**** BEGIN LOGGING AT Fri Nov 07 03:00:00 2014
Nov 07 08:01:02 <ant_work> Jay7: 83.222.111.211 / Moscow is down
Nov 07 08:01:42 <ant_work> Jay7: can you dump the site and upload it on github so we have a copy there?
Nov 07 08:11:50 <lumag> ant_work, hello
Nov 07 08:11:59 <ant_work> gm
Nov 07 08:12:33 <ant_work> tested the pxa2xx spi patch on poodle. ok. No more trace
Nov 07 08:12:48 <ant_work> still the fb remains scrambled
Nov 07 08:12:55 <ant_work> after resume
Nov 07 08:51:06 <lumag> Hmm.
Nov 07 08:51:57 <lumag> poodle uses locomo to control fb, doesn't it?
Nov 07 08:52:18 <lumag> s/fb/lcd/
Nov 07 08:53:09 <lumag> Maybe there is a bug somewhere in locomolcd driver.
Nov 07 08:54:02 <lumag> ant_work, scrambled, but does it work?
Nov 07 08:54:25 <lumag> If you draw something, print or just dd if=/dev/urandom of=/dev/fb0, does the picture change respectively?
Nov 07 08:54:43 <lumag> So, what does 'scrambled' mean?
Nov 07 08:56:18 <ant_work> let say on suspend the backlight and lcd shut off. On resume I see a black baclighted screen with a slight pattern of stripes. No clock?
Nov 07 08:57:09 <ant_work> I will try to write on fb
Nov 07 08:58:58 <ant_work> note I'm using both
Nov 07 08:59:01 <ant_work> 	locomolcd-suspend-resume.patch
Nov 07 08:59:08 <ant_work> locomo-fix-SCL-SDA.patch
Nov 07 08:59:32 <ant_work> maybe the new locomo driver fixes that
Nov 07 09:07:27 <lumag> Try removing fix-SCL-SDA
Nov 07 09:10:53 <lumag> Maybe the pxa and sa1100 resume framebuffer in a different order.
Nov 07 09:57:35 <lumag> ant_work, Today I got a 32Mb MMC card. It looks like it doesn't work with locomo :(
Nov 07 09:57:39 <lumag> I'll have more fun
Nov 07 10:59:18 <lumag> ok. Lowering speed to 2.5 Mhz makes that card to work.
Nov 07 11:12:53 <lumag> ant_work, got that strange card work quite reliably on 20Mhz.
Nov 07 11:56:12 <lumag> Hmm no. The card reports 0 bytes long :(
Nov 07 11:56:46 <ant_work> hm.. capacity changes in dmesg?
Nov 07 11:57:12 <ant_work> I had 10yrs old a dying card with such symptoms
Nov 07 11:57:55 <ant_work> it was more reliable in Zaurus than  in PC reader...
Nov 07 11:58:40 <lumag> PC reader perfectly finds 32Mbytes. Collie reports 0
Nov 07 11:58:53 <lumag> I'd assume that there is a bug in the card.
Nov 07 11:59:23 <lumag> Unfortunately that is the only MMC card I have at hand (all other are SD cards - they  are a little bit different).
Nov 07 11:59:30 <ant_work> can you try it on another Z?
Nov 07 11:59:51 <ant_work> I have as wella plethora of different SD
Nov 07 12:00:09 <ant_work> from 64MB to 32GB (sdhc)
Nov 07 12:00:32 <lumag> I will, but it won't probably matter - all other Z use 1-bit bus mode instead of using SPI :(
Nov 07 12:01:21 <lumag> It might help catch protocol errors though.
Nov 07 12:01:47 <lumag> *protocol support errors
Nov 07 12:02:02 <ant_work> or on collie with 2.4 driver...
Nov 07 12:02:52 <lumag> That might help
Nov 07 12:04:30 <lumag> What is the kernel image that I can use with original rom?
Nov 07 12:05:03 <lumag> Do you have it? Because I think I forgot to backup mine.
Nov 07 12:06:01 <ant_work> not offhand, sorry. I think I did extract it from the ospack
Nov 07 12:06:02 <lumag> I have an Ospack. Maybe I can extract it
Nov 07 12:06:06 <ant_work> ;)
Nov 07 12:06:40 <ant_work> you can put it on a CF/SD and have kexecboot point to th erootfs in mtd
Nov 07 12:06:45 <lumag> Do you remember/know where to find offsets?
Nov 07 12:08:27 <lumag> Ok, google known
Nov 07 12:08:29 <lumag> *knows
Nov 07 12:10:54 <lumag> Should I pass params in any special way?
Nov 07 12:13:15 <ant_work> I don't think it has dynamic cmdline/atags. iir cis hardcoded to boot from mtd2
Nov 07 12:13:37 <lumag> Blackscreen and nothing more.
Nov 07 12:17:48 <ant_work> I think cramfs has been added only to linux-kexecboot >= 3.10
Nov 07 12:18:02 <ant_work> which lacks mmc/sd
Nov 07 12:23:14 <lumag> cramfs?
Nov 07 12:23:41 <ant_work> which image do you have on mtd2?
Nov 07 12:26:08 <lumag> I think I only changed original zImage for the kexecboot
Nov 07 12:26:20 <lumag> Should I also load initrd?
Nov 07 12:28:48 <lumag> Kernel is alive (unplugging/replugging power toggles charger LED)
Nov 07 12:29:30 <lumag> Maybe I should spend some efforts to get collie boot from ROM in qemu
Nov 07 12:29:46 <ant_work> afair should just boot from sd/cf and mount the cramfs fs
Nov 07 12:37:13 <lumag> According to your message, old kernel might just miss atags :)
Nov 07 12:37:16 <lumag> http://permalink.gmane.org/gmane.linux.distributions.angstrom.user/3277
Nov 07 12:43:21 <ant_work> mem=64M !
Nov 07 12:44:12 <ant_work> I remember having done it with cacko / 2.6
Nov 07 12:44:32 <ant_work> but not on collie at th etime
Nov 07 12:44:41 <ant_work> argh this <space>
Nov 07 12:47:49 <lumag> mem=32 ?
Nov 07 12:48:51 <lumag> mem=64M will not work with original kernels
Nov 07 12:49:42 <ant_work> ah...actual kernel needs mem=
Nov 07 12:49:54 <ant_work> I see
Nov 07 12:50:07 <ant_work> say I have 2 collies at home...
Nov 07 12:50:25 <ant_work> 1 is 'as new'
Nov 07 12:50:27 <lumag> Anyway, neither of that helps
Nov 07 12:50:32 <ant_work> I can test on it ;)
Nov 07 12:50:37 <lumag> Heh :)
Nov 07 12:50:49 <lumag> The problem is that my MMC card is here - St. Petersburg, Russia :(
Nov 07 12:50:53 <ant_work> (actually I had a 3rd one..broken mtd ;)
Nov 07 12:51:12 <lumag> I can also try just flashing original 2.4 kernel
Nov 07 12:52:29 <ant_work> don't waste time. I'd like to build asap a new linux-kexecboot for collie with SD/MMC + cramfs + ubi stuff
Nov 07 12:52:42 <ant_work> just like for the other Z
Nov 07 12:53:30 <ant_work> even if slow, we just need to detect
Nov 07 12:53:57 <ant_work> the kernel in the rootfs will be adjusted
Nov 07 12:55:02 <ant_work> so pls tell me your suggestions for the minimal/safe settings
**** ENDING LOGGING AT Sat Nov 08 03:00:00 2014