'll find: #define MACH_TYPE_DOCKSTAR 2998 #define MACH_TYPE_SHEEVAPLUG 2097 If you want your kernel to recognize the dockstar you have to use 2998 as arcNumber in u-boot. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] jeff Doozan's uboot
Another thing: Please don't use the "reply" button to start a new thread. Use "new mail" and add the ML address. Otherwise the thread view in mail clients is wrong. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] jeff Doozan's uboot
Am Tuesday 19 April 2011 14:14:17 schrieb Davide: > Well yeah I've an update on this: > I did nothing but change the usb port from closest to RJ45 to the one nets > to the closest to RJ45. This time it boots. > > I'll investigate further on this ... AFAIK the dockstar has native USB ports and some ports provided by an additional hub. Maybe this makes a difference. > This is on a brand new dockstar (I've now 2 of these things) but the seem > to have small hardware differences ... read on > > Also Another thing I noticed: the led stopped working and things do not > look rigth in /sys Do cat /proc/cpuinfo. Do you have "Hardware: Seagate FreeAgent DockStar"? If not the "arcNumber" that gives u-boot to the kernel is wrong. IIRC arcNumber 2998 is the right one for the Dockstar. 2097 is Sheevaplug. Please search for it in Doozans forum. The arcNumber sets some board specific details like LED setup. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] jeff Doozan's uboot
Am Tuesday 19 April 2011 12:42:02 schrieb Davide: > Is there something I need to do in order to boot from a usb HDU that > differs from booting from a usb flash drive ? My Dockstar boots fine with Doozan's u-boot and the original Seagate Freeagent Go HD. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Kernel kirkwood-2.6.38.2: CONFIG_N_HDLC
Am Thu, 7 Apr 2011 19:39:16 +0100 (BST) schrieb Stuart Winter : > I'll add the kernel option before the 13.37 release - either as a > kernel rebuild, or a kernel update - which ever comes first. Thanks! Just noticed: pppoe uses about 11 % CPU with downstram fully loaded (16 MBit/s). -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Kernel kirkwood-2.6.38.2: CONFIG_N_HDLC
Hi, is there any special reason why CONFIG_N_HDLC is disabled? In x86 it's build as a module. The only reason I'm asking is because the rp-pppoe documentation says: "Synchronous PPP is safe on Linux machines with the n_hdlc line discipline. (If you have a file called "n_hdlc.o" in your modules directory, you have the line discipline.)" It also says that synchronous mode "is much easier on CPU usage" than asynchronous mode. So I've thought it's worth a try. Location: -> Device Drivers -> Character devices -> Non-standard serial port support (SERIAL_NONSTANDARD [=y]) -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] gdb python error
Am Thursday 07 April 2011 09:30:10 schrieb Thorsten Mühlfelder: > gdb shows this error on arm current: > > (gdb) run > Starting program: /usr/bin/mpd -h > [Thread debugging using libthread_db enabled] > Traceback (most recent call last): > File > "/usr/share/gdb/auto-load/usr/lib/libgobject-2.0.so.0.2800.4-gdb.py", line > 9, in > from gobject import register > File "/usr/share/glib-2.0/gdb/gobject.py", line 3, in > import gdb.backtrace > ImportError: No module named backtrace > > It works on x86_64 current. So either I'm missing a dep (but cannot find > the right one yet) or some other problem. Maybe different gdb configure > options? I have to correct myself: the x86_64 version shows the same error. On both arch it only comes up if the debugged program exits abnormally. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] gdb python error
gdb shows this error on arm current: (gdb) run Starting program: /usr/bin/mpd -h [Thread debugging using libthread_db enabled] Traceback (most recent call last): File "/usr/share/gdb/auto-load/usr/lib/libgobject-2.0.so.0.2800.4-gdb.py", line 9, in from gobject import register File "/usr/share/glib-2.0/gdb/gobject.py", line 3, in import gdb.backtrace ImportError: No module named backtrace It works on x86_64 current. So either I'm missing a dep (but cannot find the right one yet) or some other problem. Maybe different gdb configure options? -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] ARMed Slack repository with dependency support
Hi, there is a ARMed Slack repository with dependency support available at our Salix OS server: http://salix.enialis.net/arm/armedslack-current/ You can use it with a package manager like slapt-get. slapt-get and it's dependencies are available here: http://salix.enialis.net/arm/current/slapt-get/ Just install all of them and you are ready to go: slapt-get --add-keys slapt-get -u slapt-get -i whatever-you-want PS: this slapt-get version is patched to use the faster spkg package tool instead of pkgtools. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Thursday 17 March 2011 17:48:09 schrieb Stuart Winter: > On Thu, 17 Mar 2011, Davide wrote: > > I downloaded miniroot current and did the exact same and it booted !!! > > I guess there is something wrong with 13.1's kirkwood kernel image. > > NOthing wrong with 13.1's kirkwood kernel - works fine on the SheevaPlug, > Guru and OpenRD client. I don't know what the support was like for the > Dockstar for that kernel though (or if it was even in there at all!) I can confirm that I had the 13.1 kernel running on my dockstar without problems. > > The only thing I noticed wrong while system came up is: > > [ 25.669649] uncorrectable error : > > [ 25.672905] end_request: I/O error, dev mtdblock0, sector 1920 > > [ 25.678944] Buffer I/O error on device mtdblock0, logical block 240 > > [ 25.839517] uncorrectable error : > > [ 25.845664] end_request: I/O error, dev mtdblock0, sector 1920 > > [ 25.851715] Buffer I/O error on device mtdblock0, logical block 240 > > [ 25.860583] uncorrectable error : > > [ 