IRC log of #maemo for Sunday, 2012-10-07

kerio	isn't it much faster than the emmc or the usd?	00:00
DocScrutinizer05	there's not more than 256MB NAND anyway	00:00
DocScrutinizer05	and no, it's not THAT much faster	00:00
shamus	http://hackaday.com/2012/08/10/swapping-out-eee-pc-bga-chip-for-1-6-ghz-upgrade/	00:01
shamus	if you ask my thats just crazy	00:01
DocScrutinizer05	you still need to flash-erase whole pages, which kills your swap	00:01
DocScrutinizer05	performance-wise	00:01
teotwaki	shamus: what's the point, though?	00:01
DocScrutinizer05	and also regarding wear	00:01
shamus	im guessing they did it becasue they could	00:02
shamus	do they realy need a reasion	00:02
kerio	DocScrutinizer05: would it be easier to replace the whole SoC?	00:02
DocScrutinizer05	yes	00:02
DocScrutinizer05	waaaay easier	00:02
kerio	cool, let's do that then	00:02
teotwaki	has anyone run Maemo on any other SoC than the n900?	00:02
DocScrutinizer05	we already had that topic finalized some 4 weeks ago	00:02
DocScrutinizer05	we just need an OMAP3430 with better specs and better OneNAND POP	00:03
DocScrutinizer05	and somebody uttered "palm pre"	00:03
kerio	neat	00:03
DocScrutinizer05	as a possuble source for such a critter	00:03
DocScrutinizer05	but then, there's cert for xloader	00:04
DocScrutinizer05	so unless we get uBoot to properly boot maemo, we're probably still screwed	00:05
shamus	now what would be funny/sad is somone trying to get Maemo to run on a palm Vx lol	00:05
kerio	would a 3630 work fine too?	00:05
kerio	(palm pre 2)	00:05
DocScrutinizer05	prolly not	00:05
kerio	damn	00:05
DocScrutinizer05	I guess their electric specs don't match	00:06
kerio	the pre's is 500/600mhz like the n900's	00:06
DocScrutinizer05	uneducated guess	00:06
kerio	the pre plus has half a gb of ram, though	00:06
*** m4g0g has left #maemo		00:06
kerio	which would be ZOMGSOGOOD	00:06
DocScrutinizer05	kerio: download "datasheets" of omap3430 and 3630 from TI and check	00:07
DocScrutinizer05	if the pins are all the same names, odds are they could be just a drop-in replacement	00:08
DocScrutinizer05	if you find a devel board with chip socket, that claims to take both 3430 and 3630, then as well odds are good	00:09
*** ced117 has joined #maemo		00:09
*** ced117 has quit IRC		00:09
*** ced117 has joined #maemo		00:09
DocScrutinizer05	and particularly if you check kernel sources and there's no diff in noardconfig parameters for 3430 and 3630 boards, you could assume the peripherals are all the same, and thus maybe chips are compatible	00:11
DocScrutinizer05	board-config even	00:11
*** felipe` has joined #maemo		00:14
*** VDVsx has joined #maemo		00:15
*** int_ua has joined #maemo		00:16
*** bugzy has joined #maemo		00:17
*** netkat has quit IRC		00:18
*** VDVsx has quit IRC		00:20
*** Rqper has quit IRC		00:21
*** nox- has joined #maemo		00:23
jacekowski	freemangordon: only thing i don't like about it is the fact that there are thumb alternatives you are supposed to use instead	00:24
jacekowski	freemangordon: because swp + 2 switches have HUGE penalty	00:24
kerio	jacekowski: can't that code be written in C?	00:26
jacekowski	freemangordon: swp and swpb is deprecated in arm6	00:26
teotwaki	any code can be rewritten in C, doesn't mean it'll be faster.	00:27
jacekowski	kerio: not something like that	00:27
kerio	why not	00:27
kerio	?	00:27
*** bugzy has quit IRC		00:27
jacekowski	kerio: because it's thumb to arm switch and back to thumb for once	00:27
*** bugzy has joined #maemo		00:27
jacekowski	kerio: and it's using swp instruction and you can't make gcc emit that instruction	00:28
jacekowski	for two reasons	00:28
jacekowski	it's deprecated	00:28
jacekowski	and C has no understanding of mutexes	00:28
*** raandoom has joined #maemo		00:28
jacekowski	what SWP is doing is swapping two memory locations as one atomic operation	00:29
*** uen has joined #maemo		00:30
jacekowski	there are compiler specific ways of doing things	00:30
jacekowski	but those use memory barriers	00:31
jacekowski	which is the way it should have been done	00:31
jacekowski	GCC is pretty shit	00:32
kerio	:c	00:32
kerio	don't say that!	00:32
jacekowski	it's decent on x86	00:32
jacekowski	but other platforms, it makes me cry	00:32
kerio	i thought linaro wasn't that bad	00:32
jacekowski	linaro has not much to do with gcc	00:33
jacekowski	only platform where gcc generates acceptable code is x86	00:34
jacekowski	and even then it's pretty bad when compared with ICC	00:34
jacekowski	only "good" thing is, that MSVC is just as bad as gcc	00:34
kerio	heh	00:35
kerio	how does clang compare?	00:35
jacekowski	clang is bit strange thing to compare it directly	00:35
kerio	clang/llvm as a whole?	00:35
jacekowski	yeah	00:36
kerio	no, i mean, how does it compare to clang/llvm as a whole? :)	00:36
jacekowski	well, it has improved a lot	00:36
*** VDVsx has joined #maemo		00:36
*** Pali has quit IRC		00:37
kerio	which one? gcc or llvm?	00:37
jacekowski	llvm	00:37
kerio	but gcc is still better?	00:37
jacekowski	but thing is, llvm has plenty of overheads	00:37
kerio	i'm talking about arm	00:37
jacekowski	dunno about arm	00:37
jacekowski	to be honest, there is no good all round compiler for arm	00:38
kerio	:(	00:38
jacekowski	i've seen code generated by codesorcery	00:38
jacekowski	hmm	00:38
jacekowski	or codesomething	00:38
jacekowski	keil	00:38
jacekowski	and arm arm compiler	00:38
jacekowski	and all had their problems	00:38
jacekowski	though, arm compiler seemed to do least amount of fuck ups	00:39
*** VDVsx has quit IRC		00:41
freemangordon	jacekowski: ldrex / strex won;t do the job	00:41
freemangordon	I am using 4.7.2 and can use sync intrinstics	00:42
freemangordon	but they implement mutex by themselves, not MUTEX_SET	00:42
jacekowski	yeah, well, ideally you would replace whole mutex implementation fully in thumb	00:42
freemangordon	and having in mind berkeley deb uses swp/b even in it's latest code, I don't dare to rewrite mutex implementation in such a critical databse	00:43
jacekowski	but thing about bdb, in most cases there is no real need for mutexes in it	00:43
freemangordon	s/deb/db/	00:43
infobot	freemangordon meant: and having in mind berkeley db uses swp/b even in it's latest code, I don't dare to rewrite mutex implementation in such a critical databse	00:43
jacekowski	in mose use cases*	00:43
freemangordon	jacekowski: there is no penalty for going thumb/arm/thumb afaik	00:44
jacekowski	i'm pretty sure it causes pipeline flush	00:44
freemangordon	most probably. but it is the same wuth gcc intrinstics, they put data memory barriers, i've already tried them	00:45
freemangordon	and used IDA to check the code after that	00:45
jacekowski	but i think swp has higher penalty than mode switch	00:46
*** brzys has quit IRC		00:46
freemangordon	i know I can write in assembly, as memory barriers are for luti-cpu, but still	00:46
freemangordon	mode shwitch should not have any penalty besides what bx does	00:46
jacekowski	branch misprediction has pretty hefty penalty	00:47
jacekowski	and normal running system you get very very high hitrate on that	00:47
*** brzys has joined #maemo		00:48
jacekowski	but branch misprediction means that full pipeline is thrown out	00:48
jacekowski	and cortex-a8 has 13 stage pipeline	00:48
freemangordon	you mean bx will flush the pipeline?	00:48
freemangordon	but aiui there is no way to mispredict it, as it is always taken	00:49
jacekowski	BX can have condition	00:49
freemangordon	and abyway BTB is flushed on every context switch :D:D:D	00:49
*** ced117 has quit IRC		00:50
freemangordon	well, it can, but it has not in the current code	00:50
*** messerting has joined #maemo		00:50
jacekowski	ahm, you are talking about bx in your code	00:50
jacekowski	i though we are talking generally about bx	00:50
freemangordon	yeah	00:50
freemangordon	no	00:50
freemangordon	i am talking about my code	00:51
jacekowski	but in your case it will cause flush	00:51
freemangordon	will it?	00:51
freemangordon	because of swpb?	00:51
jacekowski	it's just predicting location	00:51
jacekowski	no, because instruction decoder will be running in ARM mode	00:51
freemangordon	aah, i see	00:51
jacekowski	so prefetched instructions will be decoded as arm	00:51
freemangordon	yes, you're right	00:52
jacekowski	and then you switch to thumb and only then processor realises that it's decoded them wrong	00:52
freemangordon	correct. damn, i'll have to rewrite that with lrdex/strex	00:52
jacekowski	but thing is, penalty on that is only part of it	00:53
*** OkropNick has quit IRC		00:53
jacekowski	swp is not penalty free as well	00:53
jacekowski	i'm not 100% sure if it's cached	00:53
*** int_ua has quit IRC		00:53
freemangordon	yeah, but it is a legacy from the origoinal code	00:53
freemangordon	jacekowski: i'm convinced :D	00:53
jacekowski	SWP causes processor wide bus lock	00:54
jacekowski	though, on n900 with single core it doesn't matter so much	00:54
freemangordon	no matter that code will work in it's current form, I hate when developers do it in the easy (read ineffective)way	00:55
jacekowski	but on multicore system it will stop all cores	00:55
freemangordon	yeah, i know that.hmm, actually i can rewrite that with only 3 instruction	00:56
jacekowski	and may cause cache lines to be flushed	00:56
freemangordon	ldrex,mov, strex	00:56
freemangordon	no memory barrier as we're on single core	00:57
*** grayhat has joined #maemo		00:57
jacekowski	on x86 there is memory snooping for that, but i'm not sure how ARM implements that kind of stuff exactly	00:57
freemangordon	and I am not trying to upstream that :D	00:57
grayhat	How can I reset n900 to factory default	00:57
jacekowski	grayhat: flash it	00:57
jacekowski	anyways	00:57
jacekowski	sleep to do	00:57
freemangordon	jacekowski: ldrex/strex	00:57
grayhat	jacekowski: i dont know how to do that	00:58
jacekowski	freemangordon: i'm talking hardware side	00:58
freemangordon	~flashing	00:58
infobot	methinks maemo-flashing is http://wiki.maemo.org/Updating_the_tablet_firmware	00:58
jacekowski	freemangordon: as in what happens if one core updates memory location that is cached in another core	00:58
freemangordon	jacekowski: aah, ok	00:58
freemangordon	there is cache protocol for that	00:58
grayhat	Thanks y'all	00:58
* freemangordon has forgotten the name		00:59
Cor-Ai	john?	00:59
jacekowski	freemangordon: on x86 all cores snoop memory accesses and will update or in some cases invalidate specific lines	00:59
jacekowski	anyways	01:00
jacekowski	sleep time	01:00
freemangordon	MESI	01:00
freemangordon	jacekowski: ^^^	01:00
freemangordon	night	01:00
*** niwakame\|away has joined #maemo		01:01
jacekowski	freemangordon: NUMA makes it a lot more fun than that	01:01
jacekowski	but anyways	01:01
jacekowski	sleep time	01:01
freemangordon	:)	01:01
