**** BEGIN LOGGING AT Mon Oct 21 02:59:59 2013
Oct 21 07:10:58 <lumag> hello
Oct 21 07:11:16 <Jay7> lumag: hi
Oct 21 07:11:38 <Jay7> have read your conversation with ant
Oct 21 07:12:38 <Jay7> what was decision about fb.line_width?
Oct 21 07:30:13 <lumag> no decision I think.
Oct 21 07:30:24 <lumag> I have not-accepted patch for tmiofb
Oct 21 07:31:32 <lumag> Jay7, I think there should be the same problem on pdas with w100fb.
Oct 21 07:31:38 <lumag> pxafb should be safe though.
Oct 21 07:44:28 <Jay7> lumag: is it ok to add check like  if (line_length == 0) { line_length = width * bpp; } ?
Oct 21 07:45:23 <lumag> Jay7, it might help. There is one problem. IIUC, no _set_par() means that no mode was really set up.,
Oct 21 07:45:32 <Jay7> ah..
Oct 21 07:50:03 * Jay7 afk
Oct 21 07:52:57 <ant_work> gm
Oct 21 07:53:36 <ant_work> lumag: it is no harm to keep that kernel patch for now
Oct 21 07:54:02 <ant_work> the kb / cursors are fixed now?
Oct 21 07:54:36 <lumag> Yes.
Oct 21 07:54:51 <lumag> If you mean that keys do not correspond to keyboard ;)
Oct 21 07:55:02 <ant_work> people were complaining iirc
Oct 21 07:55:41 <ant_work> is [PATCH 1/3] ALSA: pxa: slightly refactor reset handling in linux-next?
Oct 21 07:55:47 <ant_work> let me see
Oct 21 07:56:49 <ant_work> not yet
Oct 21 07:57:15 <ant_work> well, not yet in 3.12-rc6
Oct 21 07:59:16 <ant_work> I think it is better to backport all for 3.10 and wait for 3.12
Oct 21 08:21:33 <lumag> ant_work, linux-next is not updated for a month or so.
Oct 21 08:21:40 <lumag> Probably due to all conferences.
Oct 21 08:27:03 <ant_work> yes, there is an 'unofficial' git tree mentioned on the ML
Oct 21 08:27:34 <ant_work> the patch for sharpsl_param was there and now is in -rc6
Oct 21 08:27:59 <ant_work> (there are too many git trees imho ;)
Oct 21 09:40:42 <ant_work> bluelightning: hi
Oct 21 09:40:53 <bluelightning> hi ant_work
Oct 21 09:42:41 <ant_work> JaMa builder has a new failure for initramfs-kexecboot-klibc-image
Oct 21 09:42:58 <ant_work> first thing I see is  makedevs: No entry for root in search list
Oct 21 09:43:16 <ant_work> well, it was this patch  http://patchwork.openembedded.org/patch/55495/
Oct 21 09:43:43 <ant_work> we set IMAGE_DEVICE_TABLES = "" for the initramfs then we
Oct 21 09:43:44 <ant_work> mount devtmps during init
Oct 21 09:44:38 <ant_work> strangely that patch is 2 months old...
Oct 21 10:30:01 <lumag> ant_work, I had problem with devtmpfs mounting by kexecboot (I wanted to have debug output on console).
Oct 21 10:30:23 <lumag> With linux-yocto I could not get it to work till I actually put /dev/console to rootfs
Oct 21 10:30:30 <lumag> *linux-yocto-tiny-kexecboot
Oct 21 10:58:09 <ant_work> hm. strange
Oct 21 10:59:07 <ant_work> hm. strange
Oct 21 10:59:31 <ant_work> we even skip purposedly the check for console
Oct 21 11:30:34 <lumag> ant_work, we don't need console usually, because neither printk nor ttyS are included in tiny kernels.
Oct 21 11:31:22 <ant_work> well, exactly, you'd readd printk and error stuff, then change the loglevel ofc
Oct 21 11:32:06 <ant_work> the hope is to have a 'decent' kernel to start with ;)
Oct 21 11:33:17 <ant_work> lumag: in which timezone are you nowadays?
Oct 21 11:33:55 <lumag> UTC-4
Oct 21 11:35:48 <ant_work> ah, ok, East Coast
Oct 21 11:37:35 <lumag> ant_work, No. Europe/Moscow
Oct 21 11:38:00 <lumag> It's 15:37 here
Oct 21 11:42:29 <ant_work> ok. Just to say I'm hacking at my pc evening so it is typically very late there... here I don't have any git trree, anything
Oct 21 11:43:21 <lumag> ant_work, you are UTC+02, aren't you?
Oct 21 11:44:09 <ant_work> yes, for one more week +2 then UTC+1
Oct 21 12:05:10 <ant_work> lumag: one Q for you engineers...