25.863897] uncorrectable error : > > [ 25.867630] end_request: I/O error, dev mtdblock0, sector 0 > > [ 25.873419] Buffer I/O error on device mtdblock0, logical block 0 > > [ 25.879845] uncorrectable error : > > [ 25.883085] uncorrectable error : > > [ 25.886731] end_request: I/O error, dev mtdblock0, sector 8 > > > > This may be because mtdparts may be incorrect since I'm using openwrt > > second stage u-boot that may have passed wrong flash partitions to > > kernel. Hope it did not make a mess of the content. > > http://www.spinics.net/lists/arm-kernel/msg62881.html > > AFAIK it's nothing to worry about, but read up on it by following the info > in there. I have the same errors and I've assumed that they are bad block related. The link confirmed that in some way (OOB holds bad block information and some other data as well). But I've wondered why the mtd partitions are touched at bootup, even they are not mounted. Was this issue introduced with new kernels? On another device I'm using a 2.6.24 and 2.6.29 kernel and I'm writing u-boot and u-boot-env from within Linux to the NAND, but u-boot does not complain about anything (as noted in your link). -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Thursday 17 March 2011 10:17:45 schrieb Davide: > x_bootargs=console=ttyS0,115200 Seems to be OK. > x_bootcmd_kernel=fatload usb 0:1 0x640 uImage Where are you doing "usb start"? > x_bootargs_root=root=/dev/sda1 ro rootfstype=ext2 That probably should be /dev/sda2 because sda1 is your /boot partition. You have to specify this in fstab, too. > One other thing: how do I get uboot to tell kernel to load initrd ? I guess you can try like this: fatload usb 0:1 0x80 uImage fatload usb 0:1 0x110 uInitrd bootm 0x80 0x110 Which means: Load kernel to RAM address 0x80, load initrd to 0x110 and then boot both. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Wednesday 16 March 2011 12:22:18 schrieb Davide: > I had a look at jeff doozan's enhanced uboot installer script. > I look like it replaces the original uboot on mtd0, is that correct ? > > While openwrt is using original uboot to start another version so that if > you screw up you still have the original one to get things working again > without need for jtag. > Just noticed: I've used the old method from that page, that replaces /dev/mtd3 and chainloads the bootloader. But as Doozans bootloaders should be well tested now installation on mtd0 shouldn't be a problem. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Wednesday 16 March 2011 13:45:53 schrieb Davide: > > Yes, that's possible. AFAIK you just have to place the > > second u-boot to the > > NAND address where originally the kernel can be found. If > > it fails to load > > the second u-boot the first u-boot must be able to boot > > from usb to recover > > things. Otherwise without JTAG you are stuck, too. > > Not really: you can use uboot to write to load via tftp and write to nand: > That's how I recovered from my dockstar invalid kernel CRC. > For anyone intrested insructions are quite clear on this on openwrt serial > console install. Yes, forgot about the network method. Of course it has to be compiled in, too ;) > > IIRC on mikrocontroller.net the layout of the JTAG > > connector can be found. > > Yea I saw that but I'd still regard that as being more difficult then > plying with uboot from serial. Indeed. As long as I can play with serial/netconsole and a working u-boot I won't try JTAG. > > > ext2load- load binary file from a Ext2 filesystem > > > ext2ls - list files in a directory (default /) > > I tryed that but got nothing. meybe because this uboot does not support usb > ... but then again where else would the dockstar have an ext2 filesystem ? > directly on flash without any ware levelling dedicated hardware would not > be a smart thing to do. The included options are only a matter of configuration at compile time. And you are right: ext2 support without usb support seems to be useless. At least at the moment I canno imagine any useful scenario. > I'll play with both uboot and see if I can boot armedslack in some way > other then using openwrt as an initrd to start armedslack. If that fails > I'll try jeff's enhanced uboot. Some hints: Create first partition as fat on usb stick, second partition as ext4. The first will be the boot partition and contain your uImage. usb start ext2load usb 0:1 0x640 /uImage setenv bootargs root=/dev/sda2 rootfstype=ext4 run bootcmd Something like this... -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Wednesday 16 March 2011 12:22:18 schrieb Davide:
> I look like it replaces the original uboot on mtd0, is that correct ?
I'm not sure anymore. I would have to investigate the script again first.
> While openwrt is using original uboot to start another version so that if
> you screw up you still have the original one to get things working again
> without need for jtag.
Yes, that's possible. AFAIK you just have to place the second u-boot to the
NAND address where originally the kernel can be found. If it fails to load
the second u-boot the first u-boot must be able to boot from usb to recover
things. Otherwise without JTAG you are stuck, too.
IIRC on mikrocontroller.net the layout of the JTAG connector can be found.
> Just incase someone else needs it here is what the 2 versions of ubuut
> currently on my dockstar do: U-Boot 1.1.4 (Jul 16 2009 - 21:02:16) Cloud
> Engines (3.4.16)
Is this the original one?
> ext2load- load binary file from a Ext2 filesystem
> ext2ls - list files in a directory (default /)
This one can load the kernel of an ext2 file system. But I did not see any USB
capabilities.
> bootargs_root=root=/dev/mtdblock2 ro
It expects root on mtd2.
> bootcmd=nand read.e 0x80 0x10 0x4; go 0x80
It loads a kernel from NAND address 0x4 to RAM and boots it:
> CE>> boot
>
> NAND read: device 0 offset 0x10, size 0x4
>
> Reading data from 0x10 -- 0% ...