*** ced117 has joined #maemo		01:04
*** messerting has quit IRC		01:09
*** croppa has joined #maemo		01:11
*** raandoom has quit IRC		01:18
*** NIN101 has quit IRC		01:18
*** hardaker has joined #maemo		01:19
*** lxp has quit IRC		01:21
*** lxp has joined #maemo		01:23
*** _rd has quit IRC		01:24
marainein	is uboot working or not? there seems to be a lot of talk about its problems	01:32
Skry	it works	01:34
*** brzys has quit IRC		01:35
*** brzys has joined #maemo		01:35
freemangordon	Skry: or better said - if it works, it works :D	01:36
*** rcg has quit IRC		01:38
*** int_ua has joined #maemo		01:40
DocScrutinizer05	or even better: compare your hw-rev with the list of those who are known to work	01:40
*** eijk_ has quit IRC		01:41
*** int_ua has quit IRC		01:42
merlin1991	aka 2101 works everything else is a gamble :D	01:45
marainein	2204	01:46
marainein	...swell	01:46
*** twb15 has joined #maemo		01:48
*** twb15 has quit IRC		01:51
marainein	so what do 2204 owners use for a boot loader?	01:54
Skry	freemangordon: so very true, fortunately I've personally never had any problems with it.	01:57
*** VDVsx has joined #maemo		02:02
*** VDVsx has quit IRC		02:06
Hurrian	Where do people get non-2101 units? I've personally handled 6 N900s, and all of them were r2101.	02:06
entitled	how do I check which I have?	02:07
entitled	got two	02:07
marainein	Hurrian, mine came from amazon	02:10
marainein	entitled, cat /proc/component_version	02:11
*** grayhat has quit IRC		02:11
entitled	2101 and... 2101	02:15
* nox- 2204		02:15
nox-	good i never tried uboot i guess...	02:16
merlin1991	I have a 2101 ad 2104 a 2204 and a 4th where I don't know the hw revision :D	02:16
marainein	nox-, you use any kind of multi boot?	02:16
merlin1991	nox-: actually u-boot prior to some very specific version works on all of them	02:16
nox-	marainein, no	02:17
nox-	merlin1991, oh...	02:17
merlin1991	ie the uboot-pr1.3 package in extras devel works on all device I own	02:17
nox-	and which one is broken?	02:18
merlin1991	the latest and greatest by pali :D	02:18
nox-	oh well...	02:18
entitled	/proc/cpuinfo... in features the one im using daily has thumbee and the one in reserve does not?	02:21
Hurrian	Nothing currently uses ThumbEE. It's just a CPU feature seen by Linux.	02:22
entitled	due to software I've installed 8 suppose, all should support thumbee	02:22
entitled	as it's the same armv7 revision	02:23
*** VDVsx has joined #maemo		02:23
Hurrian	Again, it /supports/ ThumbEE, but nothing uses it	02:24
entitled	anyone replaced the digitizer on their n900? are all those chinese ones as good as any other?	02:24
Hurrian	It's for faster Java, but I don't recall even Android devices using it.	02:24
Hurrian	entitled, the chinese digitizers are crap.	02:25
Hurrian	IIRC there was this one seller on eBay with surplus N900 ones, but I can't find it anymore.	02:25
entitled	friend's n900 has broken touch sensor, only works in part of the screen	02:25
entitled	and mine is pretty scratched up	02:26
*** at1as has quit IRC		02:26
Hurrian	ouch,	02:27
Hurrian	these things are quite squishy. They pop up on eBay sometimes, though	02:27
*** VDVsx has quit IRC		02:27
Hurrian	it's pretty clear which ones are Nokia and which ones are ripoffs	02:27
*** at1as has joined #maemo		02:31
entitled	he got it from uk vendor but with the price being so low it's most likely one of those chinese copies	02:31
Hurrian	these things are pretty pricey, 30-50 EUR a pop.	02:33
entitled	ye	02:33
*** VDVsx has joined #maemo		02:44
*** VDVsx has quit IRC		02:49
*** LjL has joined #maemo		02:50
DocScrutinizer05	re cpuinfo: what I find a tad concerning is the >>BogoMIPS : 249.96<< on PK46 device, while stock kernel gives >>BogoMIPS : 499.92<<	02:50
DocScrutinizer05	WTH?! the other one with stock kernel and CSSU-TH also has >>BogoMIPS : 249.96<<	02:52
DocScrutinizer05	btw all three 2101	02:52
*** teotwaki has quit IRC		03:04
DocScrutinizer05	btw, on a sidenote:	03:05
DocScrutinizer05	t900:~# uptime	03:05
DocScrutinizer05	02:03:09 up 190 days, 2:58, load average: 0.00, 0.00, 0.00	03:05
SpeedEvil	:-)	03:05
DocScrutinizer05	since I got other N900 for battery charging, I won't probably boot it that often anynore ;-D	03:07
DocScrutinizer05	so one thing I can confirm: WLAN driver, kernel, and sshd are rock solid ;-)	03:08
GeneralAntilles	I can hardly get through a day without having to reboot the N950 to get back lost modem connectivity.	03:09
DocScrutinizer05	no modem on t900	03:10
DocScrutinizer05	no SIM	03:10
DocScrutinizer05	or rather...	03:10
GeneralAntilles	I want a Jolla device. :<	03:10
SpeedEvil	''	03:10
SpeedEvil	probably	03:10
DocScrutinizer05	a sim that's not valid anymore, and thus not registering	03:10
Hurrian	DocScrutinizer05, measurements are taken at boot	03:12
DocScrutinizer05	dang, *#0000# shows "V 3.2010.02-8" on t900	03:13
DocScrutinizer05	Hurrian: :nod:	03:13
DocScrutinizer05	still doesn't explain it completely to me	03:14
Hurrian	If you edit /etc/pmconfig, you can get it to bump up.	03:14
Hurrian	I think for some devices, it ramps up the CPU OPP earlier, and for some devices later, resulting in 250 and 500 MHz BOGOMIPS	03:14
DocScrutinizer05	yeah, but for both systems they run same kernel and same /etc/pmconfig	03:15
DocScrutinizer05	I wonder if one of them booted on battery while the other on charger	03:16
*** hardaker has quit IRC		03:16
*** Darkchaos2 has quit IRC		03:20
DocScrutinizer05	MEH, CSSU-TH comes with PK, no?	03:21
DocScrutinizer05	KP	03:21
Hurrian	A rebranded KP (the one on the newest -thumb comes with 51r1)	03:22
DocScrutinizer05	so, after testing the thumb device with booting from battery, this boils down to "on KP: BogoMIPS : 249.96. On stock kernel: BogoMIPS : 499.92"	03:23
SpeedEvil	weird	03:24
SpeedEvil	does (say) md5sum /dev/zero go the same speed	03:24
DocScrutinizer05	hmm	03:24
DocScrutinizer05	how would I test speed of md5sum /dev/zero ;-)	03:25
SpeedEvil	dd if=/sev/zero bs=1024 count=102400	03:25
SpeedEvil	OE something	03:26
*** VDVsx has joined #maemo		03:27
*** VDVsx has quit IRC		03:27
DocScrutinizer05	IroN900:~# time md5sum /dev/mtd4ro	03:28
DocScrutinizer05	b23b5d09162b92c0284923a7f628d2a5 /dev/mtd4ro	03:28
DocScrutinizer05	real 0m0.154s	03:28
DocScrutinizer05	t900:~# time md5sum /dev/mtd4ro	03:28
DocScrutinizer05	b23b5d09162b92c0284923a7f628d2a5 /dev/mtd4ro	03:28
DocScrutinizer05	real 0m 0.22s	03:28
SpeedEvil	add another couple of zeros, I guess	03:29
DocScrutinizer05	grr, scratch that	03:29
SpeedEvil	it's probably not significant those	03:29
SpeedEvil	then	03:29
SpeedEvil	did it have thumb patches?	03:30
Skry	real 0m0.317	03:31
DocScrutinizer05	busybox version seems to differ	03:31
DocScrutinizer05	Skry: you need to call this at least 4..5 times	03:32
DocScrutinizer05	first time is moot	03:32
DocScrutinizer05	format of my md5sum output differs between the two devices	03:32
Hurrian	Isn't MTD4 full of zeros?	03:33
Hurrian	(and roughly 2MB)	03:33
SpeedEvil	/dev/zero is a pure CPU test	03:33
DocScrutinizer05	IroN900:~# time busybox md5sum /dev/mtd4ro	03:34
DocScrutinizer05	b23b5d09162b92c0284923a7f628d2a5 /dev/mtd4ro	03:34
DocScrutinizer05	(empty line)	03:34
DocScrutinizer05	real 0m0.173s (.174 .170 .173)	03:34
DocScrutinizer05	t900:~# time busybox md5sum /dev/mtd4ro	03:35
DocScrutinizer05	b23b5d09162b92c0284923a7f628d2a5 /dev/mtd4ro	03:35
DocScrutinizer05	real 0m 0.24s (.20 .24 .23 .23 .19)	03:35
DocScrutinizer05	BusyBox v1.10.2 (Debian 3:1.10.2.legal-1osso26+0m5) vs BusyBox v1.10.2 (Debian 3:1.10.2.legal-1osso31+0cssu0)	03:36
*** brzys has quit IRC		03:37
DocScrutinizer05	so the one busybox (0cssu0, CSSU-T) is faster, has an empty line, and has 3 decimals	03:38
DocScrutinizer05	while the other is significantly slower, has no empty line and only 2 decimals	03:39
DocScrutinizer05	:-/	03:39
GGon	DocScrutinizer05: ever think of making a completely new distro for the n900?	03:40
GGon	heh	03:40
GGon	using the kde based stuff? or is the n900 too slow to handle it	03:40
DocScrutinizer05	I don't think kde has many apps that I'd like to use on N900	03:41
Skry	GGon: its slow but usable after some tweaking	03:42
DocScrutinizer05	or rather, under KDE I'd miss a lot of apps which aren't available there	03:42
GGon	like what?	03:42
GGon	telepathy? :)	03:42
SpeedEvil	dialer, ...	03:42
GGon	SpeedEvil: lol	03:43
GGon	still no easy way to access the modem for voice ?	03:43
*** marainein has quit IRC		03:43
DocScrutinizer05	no	03:43
GGon	wow	03:43
GGon	i figured that would have been figured out by now	03:43
GGon	then again i haven't used my n900 for voice comm in forever since the modem is broken anyways heh	03:44
GGon	i've been using it like a glorified nit	03:44
Hurrian	GGon, you just need to setup ofono.	03:44
GGon	Hurrian: i thought they were still having problems with ofono	03:45
*** M4rtinK has quit IRC		03:45
*** muelli has joined #maemo		03:45
GGon	with regards to the sound or something	03:45
Hurrian	It works for voice calls and texts and data, if you're using the 2.6.37-N900 Adaptation kernel.	03:45
Hurrian	There's no dialer app, so if you're going to really use it for that purpose then don't bother.	03:45
GGon	is a dialer app that difficult to make in QT?	03:45
Hurrian	You may want to try building the Nemo dialer though. It's in Qt.	03:46
GGon	just thought it would be interesting to see something running activ	03:46
GGon	that's what the kde touchy ui is called isn't it?	03:46
*** at1as has quit IRC		03:47
GGon	i haven't seen a real tablet or phone use it tho... so it must not exist :)	03:47
GGon	jk. i know people got it to half ass run on a transformer but i know i couldn't figure out how to get it on mine. or there wasn't an easy image to just flash to it in order to get it working	03:47
Hurrian	Plasma Active?	03:47
Hurrian	It runs on the N900.	03:48
*** VDVsx has joined #maemo		03:48
*** VDVsx has quit IRC		03:48
*** at1as has joined #maemo		03:50
*** VDVsx has joined #maemo		03:54
*** alex______ has joined #maemo		03:58
alex______	someone?	03:58
Cor-Ai	nope, he's not here right now!	03:59
*** teotwaki has joined #maemo		04:02
*** edheldil has quit IRC		04:03
alex______	i have o problem with u-boot	04:06
alex______	restart looping	04:06
alex______	no usb cable at hand	04:06
alex______	help	04:07
*** nox- has quit IRC		04:07
*** stardiviner has joined #maemo		04:07
*** GGon has quit IRC		04:07
*** teotwaki has quit IRC		04:14
*** GGon has joined #maemo		04:16
*** sLumPia has quit IRC		04:34
*** stardiviner has quit IRC		04:36
ketas	foo... i got n9	04:38
ketas	fuck	04:43
ketas	really	04:43
*** MacDrunk has joined #maemo		04:44
*** flo_lap has joined #maemo		04:45
*** VDVsx has quit IRC		04:45
*** muelli has quit IRC		04:45
entitled	that was weird	04:47
