Latest posts of: mgillespie

Show Posts
Pages: 1 2 [3] 4 5 ... 16

31	General Category / General Discussion / /var on SDCard, frequently fails and ends up being mounted readonly.	on: April 06, 2010, 12:00:44 AM
A couple of months back, after several posters here planted seeds of doubt in my head about internal nand wear, I picked up a transcend 4GB Class6 SDCard, and moved my /var partition to it. (as looking at iotop, that's where most of my writes were occuring, by doing just var, I don't take too much of a write performance hit that sdcard brings compared to internal nand). Anyway, several times now, it's all gone pear shaped, and several services break down. Upon investigation, my sdcard is mounted read-only. Today I managed to get in via ssh after this had happened, and got a dump of dmesg (snippet included) Any ideas what's going on? (latest cxbiker kernel) Quote mmcblk0: error -110 transferring data, sector 437848, nr 8, card status 0xc00 end_request: I/O error, dev mmcblk0, sector 437848 Buffer I/O error on device mmcblk0p1, logical block 53707 lost page write due to I/O error on mmcblk0p1 mmcblk0: error -110 sending read/write command, response 0xd00, card status 0x400d00 mvsdio mvsdio: FIFO_EMPTY bit missing end_request: I/O error, dev mmcblk0, sector 3700800 end_request: I/O error, dev mmcblk0, sector 3700808 Aborting journal on device mmcblk0p1-8. mvsdio mvsdio: FIFO_EMPTY bit missing mmcblk0: error -110 sending read/write command, response 0x400d00, card status 0x400d00 mvsdio mvsdio: FIFO_EMPTY bit missing end_request: I/O error, dev mmcblk0, sector 3678208 Buffer I/O error on device mmcblk0p1, logical block 458752 lost page write due to I/O error on mmcblk0p1 JBD2: I/O error detected when updating journal superblock for mmcblk0p1-8. EXT4-fs error (device mmcblk0p1): ext4_journal_start_sb: Detected aborted journal EXT4-fs (mmcblk0p1): Remounting filesystem read-only EXT4-fs error (device mmcblk0p1): ext4_journal_start_sb: Detected aborted journal EXT4-fs (mmcblk0p1): Remounting filesystem read-only mvsdio mvsdio: FIFO_EMPTY bit missing mmcblk0: retrying using single block read mvsdio mvsdio: FIFO_EMPTY bit missing mmcblk0: error -110 sending read/write command, response 0xd00, card status 0x400d00 end_request: I/O error, dev mmcblk0, sector 75664

32	Linux Stuff / Linux distributions / Re: The alternative Debian install method.	on: March 09, 2010, 01:22:32 PM
The kernel that gets loaded is the one in the SheevaInster 1.0 (2.6.30 I recall). But yes, once you are up and running, you can simply update it to the latest from sheeva-with-linux. This is exactly what I do, it works without issue.

33	Linux Stuff / Linux distributions / Re: The alternative Debian install method.	on: March 08, 2010, 06:24:15 AM
Quote from: meldini@ngi.it on March 08, 2010, 03:17:33 AM Hi, this is my first message. i don't understand this passage: "At the end of the build process (which should only take about 10 minutes), you will be left with a rootfs.tar.gz that can be used to overwrite the one in the sheeva installer 1.0 download." I have downloaded a prebuilt rootfs.tar.gz I have downloaded sheevaplug-installer-v1.0.tar.gz and now? Sorry for my english You don't need to use the "build process", this is if you want to use the script to build your OWN rootfs.tar.gz You have opted to download a pre-built one. You need to untar the sheevaplug-installer-v1.0.tar.gz, replace the Ubuntu rootfs.tar.gz with the debian one you have download. Then follow the instructions for SheevaInstaller.

34	Linux Stuff / Kernel / Re: 2.6.33 new release	on: March 03, 2010, 02:02:39 AM
Thanks, working here Globalscale 1.3 Sheevaplug.

35	Linux Stuff / Linux distributions / Re: The alternative Debian install method.	on: February 07, 2010, 09:30:10 AM
Use these ones: http://www.mediafire.com/?mxzzmymzyjk (Lenny) Or http://www.mediafire.com/?z2ljmnukmdc (Squeeze) Or use the script to build your own (and more uptodate) rootfs.

36	Hardware and U-Boot firmware / Hardware / Weird problem with mmc card being mounted readonly	on: February 01, 2010, 03:39:51 AM
My server has been stable for many months, however yesterday I did two things: 1/ Updated to 2.6.32.7 cxbiker kernel. 2/ Update debian squeeze packages. Last night I got a couple of errors from cron saying that /var is readonly (I have most of my system on nand, but have /var mounted on SDcard). I did a check and /var was listed as being mounted as readonly. What is weird, is the system uptime was in weeks, and it certainly wasn't readonly earlier that day (as the cron would be moaning every hour). Even stranger today, I couldn't access the system via SSH, and had to reboot it to get access. I can access it again now, and /var on mcc is rw. Anyone got any ideas what may have occured?