> 262144 bytes read: OK
> ## Starting application at 0x0080 ...
Second one:
> U-Boot 2010.09 (Mar 10 2011 - 00:51:16)
> Marvell-Sheevaplug
>
> SoC: Kirkwood 88F6281_A0
> DRAM: 128 MiB
> NAND: 256 MiB
> *** Warning - bad CRC or NAND, using default environment
> fatinfo - print information about filesystem
> fatload - load binary file from a dos filesystem
> fatls - list files in a directory (default /)
Support loading kernel of a FAT partition but not of ext2 (no problem).
> usb - USB sub-system
> usbboot - boot from USB device
It supports loading the kernel from USB devices.
> bootcmd=${x_bootcmd_kernel}; setenv bootargs ${x_bootargs}
> ${x_bootargs_root}; ${x_bootcmd_usb}; bootm 0x640; bootdelay=3
> baudrate=115200
> ipaddr=169.254.254.243
> serverip=169.254.254.254
> x_bootargs=console=ttyS0,115200
> mtdparts=orion_nand:1M(u-boot),1M@1M(second_stage_u-boot),3M@2M(kernel),32M
>@5M(rootfs),219M@37M(data) rw x_bootcmd_kernel=nand read 0x640 0x20
> 0x30
It loads a kernel from NAND address 0x30...
> x_bootcmd_usb=usb start
... starts the usb system (no idea why)
> x_bootargs_root=root=/dev/mtdblock3 rw rootfstype=jffs2
... and expects the root file system on mtd3 as jffs2.
What you can do:
Prepare a USB stick with a FAT boot partition. Place the kernel on it. Use
usbstart and fatload to load the kernel to the RAM.
Provide the right bootargs and your ARMed slack rootfs on the same USB stick.
And run bootcmd.
--
Thorsten Mühlfelder
Salix OS: www.salixos.org
___
ARMedslack mailing list
ARMedslack@lists.armedslack.org
http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] dockstar SOC unused features pin praceout
Am Tuesday 15 March 2011 09:19:13 schrieb Davide: > After some research I foud a listing of the trace out of some unused > features of the dockstar's SOC. It's in German but Google translate makes > it readable :-) > http://www.mikrocontroller.net/articles/Dockstar#PIN-Forschung As I'm German I have no problem with it :-D I remember a discussion about adding a RTC to the dockstar. Apparently they have found out a lot more now. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Monday 14 March 2011 11:51:53 schrieb Davide: > > > 4) You mentioned just renaming kernel and initrd: is > > > > that part of what > > > > > needs to be done to boot from usb ? > > > > 4 depends on the boot loader you are using and if you use > > the pre-configured > > boot commands. > > I'm using this: > http://downloads.openwrt.org/snapshots/trunk/kirkwood/openwrt-kirkwood-dock >star-u-boot.bin Which means using the seagate uboot to load > openwrt-kirkwood-dockstar-u-boot typing help when in the second uboot > environment shows usb boot capabilites with useless help on it. I don't > have the dockstar with me right nor to cut and paste a chunk of the > terminal output and the version info directly from the system. If necessary > I will supply it as soon as I get back home. OK, I have no idea what commands are activated at this u-boot. The usbload command loads the kernel from a given memory address. For better handling I'd recommend to use the ext2load command (available in Doozan's u-boot: http://jeff.doozan.com/debian/uboot/) Then you can save your kernel on a ext2/ext3 boot partition and easily access it from within linux, too. Doozan's bootloader also comes with preconfigured boot "scripts", that run the needed boot commands automatically (e.g. usbstart, ext2load, ...). You just have to use correct names for uImage and uInitrd on you ext2 boot partitions. IIRC I have changed these names in my own u-boot configuration, so someone else certainly can provide you the needed information. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Monday 14 March 2011 11:29:54 schrieb Stuart Winter: > > > 5) Are the slackware kirkwood kernel and initrd good for the dockstar > > > or do they need to be recompiled/remade ? I hav experience on zaurus > > > and know that kernels for c860 and c1000 are incompatible although the > > > hardware difference is minimal. > > > > The kirkwood kernel works fine on the dockstar. Only the LED won't work. > > Therefore you'd need a patched kernel. > > The one in -current doesn't, I think. I remember switching on the LED > support a while ago. > If not then it'll be on when I push out 2.6.38 later this week. I'm running 2.6.36.3-kirkwood on my Dockstar. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Monday 14 March 2011 10:05:34 schrieb Davide: > That's not coorect: the dockstar has 128 Mb ram and 256 or 512 Mb flash > depending on model. You are right. Mine has a 256 MB flash. There is a data partition with 219 MB that is empty and could be used for root. But still this is very small for an ARMed slack installation. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Monday 14 March 2011 08:48:23 schrieb Davide: > 1) is the mini root tarball just a xz compressed tarball of / ? Correct. Just untar on a partition of your choice, make some config file adjustments and you are ready to go. A readme file exists. > 2) how would I use the openwrt second stage boot loader to boot from a usb > drive with the mini root on it ? > > 3) I see in the help that there is a bootusb (or usbboot or something to > that meaning anyway) command: the help on it is not very helpfull so how > would I use that to boot from mini root on usb drive ? > > 4) You mentioned just renaming kernel and initrd: is that part of what > needs to be done to boot from usb ? 4 depends on the boot loader you are using and if you use the pre-configured boot commands. > 5) Are the slackware kirkwood kernel and initrd good for the dockstar or do > they need to be recompiled/remade ? I hav experience on zaurus and know > that kernels for c860 and c1000 are incompatible although the hardware > difference is minimal. The kirkwood kernel works fine on the dockstar. Only the LED won't work. Therefore you'd need a patched kernel. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Monday 14 March 2011 01:21:21 schrieb Rich: > I have not heard of anyone installing > the armedslack to the on-board flash. Probably because the on-board flash size is quite tiny. IIRC it's a 128 MB NAND flash. I think one can make ARMed slack to fit on it but there wouldn't be much room for additional packages. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] ARMed Slack on Toshiba AC100 netbook?