*** robink has quit IRC		04:47
entitled	n900 was fully charged, I turned it off and then I get this weird bzzzzz sound and it blinked purple.	04:48
*** florian has quit IRC		04:48
entitled	was connected to the charger at that moment	04:48
entitled	seems to be ok tho	04:48
* ketas curses at the security features		04:48
SpeedEvil	flickered purple?	04:48
SpeedEvil	ketas++	04:49
ketas	i got used n9 and accidentally runned track and protect... now WHAT	04:50
*** stardiviner has joined #maemo		04:50
SpeedEvil	no idea what that is.	04:50
SpeedEvil	I think t calls the ninjas.	04:50
ketas	maybe should use #meego ?	04:51
SpeedEvil	#harmattam	04:51
SpeedEvil	n	04:51
SpeedEvil	is more on topic for n9	04:52
*** robbiethe1st has joined #maemo		04:54
*** _xnt14 has quit IRC		04:55
*** _xnt14 has joined #maemo		04:57
*** uen\| has joined #maemo		05:06
*** uen has quit IRC		05:08
*** uen\| is now known as uen		05:08
*** at1as has quit IRC		05:13
*** at1as has joined #maemo		05:15
*** VDVsx has joined #maemo		05:16
*** VDVsx has quit IRC		05:16
*** dos1 has quit IRC		05:22
*** MacDrunk has left #maemo		05:24
*** at1as has quit IRC		05:27
*** at1as has joined #maemo		05:28
*** radic has quit IRC		05:32
*** radic_ has joined #maemo		05:32
*** at1as has quit IRC		05:33
*** bugzy has quit IRC		05:37
*** VDVsx has joined #maemo		05:37
*** VDVsx has quit IRC		05:37
*** iluminator105 has joined #maemo		05:47
*** robink_ has joined #maemo		05:48
iluminator105	is there an auto installer for thumb	05:53
*** marainein has joined #maemo		05:59
*** DocScrutinizer05 has quit IRC		06:04
*** DocScrutinizer05 has joined #maemo		06:04
marainein	hey, this may be more a linux/ubuntu question than a maemo one, but when i disconnect my n900's usb connection then plug it back in, it doesn't get detected again by the computer, i have to reboot it for the computer to detect the phone...is there some way to fix that?	06:18
*** Gh0sty has quit IRC		06:19
robbiethe1st	marainein, pull up a terminal on Ubuntu	06:19
robbiethe1st	run 'dmesg \| tail'	06:19
robbiethe1st	then unplug/replug, do it again	06:20
*** VDVsx has joined #maemo		06:20
*** VDVsx has quit IRC		06:20
*** Gh0sty has joined #maemo		06:20
marainein	robbiethe1st, i'm actually not seeing anything new in dmesg after i do that	06:22
marainein	nor is it showing up on lsusb	06:22
robbiethe1st	That's interesting	06:22
robbiethe1st	I wonder... check the same on the N900	06:23
marainein	it's giving me 'tw14030_usb tw14030_udb: HW_CONDITIONS 0xd0/208; link 2'	06:30
marainein	and the line before it, similar except for link 1	06:30
robbiethe1st	hm	06:32
robbiethe1st	I think that's good...	06:32
robbiethe1st	I'd try updating Ubuntu/loading a different kernel	06:32
marainein	yeah, i suspect it's a linux usb bug	06:33
robbiethe1st	version-specific bug evidently; never had the issue with Debian	06:34
*** dhbiker has quit IRC		06:37
*** erstazi has joined #maemo		06:45
*** dhbiker has joined #maemo		07:01
*** VDVsx has joined #maemo		07:02
*** VDVsx has quit IRC		07:02
*** VDVsx has joined #maemo		07:24
*** FlameReaper has joined #maemo		07:26
*** VDVsx has quit IRC		07:29
*** FlameReaper has quit IRC		07:36
*** dhbiker has quit IRC		07:37
*** FlameReaper has joined #maemo		07:37
*** iluminator105 has quit IRC		07:44
*** VDVsx has joined #maemo		07:45
*** at1as has joined #maemo		07:46
*** iluminator105 has joined #maemo		07:46
*** VDVsx has quit IRC		07:50
*** at1as has quit IRC		07:51
*** dhbiker has joined #maemo		07:51
*** at1as has joined #maemo		07:55
*** at1as has quit IRC		08:01
*** dhbiker has quit IRC		08:01
*** dhbiker has joined #maemo		08:02
*** dhbiker has quit IRC		08:03
*** VDVsx has joined #maemo		08:07
*** VDVsx has quit IRC		08:11
*** at1as has joined #maemo		08:18
*** at1as has quit IRC		08:23
*** VDVsx has joined #maemo		08:28
*** VDVsx has quit IRC		08:32
*** _PanzerSajt has joined #maemo		08:40
*** at1as has joined #maemo		08:47
*** muelli has joined #maemo		08:49
*** VDVsx has joined #maemo		08:49
*** beford has quit IRC		08:50
*** VDVsx has quit IRC		08:54
*** iluminator105 has quit IRC		08:54
*** VDVsx has joined #maemo		09:10
*** VDVsx has quit IRC		09:14
*** Pali has joined #maemo		09:23
*** jevin has quit IRC		09:29
*** DHR has quit IRC		09:31
*** muelli has quit IRC		09:34
*** alex______ has quit IRC		09:43
*** beford has joined #maemo		09:43
*** VDVsx has joined #maemo		09:53
*** VDVsx has quit IRC		09:53
_PanzerSajt	Pali, Hy! Is there any stable and safe u-boot?	10:02
_PanzerSajt	I have reflash my N900 and I don't want to ruin it with not working u-boot	10:03
*** udovdh has quit IRC		10:04
*** Pali has quit IRC		10:05
*** Pali has joined #maemo		10:07
robbiethe1st	You can always reflash no matter how badly anything screws it up	10:07
robbiethe1st	u0boot fails? reflash	10:07
robbiethe1st	It's got the flashing stub in the bootloader in a different rom chip	10:08
kerio	yeah, you can't really brick the n900, apart for hosing CAL	10:09
kerio	robbiethe1st: hi	10:09
robbiethe1st	And even then...	10:09
kerio	robbiethe1st: BM should really ignore the installed /etc/passwd	10:09
robbiethe1st	Why?	10:10
kerio	because my login shell is bash	10:10
kerio	and there's no bash in the BM root	10:10
robbiethe1st	I think my next version or patch which I've nit released just adds a symlink to /bin/sh...	10:11
*** Free-MG has joined #maemo		10:11
robbiethe1st	But I can't really do that either, see how I'm entering terminal mode?	10:11
robbiethe1st	Basically just letting the system do it	10:11
kerio	robbiethe1st: doesn't work	10:11
kerio	busybox won't do shit if ran as "bash"	10:11
kerio	bash: applet not found	10:12
robbiethe1st	symlink /bin/bash to /bin/ash which is symlinked to /bin/busybox?	10:12
kerio	yeah, the "inner" symlink doesn't exist	10:12
robbiethe1st	darn	10:12
kerio	according to busybox	10:12
kerio	you could use a shell script, i guess	10:13
robbiethe1st	hm, yeah	10:13
kerio	or ship an /etc/passwd >:(	10:13
_PanzerSajt	kerio, yes I know that we can't brick n900 but I don't want to reflash again and install all the softwares.	10:13
robbiethe1st	I just thought, however, that what I could do is check for the existance of that password-enable file	10:13
robbiethe1st	the one I use for BM password lock	10:13
robbiethe1st	if it exists, load the root's /etc/passwd, otherwise use a stock one with a default password	10:14
kerio	_PanzerSajt: that's where BM comes into play	10:14
_PanzerSajt	can you explain it to me?	10:14
*** VDVsx has joined #maemo		10:14
kerio	robbiethe1st: or just require a hashed password in it	10:15
robbiethe1st	uh...?	10:15
kerio	robbiethe1st: if you have a password there, you don't need the password from /etc/passwd	10:16
robbiethe1st	But I'm using the official login app or w/e	10:17
robbiethe1st	so I think it has to exist at /etc/passwd	10:17
kerio	hm	10:17
kerio	still, i have to use bootmenu's usb networking to ssh because i can't login in BM :(	10:18
kerio	DO YOU KNOW HOW TOUGH MY LIFE IS	10:18
kerio	_PanzerSajt: full-ish on-device backups	10:19
*** VDVsx has quit IRC		10:19
_PanzerSajt	Hmm that's nice	10:19
_PanzerSajt	how can I perform a full backup and restore?	10:19
kerio	i should probably backup CAL	10:19
robbiethe1st	Change your damn login shell to /bin/sh then?	10:19
kerio	robbiethe1st: no u	10:19
robbiethe1st	and just use bash when you need it	10:19
robbiethe1st	...Or edit bm's loader to add bash	10:19
kerio	_PanzerSajt: install BM, reboot in BM, follow semigraphical UI	10:19
robbiethe1st	Semigraphical is a good way to put it	10:20
*** Darkchaos has joined #maemo		10:20
kerio	robbiethe1st: how do i backup CAL?	10:23
robbiethe1st	I have absolutely no clue	10:23
robbiethe1st	I've never done it, or messed with it	10:23
kerio	ok, nanddumped it :)	10:26
kerio	robbiethe1st: add it in "extra functions"	10:26
robbiethe1st	paste the commands into the BM topic	10:27
robbiethe1st	and I'll read it tomorrow, or try to anyway	10:28
*** robbiethe1st has quit IRC		10:28
*** sasquatch has quit IRC		10:29
*** sasquatch has joined #maemo		10:29
jacekowski	kerio: well, recent events have confirmed that cal is not very important	10:29
*** Pali has quit IRC		10:30
Hurrian	jacekowski, someone tried wiping their cal?	10:31
jacekowski	yes	10:31
Hurrian	Interesting.	10:31
jacekowski	someone did wipe all of nand	10:31
Hurrian	Did it just rebuild?	10:31
jacekowski	no	10:31
jacekowski	but everything works without it	10:31
Hurrian	Even phone functions?	10:31
jacekowski	yes	10:31
Hurrian	Interesting.	10:32
jacekowski	rapuyama is doing phone	10:32
jacekowski	and there is separate cal in there	10:32
jacekowski	that can't be touched	10:32
jacekowski	cal in main onenand can be wiped	10:33
*** _PanzerSajt has quit IRC		10:33
jacekowski	and data there is not very important	10:33
jacekowski	fmtxd has region settings stored there	10:33
jacekowski	and bt mac	10:33
jacekowski	i'm not sure what would happen on fully wiped nand	10:34
jacekowski	as in, partition table gone as well	10:34
*** VDVsx has joined #maemo		10:36
Hurrian	well, you need to use the UBI tool to create them in the first place.	10:36
Hurrian	how you're going to access that is the question - Linux exposes /dev/mtdN, where N is a partition.	10:36
Hurrian	ubinize and burn?	10:37
*** valdyn has quit IRC		10:39
*** VDVsx has quit IRC		10:40
*** FlameReaper has quit IRC		10:44
*** timo^ has joined #maemo		10:45
*** timo^ has quit IRC		10:45
*** Timo has joined #maemo		10:45
Timo	Hi there	10:45
*** _rd has joined #maemo		10:46
Timo	I've installed Pali's Uboot (with menu) and the Kernel Power for multiboot. I ran the u-boot-update-bootmenu too of course. When I rebooted and chose the kernel power (same happens with the stock kernel), Uboot hangs on 'Starting Kernel'. How do I fix that? (I've got a working phone atm because of a reflash of the stock kernel (flash-only=kernel) :)	10:49
kerio	jacekowski: the partition table is written on the nand? :o	10:50
kerio	i thought it was just a parameter passed to the kernel by NOLO	10:50
kerio	jacekowski: also, you lose info regarding your hw rev and BT doesn't work properly	10:51
kerio	DocScrutinizer05: do you reckon nanddump/nandwrite can be used to backup/restore CAL?	10:55
kerio	also, why does on-device kernel flashing use flasher instead of nandwrite?	10:56
*** VDVsx has joined #maemo		10:57
*** VDVsx has quit IRC		11:01
*** rcg has joined #maemo		11:11
ShadowJK_	was nandwrite the one that doesn't deal that well wih badblocks?	11:11
*** valdyn has joined #maemo		11:11
*** Timo has quit IRC		11:12
*** brzys has joined #maemo		11:14
kerio	i thought it did	11:14
kerio	dd on mtdblk is the one that doesn't deal with badblocks	11:15
*** VDVsx has joined #maemo		11:18
*** OkropNick has joined #maemo		11:20
kerio	do we know how CAL works?	11:22
