IRC log of #maemo for Friday, 2013-01-04

fastaOnly because it has a faster processor and a more fancy screen?00:00
ShadowJKwirr; weird in emmc to usd case mmcblk1 shows both reads and writes00:00
*** VDVsx has quit IRC00:00
fastaThe only good place for an Android device seems to be /dev/null.00:01
fastaSpeaking of which: someone said I could test harddisk performance by doing dd if=/dev/zero of=/dev/null o.O00:01
fastaWhat has the world come to?00:02
ShadowJKwirr; error -110 in dmesg?00:02
wirrShadowJK: yep00:02
*** lnxmen has joined #maemo00:03
fastaAre the Lumia phones spyphones like the Androids and the i-devices?00:03
wirrShadowJK: there is a swap partition on mmcblk1 (default nemo image) but it is not used/active in maemo. i definitely cannot explain the reads on mmcblk100:03
fastaThat would seem a very Microsoft way of doing things; doing things exactly the same.00:03
lnxmen[  628.561950] HWRecoveryResetSGX: SGX Hardware Recovery triggered00:04
freemangordonlnxmen: power-cycle your device00:04
freemangordonthis is the long-standing bug in SGX firmware00:04
infobotfreemangordon meant: this is a long-standing bug in SGX firmware00:04
fastaDoes the N900 come with cgroups preconfigured? o.O00:04
freemangordonfasta: yes00:05
fastaDid Nokia invent them?00:05
fastaAt least, I thought they weren't _that_ old.00:05
fastaMerged in 2007.00:07
wirrShadowJK: I think the device descriptors in iostat are swapped (so mmcblk0=sd, mmcblk1=emmc). then the write effectively are page swaps00:10
lnxmenfreemangordon: thanks, it works :)00:10
freemangordonlnxmen: np00:11
wirrso far I can say the following in regards to the sandisk card:00:14
wirr- works fine in stock pr1.300:14
wirr- throws a lot fi/o errors in kp51 with default00:14
wirr- behaves way better in kp51 using ShadowJKs tweaks but still throws i/o errors from time to time00:15
wirrShadowJK: any advice on how to find optimal alignment/scheduler settings for this card?00:15
ShadowJKwell the scheduler stuff just tries to make data "easier" for the card to process, it's still possible that the vard might get data patterns it times out on00:17
wirrso far couldn't find _any_ documentation on sandisk website regarding recommended settings00:17
jon_yheh, docs? what docs?00:18
jon_ythese are highly confidential and trade secret information, yada yada yada00:18
* ShadowJK suspects everyone is just ignoring it now00:19
ShadowJKignoring timeout stuff00:20
wirrhmm... doesn't seem to be a good option if this results in I/O -110 transmission errors :-(00:21
wirrcould there be any relation to the older -110 problem with palis u-boot bootstrapping the nemo kernel?00:21
ShadowJKBut basically kp sets timeout to maximum possible, and stock kernel sets it shorter, maybe00:22
kerioDocScrutinizer05: it's00:22
infobotmethinks snowboarding is a sport that kills your n90000:22
ShadowJKSo kp should give the card the longest possible time to do stuff00:22
keriooh no, not you, someone else who was talking at the same time, sorry for the ping00:22
ShadowJKWhy it works on stock and not pk is mysterious :)00:24
*** shamus has quit IRC00:24
*** shamus has joined #maemo00:24
*** ced117 has quit IRC00:31
*** cyborg-one has quit IRC00:34
DocScrutinizer05ShadowJK: MAXINT==MININT--00:34
*** xes has left #maemo00:37
kerioDocScrutinizer05: not necessarily00:37
infobotextra, extra, read all about it, coffee is the reason the net exists, the drug of choice for a GNU generation,, /usr/share/doc/HOWTO/en-html/mini/Coffee.html, geiseri's favorite beverage00:38
DocScrutinizer05kerio: once more, verbose: setting timeout to MAX might trigger a quite common bug in the card driver00:39