Hi, I wonder If anyone has tried ARMed Slack on a Toshiba AC100 netbook? http://www.engadget.com/2010/06/21/toshibas-ac100-8-hour-smartbook-runs-android-2-1-on-a-1ghz-tegr/ It's available in Germany for 199 € and I know that it is possible to install Ubuntu on it. So anyone? :-) -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] ARM board with PCIe slot
Hi, is there any ARM board with a PCIe slot out there? ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] DHCP as default
Hi, maybe it would be a good idea to set DHCP as default for eth0 in the mini-rootfs images? It would simplify things for newcomers, especially when ARMed Slack is installed on devices without having a console. -- Thorsten Mühlfelder Salix OS: www.salixos.org ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] #armedslack IRC channel on freenode
Am Friday 18 February 2011 10:21:07 schrieb Stuart Winter: > > > check it every day or so so might see you around if you drop in! > > > > > > s. > > > > What about a jabber chat room? :-) > > What about using a Jabber client to connect to IRC ? :) I'm still using Pidgin. So no problem. But IRC is the only non jabber account left in there :-D ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] File system on my Freeagent Dockstar got corrupted
Am Friday 18 February 2011 10:20:44 schrieb Stuart Winter: > > But of course now I wonder what caused this serious fail: > > * Did the Linux system (ARMed Slack 13.1) freeze and therefore corrupt > > the file systems? > > * Did the USB connection somehow got lost/unstable? > > * Did the hard disk fail and therefore freeze the system? > > * Problem with ext4 on ARM (I never had problems with it on my main PC) > > I've had Slackware arm 13.1 running on the box I use to build the eratta > packages, since it was installed after 13.1 was released - I've had no > problems with the ext4 filesystem on it. Yes, did not have problems for a very long time and it was running almost 24/7. I've build 424 packages on this machine :) > The problem could be anywhere -- perhaps take out the disc from the usb > caddy and put it onto a bare IDE interface and try SMART? > The disc or USB caddy could be faulty. I guess it's the only thing I can do now. I hope this Seagate drive is a SATA disk, otherwise I'll have problems to connect it to the PC. It's the original Dockstar disk, does anybody now what interface is used? ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] #armedslack IRC channel on freenode
Am Wednesday 27 October 2010 19:16:14 schrieb Stuart Winter: > Hi > > Robby's setup an #armedslack channel on the irc.freenode.net IRC service > -- there had been some discussions about armedslack on #slackware, but it > was being lost in the noise. > Hopefully this might be useful -- I leave myself logged into IRC 24/7 and > check it every day or so so might see you around if you drop in! > > s. What about a jabber chat room? :-) ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] File system on my Freeagent Dockstar got corrupted
Hi, yesterday I've logged in to my Dockstar via SSH and noticed that almost all commands fail for various reasons (e.g. segmentation faults). I tried to reboot the machine, which failed, too (it never rebooted, ping was still available, ssh was not reachable anymore). So I had to unplug the power. After power up the system did not boot again. I connected the hard disk to my PC and gparted told me that the file system (ext4) of the root and home partitions could not be recognized. Only the ext2 boot partition was still available. So I did a fsck run on both partitions, which showed _a lot of_ errors. Result: * the partitions could be recognized as ext4 again. * serious data loss on the root partition, e.g. the whole etc directory was gone. About 1600 files in lost+found. * nothing in lost+found of the home partition, so probably the data is ok there. But of course now I wonder what caused this serious fail: * Did the Linux system (ARMed Slack 13.1) freeze and therefore corrupt the file systems? * Did the USB connection somehow got lost/unstable? * Did the hard disk fail and therefore freeze the system? * Problem with ext4 on ARM (I never had problems with it on my main PC) Sadly I could not run samrtctl on the USB disk because it says "unknown USB chipset" or something like that. Did someone else ever have similiar problems? For now I've formatted the rootfs again with ext4 and copied the current mini-rootfs on it. At least the Dockstar boots again. ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] U-Boot + UbiFS
Am Wednesday 16 February 2011 16:37:47 schrieb Stuart Winter: > > does anybody know if U_Boot already supports loading the kernel of a > > UbiFS partition? I've built U-Boot 2010.09 for an Atmel board and I've > > seen that there are some ubi sources included but the U-Boot binary does > > not offer any ubi related commands. > > For u-boot it seems that there are a lot of other trees around with vendor > specific stuff and other features people have needed, and that eventually > these patches get back to the root. > > You'd need to ask on the u-boot mailing list. OK, if it is for interest for anyone else. Here is the answer I got on the u-boot ML: > This means you have not yet activated > > #define CONFIG_CMD_UBI > #define CONFIG_CMD_UBIFS > > in your configuration file (and CONFIG_YAFFS2 as well for yaffs). And as far as I know the right file to add these includes is include/config.h in the u-boot sources. I could build u-boot 2010.09 with YAFFS2 support (not yet tested), it failed to build with UBIFS support. Greetings Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] U-Boot + UbiFS