*** VDVsx has quit IRC		11:23
*** raandoom has joined #maemo		11:29
*** VDVsx has joined #maemo		11:39
*** VDVsx has quit IRC		11:44
*** eijk has joined #maemo		11:46
*** raandoom has quit IRC		11:52
*** sq-one has joined #maemo		11:54
*** FlameReaper has joined #maemo		11:58
*** VDVsx has joined #maemo		12:01
*** M4rtinK has joined #maemo		12:01
*** VDVsx has quit IRC		12:05
*** Pali has joined #maemo		12:13
*** dos1 has joined #maemo		12:19
kerio	does anyone have a bit of experience with pppd?	12:21
*** VDVsx has joined #maemo		12:22
*** NIN101 has joined #maemo		12:22
*** lbt has quit IRC		12:23
Shapeshifter	Hi. So this morning my alarm didn't ring and I found my n900 in some sort of frozen state. The led is blinking as if I received an IM message and if I slide or double-tap power, the screen turns on but it remains black. I had this once before, and I'm sure replacing the battery fixes it. But I would like to know what is going on. So I'm able to ssh onto it, luckily, and I see that a process called sgx_misr is using up all CPU.	12:24
Shapeshifter	Well and I guess I found the unfixed bug report from 2010... just great.	12:25
Shapeshifter	https://bugs.maemo.org/show_bug.cgi?id=9150	12:25
povbot	Bug 9150: Device doesn't respond via UI. syslog reports HWRecoveryResetSGX: SGX Hardware Recovery triggered, sgx_misr eating all CPU	12:25
*** VDVsx has quit IRC		12:27
*** lbt has joined #maemo		12:29
*** freemangordon has quit IRC		12:31
*** dos1 has quit IRC		12:31
Shapeshifter	PowerVR is such shit.	12:32
*** dos1 has joined #maemo		12:33
DocScrutinizer05	kerio: I suggested that several times	12:40
kerio	suggested what?	12:40
DocScrutinizer05	nanddump/write	12:40
kerio	i see	12:41
DocScrutinizer05	you even find a nanddump'ed image of my CAL on my server	12:41
*** valdyn has quit IRC		12:43
DocScrutinizer05	don't forget you'll probably also need nanderase, since iirc nandwrite can't clean blocks	12:43
*** VDVsx has joined #maemo		12:43
DocScrutinizer05	but better you read the manpages etc before you start experiments	12:44
DocScrutinizer05	there's also mtd_debug iirc, maybe also worth giving it a look	12:44
DocScrutinizer05	http://processors.wiki.ti.com/index.php/Mtdutils	12:48
*** VDVsx has quit IRC		12:48
*** valdyn has joined #maemo		12:49
DocScrutinizer05	http://stackoverflow.com/questions/11279473/how-to-correctly-nandwrite-a-nanddumped-dump-with-oob	12:50
kerio	DocScrutinizer05: we don't need oob, do we	12:50
DocScrutinizer05	http://www.linux-mtd.infradead.org/doc/general.html	12:50
DocScrutinizer05	oob is out of band data iircd, it marks bad blocks or whatever	12:51
DocScrutinizer05	>>Often people consider mtdblock as general FTL layer and try to use block-based file systems on top of bare flashes using mtdblock. This is wrong in most cases. In other words, please, do not use mtdblock unless you know exactly what you are doing.<<	12:53
*** freemangordon has joined #maemo		12:53
kerio	DocScrutinizer05: backupmenu 1? :D	12:55
DocScrutinizer05	yoh	12:56
*** antero has joined #maemo		12:59
SpeedEvil	Shapeshifter: I wonder if that isn't sometimes a hardware issue.	13:00
*** VDVsx has joined #maemo		13:05
*** Hurrian has quit IRC		13:05
DocScrutinizer05	Shapeshifter: is alarmed using the tags of alarmd events?	13:07
*** flo_lap has quit IRC		13:09
*** VDVsx has quit IRC		13:09
*** bigmeow has quit IRC		13:10
*** bigmeow has joined #maemo		13:16
*** shanttu has joined #maemo		13:22
*** dos1 has quit IRC		13:22
*** lbt has quit IRC		13:23
*** VDVsx has joined #maemo		13:26
Pali	I found another bug in maemo package interface	13:27
Pali	See this: http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/wargus/2.2.7-1maemo1~fremantle/	13:27
Pali	DocScrutinizer05 ^^^	13:27
DocScrutinizer05	?	13:28
Pali	I can vote 4 times for one package :D	13:28
DocScrutinizer05	somebody already mentioned that ;-P	13:28
kerio	but that would be cheeeeeeeeeating	13:29
Pali	but now I know how to do it :)	13:29
Pali	just open 4 tabs in browser with that vote page	13:29
*** lbt has joined #maemo		13:29
*** lbt has quit IRC		13:29
*** lbt has joined #maemo		13:29
Pali	then prepare ...	13:29
DocScrutinizer05	odds are cheated packages get kicked out of extras repo and you have to start all over again ;-D	13:29
Pali	and then very quickly click on button :D	13:30
*** livelace has joined #maemo		13:30
*** VDVsx has quit IRC		13:30
Pali	and make race condition	13:30
DocScrutinizer05	haha, they forgot to include mutex to avoid race condition	13:30
Pali	funny bug :D	13:31
DocScrutinizer05	excusable bug	13:31
Pali	who is author of that maemo system?	13:31
*** bigmeow has quit IRC		13:31
DocScrutinizer05	web developers usually have no clue about race conditions and re-entrant coding	13:31
DocScrutinizer05	adding comments has same prblem, only there it's enough to click the submit button again after 10s of waiting for anything to happen	13:32
DocScrutinizer05	almost all of my comments are double	13:32
DocScrutinizer05	some triple	13:32
DocScrutinizer05	;-P	13:33
Pali	DocScrutinizer05, now you know reason :-)	13:33
DocScrutinizer05	I always knew the reason	13:33
DocScrutinizer05	I just wonder about shortest fix for all that	13:34
DocScrutinizer05	probably something along system("uniq ...")	13:34
*** sq-one has quit IRC		13:35
DocScrutinizer05	or alternatively (if that's any kind of rdb) define some cols as nodupkey	13:35
*** VDVsx has joined #maemo		13:35
DocScrutinizer05	nfc who's maintainer for that web frontend	13:37
*** Hurrian has joined #maemo		13:37
DocScrutinizer05	in 3 months, it should be clear whom to ask though	13:37
Hurrian	Lesson learned: don't resize MyDocs partition when using u-boot.	13:38
DocScrutinizer05	and I'm not any sad it's not me now, whom you'll ask about all that	13:38
Pali	btw what is state of maemo.org?	13:38
*** bigmeow has joined #maemo		13:38
Pali	nokia will turn off servers?	13:38
DocScrutinizer05	shrug likely	13:38
Hurrian	Pali, the funding ends on the 31st.	13:38
Hurrian	so yeah, unless HF moves the site, it gets cut.	13:39
DocScrutinizer05	gossip is they will stop paying services with new year	13:39
DocScrutinizer05	all that is worst case	13:39
Hurrian	Gossip? Didn't it go more like "all funding will cease on the 31st"	13:39
DocScrutinizer05	though Nokia's worst cases usually are worse than worst case scenarios	13:40
Pali	is somebody going to backup maemo.org?	13:40
DocScrutinizer05	Hurrian: nah, it's "Nokia already paid for servers and maintenance til end of 2012"	13:41
Hurrian	Pali, it's a ~10GB backup at the very least.	13:41
DocScrutinizer05	Hurrian: and there's still no definitive statement they will cease to do that in 2013-01-01	13:41
DocScrutinizer05	but it's very likely they will	13:41
Hurrian	doesn't seem likely they will - they already fired all the Linux people.	13:41
Hurrian	unless as said in that thread, Nokia will do a transition period	13:42
Pali	is nokia going to delete also maemo.gitorious.org ?	13:42
DocScrutinizer05	but it's very likely they will cease to pay	13:42
DocScrutinizer05	no way	13:42
DocScrutinizer05	Pali: ^^^	13:42
DocScrutinizer05	Hurrian: maemo.org 10GB? haha!	13:43
Hurrian	10GB for the forum	13:43
DocScrutinizer05	maemo.org incl subdomains is more likely to be something like 10TB	13:43
Hurrian	and, if we only keep newest copy of packages, another 10GB	13:44
Hurrian	plus the Nemo/Mer hosting stuff	13:44
DocScrutinizer05	ooh, tmo. Well that's a completely segregate topic, really	13:44
Hurrian	and other obscure things hosted on M.O	13:44
*** brzys has quit IRC		13:45
DocScrutinizer05	tmo is still actually a subdomain of itt	13:45
Sicelo	[ /me has just now been reminded how badly IMAP-IDLE support is needed in modest :( ]	13:46
DocScrutinizer05	technically	13:46
Hurrian	nah, the A record's pointed to whatever ITT points to.	13:46
Pali	and https://projects.maemo.org/ ? :D	13:46
Hurrian	Pali, don't forget garage	13:46
Pali	talk.maemo.org. 41377 IN CNAME forums.internettablettalk.com.	13:46
Pali	forums.internettablettalk.com. 13684 IN A 74.86.202.247	13:46
Pali	funny :-)	13:46
DocScrutinizer05	http://www.internettablettalk.com/ more funny	13:48
DocScrutinizer05	maximum funny http://maemotalk.com/	13:50
*** teotwaki has joined #maemo		13:51
*** brzys has joined #maemo		13:53
*** lbt has quit IRC		13:53
kerio	Hurrian: merlin1991 already has a mirror of the latest versions of packages in extras-devel	13:53
*** teotwaki has quit IRC		13:54
Hurrian	kerio, that's nice to know	13:56
kerio	which is seriously good for apt	13:56
kerio	fun fact: as of now, nokia can't ship an OTA update	13:56
kerio	not even if they wanted to	13:57
*** croppa has quit IRC		13:57
Hurrian	kerio, they broke their SSL cert?	13:58
kerio	Hurrian: the gpg key for the packages	13:58
Hurrian	it's expired/revoked?	13:58
*** lbt has joined #maemo		13:58
*** lbt has quit IRC		13:58
*** lbt has joined #maemo		13:58
kerio	it's expired, and HAM requires a working signing key to identify packages coming from a trusted source	13:58
Hurrian	whelp.	13:59
Hurrian	Doubt it bothers Nokia though, support has probably already long dropped off the radar.	13:59
freemangordon	jacekowski: ping	14:00
freemangordon	DocScrutinizer05: are you awake?	14:00
SpeedEvil	kerio: oops	14:00
DocScrutinizer05	freemangordon: kinda	14:01
freemangordon	http://pastebin.com/1QtcHkS7	14:01
DocScrutinizer05	30%	14:01
freemangordon	should be enough for 6 instructions :D	14:01
DocScrutinizer05	I already followed chat between you and jacekowski yesterday. And I have to say that's actually stuff way beyond my expertise	14:02
freemangordon	aah, ok :)	14:02
*** livelace has quit IRC		14:03
DocScrutinizer05	just a general comment on mutex etc: you need to check those very carefully, since you can't really test in RL, the purpose of mutexes is to avoid race conditions that are extremely hard to prove as being dealt with in a reliable manner	14:05
*** maybeHere has joined #maemo		14:05
DocScrutinizer05	IOW you can't prove there's no open window for a race left	14:05
DocScrutinizer05	except on analytic src review level	14:06
Pali	freemangordon, hi, when you have time, try look at uboot. I cannot do it, because on my n900 is uboot working...	14:06
DocScrutinizer05	any bug that slipped your attention will show up at $random time later	14:06
freemangordon	DocScrutinizer05: I am aware of that, that is why i wan't someone else to review the code.	14:07
freemangordon	Pali: I will	14:07
DocScrutinizer05	yeah, very reasonable idea	14:07
*** maybeArgh has quit IRC		14:09
DocScrutinizer05	freemangordon: another rather noobish comment: I usually see irqs=lock_irq() and restore_irq(irqs) bracketing any mutex code, so you kinda make it atomic	14:10
*** teotwaki has joined #maemo		14:10
*** teotwaki has joined #maemo		14:10
freemangordon	yeah, ldrex/strex are kinda atomic operations	14:10
