Absolutely. Different (2.6.31) kernels same message too, but is an OFW message rather than kernel. Could it be a new bad block not detected? Reinstalling should get the same problem at exactly the same block, no?
--- On Mon, 3/15/10, James Cameron quozl@laptop.org wrote:
From: James Cameron quozl@laptop.org Subject: Re: unsupported nodetype e009 To: "Yioryos Asprobounitis" mavrothal@yahoo.com Cc: fedora-olpc-list@redhat.com Date: Monday, March 15, 2010, 8:05 PM On Mon, Mar 15, 2010 at 03:29:34PM -0700, Yioryos Asprobounitis wrote:
Under F11-XO1/os11 I get during boot at the OFW screen
the message
"unsupported nodetype e009" right after the boot
arguments.
I could not find this message in OFW or kernel sources. Are you sure about the spelling?
-- James Cameron http://quozl.linux.org.au/
On Mon, Mar 15, 2010 at 10:42:40PM -0700, Yioryos Asprobounitis wrote:
Could it be a new bad block not detected? Reinstalling should get the same problem at exactly the same block, no?
Use the OFW scan-nand command. Move around looking at the blocks. See if you can see any that are unusual.
Blue is clean free space. Green and yellow are data. Red are known bad blocks. Grey at the end are the bad block table, if I recall correctly.
Tell us the address (in hex under the help text "move") if you find anything strange.
--- On Wed, 3/17/10, James Cameron quozl@laptop.org wrote:
From: James Cameron quozl@laptop.org Subject: Re: unsupported nodetype e009 To: "Yioryos Asprobounitis" mavrothal@yahoo.com Cc: fedora-olpc-list@redhat.com Date: Wednesday, March 17, 2010, 1:45 AM On Mon, Mar 15, 2010 at 10:42:40PM -0700, Yioryos Asprobounitis wrote:
Could it be a new bad block not detected? Reinstalling
should get the
same problem at exactly the same block, no?
Use the OFW scan-nand command. Move around looking at the blocks. See if you can see any that are unusual.
Blue is clean free space. Green and yellow are data. Red are known bad blocks. Grey at the end are the bad block table, if I recall correctly.
Tell us the address (in hex under the help text "move") if you find anything strange.
I actually did "test /nandflash::fixbbt" at OFW for a thorough nand check (being almost sure that something will come up). No new bad blocks just the original 10. This is something that either the test can not detect or some really weird jffs2 case.
-- James Cameron http://quozl.linux.org.au/
Hi,
I actually did "test /nandflash::fixbbt" at OFW for a thorough nand check (being almost sure that something will come up). No new bad blocks just the original 10. This is something that either the test can not detect or some really weird jffs2 case.
What happens if you install an image (e.g. build 802) that is not os11?
If the message is limited to os11, we might expect either: * os11 was created with a new version of mkfs.jffs2 that contains knowledge of a new nodetype that OFW doesn't understand, or * os11's jffs2 image is partially corrupt.
- Chris.
--- On Wed, 3/17/10, Chris Ball cjb@laptop.org wrote:
From: Chris Ball cjb@laptop.org Subject: Re: unsupported nodetype e009 To: "Yioryos Asprobounitis" mavrothal@yahoo.com Cc: "James Cameron" quozl@laptop.org, fedora-olpc-list@redhat.com Date: Wednesday, March 17, 2010, 11:26 AM Hi,
> I actually did "test /nandflash::fixbbt" at OFW for a thorough > nand check (being almost sure that something will come up). No > new bad blocks just the original 10. This is something that > either the test can not detect or some really weird jffs2 case.
What happens if you install an image (e.g. build 802) that is not os11?
If the message is limited to os11, we might expect either: * os11 was created with a new version of mkfs.jffs2 that contains knowledge of a new nodetype that OFW doesn't understand, or * os11's jffs2 image is partially corrupt.
I installed Bernie's os67 and is fine. Neighborhood is back too. It comes with q2e42d that I hadn't check before but I do not think that is related.
- Chris.
-- Chris Ball cjb@laptop.org One Laptop Per Child
A friend of mine has a G1G1 XO-1 that he got Dec 2007. I help him to install reasonably-up-to-date software - his system currently has q2e43e and os11. [BTW, he normally does not power off his XO, but leaves it sit forever with the lid closed and AC plugged in - with papers stacked on top. I've seen his system be warm to the touch.]
It is quite rare that I observe his XO being booted -- but I do not recall ever seeing the "unsupported nodetype e009" message. For the fun of it, I added the four lines at the beginning of his olpc.fth that caused the XO to be overclocked (to now 462 Mhz) -- and now I *did* see several of those messages while booting his XO. [Neither those messages, nor the overclocking, appeared to affect the operation of the XO. (My reason for applying a mild overclock was that Timidity was running the default-speed XO at more than 85% average CPU utilization - I wanted to create some "head room" by speeding up the CPU).]
IIRC, Yioryos has his XO-1 overclocked. As the above-described experience shows, it is likely that the "unsupported nodetype e009" message is being brought on by the use of overclocking.
mikus
I'm sure that over-clocking can cause this problem, however when they appeared and tested the was with a clean install and nothing attached on the XO (not even a mouse). BTW me XO-1 is over-clocked for 2 years now and is going through several reboots and OSs every day under my observation. This was a first. Also going through os67 and then back to os11 the problem does not appear anymore (but still no neighborhood in os11). Maybe 2 years of overclocking, several re-flashes multiple OSs, crashes, and experimenting, start getting the best of it :-)
--- On Fri, 3/19/10, Mikus Grinbergs mikus@bga.com wrote:
From: Mikus Grinbergs mikus@bga.com Subject: Re: unsupported nodetype e009 To: "Yioryos Asprobounitis" mavrothal@yahoo.com Cc: "James Cameron" quozl@laptop.org, fedora-olpc-list@redhat.com Date: Friday, March 19, 2010, 8:27 AM A friend of mine has a G1G1 XO-1 that he got Dec 2007. I help him to install reasonably-up-to-date software - his system currently has q2e43e and os11. [BTW, he normally does not power off his XO, but leaves it sit forever with the lid closed and AC plugged in - with papers stacked on top. I've seen his system be warm to the touch.]
It is quite rare that I observe his XO being booted -- but I do not recall ever seeing the "unsupported nodetype e009" message. For the fun of it, I added the four lines at the beginning of his olpc.fth that caused the XO to be overclocked (to now 462 Mhz) -- and now I *did* see several of those messages while booting his XO. [Neither those messages, nor the overclocking, appeared to affect the operation of the XO. (My reason for applying a mild overclock was that Timidity was running the default-speed XO at more than 85% average CPU utilization - I wanted to create some "head room" by speeding up the CPU).]
IIRC, Yioryos has his XO-1 overclocked. As the above-described experience shows, it is likely that the "unsupported nodetype e009" message is being brought on by the use of overclocking.
mikus
Interesting. Next time it happens, try under-clocking as a workaround.