Hi, does anybody know if U_Boot already supports loading the kernel of a UbiFS partition? I've built U-Boot 2010.09 for an Atmel board and I've seen that there are some ubi sources included but the U-Boot binary does not offer any ubi related commands. ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] mfloat-abi
Am Friday 04 February 2011 15:34:57 schrieb Thorsten Mühlfelder: > They say: > "Floating point performance, with or without an FPU is very much faster, > and mixing soft and hardfloat code is possible". Here is a benchmark that I've found: http://www.linuxfordevices.com/c/a/Linux-For-Devices-Articles/Why-ARMs-EABI-matters/ ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] mfloat-abi
Am Friday 04 February 2011 15:07:34 schrieb Stuart Winter: > > Hmm, I did not really see what's the default here (soft, softfp, hard)? > > Me neither - I thought that the specs there were what was used by default. > I had a further quick read about it and that isn't the case. > > > As soft/softfpe is some kind of porting floating point operations to > > integer I've assumed that Armed Slack uses mfloat-abi=hard, which > > forwards the floating point instructions to the slow kernel FP emulation. > > Yes, it's been ages since I looked at the issue -- I know the oldabi > ARMedslack used that method, but I'd forgotten about the EABI version. > > The floating point used is the same as Debian use for their armel port: > http://wiki.debian.org/ArmEabiPort They say: "Floating point performance, with or without an FPU is very much faster, and mixing soft and hardfloat code is possible". This was not possible with OABI but I've never tested this on EABI. Building libvorbis with soft, softfp & hard may enlighten me ;-) If it is possible to build these three then mixing is allowed and a cpu usage comparison can follow. ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] mfloat-abi
Am Friday 28 January 2011 12:39:34 schrieb Stuart Winter: > > - soft: only software emulation > > - softfp + mfpu=xxx: enables the compiler to use FP > > instructions/registers inside functions, but not as function > > parameters. This way, code remains compatible with 'soft' libraries > > - hard: it's a different ABI, where parameters can be passed in FP > > registers, avoiding expensive moves between core and FP registers. But > > such code is not compatible with 'soft' libraries. > > You can also do > > gcc -dumpspecs > to see what the compiler's defaults are. > Hmm, I did not really see what's the default here (soft, softfp, hard)? BTW: A test with Ogg Vorbis decoding using mpd on my Dockstar gave me these results: - using libvorbis (floating point): ~20% CPU usage - using tremor (vorbis ported to integer only): ~2% CPU usage As soft/softfpe is some kind of porting floating point operations to integer I've assumed that Armed Slack uses mfloat-abi=hard, which forwards the floating point instructions to the slow kernel FP emulation. ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] mfloat-abi
Hi, I did not really understand what's the difference between -mfloat-abi=soft and -mfloat-abi=softfp? Does "soft" always use software floating point and "softfp" could use hardware floating point if a correct -mfpu is given? And why is none of them used in Armed Slack? In my understanding it should be much faster than FPU emulation in the kernel. http://www.arm.linux.org.uk/mailinglists/faq.php#f3 Greetings Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Seagate Freeagent Dockstar price
Hi, I've bought my Dockstar some months ago here in Germany for 20 € + 5 € shipping cost. Now I've noticed that the price has raised here to 80 € + shipping. So my question is: What the cheapest price nowadays all over the world? Of course I have to check shipping cost to Germany then... Thanks for infos Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] ARMedSlack on Seagate Dockstar
Am Friday 14 January 2011 11:20:51 schrieb Stuart Winter: > > I wonder how it booted the second time? IIRC I had to comment out the > > fdisk sectionin a rc script because the Dockstar does not have an RTC and > > after reboot the system time is set to 1970 again which causes another > > fsck run (fsck is forced if umount time is later than mount time). This > > ends up in an endless loop. > > This is how to do this properly: > > cat << EOF > etc/e2fsck.conf > # These options stop e2fsck from erroring/requiring manual intervention > # when it encounters bad time stamps on filesystems -- which happens on > # the Versatile platform because QEMU does not have RTC (real time clock) > # support. > # > [options] > accept_time_fudge = 1 > broken_system_clock = 1 > EOF > Thanks! Looks like a much better and easier solution than editing the rc script ;-) ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] ARMedSlack on Seagate Dockstar
Am Tuesday 21 December 2010 03:37:12 schrieb Rich: > On the initial boot, it did boot > the kernel and started the system initialization scripts but then fdisk > the sda1 and rebooted. The second boot did complete and I got a login > prompt on the serial terminal. I wonder how it booted the second time? IIRC I had to comment out the fdisk sectionin a rc script because the Dockstar does not have an RTC and after reboot the system time is set to 1970 again which causes another fsck run (fsck is forced if umount time is later than mount time). This ends up in an endless loop. > I then checked /etc/rc.d/rc.inet1.conf file and noticed that it wasn't > configured. I edited to use dhcp > USE_DHCP[0]="yes" > rebooted and was able to ssh into the dockstar Perhaps you remember that I've told you to do so ;-) ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Thinking about changing Slackware ARM to armv5t baseline
