**** BEGIN LOGGING AT Mon Jun 05 03:00:03 2017
Jun 05 11:08:23 <ant_work> VartiWork, try that 3.10 pls  https://github.com/LinuxPDA/linux-kexecboot/blob/master/zaurus/kernel-3.10.35/installkit-akita.tar.gz
Jun 05 11:08:45 <ant_work> next will be readding the 'lost' 3.2
Jun 05 11:09:17 <VartiWork> will do
Jun 05 11:09:53 <ant_work> in theory 3.10 has the fixes for the breakaghes signaled by dromede after 3.7
Jun 05 11:10:34 <ant_work> stiil, on my c860, SD/MMC detection is not 100%correct
Jun 05 11:10:43 <VartiWork> hmm, wasn't the related patch still to be merged?
Jun 05 11:11:10 <ant_work> maybe my c860 is getting old, or the caqrds
Jun 05 11:11:20 <ant_work> let me do countercjheck on other Z
Jun 05 11:11:39 <ant_work> anyway, this 3.10 is not 'OK' in my opinion
Jun 05 11:11:59 <ant_work> the cards are not detected 4 out 5 times
Jun 05 11:12:03 <VartiWork> or was it merged, then for the 4.x kernel they reverted to the post-3.7, broken code?
Jun 05 11:12:46 <ant_work> it seems 3.8 was ok
Jun 05 11:12:48 <VartiWork> ok... in case the card won't be detected at the first time, I'll check more times
Jun 05 11:12:48 <ant_work> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/arch/arm/include/asm/assembler.h?h=v3.8&id=1ecec696c8bb9b4cefb09495d81d081d1c81b578
Jun 05 11:13:03 <ant_work> and
Jun 05 11:13:04 <ant_work> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/video/backlight/corgi_lcd.c?h=v3.8&id=1581b088fc91cbb974ad2b98431a8ecabb4852ee
Jun 05 11:13:14 <ant_work> both in 3.8
Jun 05 11:13:19 <VartiWork> try to do a fsck + f3 on that SD on a PC
Jun 05 11:13:42 <ant_work> heh, it's not that ewasy..I did it already
Jun 05 11:13:50 <ant_work> it is th ekernel imho
Jun 05 11:14:02 <VartiWork> ok
Jun 05 11:14:04 <ant_work> 3.2 was much quicker detecting cards
Jun 05 11:14:24 <ant_work> now there are things like 'deferred probe' in kernel...
Jun 05 11:14:42 <ant_work> a.k.a. take your time....
Jun 05 11:14:49 <ant_work> not good for bootloaders :)
Jun 05 11:15:13 <VartiWork> nope :)=
Jun 05 11:15:29 <ant_work> anyway, let see if 3.10 boots on akita =)
Jun 05 11:17:57 <VartiWork> will let you know
Jun 05 11:20:31 <VartiWork> these days I'm testing kexecboot 2.6, booting a 4.12 kernel on mtd1
Jun 05 11:22:11 <ant_work> yes, but it's a band-aid...
Jun 05 11:22:43 <VartiWork> I know, but at least this way I can test alarmz on Akita, while waiting for a definitive kexec
Jun 05 11:22:57 <VartiWork> I see that the zImage can be stored in the root directory, but does boot.cf have to be stored in /boot?
Jun 05 11:23:04 <VartiWork> boot.cfg
Jun 05 11:23:58 <ant_work> yes
Jun 05 11:24:02 <VartiWork> or should both zImage and boot.cfg be in /boot?
Jun 05 11:25:46 <ant_work> #ifdef USE_ZIMAGE
Jun 05 11:25:46 <ant_work> 	PREPEND_MOUNTPATH("/boot/zImage"),
Jun 05 11:25:46 <ant_work> 	PREPEND_MOUNTPATH("/zImage"),
Jun 05 11:25:46 <ant_work> #endif
Jun 05 11:25:54 <ant_work> same for uImage
Jun 05 11:26:24 <VartiWork> aha thanks
Jun 05 11:27:43 <VartiWork> also, how should be the home partition called by root in the APPEND label? /dev/mtd2, /dev/mmcblk0p2 or /dev/sda2, or it doesn't matter?