wirrDoc: is there an easy way for me to verify without to compile a custom kernel (don't have the sdk)?00:39
DocScrutinizer05usually called one-off-bug00:39
*** lnxmen has quit IRC00:40
DocScrutinizer05wirr: sorry, I never looked into that driver00:40
*** eijk_ has joined #maemo00:40
DocScrutinizer05a modinfo will tell a bit about allowed module parameters, if any00:40
DocScrutinizer05I gather that module is no module though, it's probably monolithic00:41
*** eijk has quit IRC00:41
DocScrutinizer05aka built-in to kernel00:41
DocScrutinizer05if it were a module, you could try unloading it and force-loading the stock kernel's driver module instead00:42
wirrmhm... I'll check that out00:43
DocScrutinizer05wirr: otoh I'm rather sure pali and other devels here will be happy to compile test kernels for you00:43
wirrDoc: cool, that sounds generous :-)00:43
DocScrutinizer05at least pali should be interested in finding out complete info about a possible flaw/bug in his powerkernel00:44
DocScrutinizer05freemangordon might be willing to help as well00:44
DocScrutinizer05wirr: btw doc: is not highlighting here (otherwise I'd get ~500 flase positives per day)00:45
DocScrutinizer05xchat should allow tab-completion of nicknames00:46
wirrDocScrutinizer05: thanks00:47
DocScrutinizer05np :-)00:47
*** sq-one has quit IRC00:47
*** luf has quit IRC00:48
*** NishanthMenon has quit IRC00:48
wirrDocScrutinizer05: while I'm restoring my crypto-backup could you tell me a little something about what's to be expected from maemo's future?00:49
wirrI mean considering the current transition of can we expect a boost in activity or is cssu-development rather slowing down due the lack of new hardware?00:51
keriowirr: the transition shouldn't really change things00:54
kerioat best we'll get a bit more control over the repos00:54
kerioso things should work a bit better, i suppose00:54
wirrkerio: so far that sounds goods00:55
kerioyep :D00:55
keriothe lack of new hardware is annoying, yes00:56
wirrwhats the stopper in porting maemo to other hardware? i mean the n9 seems ideal considering its omap00:57
keriothe closed blobs00:58
kerioand all the closed stuff00:58
*** pvanhoof has quit IRC00:58
wirrarrgh... it's always the same problem :-(00:59
kerioalso, maemo kinda sucks00:59
keriosomething like cordia HD would be better00:59
wirrwhy's that?00:59
kerioor, hell, just debian with hildon desktop00:59
wirroh yeah!00:59
* wirr agrees strongly00:59
wirris it already possible to run hildon ontop of debian?01:00
infobotfrom memory, closed is or
keriomessybox, upstart, the fucking FUBAR partitioning, the billion ways to run daemons...01:00
keriowirr: well, it's open source01:00
kerioand the cordiahd guys patched it to support resolutions other than 800x480 or 480x80001:00
wirrkerio: ok, I rephrase: is it already possible for non-developer?01:00
kerioit *should* be just a x11 window manager01:01
DocScrutinizer05kerio: the other sucking parts could get fixed easily in a spinoff, since any maemo on N9 wouldn't be fremantle anyway01:02
keriomy point is that we don't really want maemo, we want a linux distro with hildon-desktop01:02
DocScrutinizer05well, no01:02
DocScrutinizer05I also want ICD for example01:03
kerioi wonder why the openpandora guys didn't go with hildon-desktop too01:03
DocScrutinizer05the status bar01:03
DocScrutinizer05the dialer01:03
kerioICD is kind of a piece of crap though01:03
kerioand can't manage more than one connection at a time, at least with the current UI01:03
DocScrutinizer05there's a lot more than HD that defines awesomness of fremantle01:03
