**** BEGIN LOGGING AT Sat Jul 12 02:59:59 2014 Jul 12 03:13:43 Kilroo: yeah, a garage account is a mandatory prerequisite to receive a ballot Jul 12 06:40:18 A wonderful fine good Morning (UGT) on this gorgeous Saturday (CEST) Morning, Jul 12 08:14:37 cmd: Today is Pungenday, the 47th day of Confusion in the YOLD 3180 Jul 12 10:32:27 "All telephony functions, including emergency calls, are disabled due to a communication error. To recover, you might have to reboot the device" Jul 12 10:32:37 springfix didn't help :( Jul 12 11:36:34 Eh, apparently it didn't help its inventor either: http://talk.maemo.org/showpost.php?p=1325947&postcount=29 Jul 12 12:37:38 what the heck is /sys/class/backlight/ac*/bl_power? Jul 12 12:38:17 lol Jul 12 12:38:26 You won't believe me this Story: Jul 12 12:38:38 I have my IBM Thinkpad T42 since last September. Jul 12 12:38:44 Now i found the Thinklight. =^.^= Jul 12 12:38:57 Sorry for OT. Jul 12 12:39:11 ?? Jul 12 12:41:42 I said sorry. Jul 12 12:41:49 Weitermachen. ;-) Jul 12 12:42:13 I didn't get a single word of what you told Jul 12 12:42:53 It's ok. Jul 12 12:42:54 Just forget it. Jul 12 12:43:00 aha Jul 12 12:44:58 APic is... special Jul 12 12:45:08 Thanks for the Compliment. Jul 12 12:48:13 I'd still want to know what the heck is /sys/class/backlight/ac*/bl_power Jul 12 12:49:00 GL and hF finding it out. Jul 12 12:49:33 could you please stop flooding channel with gbberish resp noise? Jul 12 12:49:59 DocScrutinizer05: Maybe it's a part of the acpi backlight api ? Jul 12 12:50:55 quite possible Jul 12 12:52:10 I tried making backlight come up while screen locked, so finally I wrote 1 to where been 0. Sticky :-/ and worse: backlight went inoperable, had to reboot Jul 12 12:53:12 fun detail: after eboot I cehcked content of bl_power when screen lit, still 0 Jul 12 12:53:21 reboot, that is Jul 12 12:54:03 so I really wonder wtf it's doing. Selfdestruction? Jul 12 12:54:13 DocStrutinizer05: Sure Thing. Jul 12 12:55:11 I wrote 100 to it which also been sticky after 1, and then it didn't accept 1 anymore Jul 12 12:55:47 seems like it accepts one integers that are >$current_value Jul 12 12:56:02 s/ one/ only/ Jul 12 12:56:03 DocScrutinizer05 meant: seems like it accepts only integers that are >$current_value Jul 12 12:56:27 Sounds like a buggy backlight driver or something. Jul 12 12:56:49 :nod: sounds fishy Jul 12 12:57:42 blub blub Jul 12 12:57:49 lol Jul 12 12:58:02 this is where we get kicked i think Jul 12 12:58:16 https://wiki.ubuntu.com/Kernel/Debugging/Backlight "bl_power has no effect, what is it supposed to do?" :P Jul 12 12:58:18 DocScrutinizer05: are you sure it's a good idea to tinker with stuff like that? Jul 12 12:58:25 yes Jul 12 12:58:29 like, what if it actually damages things? Jul 12 12:58:40 then it's borkeded Jul 12 12:58:56 wouldn't it be better to figure that out *before* tinkering with it? Jul 12 12:59:06 meh, no fun Jul 12 13:00:39 actually when there was yet another self destruction button besides the known LP5523, CPUvoltage, CPUclockfreq, and nandtest, then damn sure I would wanna know Jul 12 13:00:47 Heh, if it's revealed in /sys, it shouldn't really break stuff :P ... That is, unless you touch the debug stuff. Then werewolves and sparkly vampires will come in the night and eat you :P Jul 12 13:01:00 nandtest isn't a self-destruct button Jul 12 13:01:06 it's not even close to being fatal Jul 12 13:01:16 depends Jul 12 13:01:18 the n900 is way too