#qi-hardware IRC log for Saturday, 2015-01-10

wpwraknot if you look at catch 22 as a means to prevent change :)00:00
wpwrak"you must be sane to be allowed to leave, but if you were sane, you wouldn't be here"00:00
DocScrutinizer05sounds about right01:37
DocScrutinizer05I wonder if Rigol scope runs sth like a VNCserver01:39
whitequarkwpwrak: not a ban06:09
whitequarkthe law makes it legal to forbid such people from getting a license06:09
whitequarki.e. it is at discretion of the doctor06:09
whitequarkwill likely be used for selective enforcement06:09
wpwrakso at the level of that law it sounds pretty harmless. i would kinda expect to find such things everywhere.06:12
wpwrakand yes, selective enforcement is of course an issue. again, everywhere ...06:13
whitequarkrussia has a history of punitive psychiatry...06:14
whitequarkso, no, I do not see it as harmless06:14
whitequarknot to mention the timing of the law06:14
wpwrak(harmless) i mean that you probably have a provision in any country that you can be declared unfit to drive for medical conditions06:16
wpwrakif it's not some blanket "for whatever reason" (to be defined by some rules governed my the entity in charge of administering licenses)06:16
whitequarkdepends on context. I mean, there always was a provision that you can't drive if you're blind or can't use the controls06:17
whitequarkhowever, adding a provision on "behavioral disorders" right in the middle of crackdown on lgbt is uh06:17
wpwrakat least insensitive :)06:18
mththe message seems to be that if you don't conform to the norm, they want you to keep your mouth shut about it06:18
whitequarkof course06:19
wpwrakluckily, in the free world, we regularly share our most intimate details, especially those of a sexual nature, with the world at large and only experience friendly acceptance and support ;-)06:21
wpwrakthe difference seems to be more along the lines of people in .ru staying in their closets because they fear bad things will happen while people in other countries stay in their closets because they fear bad things may happen06:23
wpwrak(for whatever the closet may be)06:23
mththere are different meanings to "sexual"... setting orientation to "gay" in a profile is not the same as posting bedroom videos06:29
mthbut in either case, I don't think it's any business of the state06:29
whitequarknote that the russian state does not actually care, it needs a scapegoat06:30
wpwraki'd agree with that. but it seems they just copied their list of problematic conditions from the WHO. they may not even have paid much attention to the details.06:30
wpwrak(or maybe it's just included by reference)06:31
mthmy guess is they just quoted the WHO as a "look, we didn't make this up" excuse06:31
whitequarkslipping something problematic hidden in a larger packet of law is a very common tactic06:31
wpwraki.e., i can easily see how you end up with this constellation if all you want is to say that driving licenses can be denied if 1) the applicant suffers a recognized medical condition and 2) a doctor says it's bad enough to be a reason to not let that person drive.06:32
whitequarkfrankly, I don't care whether the intent was malicious or not. the result *will* be used maliciously, as evident from, say, how it happens in USA06:33
whitequarkwhere similar laws *are* selectively enforced against even the same groups06:33
whitequarklack of knowledge of law does not absolve you of responsibility. then how lack of understanding of consequences absolves the lawmaker from it?06:34
wpwrakyeah, it's probably not the most clever law. i wonder if the lawmakers went "oops" after they realized the implications.06:34
wpwrakhehe ;-)06:34
whitequark"oops"? more like "yay"06:34
wpwrakhehe :)06:35
whitequarkI'm not being sarcastic, transphobic and proud of it MPs are abundant06:35
wpwrakperhaps a general lack of neutral viewpoint in those folks ?06:41
wpwrakof course, kant would approve :) "if you don't like someone's idiosyncrasies, then there should be a universal law against them"06:44
wpwrakbut i guess such things will even themselves out in a couple of generations if russia can avoid any major societal upset for that long06:50
wpwrakan aggressive stance against minorities is quite typical in societies that have a recent history of distress06:51
wpwrakwhere "minorities" is any kind of "outlier" perceived as negative. of course, the definition of "negative" can vary06:54
wpwrake.g., in russia it's probably quite okay to be a flamboyant atheist. now try that in the caliphate ...06:55
qi-bot[commit] Werner Almesberger: gui_canvas.c (key_press_event): make / rotate through packages as well (master) http://qi-hw.com/p/fped/5fdea0e08:33
kyakdespite all, the title couldn't be more yellow08:59
kyakof course, they had to mention putin08:59
kyakfrankly, the more western media/politics bash Putin _personally_, the more support he will get in Russia09:06
kyaki don't know if this is a pecularity of russians or if the same psyhologically will work in any other country09:07
larscand he probably knows that09:16
viricit's a bit similar with president Mas in Catalonia09:21