keriothe status bar is part of hildon01:04
kerionot just hildon-desktop though, that's true01:04
DocScrutinizer05not of HD though01:04
keriowhy the dialer?01:04
DocScrutinizer05because it works01:04
kerioit makes phone calls, what's so good about it?01:04
keriothe address book system is really good, that's true01:05
DocScrutinizer05that it makes phonecalls01:05
keriomeh, phone calls are overrated :P01:05
keriobesides, couldn't we grab the openmoko stuff for that?01:05
ShadowJKDocScrutinizer05: it's 4 bits in a 32 bit register. 15 (1111) is reserved, 14 is longest timeout, and pk hard codes that, while the original code gets a timeout requested from higher level mmc code01:05
ShadowJKAnd then sets timeout to closest available of the 1501:06
ShadowJKAnother patch I found online sets it to twice the requested timeout (but still limited to 14 max, which is max the omap hw supports)01:07
*** e-yes has quit IRC01:08
wirrShadowJK: so what do you suggest for my case?01:13
ShadowJKI have no idea :(01:14
*** e-yes has joined #maemo01:19
*** dhbiker has joined #maemo01:22
*** shamus has quit IRC01:33
*** shamus has joined #maemo01:33
*** eijk_ has quit IRC01:35
*** eijk_ has joined #maemo01:35
DocScrutinizer05ShadowJK: SiErr?01:44
DocScrutinizer0514 defective in silicon01:45
* DocScrutinizer05 also rises eyebrows on "15: reserved"01:46
DocScrutinizer05I'd build a PK module that hardcodes 7 instead 1401:48
DocScrutinizer05another one with 1301:48
DocScrutinizer05hell, if the 'patch' in PK that makes this adata card work is that it uses an hardcoded 14 instead a value it gets from card...01:49
DocScrutinizer05could we enable some debug mode in that driver, to see what wirr's 'defective' card requests for that timeout value?01:50
DocScrutinizer05raw value, not after trimming01:51
DocScrutinizer05maybe the card's controller depends on that timeout? to save some internal timers for whatever01:52
DocScrutinizer05after all flash-erasure of a page is some timed process01:53
wirrDocScrutinizer05, ShadowJK may this be related:
ShadowJKThat's a circuit board design thing... if Nokia did it wrong we'd see issues on many many devices and cards01:54
DocScrutinizer05wirr: excellent find01:55
wirrDocScrutinizer05, ShadowJK: I assume you're aware of the TI OMAP errata document of 2010? There's something bout "MMC: Multiple Block Read Operation Issue"02:00
*** lxp has joined #maemo02:01
DocScrutinizer05anyway N18 doesn't show up as a pin name on OMAP3530 of N90002:01
wirrDocScrutinizer05, ok02:01
*** florian has quit IRC02:02
DocScrutinizer05seems to indicate it's actually NC02:02
* DocScrutinizer05 bookmarks this gem anyway, it's an excellent example of the daily WTF you're facing when building hw02:03
wirrDocScrutinizer05, you're an electronics/hw guy?02:03
*** int_ua has quit IRC02:04
infobotsomebody said joerg was a HW-developer and engineer of Openmoko, usually known as DocScrutinizer02:04
*** NIN101 has quit IRC02:04
*** lxp1 has quit IRC02:05
infobotrumour has it, docscrutinizer is jOERG, a HW-developer and engineer of Openmoko02:06
wirrI see. :-)02:06
DocScrutinizer05I'd actully say senior EE02:07
wirrso why don't you build a new device for us ;-)02:07
wirrdoes this ring a bell? "The maximum block size supported by the host controller is 1024 bytes. This value is hard-coded in the02:07
wirrregister MMCi.MMCHS_CAPA[17:16] MBL field and cannot be changed.02:07
DocScrutinizer05because it costs "a significan slice of a million", as SpeedEvil likes to put it02:08
DocScrutinizer05(ring a bell) well, it for sure should show up in driver somehow02:09