freemangordon	they put a memory observer to the address in question	14:11
DocScrutinizer05	well, so you say the rest of code is reentrant?	14:11
DocScrutinizer05	or at least thread-safe	14:11
freemangordon	reenterant? NFC, I didn't write it :)	14:11
DocScrutinizer05	consider something else messing with virtually anything it likes, at any arbitrary point during execution of your code. Will it cause trouble?	14:12
freemangordon	but it should be thread safe, after all this is DB	14:12
freemangordon	no	14:13
*** VDVsx has quit IRC		14:13
DocScrutinizer05	aah I see you're userland process, you can't do any lock_irq()	14:14
freemangordon	DocScrutinizer05: http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dht0008a/CJHIGEFG.html	14:14
freemangordon	WTF	14:14
freemangordon	wrong link, but don;t ask me why	14:14
freemangordon	DocScrutinizer05: google for ldrex/strex	14:15
DocScrutinizer05	:nod:	14:15
DocScrutinizer05	http://infocenter.arm.com/help/topic/com.arm.doc.ddi0344k/ch08s05s03.html?resultof=%22%6c%64%72%65%78%22%20	14:18
freemangordon	yeah, those are the instructions	14:19
DocScrutinizer05	not only instructions but also a general concept how to use them	14:19
kerio	oh ARM <3	14:20
freemangordon	yeah,but difference with my code is that I am not trying to aquire the lock, but to set a value	14:20
DocScrutinizer05	hm?	14:20
freemangordon	MUTEX_SET is called from the outside for() loop who checks the result	14:20
DocScrutinizer05	aiui MUTEX_SET is actually acquiring a lock	14:21
freemangordon	yes, but the check if the lock is aquired is not in MUTEX_SET	14:21
freemangordon	the example on ARM site does that too, i.e it does a spinlock	14:22
freemangordon	not a mutex	14:22
*** valdyn has quit IRC		14:22
*** flo_lap has joined #maemo		14:23
DocScrutinizer05	yeah, the BNE try	14:23
freemangordon	the same do gcc sync intrinstics	14:23
DocScrutinizer05	just change that to a return	14:23
freemangordon	DocScrutinizer05: not so simple, as I have to return correct result to the upper for() loop	14:24
DocScrutinizer05	so your upper layer code gets result OK/FAIL from call to MUTEX_SET	14:24
freemangordon	i.e. AIUI I have to return 1 from MUTEX_LOCK only if previous tsl value was 0 (i.e. mutex succesfully acquired) and strex ioperation succeeded (we were able to store 1 to tsl)	14:25
freemangordon	DocScrutinizer05: yes	14:25
*** valdyn has joined #maemo		14:29
*** phryk has joined #maemo		14:29
*** Wnt has joined #maemo		14:30
*** livelace has joined #maemo		14:31
*** ChanServ has quit IRC		14:35
*** Psotnick_ has joined #maemo		14:37
*** Psotnick has quit IRC		14:39
freemangordon	Pali: ping	14:41
Pali	freemangordon, pong	14:41
freemangordon	Pali: which .deb to use?	14:41
freemangordon	or to build mine?	14:41
Pali	what debs?	14:41
Pali	of uboot?	14:42
freemangordon	for u-boot	14:42
Pali	clone git tree	14:42
Pali	branch pali	14:42
freemangordon	yes, that is what I am doing	14:42
freemangordon	dpkg-buildpackage?	14:42
freemangordon	Pali: does it fail on all 2204s?	14:43
DocScrutinizer05	prolly	14:44
teotwaki	gvim src/common/clients/modulemanager.cpp	14:44
freemangordon	Pali: Unmet build dependencies: kernel-bootimg :(	14:44
teotwaki	wrong window	14:44
*** ChanServ has joined #maemo		14:44
*** brooks.freenode.net sets mode: +o ChanServ		14:44
freemangordon	I am with -kernel-cssu3 on the device, i.e.thumb thingie	14:45
*** dos1 has joined #maemo		14:45
freemangordon	installed kernel-cssu-bootimg, didn't help	14:45
Pali	freemangordon, download some zImage (not fiasco) and put it in root dir	14:45
Pali	then edit debian/rules	14:45
freemangordon	ok	14:46
Pali	and find /boot/zImage-2.6.28-omap1	14:46
Pali	change it to your path of zImage	14:46
Pali	this image will be attached to uboot	14:46
Pali	and then ignore build dependency	14:46
freemangordon	ok, got it	14:46
freemangordon	Pali: only uboot-flasher, ain't?	14:50
Pali	u-boot-tools + u-boot-flasher	14:50
freemangordon	ok	14:51
Pali	flasher has fiasco image (combined uboot+zImage)	14:51
Pali	-tools has shell script + mkimage for generating bootmenu	14:51
freemangordon	Pali: onenand read 0 0 4096 gives total chip size (0x0) exceeded :D	14:55
freemangordon	normal?	14:55
Pali	I do not know	14:55
Pali	I did not tested onenand uboot command yet	14:55
Pali	I only saw support, so I enabled it in -15	14:55
freemangordon	how to disable it?	14:56
Pali	there is also ubifs support, but due big size, I disabled it	14:56
freemangordon	in config file?	14:56
Pali	comment all ONENAND in include/configs/nokia_rx51.h	14:57
freemangordon	what about mdtparts?	14:58
Pali	you can comment them too	14:59
Pali	but you need to add "#define MTDPARTS_DEFAULT"	14:59
Pali	because there is env variable "mtdparts=" MTDPARTS_DEFAULT	15:00
freemangordon	but it is there	15:00
freemangordon	#define MTDPARTS_DEFAULT"mtdparts=onenand:" \	15:00
Pali	replace it with "#define MTDPARTS_DEFAULT"	15:00
freemangordon	why do I need that?	15:00
*** ChanServ sets mode: +v infobot		15:00
freemangordon	you mean an empty define?	15:01
Pali	for disabling mtd code	15:01
Pali	yes	15:01
Pali	empty define	15:01
DocScrutinizer05	ChanServ: ?	15:01
freemangordon	Pali: h,, how to flash fiasko image?	15:05
Pali	via usb or on device?	15:05
freemangordon	usb	15:05
freemangordon	-F instead of -k?	15:05
Pali	via usb -F	15:05
*** Canseco has joined #maemo		15:05
Pali	on device use: fiasco-image-update <fiasco_image>	15:05
freemangordon	Pali: it boots, with mtd and onenand disabled	15:06
DocScrutinizer05	on 2204?	15:07
Pali	can you enabled onenand & mtd and test again	15:07
freemangordon	yep	15:07
freemangordon	Pali: i tested before, what is the idea?	15:07
DocScrutinizer05	messing with MTDPARTS will fsckup quite some bits, as it's the partitioning the BL is passing to kernel as well, AIUI	15:08
Pali	so with enabled onenand not working and with disabled working?	15:08
freemangordon	yes	15:08
freemangordon	onenand and mtdparts	15:08
freemangordon	disabled	15:08
Pali	DocScrutinizer05, mtdparts in uboot are only for uboot	15:08
DocScrutinizer05	so: messed up MTDPARTS -> no mounting of any mtd-based fs	15:08
Pali	partition table for kernel is in omap table	15:08
freemangordon	problem is with onenand, not with mtd AIUI	15:08
Pali	atag	15:08
Pali	freemangordon, ok	15:09
Pali	ping merlin1991	15:09
freemangordon	as it is onenand to give device size 0	15:09
* freemangordon is afk for a while		15:09
Pali	freemangordon, can you sent your compiled uboot to merlin1991 for testing?	15:09
freemangordon	sure. fiasko image?	15:10
Pali	fiasco is enought	15:10
freemangordon	aah, no, it is for cssu3 kernel, will it be ok/	15:10
*** Free-MG has quit IRC		15:10
freemangordon	?	15:10
DocScrutinizer05	who's been that guy who asked yesterday about working uBoot for 2204?	15:11
freemangordon	panzerSomething iirc	15:11
Pali	_PanzerSajt	15:11
Pali	Sicelo	15:11
DocScrutinizer05	mimamoomomo	15:11
Pali	and other people from TMO too	15:11
*** DHR has joined #maemo		15:12
*** Free-MG has joined #maemo		15:12
* DocScrutinizer05 fails to recall right searchterm for grep in backlog, said sth like "check your hw-rev against list of those known to work"		15:13
DocScrutinizer05	marainein: ^^^^	15:14
*** dhbiker has joined #maemo		15:14
DocScrutinizer05	well, -o +r +n	15:15
marainein	DocScrutinizer05, what?	15:15
DocScrutinizer05	and we're from mimamoomomo to marainein	15:15
DocScrutinizer05	marainein: seems freemangordon created a uBoot working on 2204	15:15
marainein	ahh excellent	15:15
marainein	this is a new version?	15:15
DocScrutinizer05	would you like to help testing?	15:16
*** jhb has joined #maemo		15:16
marainein	yeah, but i gotta go to sleep	15:17
DocScrutinizer05	np	15:17
freemangordon	Pali: #define PISMO1_NAND_SIZEGPMC_SIZE_128M	15:18
freemangordon	why is that?	15:18
fasta	"However, the package may create empty directories below /usr/local so that the system administrator knows where to place site-specific files. These are not directories in /usr/local, but are children of directories in /usr/local. These directories (/usr/local/*/dir/) should be removed on package removal if they are empty."	15:19
fasta	Debian people and their specifications...	15:19
Pali	freemangordon, I do not know this comes from original ali's uboot port	15:19
*** Milhouse has joined #maemo		15:19
fasta	It's written more like a riddle, then an actual specification.	15:19
fasta	than*	15:19
Pali	there is more code which was not changed...	15:20
*** joga has quit IRC		15:20
fasta	Too bad it doesn't say where the first /usr/local/foodir comes from.	15:20
*** joga has joined #maemo		15:20
DocScrutinizer05	hehe	15:21
DocScrutinizer05	~debian fasta	15:21
* infobot tells fasta to RTFM!!!! GAH!!! HELL FIRE AND BRIMSTONE!!!! BURN!!! DIE!!! MWAHAHAHAHA!!!!		15:21
fasta	DocScrutinizer05: in all your wisdom, what do you run?	15:21
DocScrutinizer05	openSuse	15:22
fasta	DocScrutinizer05: is there because you are German?	15:22
fasta	that*	15:22
DocScrutinizer05	though I don't give much about what distro I use, since I have to tweak them anyway	15:22
fasta	DocScrutinizer05: or is there some good technical reason behind your choice?	15:22
freemangordon	Pali: looks like too much code is missing to enable onenand	15:22
Pali	freemangordon, first we need if disabling onenand fixed booting...	15:23
fasta	DocScrutinizer05: how do you manage your tweaks?	15:23
DocScrutinizer05	fasta: originally (like: in 1995+-) it was because suse been local for me. Now it's because openSuse is closely linked to SLES/SLED	15:23
Pali	I think that in -12 was still onenand code disabled	15:23
freemangordon	http://merlin1991.at/~freemangordon/u-boot/	15:24
DocScrutinizer05	(manage tweaks) with lots of pain felt in the arse	15:24
DocScrutinizer05	;-D	15:24
freemangordon	Pali: but attached kernel is kernel-cssu3	15:24
fasta	DocScrutinizer05: I mean: let's say you have to reinstall a new work station, how will you do that?	15:24
Pali	freemangordon, ok	15:24
DocScrutinizer05	U usually avoid to do that	15:25
DocScrutinizer05	I*	15:25
freemangordon	Pali: so we need merlin1991	15:25
fasta	DocScrutinizer05: so, you dd everything?	15:25
Pali	yes, somebody who has non working uboot	15:25
fasta	DocScrutinizer05: e.g. harddisk upgrades, etc.	15:25
freemangordon	yay, ubuntu still boots	15:25
Pali	ping Sicelo	15:25
fasta	DocScrutinizer05: I wrote some scripts to write to /etc files, select the right alternative for various things, etc., according to my needs.	15:26
fasta	DocScrutinizer05: so, when I run that script, I have morphed a completely useless machine into something that does something useful.	15:26
DocScrutinizer05	IF I have to, it's: install openSuse_x, go thru all the packages manually in installer, opt-out/in a lot. then have two weeks minimum of days filled with tweaking all sorts of stuff, like global kbd shortcuts, themes, mail, $arbitrary-systems	15:26