virica right wing man that never was pro-independence, until last elections. He managed to convince enough people, and Spanish media proclaim everywhere that "Mas is the messiah of independence"09:22
viricwhile Mas has always been the major ballast in the independence movement.09:23
viricAs the Spaniards attack Mas personally, he gets more support from the pro-independence, while he is the ballast towards independence.09:24
wpwrakhmm, principal monitor going into suspend all the time. interestingly, another monitor on the same card doesn't. but i think the card is still trying to tell me something there ... like, that's it's envious of the recent retirement of its long-time neighbour.09:24
kyakwpwrak: does it going suspend correlate with political discussions on this channel? :)09:27
wpwrakhmm. now that you mention it. maybe it got some separatist ideas ...09:27
kyakyour monitor is corrupted, you should through it away. Or, and your video card, too09:33
kyakshit, you just started loving them even more..!09:34
larschm, I'm thinking that the TCU thingy is a CPU bug16:47
larscor compiler16:48
larscThis is what the clockevent IRQ handler looks like https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/mips/jz4740/time.c#n4616:48
larscnotice the jz4740_timer_disable() if the mode is not periodic16:48
larscif I remove that the timer does not get disabled when the watchdog is enabled16:49
larscbut if I replace the disable() with a printk I never see the printk either16:49
larscputting some random other instructions before or after the disable also fixes the bug16:50
larscand the bug only happens if TIMER_CLOCKEVENT is 016:51
larscthe asm http://pastebin.com/3fgSC28h16:52
larscyep, that's a compiler bug16:59
larscIt always writes to JZ_REG_TIMER_ENABLE_CLEAR17:01
larscif mode is != CLOCK_EVT_MODE_PERIODIC17:01
larscit reads the current value from the register17:01
larscand then writes it back17:01
larscthing is the register is not readable17:01
larscand reading it will return garbage17:01
larscI guess it aliases with the WATCHDOG enable register17:02
larscsince the bug only happens when the watchdog is enabled17:02
larscso it reads back 1 and writes 1 to the register and hence disables the TCU017:02
larscscrew you gcc!17:03
wpwrakseems that it's gcc who's doing the screwing this time :)17:03
wpwrakand not on itself :)17:03
larscbasically what the asm translates to is17:05
larscval = readb(JZ_REG_TIMER_ENABLE_CLEAR); if (cd->mode == CLOCK_EVT_MODE_PERIODIC) val = 1; writeb(val, JZ_REG_TIMER_ENABLE_CLEAR);17:05
mthwhich version of GCC are you using?17:06
larsc'(OpenWrt/Linaro GCC 4.6-2012.12 r37492) 4.6.4'17:06
mthif you have a minimal test case, I can compile it on 4.9.1 to check whether the issue still exists17:10
mthactually 4.9.2 is out, but I haven't built that yet17:11
larschttp://pastebin.com/zH2enADr with -O217:19
mthlarsc: http://www.treewalker.org/temp/clockbug.s17:21
larscyour webserver does not respond17:23
larsclooks better17:25
larscno ldu17:25
larscGuess it's time to update the toolchain17:27
pcercueithat explains why you had problems with the watchdog driver that we didn't have17:34
larsc4.8 looks like it works17:52
larscoh no, started make -j and forgot the number17:54
larsclooks like it's not coming back to life18:05
wpwrakvictory ! low-cost solution: find a DVD-D cable to use my one spare DVI port instead of the failing VGA18:10
wpwrakerr, DVI-D18:11
wpwrakwhy can't they choose less freudian acronyms ?18:11
wpwrakthanks to the internet of things, we'll soon be able to have this: https://www.youtube.com/watch?v=DJklHwoYgBQ18:43
DocScrutinizer05pcercuei: ((neo-nazi demonstrations like in the east of Germany)) last PEGIDA in Dresden, iirc 16k. Today: 35k *against* PEGIDA22:31
DocScrutinizer05larsc: this doesn'T already spound like compiler bug. Rather I'd check my macros. They might implement bitset/bitclear on a register by readback of the register and (re)setting the bit in "Accumulator", then write back to register. That's how such stuff usually is implemented, either on opcode level or on macro level or on level of the uppermost sourcecode. And obviously that's BAD[TM] when the register is write-only22:42
DocScrutinizer05you regularly need a shadow variable mirroring the content of w/o register, do all the bit-toggling on this variable, and after each change write the var to the register22:44
eintopfwpwrak: I already thought about connecting my toilet flush into the IPv6 world22:44
DocScrutinizer05larsc: even when the compiler uses a read-modify-write opcode for the write-only register (which obviously is illegal and an according note would show up in manual about that w/o register), it's not the compiler's fault. You need to tell the compiler about the very special case that this particular register is not readable. By whatever means, worst case by using inline assembler22:55
eintopfthe use case would be: you sitting on the toilet (you know it will take longer) sometimes you need the toilet flush and turn around.23:14
eintopfWith smartphone in your hand you can push a big button :)23:14
eintopfinstead of turning around and search the flush key23:14
eintopfs/you need the toilet/you need to hit the toilet/23:15
--- Sun Jan 11 201500:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!