partition table of uSD HC deleted on suspend/resume

[ChrisPZV]

I tried to run actual SHR-T on 4GB uSD-HC card (OM FR GTA02), booting from Qi and it seems to kill the partition table of the card so at least at next reboot Qi cannot find partion on SD card to boot from. I asume it's the same with SHR-U 'cause both are running kernel 2.6.29.rc3. This bug seems to be related to this one:  https://docs.openmoko.org/trac/ticket/1802
There it is called fixed, but using the fix on /etc/apm/suspend.d/00sd_idleclk and /etc/apm/resume.d/00sd_idleclk (making both 0755 after that) does not work for me.
Actually I'm trying to run SHR-U installed to NAND and using the 'bloody' hack of polarbear ( https://docs.openmoko.org/trac/ticket/1802#comment:3). I will report here soon if that will work.

[ChrisPZV]

Test SHR-U on uSD, the same behavior.
But running SHR-T installed with the instructions from  http://wiki.openmoko.org/wiki/Configure_SHR_for_German-speaking_use (It runs from NAND and saves maps, geocaches etc. on uSD) works well for 2 days now without any dd-ing partition table or having extra resume.d- or suspend.d-scripts.

[mrmoku]

We use a non-kms kernel now, which is said to fix uSD corruption... please reopen if that still happens.