fasta	DocScrutinizer05: yeah, I used to do that.	15:27
fasta	DocScrutinizer05: now every change I make is made programmatically.	15:27
DocScrutinizer05	on suse, there's autoyast, you'll like that if you get the catch of it	15:27
fasta	DocScrutinizer05: I still don't have a solution to manage Firefox extensions, though.	15:27
fasta	DocScrutinizer05: firefox had an API to install extensions.	15:27
fasta	But they threw it out.	15:27
fasta	So much for 'unleash the power of the web'.	15:28
fasta	Thunderbird also is completely obnoxious because of their contracting with Microsoft.	15:28
DocScrutinizer05	http://users.suse.com/~ug/AutoYaST_FAQ.html	15:29
fasta	Everytime you upgrade, they set some things by default.	15:29
fasta	Anyway, nothing which a little scripting cannot defeat, but imagine doing that by hand every single time.	15:29
fasta	Otherwise thunderbird is great.	15:29
fasta	Not a single crash in over a year or so.	15:29
fasta	DocScrutinizer05: autoyast sounds nice	15:30
DocScrutinizer05	a friend of mine has to maintain literally 10,000s of clients, and he's all full of joy about autoyast	15:31
DocScrutinizer05	(he's been a notorious yast-hater in former times ;-d)	15:32
DocScrutinizer05	;-D*	15:32
fasta	DocScrutinizer05: but autoyast seems to be only for installation, not maintenance.	15:32
DocScrutinizer05	now he defends suse against almost everything else	15:32
DocScrutinizer05	yes, autoyast is a mere installation+deployment tool	15:33
fasta	DocScrutinizer05: Via NixOS you can do similar things, but do maintenance too.	15:33
kerio	pff	15:33
kerio	just debootstrap the clients	15:33
kerio	>:D	15:33
DocScrutinizer05	for general local admin there's yast at large	15:33
DocScrutinizer05	for admin of 1000s of clients, there are shell tools	15:34
DocScrutinizer05	;-P	15:34
fasta	It would be interesting to see how long it would take to install 2000 packages with the various tools.	15:34
fasta	I think Nix would come in first, assuming the packages existed!	15:34
fasta	Apt would come in second.	15:35
fasta	And Fedora last.	15:35
fasta	Suse uses the same tools as Fedora, right?	15:35
fasta	I mean 'rpm'.	15:35
*** Darkchaos2 has joined #maemo		15:35
DocScrutinizer05	yast now does extremely smart stuff during install time: it simply copies a "disk image" that's closest to what you "ordered", then only adds / removes the delta of pkgs	15:36
*** Darkchaos has quit IRC		15:36
DocScrutinizer05	yes, suse is an RPM system	15:36
fasta	How is it then different from e.g. Fedora?	15:37
DocScrutinizer05	though you actually tend to forget about which pkg paradigm it's using, when you use yast and zypper	15:37
fasta	Are you saying it also understands debs?	15:37
trx	X-Fade ping	15:37
DocScrutinizer05	fasta: how am I supposed to answer that question?	15:37
fasta	DocScrutinizer05: the Fedora question?	15:38
DocScrutinizer05	fasta: I think it does	15:38
freemangordon	Pali: Any clue why both PISMO1_NAND_SIZE and PISMO1_ONEN_SIZE are defined?	15:38
DocScrutinizer05	yes	15:38
fasta	DocScrutinizer05: oh, well, I don't know. Perhaps Google knows.	15:38
fasta	The only unstable part on my system is my damn graphics card.	15:38
DocScrutinizer05	I could pipe result of a diff here?	15:38
Pali	freemangordon, I think that both defines are needed for onenand code...	15:38
Pali	but I did not played with onenand yet	15:39
Pali	so I do not know	15:39
freemangordon	hmm, but why 128M?	15:39
DocScrutinizer05	erase page?	15:39
DocScrutinizer05	bank size?	15:39
DocScrutinizer05	makes for a nice two banks, for 256M	15:40
freemangordon	that should be total size	15:40
freemangordon	AIUI	15:40
DocScrutinizer05	then maybe it's words?	15:40
freemangordon	as there is no bank_size or something	15:40
freemangordon	hmm, could be, lemme check in syslog	15:40
Pali	ask on #u-boot	15:42
freemangordon	OneNAND Manufacturer: Numonyx (0x20)	15:43
freemangordon	;)	15:43
DocScrutinizer05	LOL	15:43
DocScrutinizer05	Pali: hah, nice catch with #u-boot	15:44
freemangordon	2101: OneNAND Manufacturer: Samsung (0xec)	15:44
kerio	DocScrutinizer05: to be fair with 1000s of clients you probably want to do one installation and then clone it	15:44
freemangordon	Pali: ^^^	15:44
DocScrutinizer05	kerio: or use autoyast to do that for you	15:45
Pali	nice catch!	15:45
DocScrutinizer05	freemangordon: of course	15:45
*** dafox has joined #maemo		15:45
Hurrian	Pali, I've been wanting to ask for a while - the FAT32 driver for uboot is troublesome when dealing with un-fresh filesystems	15:45
kerio	freemangordon: how do i check my onenand manufacturer?	15:45
Pali	Hurrian, I know that. easy way is to backup data from MyDocs, then mkfs.msdos and copy back	15:46
*** Luke-Jr has quit IRC		15:46
*** Luke-Jr has joined #maemo		15:47
Hurrian	Indeed it does - any corruption/alteration of the table breaks the driver.	15:47
Hurrian	One problem though - on systems with CSSU-Thumb.	15:48
Hurrian	The system requires a Thumb kernel, but u-boot uses PR 1.3 kernel.	15:48
freemangordon	kerio: look in dmesg log :P	15:48
Hurrian	How about storing uboot in kernel MTD partition, and the real "flashed" kernel in MTD4?	15:49
Hurrian	MyDocs partition breaking for arbitrary reason = need to flash CSSU-Thumb kernel with flasher	15:50
DocScrutinizer05	freemangordon: share grep line ;-D	15:50
freemangordon	"OneNAND Manufacturer:" is what you need, i didn't use grep	15:51
freemangordon	though nymonix should be supported too	15:51
kerio	OneNAND Manufacturer: Numonyx (0x20)	15:51
kerio	2104	15:51
DocScrutinizer05	kerio: grep OneNAND /var/log/syslog	15:52
freemangordon	kerio: does yours boot with latest u-boot?	15:52
kerio	idk	15:52
freemangordon	Pali: dou you have such device?	15:52
Pali	0x40	15:53
DocScrutinizer05	freemangordon: >>Muxed OneNAND 256MB 1.8V 16-bit (0x40)<< here you are: 16bit	15:53
DocScrutinizer05	--<128M	15:53
DocScrutinizer05	-->*	15:53
freemangordon	yes, i got it	15:53
Pali	Samsung (0xec)	15:53
freemangordon	Pali: ox40?	15:53
freemangordon	aah	15:53
freemangordon	Pali: only one device?	15:53
Pali	0x40 same as Doc ^^^^^^	15:53
Pali	yes	15:53
freemangordon	Pali: ox40 is not the manifacturer, it should be either ec(sammie) or 20(nymonix)	15:54
Pali	Samsung (0xec)	15:54
kerio	Muxed OneNAND 256MB 1.8V 16-bit (0x40)	15:55
freemangordon	ok	15:55
Pali	and Muxed OneNAND 256MB 1.8V 16-bit (0x40)	15:55
DocScrutinizer05	mtdinfo FWIW	15:56
DocScrutinizer05	>>Device is writable: true<< ;-P	15:58
DocScrutinizer05	toldya there's a way to (un?)lock mtd partitions (or even pages?)	15:58
DocScrutinizer05	big PITA with HARM and openmode	15:59
DocScrutinizer05	since the partition/page with device lockcode is locked	15:59
DocScrutinizer05	in openmode	15:59
* DocScrutinizer05 idly wonders if it could get unlocked		16:00
DocScrutinizer05	kerio: you might also want to have a look at mtd_debug --help, not only nanddump/nandwrite	16:01
*** ZogG_laptop has quit IRC		16:01
Pali	freemangordon, in uboot are functions to lock onenand	16:02
Pali	see drivers/mtd	16:02
DocScrutinizer05	sure	16:03
Pali	onenand_unlock_all in drivers/mtd/onenand/onenand_base.c	16:03
freemangordon	Pali: i saw that, the question is why it works for samsung	16:03
freemangordon	Pali: hmm, does it?	16:04
Pali	maybe special driver, see: drivers/mtd/onenand/samsung.c	16:04
freemangordon	Pali: can you enter uboot console and check if onenand commands work?	16:04
*** FlameReaper-PC has joined #maemo		16:06
*** archeyDevil has quit IRC		16:07
Pali	not working	16:08
freemangordon	seems like drivers/mtd/onenand/samsung.c is for some samsung bord	16:08
*** ZogG_laptop has joined #maemo		16:08
*** ZogG_laptop has quit IRC		16:08
*** ZogG_laptop has joined #maemo		16:08
freemangordon	*board	16:08
*** ZogG_laptop has quit IRC		16:08
Pali	# onenand read 0x86008000 0 256	16:09
Pali	it write total chip size (0x0) exceeded	16:09
freemangordon	the same here	16:09
freemangordon	chip size is 0x0	16:09
Pali	and qemu wrote:	16:10
Pali	OneNAND read: offset 0x0, size 0x256	16:10
Pali	0 bytes read: OK	16:10
DocScrutinizer05	I doubt this onennand stuff is configured correctly for our hw	16:10
freemangordon	:nod:	16:10
DocScrutinizer05	I even dunno why we need it and for what	16:10
freemangordon	no way to replace nolo	16:11
DocScrutinizer05	I guess OneNAND has some special extended properties over 'normal' nand, and that OneNAND stuff is to handle those	16:11
Pali	if we decrease size (-mthumb) of uboot we can replace nolo	16:11
freemangordon	if ((off + size) > mtd->size) {	16:12
freemangordon	printf("total chip size (0x%llx) exceeded!\n", mtd->size);	16:12
*** ZogG_laptop has joined #maemo		16:12
*** ZogG_laptop has quit IRC		16:12
*** ZogG_laptop has joined #maemo		16:12
DocScrutinizer05	well, until it comes to flashing	16:12
*** ZogG_laptop has quit IRC		16:12
DocScrutinizer05	Pali: ^^^	16:12
DocScrutinizer05	(replace NOLO)	16:12
Pali	DocScrutinizer05, uboot has (not yet merged) patches for standard flashing protocol	16:13
Pali	and we still can use cold-flash for flashing	16:13
DocScrutinizer05	NOLO is a PITA, since it does a lot of stuff that's not documented at all, and no source available (obviously, but for not so obvious reasons)	16:13
DocScrutinizer05	sure	16:14
DocScrutinizer05	another PITA	16:14
Pali	DocScrutinizer05, in qemu I can run xloader --> uboot_spl	16:14
Pali	and uboot_spl wrote me hello message	16:14
Pali	and thats all	16:14
DocScrutinizer05	:nod:	16:14
Pali	because onenand is not implemented in uboot_spl	16:14
DocScrutinizer05	I think there's some pretty proprietary stuff regarding gfx init et al in NOLO	16:14
Pali	but it can write message to serial console - so it is working	16:14
DocScrutinizer05	what's uboot_spl?	16:15
*** ZogG_laptop has joined #maemo		16:15
*** ZogG_laptop has quit IRC		16:15
Pali	freemangordon, see arch/arm/include/asm/arch-omap3/mem.h line 340	16:15
kerio	DocScrutinizer05: what are the not so obvious reasons?	16:15
Pali	DocScrutinizer05, uboot spl is replacement for open source TI x-load	16:16
freemangordon	will compile with CONFIG_MTD_DEBUG	16:16
DocScrutinizer05	kerio: NFC, maybe the supposed proprietary GFX stuff	16:16
Pali	it is loader which load & run bootloader from RAM	16:16
kerio	:c	16:16
DocScrutinizer05	or simply Nokia being... Nokia, as usual?	16:16
kerio	fukken nokia	16:17
* kerio shakes his fist at Elop		16:17
kerio	i should make that into a "/nokia" command in xchat	16:17
DocScrutinizer05	Pali: AIUI this xloader has to carry proper signature, or ROMBL will not load&start it	16:17
Pali	#define <PART>_GPMC_CONFIG<x> <value>	16:18
Pali	ONNAND - Samsung One NAND	16:18
Pali	MNAND - Micron Large page x16 NAND	16:18
Pali	DocScrutinizer05, but we can replace nolo with uboot_spl + uboot	16:18
freemangordon	cmt?	16:18
DocScrutinizer05	that's strange then, if we can do this	16:18
