**** BEGIN LOGGING AT Sat Nov 30 02:59:59 2013 Nov 30 10:19:33 lumag: unsigned long timeo = jiffies + HZ; Nov 30 10:20:38 zaurus-non-hardware.cfg:CONFIG_NO_HZ=y Nov 30 10:20:38 zaurus-non-hardware.cfg:CONFIG_HZ=100 Nov 30 10:21:49 maybe timeout is miscalculated here ? Nov 30 16:43:35 hi lumag_ Nov 30 16:56:03 * Jay7 have just installed oe-core on notebook Nov 30 16:56:25 now checking buildability of core-image-sato Nov 30 16:56:46 ant_home: do we have any gui working on clamshells? Nov 30 17:02:17 hi Jay7 Nov 30 17:02:26 try akita first Nov 30 17:02:40 what image/ Nov 30 17:02:45 sato Nov 30 17:02:53 or opie... Nov 30 17:03:20 is oe-core + meta-hh enough for sato? Nov 30 17:04:34 yes + meta-initramfs Nov 30 17:06:37 well, let's see Nov 30 17:08:09 you may want to add xinput-calibrator to XSERVER var but seems a bit broken these days... Nov 30 17:09:08 what is BBLAYERS_NON_REMOVABLE variable? Nov 30 17:09:43 should it have same content as BBLAYERS? Nov 30 17:10:14 is just for oe-core, is sel-added iirc Nov 30 17:10:19 *self Nov 30 17:29:08 hmm.. Nov 30 17:29:33 got error from gcc build about libmpc.so.3 Nov 30 17:53:35 error not yet seen... Nov 30 18:06:29 build restart was fixed this.. strange Nov 30 18:46:58 ok Nov 30 18:48:00 I need *silence* to study that damned CFI code and compare with the flow-chart on the tech sheet.. Nov 30 18:48:10 maybe this night... Nov 30 18:48:55 ant_home: btw, how is collie progress? Nov 30 18:49:13 is it usable now (w/o nand)? Nov 30 18:49:36 yes, MMC/SD driver is done (lumag) Nov 30 18:49:46 ok with CFI Nov 30 18:49:49 *CF Nov 30 18:49:53 heh Nov 30 18:50:21 so it's possible now to boot from CF/SD and current OE? Nov 30 18:50:45 SD is still unmerged Nov 30 18:50:45 *with current OE Nov 30 18:51:02 and what about framebuffer? Nov 30 18:51:10 fixed :) Nov 30 19:01:44 Jay7: if you don't know what to do while building, pls help us reading http://www.datasheetarchive.com/dl/Datasheet-019/DSA00340384.pdf Nov 30 19:02:07 oh :) Nov 30 19:02:33 not all.. Nov 30 19:02:40 just compare the routine http://lxr.free-electrons.com/source/drivers/mtd/chips/cfi_cmdset_0001.c Nov 30 19:03:37 at 4.11 and 4.12 (flow charts) Nov 30 19:04:55 afais the code unifies the case of resuming from erase and page-write Nov 30 19:05:16 but maybe there is a 0xf0 too much..instead of a 0xff Nov 30 19:05:23 well, it's dense matter ;) Nov 30 19:05:24 bbl Nov 30 21:50:48 max_leb_cnt too low (322 needed) Nov 30 21:51:00 do_rootfs failed for collie Nov 30 21:51:40 ah, is ok, the image is just too ig for our 113 EB Nov 30 21:53:20 for core-image-base : Error: max_leb_cnt too low (144 needed) Nov 30 21:54:01 kernel size and mtd-tools can be removed once collie 'works' Nov 30 21:54:12 well, you understand... Nov 30 22:13:05 sure Nov 30 22:13:17 minimal image built ok Nov 30 22:13:44 now after some patches I can put the ubifs image on collie but kexecboot 3.10-0.6 doesn't detect it :/ Nov 30 22:13:57 well, it detects but can't launch it Nov 30 22:14:04 any diagnostic messages? Nov 30 22:14:26 at which stage it fails? Nov 30 22:14:44 can't mount device.. Nov 30 22:14:49 hm.. Nov 30 22:15:05 stay on 2.6.31 for collie Nov 30 22:15:28 did you tried to mount by hands? Nov 30 22:16:06 not yet Nov 30 22:16:08 root@collie:~# ubiformat /dev/mtd2 Nov 30 22:16:18 root@collie:~# ubiattach -p /dev/mtd2 Nov 30 22:16:29 root@collie:~# ubimkvol /dev/ubi0 -N rootfs -m Nov 30 22:16:41 root@collie:~# ubiupdatevol /dev/ubi0_0 /rootfs.ubifs Nov 30 22:16:45 exit clean Nov 30 22:17:06 same kernel version? Nov 30 22:17:14 w/out hack I'm using it was Nov 30 22:17:15 root@collie:~# ubiupdatevol /dev/ubi0_0 /rootfs.ubifs Nov 30 22:17:15 sa1100: Chip not ready for buffer write. Xstatus = 80808080, status = 80808080 Nov 30 22:17:15 UBI error: ubi_io_write: error -62 while writing 64 bytes to PEB 87:64, written 0 bytes Nov 30 22:17:15 UBI error: vol_cdev_write: cannot accept more 13824 bytes of data, error -62 Nov 30 22:17:15 