DocScrutinizer05something that's hardcoded in hw needs also to get hardcoded to the driver02:09
wirrhmmm... so that's not related to /sys/block/mmcblk1/queue/nr_requests?02:10
*** eijk_ has quit IRC02:10
DocScrutinizer05err nope, that's number of entries in IO scheduler queue afaik02:10
DocScrutinizer05IOW number of pending IO02:10
DocScrutinizer05should be unrelated to any blocksize02:11
* DocScrutinizer05 waves and heads out for a nice beer02:11
DocScrutinizer05next pub02:11
wirrDocScrutinizer05, bye then and cheers02:12
DocScrutinizer05wirr: (EE) I'm available btw, if you got some interesting project and the money to pay me :-)02:12
wirrDocScrutinizer05, I'm in network security, but who knows...02:13
*** florian has joined #maemo02:14
DocScrutinizer05wirr: btw I contributed a bit to gta0402:17
wirrthat's what i thought when you said openmoko...02:18
*** lxp1 has joined #maemo02:18
*** robbiethe1st has joined #maemo02:19
DocScrutinizer05gta04 basically unrelated to OM02:19
*** lxp2 has joined #maemo02:19
*** lxp has quit IRC02:19
wirrdidn't know that02:19
DocScrutinizer05it's an awesome incredible "one man project" courtesy Nikolaus Schaller, and recycles a few components of openmoko phones02:21
DocScrutinizer05I never felt like the latter was an extremely brilliant idea02:21
DocScrutinizer05could as well have used screen and case of N90002:22
wirrwould have been a more practical form factor...02:22
DocScrutinizer05or arbitrary other case and LCD02:22
*** lxp1 has quit IRC02:22
wirrbut if he's still on it this might happen for a next release?02:23
DocScrutinizer05i'm not involved in that project anymore02:24
wirrI see02:24
*** muellisoft has quit IRC02:26
*** lxp has joined #maemo02:28
*** Hurrian has joined #maemo02:29
*** lxp2 has quit IRC02:30
*** Hurrian has quit IRC02:33
*** gry has joined #maemo02:37
*** Hurrian has joined #maemo02:40
*** lxp1 has joined #maemo02:40
*** lxp has quit IRC02:43
*** Fosus has joined #maemo02:52
*** ZogG_laptop has joined #maemo02:53
*** ZogG_laptop has quit IRC02:53
*** ZogG_laptop has joined #maemo02:53
*** Fosus has left #maemo02:55
*** wirr has quit IRC02:56
*** tank-man has joined #maemo02:57
*** Hurrian has quit IRC03:00
*** lxp has joined #maemo03:01
*** lxp1 has quit IRC03:04
*** unclouded has quit IRC03:05
*** lxp1 has joined #maemo03:08
*** florian has quit IRC03:09
*** lxp has quit IRC03:12
*** Hurrian has joined #maemo03:13
*** unclouded has joined #maemo03:17
*** Theorbo has quit IRC03:18
*** Darkchaos2 has joined #maemo03:30
*** Darkchaos has quit IRC03:31
*** mvp__ has quit IRC03:33
*** mvp_ has quit IRC03:39
*** guampa has quit IRC03:45
*** zhxt has joined #maemo03:51
*** Darkchaos2 has quit IRC03:51
*** Aoyagi has left #maemo04:02
*** jonwil has joined #maemo04:03
*** guampa has joined #maemo04:04
*** pavi has joined #maemo04:05
*** M4rtinK has quit IRC04:09
*** MetalGearSolid has joined #maemo04:18
*** MetalGearSolid has quit IRC04:30
*** flx_ has joined #maemo04:42
*** realityg1ps has joined #maemo04:43
*** darkschneider2 has joined #maemo04:43
*** mickname_ has joined #maemo04:43
*** Scorcere1 has joined #maemo04:43
*** Viltzu_ has joined #maemo04:43
*** gry- has joined #maemo04:44
*** guampa has quit IRC04:44
*** frals_ has joined #maemo04:44
*** GAN900 has joined #maemo04:44
*** eqs has joined #maemo04:44
*** gry has quit IRC04:44
*** realitygaps has quit IRC04:44
*** lbt has quit IRC04:44
*** retro|cz has quit IRC04:44
*** ruskie has quit IRC04:44
*** darkschneider has quit IRC04:44
*** phryk has quit IRC04:44
*** frals has quit IRC04:44
*** flux has quit