37	Linux Stuff / Kernel / Re: Kernel 2.6.32 Crypto	on: January 18, 2010, 02:36:43 PM
Quote from: DanielS on December 23, 2009, 04:44:29 AM Just found out that all you have to do is select a cipher which has a drivername that starts with mv-, and you'll get hardware support. Can you eleborate? You have hardware crypto working? What with? How?

38	Hardware and U-Boot firmware / Hardware / Re: crypto accelaration	on: January 18, 2010, 02:35:41 PM
The only words I understood of that last post where "hello"...

39	General Category / General Discussion / Re: SheevaPlug Installer with updated kernel, uboot, distro ?	on: January 15, 2010, 05:27:08 PM
It's pretty hard to totally brick a plug. Those that are tinkering because they know how to get things back. There are documented and well tested methods of updating your uboot, kernel, rootfs, but this is not one of them (yet..) Off skiing next week (French Alps), so unlikely to have tinker time until I get back..

40	General Category / General Discussion / Re: SheevaPlug Installer with updated kernel, uboot, distro ?	on: January 14, 2010, 09:58:04 AM
Quote from: superpat on January 14, 2010, 08:50:19 AM Hi. Well! I just looked at uboot-nand-custom.txt That WAS set to orion_mtd # NAND flash partitions. If using kernel >= 2.6.30 then replace orion_nand with orion_mtd mtdpartitions mtdparts=orion_mtd:0x400000@0x100000(uImage),0x1fb00000@0x500000(rootfs) I changed it to orion_nand and did a complete re-install. This time I got the familiar kernel exception crash. I'm REALLY confused now, as I just downloaded the file from the same URL I PM'd you and mine says: (I changed the comment to as it seems that the 2.6.30 comment was misleading). Are you 100% sure you are not using an old copy? Quote # NAND flash partitions. If you are having problems, try replacing orion_nand with orion_mtd mtdpartitions mtdparts=orion_nand:0x400000@0x100000(uImage),0x1fb00000@0x500000(rootfs)

41	General Category / General Discussion / Re: SheevaPlug Installer with updated kernel, uboot, distro ?	on: January 14, 2010, 08:11:23 AM
I just check (as I was pretty sure I set it to orion_nand) and it is already orion_nand, however the putty.log shows: "uboot/uboot-env/fw_setenv console console=ttyS0,115200 mtdparts=nand_mtd:0xc0000@0\(uboot\)ro,0x1ff00000@0x100000\(root\)" So perhaps it's also needs setting somwhere else, or somethings not using the mtdpartitions variable... EDIT: Try changing the entry in uboot-dflt.txt also from orion_mtd to orion_nand I don't know offhand what this file does as part of the install process... But it's worth a shot... I don't otherwise see where it picked up orion_mtd from (unless it's an old uboot environet variable or something).

42	General Category / General Discussion / Re: SheevaPlug Installer with updated kernel, uboot, distro ?	on: January 14, 2010, 04:46:49 AM
Either it's a corrupted upload, or I did something silly like compress it as a bzip2 but called it a gzip... Can you try it with: tar -xvjf and tar -xvzf It could just be something simple like that. if not, i'll reupload it later.

43	General Category / General Discussion / Re: SheevaPlug Installer with updated kernel, uboot, distro ?	on: January 13, 2010, 11:49:35 AM
It's using the sheevainstaller 1.0 initrd. I am currently uploaded a tweaked version of the installer than has roosters suggestion implemented, if anyone wants to be a new tester.... http://www.mediafire.com/?sharekey=04c9dc1dbf3b7c15ab1eab3e9fa335caff81ba820fb48529

44	General Category / General Discussion / Re: SheevaPlug Installer with updated kernel, uboot, distro ?	on: January 12, 2010, 02:41:01 PM
Hmm, so regardless of what kernel is used, it's orion_nand. The note the txt file is wrong. Does this explain why superpat could not get this to work? Any ideas how to progress this is a fully working, updated debian installer pack? I'm lost....

45	General Category / General Discussion / Re: SheevaPlug Installer with updated kernel, uboot, distro ?	on: January 12, 2010, 06:31:47 AM
Looks like we have found the same problem as being discussed here: http://plugcomputer.org/plugforum/index.php?topic=717.msg7359#msg7359 Can you try it out with an older uboot to try and eliminate what is going on? (try it with both orion_nand and orion_mtd)

Pages: 1 2 [3] 4 5 ... 16