Am Tue, 4 Jan 2011 15:00:10 + (GMT) schrieb Stuart Winter : > > > Armed works fine on device with 32MB of RAM. > > Heh. It's called "ARMedslack" or "Slackware ARM". Every time I see > "Armed" I get images of men in balaclavas committing a bank > robbery ;-) > > Yes 32MB will work for minimal services like dhcp, dns and stuff like > that, but IME not much else! Yep, 32 MB is enough. See one of my devices (it uses 10 MB RAM so there should be enough left for some small services): r...@slackware:~# free -m total used free sharedbuffers cached Mem:60 29 31 0 0 18 -/+ buffers/cache: 10 49 Swap:0 0 0 r...@slackware:~# ps ax PID TTY STAT TIME COMMAND 1 ?Ss 0:21 init [3] 2 ?S 0:00 [kthreadd] 3 ?S 0:00 [ksoftirqd/0] 4 ?S 0:00 [events/0] 5 ?S 0:00 [khelper] 8 ?S 0:00 [async/mgr] 83 ?S 0:00 [sync_supers] 85 ?S 0:00 [bdi-default] 87 ?S 0:00 [kblockd/0] 98 ?S 0:00 [khubd] 118 ?S 0:00 [kswapd0] 119 ?S 0:00 [aio/0] 287 ?S 0:00 [mtdblock0] 292 ?S 0:00 [mtdblock1] 316 ?S 0:00 [yaffs-bg-1] 361 ?Ss 0:00 /usr/sbin/syslogd 392 ?Ss 0:00 /usr/sbin/klogd -c 3 -x 426 ?Ss 0:00 /sbin/dhcpcd -t 10 eth0 500 ?Ss 0:00 /usr/sbin/sshd 505 ?Ss 0:00 /usr/sbin/ntpd -g -p /var/run/ntpd.pid 510 ?Ss 0:00 /usr/sbin/crond -l10 512 ttyS0Ss+0:00 /sbin/agetty -L ttyS0 115200 vt100 513 ttyS1Ss+0:00 /sbin/agetty -L ttyS1 115200 vt100 514 ?Ss 0:00 sshd: r...@ttyp0 516 ?S 0:00 [flush-31:1] 517 ttyp0Ss 0:00 -bash 529 ttyp0R+ 0:00 ps ax r...@slackware:~# cat /proc/cpuinfo Processor : ARM926EJ-S rev 5 (v5l) BogoMIPS: 87.24 Features: swp half thumb fastmult edsp java CPU implementer : 0x41 CPU architecture: 5TEJ CPU variant : 0x0 CPU part: 0x926 CPU revision: 5 Hardware: Olimex SAM9-L9260 Revision: Serial : ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] ARMedSlack on Seagate Dockstar
Am Thursday 09 December 2010 14:48:25 schrieb Stuart Winter: > > I have a serial cable but I can't bear debian works without it. :-) > > Hang on a minute. So you've got a serial cable... > use it, figure out where the boot process stops (if at all). > Then once we know what the problem is, it can most likely be fixed or > worked around. Within 90 % a serial cable to RS232 connector won't work with the Dockstar. You need a TTL to RS232 level converter. Search for it on ebay and you'll find some cheap TTL to USB interfaces from China. ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] How to compile the whole Slackware ARM?
Am Thursday 18 November 2010 11:41:57 schrieb Carlo Caione: > Here some pointers (not mine), about performance for different compilers > (but same parameters): > > http://hardwarebug.org/2009/08/05/arm-compiler-shoot-out/ > Thanks for providing this link. As I can see there FLAC uses integer and Vorbis uses float. So rebuilding the libvorbis and flac packages should be the easiest method to get some test tools for different compiler optimizations. Does anybody want to do it? ;-) ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] How to compile the whole Slackware ARM?
Am Thursday 18 November 2010 10:51:47 schrieb Carlo Caione: > Ok, I hijack the discussion. > As I explained before in my previous email the idea is to recompile > armedslack with the right optimization for beaglebord (armv7 + neon + > cortex-a8 optimization) How much speed improvement can be gained by this, e.g. at Vorbis/Flac encoding or at compiling? ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Armedslack on a dockstar :-)
The Dockstar works fine with the Armed Slack kernel. See my message from 29.08.2010: > OK, I've received my Seagate Freeagent Dockstar. I've flashed a new > U-Boot loader to it. A description can be found here: > http://www.plugapps.com/index.php5?title=PlugApps:Pogoplug_Setboot > http://plugapps.com/os/pogoplug/uboot/ > > The board boots fine when the kirkwood kernel is used, that comes with > the mini rootfs. Here is the cpuinfo: > > cat /proc/cpuinfo > Processor : Feroceon 88FR131 rev 1 (v5l) > BogoMIPS: 1192.75 > Features: swp half thumb fastmult edsp > CPU implementer : 0x56 > CPU architecture: 5TE > CPU variant : 0x2 > CPU part: 0x131 > CPU revision: 1 > > Hardware: Marvell SheevaPlug Reference Board > Revision: > Serial : I've installed the new U-Boot. Make sure the kernel is called uImage and the initrd is called uInitrd. Am Wednesday 15 September 2010 11:15:11 schrieb Stuart Winter: > > Recently I'm being in touch with the Seagate Dockstar which is a > > promising device based on Kirkwood Soc: > > http://xkyle.com/2010/07/19/the-seagate-dockstar-a-very-cool-linux-device > >/ > > This looks like an interesting device. > > You'll probably need to build a new > kernel to begin with patched with support for the Dockstar, because I > don't see anything about the "Dockstar" in the "kirkwood implementations" > in the Kernel armedslack ships. There might be support in linux 2.6.36, > in which case I'll add it in when I next update the kernel. > > Let us know how you get on with it. Good luck! :) > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack -- Dipl.-Ing. (FH) Thorsten Mühlfelder Hard- und Softwareentwicklung ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Setting up distcc for cross compiling
Am Tuesday 31 August 2010 12:03:15 schrieb Stuart Winter:
> > is there a good guide how to set up distcc to build to help building arm
> > packages on my x86_64 and i486 hosts?