IRC04:44
*** EgS has quit IRC04:44
*** mickname has quit IRC04:44
*** Viltzu has quit IRC04:44
*** till- has quit IRC04:44
*** GeneralAntilles has quit IRC04:44
*** felipe` has quit IRC04:44
*** Scorcerer has quit IRC04:44
*** GAN900 is now known as GeneralAntilles04:44
*** till has joined #maemo04:44
*** flx_ is now known as flux04:44
*** lbt has joined #maemo04:45
*** gry- is now known as gry04:45
*** thedead1440 has quit IRC04:46
*** yabez has joined #maemo04:49
*** yabez has quit IRC04:53
*** ruskie has joined #maemo04:59
*** maybeWTF has joined #maemo05:00
*** maybeHere has quit IRC05:04
*** user_ has joined #maemo05:05
*** thedead1440 has joined #maemo05:06
*** user_ is now known as sou05:07
*** sou is now known as sourav05:07
*** user_ has joined #maemo05:10
*** thomasjfox has quit IRC05:18
*** retro|cz has joined #maemo05:18
*** phryk has joined #maemo05:19
*** user_ has quit IRC05:31
*** rm_you has joined #maemo05:51
*** rm_you has joined #maemo05:51
*** rm_you has quit IRC05:52
*** lxp has joined #maemo06:01
*** DocScrutinizer05 has quit IRC06:03
*** DocScrutinizer05 has joined #maemo06:03
*** lxp1 has quit IRC06:03
*** Theorbo has joined #maemo06:21
*** robbiethe1st has quit IRC06:25
*** uen| has joined #maemo06:29
*** auenf has quit IRC06:31
*** githogori has quit IRC06:31
*** auenf has joined #maemo06:31
*** uen has quit IRC06:33
*** uen| is now known as uen06:33
*** shamus has quit IRC06:38
*** shamus has joined #maemo06:38
*** Theorbo has quit IRC06:43
*** githogori has joined #maemo06:46
*** Theorbo has joined #maemo06:46
*** shamus has quit IRC06:49
*** shamus has joined #maemo06:49
*** sourav has quit IRC06:50
*** bugzy has quit IRC06:57
*** nox- has quit IRC06:59
*** bugzy has joined #maemo07:07
*** ArGGu^^ has quit IRC07:13
*** ron0062000 has joined #maemo07:23
*** InvalidHandle has quit IRC07:25
*** InvalidHandle has joined #maemo07:28
*** ron0062000 has quit IRC07:45
*** user_ has joined #maemo08:01
*** jonwil has quit IRC08:11
*** LaoLang_cool has joined #maemo08:14
*** croppa has joined #maemo08:15
*** LaoLang_cool has quit IRC08:18
*** user_ is now known as sourav08:22
*** hurbu_ has joined #maemo08:22
*** hurbu__ has quit IRC08:25
*** zhxt has quit IRC08:26
*** zhxt has joined #maemo08:31
*** sourav has quit IRC08:32
*** XATRIX has joined #maemo08:41
*** tank-man has quit IRC08:41
*** thedead1440 has quit IRC08:47
*** VDVsx has joined #maemo09:00
*** khertan has quit IRC09:19
*** khertan has joined #maemo09:19
*** khertan has joined #maemo09:20
*** geaaru has joined #maemo09:41
*** Martix has quit IRC09:46
*** pvanhoof has joined #maemo09:49
*** pcfe has quit IRC10:06
*** pcfe has joined #maemo10:10
*** pcfe has quit IRC10:10
*** pcfe has joined #maemo10:10
*** s33k3rgr has joined #maemo10:18
*** shamus has quit IRC10:20
*** shamus has joined #maemo10:20
*** crope has quit IRC10:23
*** crope has joined #maemo10:24
*** eijk has joined #maemo10:26
*** sasquatch has quit IRC10:27
*** s33k3rgr has quit IRC10:27
*** ludens has quit IRC10:28
*** jonwil has joined #maemo10:29
*** cyborg-one has joined #maemo10:31
*** Martix has joined #maemo10:35
*** Martix has quit IRC10:35
*** Pali has joined #maemo10:38
*** Pali has quit IRC10:43
*** ludens has joined #maemo10:46
*** Pali has joined #maemo10:46
*** sasquatch has joined #maemo10:47
*** mhlavink has quit IRC10:53
*** mhlavink has joined #maemo10:53
*** Gh0sty has quit IRC10:54
*** thedead1440 has joined #maemo10:55
*** dhbiker has quit IRC10:55
*** Gh0sty has joined #maemo10:55
*** thedead1440 has quit IRC10:59
*** OkropNick has joined #maemo11:03
*** Neutron19 has