On Fri, 2015-08-07 at 17:15 +0200, Hans de Goede wrote: > Hi, > > On 07-08-15 12:19, Clive Messer wrote: > > On Fri, 2015-08-07 at 10:46 +0200, Hans de Goede wrote: > > > > > I've been running with that fix for a while now and I'm seeing no > > > such issues. > > > > Patch was applied to F22 4.1.3-201. Didn't managed to boot that > > successfully due to problems reading/writing sdcard and timeouts. > > No > > such issues booting with a "stock" 4.1.3-201. That was on > > BananaPro. > > I'll try again with Cubietruck today. (And apart from the fsck > > required, no problems immediately booting the stock kernel after > > the > > boot failures due to mmc timeouts with the patched kernel.) > > Weird. I'll keep this in mind when there is more clarity about a > "proper" fix for this upstream. Hans, I've also booted the kernel with your sunxi-mmc patch on CubieTruck and BananaPi, since my initial testing with BananaPro. Even if it doesn't hang on boot, while accessing SDCARD, sooner or later.... ** 61 printk messages dropped ** [ 105.010500] mmcblk0: error -5 sending status command, retrying ** 36 printk messages dropped ** [ 105.020100] mmcblk0: error -5 sending status command, aborting ** 31 printk messages dropped ** [ 105.028429] mmcblk0: error -5 sending status command, retrying ** 43 printk messages dropped ** [ 105.037543] mmcblk0: error -5 sending status command, retrying ** 38 printk messages dropped ** [ 105.046260] mmcblk0: error -5 sending status command, retrying ** 38 printk messages dropped ** [ 105.055028] mmcblk0: error -5 sending status command, retrying ** 33 printk messages dropped ** [ 105.063290] mmcblk0: error -5 sending status command, retrying I rebuilt with original sigint patch ( https://groups.google.com/forum/#!topic/linux-sunxi/v6Ktt8lAnw0) where the check is only being performed for sdio access from wifi driver rather than everything that goes through sunxi mmc driver.... No issues. Have put many Gigs of data through the wifi connection on the BananaPro. Sorry, don't have any more time to look into this and as you say, probably a non-issue, if neither is the actual "fix" that will make it upstream anyway. Clive -- Clive Messer <clive.m.messer@xxxxxxxxx> _______________________________________________ arm mailing list arm@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/arm