>
> Yep. Go back to the FTP site and download the script "dbuild" - I just
> put it there.
OK, had time to take a look at it. Please correct me if I'm wrong:
You are creating gcc links to distcc in a temp dir, so that distcc can be
called by usual command "gcc":
cd /tmp/DISTCC
ln -vfs /usr/bin/distcc gcc
ln -vfs /usr/bin/distcc cc
ln -vfs /usr/bin/distcc g++
ln -vfs /usr/bin/distcc c++
ln -vfs /usr/bin/distcc arm-slackware-linux-${EABI}c++
ln -vfs /usr/bin/distcc arm-slackware-linux-${EABI}g++
ln -vfs /usr/bin/distcc arm-slackware-linux-${EABI}gcc
After and to make sure distcc is called instead of original gcc you place the
directory with distcc symlinks at the beginning of the PATH. So it's not
needed to specify CC in Makefiles:
PATH=/tmp/DISTCC:$PATH
Before distcc can be called it needs info what hosts to contact and the number
of jobs the host should run:
export DISTCC_HOSTS="hostnameA/3 hostnameB/1 hostnameC/1"
And I probably have to do the same trick with the PATH on the host machines so
that the cross compiler is run?
___
ARMedslack mailing list
ARMedslack@lists.armedslack.org
http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Setting up distcc for cross compiling
Am Tuesday 31 August 2010 14:27:31 schrieb Christophe Lyon: > > Did you use distcc there, too? What about moving all jobs to the other > > hosts? I know there are still tasks left for the build host (linking?), > > but no idea how much RAM they need. > > But I've already thought that RAM will be a problem. Perhaps there may be > > some tricks that can help, e.g. > > http://hostingfu.com/article/compiling-with-gcc-on-low-memory-vps > > Why wouldn't you completely cross-build? I mean you don't have to link > on your ARM box given that you have already installed cross-toolchains > on your x86 hosts. How would I make sure that all dependencies will be found then? I guess I have to set up some kind of chroot with the arm packages installed in? So far I've only cross compiled programs without any dependencies or kernels. > > Yes, I wanted to buy a SheevaPlug, but my prefered model seemed to be out > > of stock at the moment. Furthermore the Dockstar was so pretty cheap (20 > > €), that I just wanted to give it a try. And it looks fine in my living > > room with the 2.5" disk attached ;-) hdparm gives me a speed of 26 MB/s > > which is quite OK for an USB drive. > > Too bad shipping costs are so high between Germany & France :-( I've seen it somewhere in France for 19 €, too. But IIRC this was one of these one-day-offers. ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Setting up distcc for cross compiling
Am Tuesday 31 August 2010 12:03:15 schrieb Stuart Winter: > > is there a good guide how to set up distcc to build to help building arm > > packages on my x86_64 and i486 hosts? > > Yep. Go back to the FTP site and download the script "dbuild" - I just > put it there. Thanks, I'll have a look there today evening when I'm home. > > PS: Could the 128 MB RAM on my Dockstar be a problem, even if I use > > distcc? > > Depends what you're compiling, but quite possibly, yes. > For example - compiling ruby on 256MB RAM took days because it was > constantly in swap; and building KDE is pretty much a no go. Did you use distcc there, too? What about moving all jobs to the other hosts? I know there are still tasks left for the build host (linking?), but no idea how much RAM they need. But I've already thought that RAM will be a problem. Perhaps there may be some tricks that can help, e.g. http://hostingfu.com/article/compiling-with-gcc-on-low-memory-vps > In my opinion, for a build host, the best bet at the moment is a > SheevaPlug or one of the *Plug family -- preferably with an eSATA disc > attached. Yes, I wanted to buy a SheevaPlug, but my prefered model seemed to be out of stock at the moment. Furthermore the Dockstar was so pretty cheap (20 €), that I just wanted to give it a try. And it looks fine in my living room with the 2.5" disk attached ;-) hdparm gives me a speed of 26 MB/s which is quite OK for an USB drive. ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Setting up distcc for cross compiling
Hi, is there a good guide how to set up distcc to build to help building arm packages on my x86_64 and i486 hosts? What I have right now: I've compiled the cross toolchain from ftp://ftp.armedslack.org/armedslack/armedslack-devtools/x-toolchain/ and installed it in /opt/arm/$LOCALARCH I've installed Armed Slack on my Seagate Freeagent Dockstar (1.2 GHz, 128 MB RAM). It should act as building host, so I've installed the needed development packages and already have built some packages there for testing purposes. Works fine, but slow so far. So what would I do next? All distcc how-tos I've found so far are for Gentoo or Debian and setup seems to differ greatly from my purpose. Usually I'd just try and error, but to save some time I first ask here this time. PS: Could the 128 MB RAM on my Dockstar be a problem, even if I use distcc? ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Error in perl package
There is an error in the perl package: mv: cannot stat `usr/lib/perl5/5.10.0/arm-linux-gnueabi-thread-multi/perllocal.pod.new': No such file or directory WARNING: Post-installation script failed. (256) Actually the dir should be usr/lib/perl5/5.10.1 ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