quit IRC11:03
*** Neutron19 has joined #maemo11:04
*** zogg_ has quit IRC11:05
*** RP_ has quit IRC11:05
*** Jaffa has quit IRC11:05
*** Dragnslicer has joined #maemo11:05
*** ds3 has quit IRC11:05
*** q_plaz has quit IRC11:05
*** Jaffa has joined #maemo11:05
*** lool has quit IRC11:06
*** piggz has quit IRC11:06
*** lool has joined #maemo11:06
*** Dragnslcr has quit IRC11:08
*** eMHa has quit IRC11:08
*** dhbiker has joined #maemo11:11
*** thedead1440 has joined #maemo11:11
*** thedead1440 has quit IRC11:16
*** RP_ has joined #maemo11:17
*** zogg_ has joined #maemo11:18
*** ds3 has joined #maemo11:18
*** q_plaz has joined #maemo11:19
*** user_ has joined #maemo11:21
*** user_ is now known as sourav11:21
*** topro has joined #maemo11:31
*** retro|cz has quit IRC11:38
*** maybeWTF has quit IRC11:43
*** thetet has joined #maemo11:47
*** dhbiker has quit IRC11:53
*** zhxt has quit IRC11:56
*** dhbiker has joined #maemo11:58
*** krau has quit IRC12:00
*** maybeHere has joined #maemo12:03
*** krau|away has joined #maemo12:05
*** pavi has quit IRC12:07
*** sourav has left #maemo12:10
*** thetet has quit IRC12:16
*** thetet has joined #maemo12:16
*** thetet has quit IRC12:20
*** retro|cz has joined #maemo12:22
*** lizardo has joined #maemo12:26
*** Macor has quit IRC12:26
*** thetet has joined #maemo12:27
PaliDocScrutinizer05, extras builder still not working12:30
Palisee fremantle build queue:
*** unclouded has quit IRC12:38
*** piggz has joined #maemo12:41
*** thedead1440 has joined #maemo12:49
*** M4rtinK has joined #maemo12:53
*** retro|cz has quit IRC12:53
*** flo_lap has joined #maemo13:03
*** rcg has joined #maemo13:05
*** retro|cz has joined #maemo13:06
*** flo_lap is now known as florian13:07
*** ArGGu^^ has joined #maemo13:10
*** kolp has joined #maemo13:13
*** rcg has quit IRC13:19
*** Darkchaos has joined #maemo13:26
*** rcg has joined #maemo13:27
*** rcg has quit IRC13:28
*** rcg has joined #maemo13:30
*** dhbiker has quit IRC13:31
*** dhbiker has joined #maemo13:32
*** zhxt has joined #maemo13:49
*** M4rtinK has quit IRC13:55
*** thetet has quit IRC13:58
*** LaoLang_cool has joined #maemo13:59
*** strohalm has quit IRC14:01
*** brzys has quit IRC14:02
*** eMHa has joined #maemo14:05
*** brzys has joined #maemo14:09
*** LaoLang_cool has quit IRC14:11
*** LaoLang_cool has joined #maemo14:13
*** topro has quit IRC14:15
*** user_ has joined #maemo14:15
*** user_ is now known as sourav14:16
*** ced117 has joined #maemo14:21
*** strohalm has joined #maemo14:21
*** rcg has quit IRC14:26
*** thomasjfox has joined #maemo14:27
*** Darkchaos has quit IRC14:28
*** Darkchaos has joined #maemo14:28
*** LaoLang_cool has quit IRC14:30
*** xmlich02 has joined #maemo14:37
*** NIN101 has joined #maemo14:37
*** anthony__ has joined #maemo14:39
*** landho has quit IRC14:43
*** landho has joined #maemo14:43
*** sourav has left #maemo14:44
*** Theorbo has quit IRC14:51
*** felipe` has joined #maemo14:54
DocScrutinizer05Pali: Bergie just informed me that they are about to transfer autobuilder to a VM to keep it running, >>albeit probably very slow<<14:57
PaliDocScrutinizer05, ok, do you know when it will work again?14:58
DocScrutinizer05trying to get detailed info14:58
DocScrutinizer05Niels / X-Fade doing that moving to VM15:00
*** Pali has quit IRC15:01
*** Pali has joined #maemo15:01
*** Theorbo has joined #maemo15:02
*** Hoolxi has joined #maemo15:02
DocScrutinizer05main issue for now is DNS, since Nokia not exactly responsive to any inquiries regarding that. Probably they simply don't answer Nemein's requests until the lawyers finished the paperwork with HiFo board15:02