Oct 21 12:08:42 <ant_work> about the flash chip on collie, manual says LH28F640BFHE-PTTL90 (64M 4Mx16)
Oct 21 12:09:15 <ant_work> but the old deprecated driver said
Oct 21 12:09:18 <ant_work> This driver only supports 4x1 arrangement of chips.
Oct 21 12:09:45 <ant_work> her we have 2 chips each 16bit bus-wide
Oct 21 12:14:15 <ant_work> is it possible that Sharp has wired it as 4 32Mbit chips?
Oct 21 12:14:22 <lumag> According to docs I have here (unfortunately I can't find full service manual, only some old prints from it), collie has 2 chips. One occupying D0:D15, second sitting on D16:D31
Oct 21 12:14:33 <ant_work> I have the picture ;)
Oct 21 12:15:36 <ant_work> http://www.tekprosystems.com/zaurus/internals.htm
Oct 21 12:16:02 <ant_work> short sheet: http://www.sharpsme.com/memory/flash-memory/LH28F640BFHE-PTTL90
Oct 21 12:16:13 <ant_work> pdf manual in the web
Oct 21 12:17:46 <ant_work> Serv. Manual per e-mail, tell me which one
Oct 21 12:19:32 <ant_work> about the flash, all makes sense:  64M Top Boot Block Flash
Oct 21 12:19:35 <lumag> sorry, can't parse last line.
Oct 21 12:19:56 <lumag> http://pdf1.alldatasheet.com/datasheet-pdf/view/166457/SHARP/LH28F640BFHG-PBTLZ7.html
Oct 21 12:25:31 <ant_work> http://datasheet.octopart.com/LH28F640BFHEPTTL90-Sharp-Microelectronics-datasheet-11836939.pdf
Oct 21 12:32:05 <ant_work> lumag: when we talk about interleave, are we talking about the number of the chips on the bus or the nr of 'planes' in a chip?
Oct 21 12:32:27 <lumag> Number of chips.
Oct 21 12:32:35 <ant_work> I thought I did understand but the Sharp driver is confusing me.. why 4 ?
Oct 21 12:32:54 <lumag> Note, I didn't check the datasheet. It might be that the chip internally has two chips.
Oct 21 12:33:00 <ant_work> ah
Oct 21 12:33:56 <ant_work> I'll have to retry the CFI detection and see debug :/
Oct 21 12:35:03 <ant_work> what I can say, 16bit x2 chips is working, at least in read-only
Oct 21 12:36:18 <ant_work> I think something is wrong with the voltages required to unlock
Oct 21 12:36:38 <ant_work> because those nor get sectors locked on boot
Oct 21 12:49:49 <lumag> PA18 of scoop controls #WP bit (note active low here, so low - write protect, high - write enable)
Oct 21 12:50:14 <lumag> Vpp seems to be hardwired to VCC3, so voltages should be ok.
Oct 21 12:57:06 <ant_work> the old driver did unlock:
Oct 21 12:57:08 <ant_work> http://www.infradead.org/pipermail/linux-mtd/2005-November/014284.html
Oct 21 13:00:53 <lumag> ok. One of the batteries gave collie life
Oct 21 13:02:56 <ant_work> btw, is it possible that one is now exhausted after 10yrs?
Oct 21 13:03:35 <ant_work> the driver is still in debug and iirc the readings were bad on my collie :/
Oct 21 13:05:14 <ant_work> err.. voltage now  0
Oct 21 13:05:16 <ant_work> http://paste.debian.net/59988/
Oct 21 13:06:03 <ant_work> I hope it is a bug ;) I have some patches seen around (2.6.31) for collie battery
Oct 21 13:06:30 <ant_work> I'll send that stuff to you so you'll decide if interesting or not
Oct 21 13:08:10 <lumag> No-no-no please let me finish tosa for now...
Oct 21 13:08:12 <lumag> :)
Oct 21 13:08:31 <ant_work> well, the driver should be almost the same...
Oct 21 13:08:35 <lumag> Strange, serial console is not working.
Oct 21 13:09:11 <ant_work> collie or tosa now?
Oct 21 13:10:07 <lumag> collie
Oct 21 13:10:59 <lumag> 2.4 dmesg doesn't show any useful information :(
Oct 21 13:12:29 <ant_work> are you sure console was 115200 for 2.4?
Oct 21 13:12:45 <ant_work> can't remember but maybe was 9600
Oct 21 13:13:05 <lumag> 9600
Oct 21 13:13:22 <lumag> sorry, got to go now
Oct 21 13:13:25 <ant_work> cu
**** ENDING LOGGING AT Tue Oct 22 02:59:58 2013