On 27.02.2012, at 20:05, Guillaume Gardet wrote:
Le 27/02/2012 19:10, Alexander Graf a écrit :
On 02/27/2012 07:08 PM, Guillaume Gardet wrote:
Le 26/02/2012 21:11, Arnd Bergmann a écrit :
On Sunday 26 February 2012, Alexander Graf wrote:
On 26.02.2012, at 10:25, Arnd Bergmann wrote:
On Saturday 25 February 2012, Marcus Schäfer wrote: > The start sectors are aligned to be %8 clean and the to the cylinder > boundary. So far I did not had any trouble with that setup on 4G > SD cards (did not try on 8G cards though). For MLO loading I think > only the start address might matter... not sure Aligning to 8 sectors is not going to help at all, that is much smaller than the normal erase block size (4MB) and also smaller than typical page sizes (8kb or 16kb). If you want good performance, you should always align to 4MB or more. Using the regular geometry (255/63/1024), that will never be cylinder aligned, which is ok. Don't try to do cylinder alignment. Hrm - does that also have any implications on not being able to boot from the beagleboard? :) I tried the beagle image of today and it does not help.
fdisk output: (only the end of 2nd partition seems to have moved and is no more Linux LVM type but Linux type)
So if you remove the second partition, it works?
No. My previous trick no more works. Well, I tried another thing. I dd if=/dev/zero to erase the first sectors to be sure to have a good card to start and then dd the raw image. Then I tried to boot and get a "60" on the serial port which means no suitable MLO where found where it is ecpected!
Hrm - weird. Could you please try to dd if=/dev/zero the first couple MB and then just dd MLO straight into offset 128kb? Try again for offset 0. According to the spec (http://www.ti.com/litv/pdf/spruf98v, page 3475) offsets 0 and 128kb should work and be checked before the MBR or file system headers get analyzed.
***********************************************************
Disk /dev/sdb: 3941 MB, 3941597184 bytes 122 heads, 62 sectors/track, 1017 cylinders, total 7698432 sectors Units = sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x929ea0a6
Device Boot Start End Blocks Id System /dev/sdb1 * 2048 309247 153600 83 Linux /dev/sdb2 309248 1368063 529408 83 Linux
***********************************************************
I only know of one bug on beagleboard where an old xloader requires the boot partition to start at sector 63 independent of the geometry. It is well possible that there are other bugs though. Which x-loader is currently used for openSUSE?
X-loader is MLO and you don't even get that far, so I don't think that's the issue you're hitting here. It looks more like a bootrom bug.
Yes, MLO is an x-loader signed. Could you give me the sources so that I can have a look at it, please?
Sure - all of openSUSE is public: https://build.opensuse.org/package/files?package=x-loader-omap3beagle&project=openSUSE%3AFactory%3AARM
I tried the openSUSE MLO on a FAT partition and it boot on it since the "60" is no more shown but I get nothing on the serial port which means MLO is booted but crashed somewhere.
If you want info about MLO (for DM3730, means beagleboard xM version), have a look on this PDF from TI (Pages 3593+): http://www.ti.com/lit/ug/sprugn4o/sprugn4o.pdf
Yeah, that's in line with the document I was reading :) • Raw mode: In raw mode, an image can be at offset 0 or 128KB and must not be bigger than 128KB. Raw mode is detected by reading sector 0 and sector 256. The content of these sectors is verified for the presence of a TOC structure. In the case of a GP device, a configuration header (CH) must be in the first sector followed by a GP header. The CH can be void (contains only a CHSETTINGS item for which the valid field is 0), as described in Section 26.4.8.2, Configuration Header. Image data is read directly from continuous sectors of a card. If raw mode is not detected, file system mode is checked.
I think, there is a problem on the MLO file itself and maybe also on the MLO location for the raw mode.
Maybe we just need a different MLO for the xM beagleboard? Alex -- To unsubscribe, e-mail: opensuse-arm+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-arm+owner@opensuse.org