*** etrunko has joined #maemo15:03
*** kolp has quit IRC15:05
*** valeriusL has quit IRC15:05
DocScrutinizer05ooh, and Nokia gave clearance for migration budget ~2 work weeks before new year, only then Nemein could start to do anything about it, so delay is all with Nokia, not Nemein15:08
*** rcg has joined #maemo15:08
DocScrutinizer05almost nobody inside Nokia left over who would even know what's maemo doesn't exactly help either15:10
*** valeriusL has joined #maemo15:11
*** kolp has joined #maemo15:12
*** wirr has joined #maemo15:15
*** kolp has quit IRC15:18
*** krau|away is now known as krau15:19
*** MetalGearSolid has joined #maemo15:25
*** user_ has joined #maemo15:36
*** user_ has quit IRC15:42
*** chenca has joined #maemo15:43
thomasjfoxDocScrutinizer05: thanks for the info, I was really wondering what's going on with autobuilder15:45
DocScrutinizer05Niles had to disable/stop it so the old server and the VM won't get out of sync15:46
DocScrutinizer05we still have no control over DNS15:46
*** Pali has quit IRC15:48
*** Pali has joined #maemo15:52
*** SmilybOrg has joined #maemo16:09
*** SmilybOrg is now known as Guest3820816:10
*** SmilyOrg has quit IRC16:13
*** Skry has quit IRC16:21
*** Skry has joined #maemo16:23
*** NishanthMenon has joined #maemo16:25
*** straind has quit IRC16:26
*** straind` has joined #maemo16:26
*** sourav has joined #maemo16:37
souravhi everyone, can anyone help me in setting up  ubuntu lucid desktop VM for fremantle in windows16:39
*** xes has joined #maemo16:39
*** sq-one has joined #maemo16:41
*** jonwil has quit IRC16:41
*** landho has quit IRC16:43
*** landho has joined #maemo16:44
*** Aoyagi has joined #maemo16:46
*** rcg has quit IRC16:49
*** shanttu has joined #maemo16:50
*** guampa has joined #maemo16:52
*** kolp has joined #maemo16:52
*** MrOpposite has quit IRC16:54
*** strohalm has quit IRC16:56
*** strohi has joined #maemo16:56
*** strohi has joined #maemo16:56
*** MrOpposite has joined #maemo17:03
*** xes has quit IRC17:06
*** retro|cz has quit IRC17:08
*** sourav has quit IRC17:08
*** Hoolxi has quit IRC17:10
*** Helloer has joined #maemo17:10
keriohrmpf, it's the second time that sourav asked that question17:11
keriobut really, how hard can it be to download virtualbox and adding a new VM with an existing drive?17:12
kerio*and to add17:12
*** VDVsx has quit IRC17:18
*** retro|cz has joined #maemo17:22
Skryapparently it can be quite hard17:26
SkryI can understand asking once, but doing it repeatedly over few days instead of just googling about it or figuring it out by himself...17:31
freemangordonhe esked me too, just an hour ago :D:D:D17:31
Skryhe msg'd me too, though I never quite got what he was even asking17:32
*** VDVsx has joined #maemo17:38
*** pvanhoof has quit IRC17:47
*** sn has joined #maemo17:47
*** sn is now known as Guest268617:48
*** Guest2686 has left #maemo17:48
*** shamus has quit IRC17:48
*** shamus has joined #maemo17:49
*** ekze has quit IRC17:53
*** ekze has joined #maemo17:54
freemangordonthedead1440: hmm, what is that "However, Rob Baeur (SD69) has now informed me that Harmattan might not be involved in the Board's discussions with Nokia." supposed to mean?17:57
thedead1440freemangordon: Let's clarify that the attached list is the one you are referring to.17:58
thedead1440It has files which are on merlin1991's server and are on  So those files should be deleted I believe before the list is forwarded to Nokia.17:58
thedead1440I think we are unlikely to get Harmattan files now but you can try I guess.17:58
freemangordonbut what will happen with Harmattan then?!?17:59