Pali	for unknown reason in qemu this not working: xloader -> nolo	16:19
Pali	s/nolo/uboot/	16:19
infobot	Pali meant: for unknown reason in qemu this not working: xloader -> uboot	16:19
Pali	but xloader -> uboot_spl working	16:19
*** ZogG_laptop has joined #maemo		16:19
DocScrutinizer05	xloader->* should basically be no problem, since Nokia xloader for N900 doesn't check signature of * afaik	16:20
DocScrutinizer05	ROMBL however checks signature of $xloader	16:21
DocScrutinizer05	jacekowski explained to me that in N900 xloader and NOLO are integrated into one chunk	16:21
DocScrutinizer05	iirc	16:21
Pali	DocScrutinizer05, if you boot nolo without valid kernel image and you have serial console, you will see that nolo has command line	16:22
DocScrutinizer05	ROMBL only checking signature of first part of that chunk that contains the xloader code, I guess	16:22
Pali	and nolo has separate partition for xloader and separate partition for nolo	16:22
Pali	nolo & xloader are separated	16:23
Pali	so omap rom checking only xloader	16:23
DocScrutinizer05	err, they live in mtd0, no?	16:23
Pali	yes	16:23
Pali	but partition layout comes from nolo	16:23
Pali	and nolo sent different layout to kernel (and different has itself)	16:24
Pali	but when you flashing n900, you need to flash xloader and nolo together	16:24
DocScrutinizer05	yep	16:24
*** ZogG_laptop has quit IRC		16:24
Pali	NOLO protocol has message for load xloader, message for load nolo and message for flash xloader+nolo	16:24
Pali	there is no message for flash only xloader	16:25
DocScrutinizer05	since it's not possible	16:25
Pali	I played with flashing, you know that I rewritten 0xFFFF	16:25
*** eijk has quit IRC		16:25
DocScrutinizer05	:nod:	16:25
*** sq-one has joined #maemo		16:25
Pali	btw, what is license of xloader?	16:25
Pali	redistributable or not?	16:26
DocScrutinizer05	errr	16:26
DocScrutinizer05	nfc	16:26
DocScrutinizer05	I always thought it's an open reference implementation from TI	16:26
Pali	not, nokia xloader is closed	16:26
Pali	and is not same as TI xload	16:26
DocScrutinizer05	sure, by nokia	16:26
DocScrutinizer05	well, Nokia tweaked the TI reference implementation. But what does that mean?	16:27
DocScrutinizer05	what's been the question? ;-)	16:27
DocScrutinizer05	are you asking if anybody will get upset when you redistribute xloader binary? Answer: probably not	16:28
DocScrutinizer05	except the FOSS purists of course, who will start bitching at you to disclose the sourcecode ;-P	16:29
jon_y	imma wait for the FOSS Inquisition	16:30
jon_y	Grand Inquisitor Stallman says court is in session	16:30
*** ZogG_laptop has joined #maemo		16:30
*** ZogG_laptop has quit IRC		16:31
*** ZogG_laptop has joined #maemo		16:31
jon_y	sadly the xloader source likely died with the N900 in Nokia	16:31
DocScrutinizer05	ask jacekowski, he knows a lot more about all that stuff than I do	16:32
*** at1as has quit IRC		16:32
freemangordon	Pali: onenand phys base is 0x04000000 according to syslog, in config file #define CONFIG_SYS_ONENAND_BASEONENAND_MAP	16:32
freemangordon	#define ONENAND_MAP0x20000000/* OneNand addr */	16:32
DocScrutinizer05	strange stuff	16:33
*** ALoGeNo has quit IRC		16:34
*** ALoGeNo has joined #maemo		16:36
*** ALoGeNo has joined #maemo		16:36
*** FlameReaper-PC has quit IRC		16:37
*** guestx has joined #maemo		16:39
*** guestx has quit IRC		16:45
*** toxarisswe has joined #maemo		16:45
*** ALoGeNo has quit IRC		16:48
DocScrutinizer05	>>Few developers make me so angry as WhatsApp's ones. They just took XMPP, made a couple of changes so that it does not work with normal clients, forgot about any kind of security and call it a day. Their biggest idea is using phone numbers as identifiers and marketing their app as an SMS replacement instead of an internet chat. F*ck them. -- DMiax@Slashdot<< :-D +1000	16:52
DocScrutinizer05	(from don_falcone@tmo signature)	16:53
*** teotwaki__ has joined #maemo		16:53
Sicelo	lol	16:54
*** krayon has joined #maemo		16:56
*** ToJa92 has quit IRC		17:02
*** ALoGeNo has joined #maemo		17:05
*** cyborg-one has quit IRC		17:05
*** bigmeow has quit IRC		17:06
*** dos1 has quit IRC		17:08
*** dos1 has joined #maemo		17:09
*** bigmeow has joined #maemo		17:14
*** ToJa92 has joined #maemo		17:18
*** schen has joined #maemo		17:22
*** cyborg-one has joined #maemo		17:30
*** VDVsx has joined #maemo		17:33
*** kimitake_idle has quit IRC		17:33
*** kimitake has joined #maemo		17:35
*** valerius has quit IRC		17:36
*** x29a has quit IRC		17:46
*** Canseco has quit IRC		17:47
*** at1as has joined #maemo		17:48
*** valerius has joined #maemo		17:55
*** fuz_ has quit IRC		17:55
*** flo_lap is now known as florian		18:00
*** fuz_ has joined #maemo		18:05
kerio	oooh, i realized that my neighbour has WPS enabled on his wifi network	18:11
Sicelo	lol	18:13
Sicelo	naughty thoughts? :P	18:13
*** at1as has quit IRC		18:14
*** valerius has quit IRC		18:15
kerio	maybe	18:17
*** SmilybOrg has joined #maemo		18:19
*** SmilybOrg is now known as Guest77694		18:19
*** dhbiker has quit IRC		18:22
*** SmilyOrg has quit IRC		18:22
*** hardaker has joined #maemo		18:23
*** valerius has joined #maemo		18:24
ShadowJK_	does that make it easier to crack?	18:26
*** hardaker has quit IRC		18:28
*** Free-MG has quit IRC		18:29
*** disco_stu has quit IRC		18:33
peetah	ShadowJK_: from wikipedia, theoretically 10,000,000 possibilities, actually 11,000 attempts to be sure to have an access	18:34
*** dafox has quit IRC		18:35
kerio	ShadowJK_: WPS is so dumb	18:35
kerio	you have an 8-digit pin	18:35
kerio	but the first 4 digits are checked independently	18:35
kerio	and the last one is a checksum	18:35
*** teotwaki__ has quit IRC		18:41
*** robink__ has joined #maemo		18:44
*** robink_ has quit IRC		18:45
*** valerius has quit IRC		18:46
*** timeless_ has quit IRC		18:46
*** bigmeow has quit IRC		18:49
*** timeless_ has joined #maemo		18:51
*** Skry has quit IRC		18:51
*** sunny_s has quit IRC		18:52
*** VDVsx has quit IRC		18:53
*** bigmeow has joined #maemo		18:54
ShadowJK_	wtf lol	18:57
ShadowJK_	can you disable pin and only allow buttan	18:57
kerio	i don't think so, idk	18:57
*** valerius has joined #maemo		19:03
*** XDS2010 has quit IRC		19:03
*** mvp_ has joined #maemo		19:05
Pali	ping merlin1991	19:06
*** XDS2010 has joined #maemo		19:06
*** Skry has joined #maemo		19:06
Pali	ping Sicelo	19:06
freemangordon	hehe	19:07
Sicelo	hey Pali	19:07
Pali	Sicelo, can you test freemangordon build of uboot?	19:08
Sicelo	where is it?	19:08
freemangordon	http://merlin1991.at/~freemangordon/u-boot/	19:08
freemangordon	Sicelo: kernel-cssu3 ;)	19:09
freemangordon	attached kernel that is	19:09
freemangordon	Pali: do you know where in the kernel code is onenand init for rx51	19:10
freemangordon	i need base address and CSn	19:10
Pali	try drivers/mtd/onenand	19:10
Sicelo	brb then.. n900 is my only connection	19:11
freemangordon	Sicelo: ^^^	19:11
*** Zahra has joined #maemo		19:12
merlin1991	Pali: pong	19:12
Pali	or try arch/arm/mach-omap2/board-rx51-flash.c + board-n800-flash.c	19:12
freemangordon	already tried	19:12
Pali	(yes rx51 calling n800 functions :D)	19:12
freemangordon	no, they are just declared	19:12
Pali	merlin1991, can you test freemangordon uboot build?	19:12
Pali	freemangordon, parsing omap parition table is implemented in some n8xx function	19:14
Pali	and that omap atag is same for n8xx and n900	19:15
freemangordon	yes, but i miss HW settings	19:15
Pali	btw, what is this code doing?: return &rx51_sdrc_params[0];	19:15
merlin1991	freemangordon: so you have kernel cssu3 attached ??	19:18
freemangordon	yes	19:18
merlin1991	why does it say +zImage-2.6.28-omap1 then? :D	19:18
freemangordon	NFC, ask pali	19:19
Pali	merlin1991, because debian/rules has hardcoded that name	19:19
freemangordon	:P	19:19
merlin1991	hehe	19:19
merlin1991	okay I'll reboot and flash	19:20
freemangordon	Pali: r = gpmc_cs_request(c->gpmc_cs, ONENAND_IO_SIZE, &c->phys_base);	19:21
merlin1991	freemangordon, Pali: it boots	19:22
Pali	nice :-)	19:22
merlin1991	so what did you guys change? :D	19:22
freemangordon	disabled onenand and mtdparts support	19:23
merlin1991	but -12 at least had no onenand support, no idea about mtdparts	19:23
freemangordon	merlin1991: 2204 uses different onenand chip	19:24
merlin1991	yeah but to my understanding, in the -12 image I got from pali there was no onenand support?	19:24
freemangordon	NFC	19:25
Sicelo	Pali: freemangordon .. not booting. after i choose 'Attached Kernel' , device reboots in about 5 seconds	19:26
Sicelo	watchdogs maybe?	19:26
merlin1991	Sicelo: boot a not attached kernel	19:26
merlin1991	attached is the one from cssu-thumb	19:26
merlin1991	where you probably have no modules	19:26
freemangordon	Sicelo: do you have cssu-thumb installed?	19:26
Sicelo	nop	19:26
freemangordon	well, then you don't have kernel-cssu3 as well :P	19:27
freemangordon	Pali: and deffinitely base address is 0x04000000	19:27
Pali	ok	19:28
Pali	are you going to add onenand support?	19:28
freemangordon	will try at least	19:28
Sicelo	Pali: you still have a .deb of -11 around?	19:30
Pali	yes	19:30
Pali	no deb, but only image	19:30
Pali	Sicelo: http://atrey.karlin.mff.cuni.cz/~pali/u-boot/old/	19:32
Sicelo	thanks	19:32
Pali	DocScrutinizer05, freemangordon, is driver for n900 led too complicated or is easy to port it to uboot?	19:34
DocScrutinizer05	basically easy	19:35
Pali	sometimes it is really bad when led is turned on	19:35
DocScrutinizer05	though a tad fsckdup (based on LP5521 properties)	19:35
Pali	and display backlight?	19:35
merlin1991	hehe now we need led and charge support in u-boot, then we can go FUCK act_dead we do proper battery recovery :D	19:35
Pali	sometimes nolo turned display off...	19:35
DocScrutinizer05	but maybe you just want to talk to LP5523 via direct I2C rather than porting full driver for that	19:35
Pali	DocScrutinizer05, I think that uboot needs only to turn led off... nothing more	19:36
* DocScrutinizer05 just muses about how things/own motivations become more clear in hindsight		19:37
Pali	merlin1991, I can write charge support	19:37
DocScrutinizer05	http://talk.maemo.org/showpost.php?p=1277663&postcount=116	19:37
Pali	but first I want to shred -u BME	19:37
merlin1991	:D	19:37
merlin1991	btw are you goign to upstream the n900 u-boot changes Pali?	19:39
Pali	freemangordon, are you sure that all closed maemo applications (except hald-addon-bme) using libbmeipc only for temperature?	19:39
Pali	merlin1991, yes, I want to upstream n900 uboot	19:40
Pali	I already sent more patches, but they was not merged	19:40
Pali	(due to some changes in uboot core for relocation)	19:40
merlin1991	hm is there a dedicated mailinglist for u-boot?	19:41
DocScrutinizer05	Pali: (using libbmeipc) maybe that's really hard to tell	19:42