OK, I've received my Seagate Freeagent Dockstar. I've flashed a new U-Boot loader to it. A description can be found here: http://www.plugapps.com/index.php5?title=PlugApps:Pogoplug_Setboot http://plugapps.com/os/pogoplug/uboot/ The board boots fine when the kirkwood kernel is used, that comes with the mini rootfs. Here is the cpuinfo: cat /proc/cpuinfo Processor : Feroceon 88FR131 rev 1 (v5l) BogoMIPS: 1192.75 Features: swp half thumb fastmult edsp CPU implementer : 0x56 CPU architecture: 5TE CPU variant : 0x2 CPU part: 0x131 CPU revision: 1 Hardware: Marvell SheevaPlug Reference Board Revision: Serial : Am Tue, 24 Aug 2010 22:24:44 +0200 schrieb Thorsten Mühlfelder : > I've ordered one of these Dockstars now. I cannot get any cheaper > thing here (25 € incl. shipping). Let's see what I can do with it. > > Am Thu, 19 Aug 2010 00:39:19 -0600 > schrieb Tyler T : > > > > what CPU/RAM/flash does the Seagate Dockstar have? > > > > 1.2GHz ARMv5TE (Kirkwood), 128MB DDR RAM, 256MB flash. > > ___ > > ARMedslack mailing list > > ARMedslack@lists.armedslack.org > > http://lists.armedslack.org/mailman/listinfo/armedslack > > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
I've ordered one of these Dockstars now. I cannot get any cheaper thing here (25 € incl. shipping). Let's see what I can do with it. Am Thu, 19 Aug 2010 00:39:19 -0600 schrieb Tyler T : > > what CPU/RAM/flash does the Seagate Dockstar have? > > 1.2GHz ARMv5TE (Kirkwood), 128MB DDR RAM, 256MB flash. > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Hi laprjns, what CPU/RAM/flash does the Seagate Dockstar have? Am Wed, 18 Aug 2010 20:04:57 -0400 schrieb Richard Lapointe : > I have a Dockstar coming directly from Seagate ($39 USD) and would > like to run ARMedslack on it. Is there a single site that provides > instructions on how I would go about this? Since I am using Salix on > my desktops, I want to replace my N2SLU running SlugOS with the > Dockstar running Slackware and then build some packages using Salix > packaging took, slkbuild. > > Thanks > > Rich Lapointe (laprjns) > > > On 08/18/2010 02:27 PM, Tyler T wrote: > >> Now I wonder if I should buy an IGEPv2 board. It seems very > >> promising. This indeed would be a nice thing to start a Salix > >> port. Does anybody have experience with this one? > >> > > Don't know anything about that but here's a forum posting about > > cheap ARM devices: > > http://plugapps.com/forum/viewtopic.php?f=8&t=212. I have a > > DockStar myself; armedslack runs great on it (from a USB flash > > drive). I do have to chainload a better u-Boot though, detailed > > here: http://jeff.doozan.com/debian/uboot/. You'll also want a TTL > > serial console cable (http://ahsoftware.de/dockstar/). > > > > I wonder if USB DisplayLink hardware could work with one of these > > devices to create a "nettop" computer. > > ___ > > ARMedslack mailing list > > ARMedslack@lists.armedslack.org > > http://lists.armedslack.org/mailman/listinfo/armedslack > > > > > ___ > ARMedslack mailing list > ARMedslack@lists.armedslack.org > http://lists.armedslack.org/mailman/listinfo/armedslack ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Saturday 14 August 2010 18:20:45 schrieb Tyler T: > Hello and welcome! There doesn't seem to be a lot of traffic on this > list, I guess because Armedslack "just works". Salix looks perfect for > netbooks, maybe an ARM port of it is in order. ;) Now I've managed to get Armed Slack working on an Olimex SAM9-L9260 Board (http://www.olimex.com/dev/sam9-L9260.html). Got this thing very cheap from eBay. It features a 180 MHz Atmel AT91SAM9260 CPU, 64 MB RAM and 512 MB Flash. I've used some Atmel tool to program boot loader and kernel to the NAND and then I could boot the mini rootfs from an USB stick. Biggest problem was: The Olimex board is supported by vanilla kernel, but it won't work unless you'll add the patch from Olimex homepage, too. This mademe headaches for some time :-) BTW: I've used the toolchain from Codesourcery to compile the kernel. Now I wonder if I should buy an IGEPv2 board. It seems very promising. This indeed would be a nice thing to start a Salix port. Does anybody have experience with this one? I did not do any more research on the IGEPv2, but perhaps somebody knows how to program a boot loader and a kernel on it? It features a micro SD slot which I would use as data or even as root partition. Regards Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
Re: [ARMedslack] Hello
Am Sat, 14 Aug 2010 10:20:45 -0600 schrieb Tyler T : > Hello and welcome! There doesn't seem to be a lot of traffic on this > list, I guess because Armedslack "just works". Salix looks perfect for > netbooks, maybe an ARM port of it is in order. ;) Yes, this probably is something I'll try to do in the future. But first I have to check how to build a suitable cross compilation environment. I've already built small tools, that did not need any deps, but I've never build complete packages, yet. So are there any information how the armed slack packages are build? Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
[ARMedslack] Hello
Just wanted to say a short hello to the list. I'm one of the developers at Salix OS (salixos.org) and as I'm in contact with Linux based ARM devices at work I'm just interested in what's going on here. Probably I'll buy a little ARM board for playing with it at home, because I'm tired of having to use Debian ARM :-D Greetings Thorsten ___ ARMedslack mailing list ARMedslack@lists.armedslack.org http://lists.armedslack.org/mailman/listinfo/armedslack