resilient, it's no fun Jul 12 13:01:40 you actually can wear out NAND with nandtest Jul 12 13:01:49 simly do a million cycles Jul 12 13:02:07 actually i don't know if you can safely invoke coldboot without knowing the hardware rev for sure Jul 12 13:02:13 N900 > * Jul 12 13:02:14 first 4 blocks already suffice iirc Jul 12 13:02:16 Wait. Jul 12 13:02:19 Neo900 > N900 Jul 12 13:02:21 ;-) Jul 12 13:02:34 omg dd is a self-destruct button because you can wear down hard drives with it Jul 12 13:02:40 lol Jul 12 13:02:58 actually Jul 12 13:03:00 hard drives? hardly Jul 12 13:03:03 do you need root access for O_DIRECT? Jul 12 13:03:09 SSD maybe, yes Jul 12 13:03:51 You could attach an USB HD Jul 12 13:04:00 W/ the genial USB-Power. Jul 12 13:04:02 I'd rather think hdparm is more of a self destruct for HDD Jul 12 13:04:06 mmh Jul 12 13:04:20 can't ssds last pretty much as long as hdds? Jul 12 13:04:22 modern ones, at least Jul 12 13:04:44 I don't think so Jul 12 13:05:04 under normal usage pattern maybe Jul 12 13:05:53 but not when you carefully craft some pattern that deliberately creates one page erase after the other and defeats wear leveling Jul 12 13:06:23 HDDs however might not survive even 10000 headpark events Jul 12 13:06:39 some HDDs do Jul 12 13:06:42 but for that you wanna use hdparm Jul 12 13:07:08 DocScrutinizer05: modern SSDs won't let you "defeat wear leveling" Jul 12 13:07:15 depends Jul 12 13:08:19 when you completely fill the critter, then it starts having a hard time Jul 12 13:08:59 sure it still can have a 10% shadow free pages Jul 12 13:09:38 there are other ways to completely nuke such peripherals: they all ahve firmware Jul 12 13:09:51 that's cheating, though Jul 12 13:10:09 kerio ftw Jul 12 13:10:22 you won't ask when your hw went south, whether it was cheating or not Jul 12 13:10:38 mh Jul 12 13:10:54 would you consider it a flaw of the hardware if it's possible to write a botched firmware and make it stop working? Jul 12 13:11:23 there's no such thing like cheating in destruction. Either dead or failed to kill Jul 12 13:11:29 like, without a hardware programmer Jul 12 13:11:45 actually no Jul 12 13:16:11 I like You very much, kerio. ♥ Jul 12 13:16:54 ty Jul 12 14:55:36 to finally answer my own question: https://www.kernel.org/doc/Documentation/ABI/stable/sysfs-class-backlight Jul 12 14:56:02 I can confirm that writing 1 to bl_power shuts off backlight Jul 12 14:56:12 perfect sense Jul 12 14:56:32 right Jul 12 14:56:33 yeah, Jul 12 14:57:23 I also can confirm that bl_backlight doesn't like 1 as value, it seems to stick Jul 12 14:58:31 maybe when you write 4 instead, which seems the correct value for "off" instead of 1, then *maybe* it would accept 0 after that to get re-enabled. I'm honestly too bored to even ponder trying that right now Jul 12 14:59:14 i want to know! Jul 12 14:59:17 who the heck wrote THIS pile of crap? Jul 12 15:01:03 kerio: I suggest echo 4 echo 0 echo 1 echo 0 echo 4 echo 0 Jul 12 15:01:56 0 - FB_BLANK_UNBLANK Jul 12 15:02:00 1 - FB_BLANK_NORMAL Jul 12 15:02:08 2 - FB_BLANK_VSYNC_SUSPEND Jul 12 15:02:12 3 - FB_BLANK_HSYNC_SUSPEND Jul 12 15:02:16 LOL Jul 12 15:02:16 4 - FB_BLANK_POWERDOWN Jul 12 15:02:21 ok Jul 12 15:02:22 from fb.h Jul 12 15:02:25 which is funny :D Jul 12 15:02:30 indeed Jul 12 15:03:19 why the heck is this /bl_power and not /bl_blank? Jul 12 15:03:31 or even /fb_blank Jul 12 15:03:40 bl_poweroff **** ENDING LOGGING AT Sun Jul 13 02:59:59 2014