Jun 05 11:28:15 <VartiWork> in my case, it's in the second partition in the SD, the first hold the 4.12 kernel only
Jun 05 11:28:22 <ant_work> btw
Jun 05 11:28:23 <ant_work> #define BOOTCFG_PATH MOUNTPOINT "/boot/boot.cfg"
Jun 05 11:35:03 <ant_work> VartiWork, sorry, on the phone...
Jun 05 11:35:27 <VartiWork> np :)
Jun 05 11:42:02 <ant_work> VartiWork, you don't specify home on commandline, just root and initrd
Jun 05 11:42:10 <ant_work> home is set in fstab
Jun 05 11:42:32 <VartiWork> soory, I meant root, not home
Jun 05 11:42:49 <VartiWork> soory = sorry
Jun 05 11:43:33 <ant_work> smthg like APPEND=root=/dev/mtd2 rootfstype=jffs2
Jun 05 11:45:26 <ant_work> kexecboot lists the output of /proc/partitions so you see th eblock devices
Jun 05 11:45:36 <VartiWork> ok, so mtd2
Jun 05 11:45:51 <VartiWork> and in my case, rootfstype=f2fs
Jun 05 11:46:00 <ant_work> yes
Jun 05 11:47:13 <VartiWork> greguu mentioned to use real_root, but from what I have read you can use it only if you use initrd: https://unix.stackexchange.com/questions/16098/description-of-linux-kernel-boot-parameters-real-root-and-cdroot
Jun 05 11:48:06 <VartiWork> in the alarmz, noinitrd is set inside APPEND, so I guess it won't work. I'll try to set root= directly, if that won't work I'll try to force pivoting as you suggested
Jun 05 11:48:22 <ant_work> you can do both ways
Jun 05 11:48:33 <ant_work> with or without initrd
Jun 05 11:48:54 <ant_work> here you don't need it, because the filesystem drivers are compiled in kernel
Jun 05 11:49:07 <ant_work> so it is enough the APPEND= like above
Jun 05 11:49:51 <VartiWork> ok
Jun 05 11:49:53 <ant_work> alternatively, you use a small kernel with fs compiled as modules, load an initramfs/initrd using append, this one contains the drivers and the init script
Jun 05 11:50:14 <ant_work> same if you have encrypted partitions: you need a middle layer
Jun 05 11:50:33 <VartiWork> I see
Jun 05 11:50:44 <ant_work> not our case with zaurus, you know where and which fs
Jun 05 11:51:54 <ant_work> if you want to have the kernel in mtd1 you need to shrink it and use an 'initrd', like Sharp did
Jun 05 11:52:09 <ant_work> then you have fixed boot
Jun 05 11:54:22 <ant_work> VartiWork, btw, real_root must be a script option, like pivot_root or such
Jun 05 11:54:32 <ant_work> it is not a kernel boot param
Jun 05 11:54:35 <ant_work> see https://www.kernel.org/doc/Documentation/admin-guide/kernel-parameters.txt
Jun 05 12:05:34 <VartiWork> if I understood correctly, you'd use in as e.g. APPEND=root=/dev/(partition1) rootfstype=f2fs real_root=/dev/(partition 2)
Jun 05 12:06:45 <VartiWork> but "real_root" itself is a script, and the kernel interpretates this as "run the real_root script"?
Jun 05 12:51:59 <ant_work> kernel doesn't know about it, is ignored
Jun 05 12:52:13 <ant_work> but is probably interpretated by the init
Jun 05 12:52:32 <ant_work> like linuxrc
Jun 05 12:53:09 <ant_work> see i.e. https://blog.kubovy.eu/2012/08/24/initramfs-for-encrypted-root-using-gpg-encrypted-random-keys/
Jun 05 12:53:22 <ant_work> (first google match...)
Jun 05 13:11:32 <VartiWork> ok
**** ENDING LOGGING AT Tue Jun 06 03:00:02 2017