merlin1991	ah found it	19:42
merlin1991	sometimes I hate google, I search for u-boot and it gives me a billion entries for the submarine thanks to my "german speaking" ip	19:42
DocScrutinizer05	Pali: I could figure some of those closed maemo apps will do some exceptional stuff in exceptional situations, like overheated device etc	19:42
Pali	merlin1991, last patch series is here: http://article.gmane.org/gmane.comp.boot-loaders.u-boot/130370/	19:43
DocScrutinizer05	emergency call	19:43
DocScrutinizer05	merlin1991: the joy of silly pkg naming	19:43
*** markinfo has joined #maemo		19:44
*** valerius has quit IRC		19:45
*** int_ua has joined #maemo		19:47
*** int_ua has quit IRC		19:47
*** wmarone has quit IRC		20:04
*** wmarone has joined #maemo		20:04
*** wmarone has quit IRC		20:06
*** kimitake has quit IRC		20:06
*** dos1 has quit IRC		20:09
*** skope has quit IRC		20:10
*** wmarone has joined #maemo		20:11
freemangordon	Pali: it is above my expertize to enable onenand support, you should ask for some help on #u-boot, sotty	20:12
freemangordon	*sorry	20:12
Pali	ok	20:12
freemangordon	and looking at kernel code we just miss support for our device	20:12
freemangordon	Pali: see drivers/mtd/onenand/omap2.c	20:13
*** dhbiker has joined #maemo		20:16
*** romaxa_home has joined #maemo		20:19
*** romaxa has joined #maemo		20:19
*** schen has quit IRC		20:22
*** retro\|cz has joined #maemo		20:24
*** kimitake has joined #maemo		20:26
*** sq-one has quit IRC		20:30
*** dhbiker has quit IRC		20:32
*** schen has joined #maemo		20:36
*** vblazquez has quit IRC		20:37
*** vblazquez has joined #maemo		20:47
* ShadowJK_ made a small script to killall -STOP fennec if loadavg goes over 3.00		20:50
ShadowJK_	merlin1991; it does that to me too	20:50
*** disco_stu has joined #maemo		20:53
*** dhbiker has joined #maemo		20:53
kerio	ShadowJK_: but... what if you're using it? D:	20:53
ShadowJK_	i need to install gcc and rewrite script in gcc	20:53
ShadowJK_	uh, in c	20:53
ShadowJK_	loadvg jupmed to 9 all of sudden	20:54
DocScrutinizer05	ShadowJK_: mind to share the script?	20:57
ShadowJK_	it's a "one-liner"	20:57
ShadowJK_	I need to stop fennec and script to copypaste it	20:58
DocScrutinizer05	lol	20:59
DocScrutinizer05	(one-liner) guessed as much	21:00
DocScrutinizer05	anyway it's same direction as my idea of general system sanity monitor	21:00
DocScrutinizer05	so I might take it for inspiration	21:01
ShadowJK_	from memory: while true; do L=$(sed s/\\\./\ / </proc/loadavg \| awk '{print $1}' ) ; if [ $L -gt 2 ] ; then killall -STOP fennec ; else killall -CONT fennec ; fi ; sleep 1 ; done	21:03
*** toxarisswe has quit IRC		21:03
DocScrutinizer05	thanks	21:08
* DocScrutinizer05 still wonders how to evaluate per-process cpu usage		21:09
DocScrutinizer05	and other top-alike stuff	21:09
DocScrutinizer05	prolly I have to learn more about /proc///*	21:10
*** dafox has joined #maemo		21:10
*** FIQ has quit IRC		21:10
*** dafox is now known as Guest44645		21:10
DocScrutinizer05	hell, I could check src of top, or procps	21:10
*** FIQ has joined #maemo		21:11
*** FlameReaper has quit IRC		21:11
*** dos1 has joined #maemo		21:14
ShadowJK_	or Documentation/ in kernel src	21:16
ShadowJK_	/proc/pid/stat, but I forget what the numbers are ;p	21:16
ShadowJK_	and /proc/stat systemwide	21:17
freemangordon	jacekowski: ping	21:22
DocScrutinizer05	yeah, /proc/<PID>/stat	21:23
DocScrutinizer05	thanks	21:23
DocScrutinizer05	http://stackoverflow.com/questions/1420426/calculating-cpu-usage-of-a-process-in-linux	21:24
kerio	heh, TIL that the unit of cpu time in linux is called a jiffy	21:26
*** stardiviner has quit IRC		21:33
*** livelace has quit IRC		21:39
DocScrutinizer05	TIL?	21:53
DocScrutinizer05	~wtf til	21:54
infobot	Gee... I don't know what til means...	21:54
DocScrutinizer05	~dict til	21:54
infobot	Dictionary 'til' (3): Sesame. [Sometimes written {til}.] [1913 Webster] {Teel oil}, sesame oil. [1913 Webster] ;; \Til\, prep. & conj. See {Till}. [Obs.] --Chaucer. [1913 Webster] ;; Tech Info Library (Apple, WWW).	21:54
*** valerius has joined #maemo		21:55
DocScrutinizer05	jiffy is age old time unit in all *nix	21:55
DocScrutinizer05	there actually even is a way to find out about jiffies/second on your particular system - just I forgot how	21:56
DocScrutinizer05	basically a jiffy is one heartbeat of the hw-clock used to drive systemclock	21:58
DocScrutinizer05	man:/usr/share/man/man7/time.7.gz	22:01
*** Zahra has quit IRC		22:01
DocScrutinizer05	or man 7 time	22:01
kgu_	What is in scope for the next CSSU?	22:02
*** kgu_ is now known as kgu		22:05
kerio	kgu: a lot more flamewars	22:08
kerio	also a way to choose optional package blocks	22:08
merlin1991	kgu: well firstly, stable or testing?	22:08
merlin1991	in testing we have updated openss, camera-ui and possibly sqlite3	22:09
merlin1991	openssl	22:09
Sicelo	and in stable?	22:10
merlin1991	ask chemist :D	22:11
kgu	I meant in testing	22:11
*** valerius has quit IRC		22:11
*** messerting has joined #maemo		22:14
*** messerting has quit IRC		22:17
kgu	Is there an alternative to the calendar widget? I would love to be able to automatic switch profile according to meetings in calendar	22:17
DocScrutinizer05	kgu: I think that should be feasible with original calendar, some interface process to pick out the meeting dates, and switch profiles accordingly	22:20
DocScrutinizer05	btw I recall somebody doing exactly what you're interested in, quite some time ago	22:21
merlin1991	there is a packge / python script for that somewhere	22:21
*** robink__ has quit IRC		22:21
kgu	ohh cool - do you know where to find it?	22:21
DocScrutinizer05	search packet interface for "calendar" or "profile"	22:22
DocScrutinizer05	might help	22:22
kgu	in HAM?	22:24
DocScrutinizer05	http://maemo.org/packages/	22:25
*** robink_ has joined #maemo		22:26
*** ced117 has quit IRC		22:30
*** valerius has joined #maemo		22:31
merlin1991	hm forgot that I have to assess the qt4 merge request	22:35
Pali	merlin1991, I remember that MAG compiled new version of pulseaudio for maemo	22:36
Pali	do you know reason why he want to update pulseaudio?	22:36
merlin1991	no idea	22:36
Pali	MAG is offline now...	22:37
kgu	maemo-profiler 0.0.1 could be the one, but just tried to install it, but no program icon shows up..	22:39
*** jhb has quit IRC		22:46
kgu	ahh, it shows up as 'application' - not so intuitive. But no functions in this app, so cant be the one	22:46
*** Dibblah has quit IRC		22:50
*** Dibblah_ has joined #maemo		22:50
*** Guest44645 is now known as dafox		22:55
*** Plnt_ is now known as Plnt		22:58
*** croppa has joined #maemo		22:59
DocScrutinizer05	kgu: http://talk.maemo.org/showthread.php?t=72876	22:59
DocScrutinizer05	kgu: google maemo calendar profile: ~15th hit	23:00
DocScrutinizer05	nah, 4th	23:00
Jaffa	GeneralAntilles: ping	23:02
*** eijk has joined #maemo		23:06
kgu	DocScrutinizer05: perfect, you are the man!	23:08
DocScrutinizer05	kgu: yw	23:08
kgu	but what to do, when it returns this error: ImportError: No module named mafw	23:08
*** NIN101 has quit IRC		23:09
*** teotwaki has quit IRC		23:11
*** nlpplz has joined #maemo		23:14
DocScrutinizer05	http://maemo.org/packages/view/python-mafw/ maybe?	23:15
DocScrutinizer05	kgu: ^^^	23:17
kgu	sounds like the one - will give it a try	23:20
kgu	hmm, can't install	23:21
*** antero has quit IRC		23:23
DocScrutinizer05	hm?	23:30
DocScrutinizer05	either install with apt-get when you got extras-devel repo enabled. Or download the .deb from the pkg site, and install with dpkg (or simply by clicking it in microB browser)	23:31
kgu	it says incompatible package	23:31
DocScrutinizer05	o.O	23:31
DocScrutinizer05	you downloaded the i386?	23:31
kgu	no, the armel	23:32
DocScrutinizer05	nfc what's wrong then. Who says that?	23:32
kgu	HAM	23:32
ketas	damn, is there SANE phone on earth	23:32
DocScrutinizer05	check problems tab	23:32
kgu	problems tab is exactly 'incompatible package'	23:33
DocScrutinizer05	use apt-get or dpkg approach	23:33
DocScrutinizer05	HAM is not really meant for installing stuff like that	23:33
kgu	your wish is my command ;)	23:33
DocScrutinizer05	I'm actually amazed you even see this python binding in HAM	23:34
DocScrutinizer05	maybe python bindings are not exactly libraries like arbitrary .so	23:35
DocScrutinizer05	in HAM world	23:35
DocScrutinizer05	but anyway, use dpkg (or apt-get)	23:35
kgu	yeah, it works with apt-get! Thx again DocScrutinizer05. Now I will just have to figure out how to use the script..	23:35
*** valerius has quit IRC		23:36
jacekowski	freemangordon: pong	23:41
DocScrutinizer05	kgu: aiui this python script has to get started once, preferrably during boot, and waits in background for next meeting	23:41
DocScrutinizer05	I'd probably use a .desktop file for autostarting it, see modest.desktop etc	23:42
*** nox- has joined #maemo		23:42
DocScrutinizer05	there are several .desktop that don't show any icon	23:42
DocScrutinizer05	just to autostart stuff	23:42
DocScrutinizer05	the shebang line of that script and the run-standalone.sh in there makes me think though that the original author planned to start it via initscript form /etc/init.d	23:43
DocScrutinizer05	ask merlin1991;-D I see him mentioned in that scriptie	23:44
kgu	ok, I will try to drop it in init.d	23:45
kgu	but it should be run as user, right?	23:46
DocScrutinizer05	run-standalone.sh will take care	23:49
DocScrutinizer05	afaik	23:49
DocScrutinizer05	I'm musing about this: sys.path.insert(0, '/usr/lib/switchProfByMeeting')	23:53
merlin1991	that's for loading python modules	23:54
merlin1991	python has a fixed path to load modules from	23:54
merlin1991	also anything can be visible in ham as long as the package section is in user/*	23:54
DocScrutinizer05	I for sure don't have that dir, nor any python modules in it	23:54
freemangordon	jacekowski: sorry to pester, would you mind to look at this http://pastebin.com/1QtcHkS7	23:55
*** shanttu has quit IRC		23:55
DocScrutinizer05	(ham) I know	23:55
*** valerius has joined #maemo		23:56
DocScrutinizer05	sys.path.insert(0, '/home/user')	23:56
DocScrutinizer05	o.O	23:56
DocScrutinizer05	^^^ also for loading python modules?	23:56
kerio	if you're dicking around with sys.path, you're doing it wrong	23:57
*** perlite_ has joined #maemo		23:57
DocScrutinizer05	kerio: pretty please go to http://talk.maemo.org/showthread.php?t=72876 and fix it, then	23:57
kgu	A nice little icon just showed up in the status window :)	23:58
kerio	i don't even have a switchProfByMeeting	23:58
kgu	now I just have to fidle around with the profile names, that should be used	23:58
*** perlite has quit IRC		23:58
*** perlite_ is now known as perlite		23:59
*** MacDrunk has joined #maemo		23:59

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!