> Very cool.. I'm a little surprised about v3.13.x being the base. Thought > Xenomai & RT was usually patched against even releases? My plan with > v3.13.x was just a quick release, and switch to v3.14.x asap and then just > keep updating v3.12.x/v3.14.x while dropping v3.13.x. Maybe i better not > do that too quickly.. >
Thanks, but better do it as usual, because I am really second-class developer: not perfect in kernel development (I spend much more time reading docs & sources than coding & trying), also I drive my company and do the ipipe work in very little spare time . As far as I noticed: 1) 3.13.4 with IPIPE patches I modified is generally more stable than 3.12 without basically no change in IPIPE part (I've noticed some changes in between in MM). 2) With IPIPE patches mmc subsystem when using external SDCARD is quite unstable, but the problem is I suppose not in IPIPE patches but in mmc subsystem itself. In Your "vanilla" 3.13.4 + patchset applied by Your build scripts there are also sometimes problems with mmc stability, but I think because much lower mmc irq latencies it is not so easily visible, but also (infrequently) happens. I check it following way: after booting I run quick & simple random FS read (I use ext4): ls -lR / | grep \?\?\? when during the read anything is flawed I see some signatures with permissions printed as ?????? There are also some modifications which could be applied to Your 3.13 tree from Beagle's 3.8: JTAG (debugss) clock enable etc... regards, T. > > Regards, > > -- > Robert Nelson > http://www.rcn-ee.com/ > -- For more options, visit http://beagleboard.org/discuss --- You received this message because you are subscribed to the Google Groups "BeagleBoard" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