ubiupdatevol: error!: cannot write 13824 bytes to volume "/dev/ubi0_0" Nov 30 22:17:17 error 62 (Timer expired) Nov 30 22:17:21 root@collie:~# Nov 30 22:17:56 I expect the same failure if you try with 3.10 kernel Nov 30 22:26:07 root@collie:~# ubiattach -p /dev/mtd2 Nov 30 22:26:07 UBI error: ubi_read_volume_table: the layout volume was not found Nov 30 22:26:09 hmm Nov 30 22:29:42 UBIFS error (pid 294): validate_sb: min. I/O unit mismatch: 2048 in superblock, 8 real Nov 30 22:29:43 UBIFS error (pid 294): validate_sb: bad superblock, error 1 Nov 30 22:31:34 so it works with empty volumes.... Nov 30 22:31:35 root@collie:~# ubiupdatevol -t /dev/ubi0_0 Nov 30 22:31:35 root@collie:~# mount -t ubifs ubi0:rootfs /media/ Nov 30 22:31:35 root@collie:~# mount Nov 30 22:31:35 rootfs on / type rootfs (rw) Nov 30 22:31:36 devtmpfs on /dev type devtmpfs (rw,relatime,size=27748k,nr_inodes=6937,mode=755) Nov 30 22:31:38 proc on /proc type proc (rw,relatime) Nov 30 22:31:40 sysfs on /sys type sysfs (rw,relatime) Nov 30 22:31:42 debugfs on /sys/kernel/debug type debugfs (rw,relatime) Nov 30 22:31:44 tmpfs on /run type tmpfs (rw,nosuid,nodev,mode=755) Nov 30 22:31:46 tmpfs on /var/volatile type tmpfs (rw,relatime) Nov 30 22:31:48 devpts on /dev/pts type devpts (rw,relatime,gid=5,mode=620) Nov 30 22:31:52 ubi0:rootfs on /media type ubifs (rw,relatime) Nov 30 22:31:54 root@collie:~# Nov 30 22:34:20 if you write smthg it gets borked... UBIFS error (pid 302): ubifs_read_node: bad node type (255 but expected 2) Nov 30 22:39:00 heh..I've repaired it one side but borked on the other... Nov 30 22:39:57 Jay7: it's like talking with a remote terminal / black box Nov 30 22:40:06 you just check the status register Nov 30 22:40:09 hehe.. Nov 30 22:40:46 I start to suspect there might be some hw issue/bugged code/missing errata Nov 30 22:41:42 the code was written with another chip in mind (strataflash, like on h3600) Nov 30 22:42:26 which was needing many 'fixups' Nov 30 22:51:08 well.. good luck with your investigations Nov 30 22:51:12 * Jay7 need some sleep Nov 30 22:51:45 now reading some notes, about initial misdetection Nov 30 22:51:48 "Due to the system noise or any noise upon turning Nov 30 22:51:48 on/off the power supply, an unforeseen rewriting may Nov 30 22:51:48 be executed or the status register read mode may be Nov 30 22:51:48 invoked by the misrecognition of command." Nov 30 22:53:09 "By writing the Nov 30 22:53:09 READ ARRAY (FFH) command three times or more in Nov 30 22:53:09 order to initialize a device after turning on the power Nov 30 22:53:09 supply or changing RST#(RP#) from "low" to "high", Nov 30 22:53:09 unforeseen rewriting due to misrecognition of Nov 30 22:53:09 command can be prevented. Or you can return to the Nov 30 22:53:11 array data read mode from the status register read Nov 30 22:53:13 mode (Figure 3). The READ ARRAY (FFH) command Nov 30 22:53:15 needs to be issued three times or more because some Nov 30 22:53:17 of SHARP flash memories have a three-cycle Nov 30 22:53:19 command." Nov 30 22:53:38 so..three times or more... Nov 30 23:13:59 hm.. silly error for NOR Nov 30 23:14:01 UBIFS error (pid 442): validate_sb: min. I/O unit mismatch: 2048 in superblock, 8 real Nov 30 23:14:56 Minimum input/output unit size: 1 byte Nov 30 23:15:56 http://lists.infradead.org/pipermail/linux-mtd/2011-January/033794.html Nov 30 23:16:04 heh, known bug ;) Nov 30 23:18:25 but hey.. ubiformat: warning!: 31 of 113 eraseblocks contain non-ubifs data even after ubiformat...what's wrong? Nov 30 23:24:41 ok, about io size: http://patchwork.ozlabs.org/patch/244130/ Dec 01 00:09:59 sigh.. and jffs2 once flashed.. jffs2: jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x00c60020: 0x8080 instead Dec 01 00:10:18 chip is in FL_STATUS Dec 01 00:10:59 afair at the end of a page/buffer operation 0xff needs to be sent Dec 01 00:11:12 will check this tomorrow Dec 01 00:11:16 gn **** ENDING LOGGING AT Sun Dec 01 02:59:58 2013