freemangordonthe fuck17:59
DocScrutinizer05sd69 honestly needs help with the technical implications of the migration18:05
DocScrutinizer05it's a tad unfortunate that probably neither side (board, nokia's lawyers) has full expertise about what maemo means and includes, on a technical level18:07
DocScrutinizer05neither for HARM18:07
DocScrutinizer05actually for HARM even I am basically clueless18:08
DocScrutinizer05council got qwazix though, who seems pretty active and up to date with harmattan details18:08
wirrShadowJK: I've found some more info about the troubling card: Allocation units are as big as 12M, Write Size is 64K and Page Size is 16K. This seems unusually high to me for an SDHC-card. Could this have any influence on the IO errors?18:10
*** landho has quit IRC18:12
*** landho has joined #maemo18:13
DocScrutinizer05[general notice] autobuilder considered down according to read-only notice on top of any page18:15
DocScrutinizer05please any of you following tmo closely, would you point your humble councilor [me] to any threads/posts relevant to council that I should notice (particularly those related to migration/backup). Thanks.18:17
ShadowJKwirr: yes, it could. Biggest AU I've seen so far is 8M. A 32g sandisk I recentlyish purchased is 4M18:22
ShadowJKthe max timeout is approx 2s iirc, read-mod-write of 12M is going to take 2 sec at class 6 speeds...18:23
ShadowJKor slightly more :)18:23
wirrShadowJK, I see... so this would mean ideally by optimizing alignment and fs options speeds _could_ be increased to stay below timeout?18:24
ShadowJKit'd mitigate it, I imagine18:24
*** thomasjfox has quit IRC18:25
wirrdo you know of any other configurable or hardcoded options which could be tweaked?18:25
*** XATRIX has quit IRC18:26
DocScrutinizer05is there a way to completely disable, or retrigger, the timeout?18:26
DocScrutinizer05also AIUI an allocation unit is a chunk of flash space that's supposed to get written in one operation? could we just ignore the 12M AU the card requests and go for smaller size?18:29
*** Guest38208 is now known as Smily18:31
*** sq-one has quit IRC18:33
ShadowJKAU is totally hidden from host18:40
ShadowJK(except on newer cards they do provide hint as to what optimal would be)18:40
ShadowJKAU, as i understand it, is smallest physically erasable block.18:41
ShadowJKwrite size 64k sounds like smallest writable block, where the destination must be erased, is 64k18:42
ShadowJKwhere destination must be already erased18:43
*** stardiviner has joined #maemo18:43
wirrShadowJK, I agree with that. do you know how i can configure the write size on linux?18:44
ShadowJKWell we don't have a filesystem that is flash aware18:45
Apicjffs rules18:45
ShadowJKbut maybe tinkering with the raid stripe/stride options in ext* and the ext* block size?18:45
ShadowJKjffs sucks in that it keeps all filesystem metadata in ram18:48
wirrI've already adjusted stripe/stride to page size18:48
wirrbut write size is 4 times page size... so this not possible without flash-aware fs?18:48
ShadowJKand has to scan entire device on mount18:49
ShadowJKnilfs2 might be more sensible, and some version of it seems to  be included in kp18:49
ShadowJKcan stripe/stride be adjusted to 64k?18:49
*** shamus has quit IRC18:49
*** shamus has joined #maemo18:50
wirrShadowJK, possibly, haven't tried yet18:51
*** snx has joined #maemo18:51
ShadowJKI vaguely remember playing with this when I set up my adata card for my N80018:51
ShadowJKbut I forget the outcome :)18:52
*** snx has quit IRC18:55
ShadowJKwirr: does this card have a UHS logo on it too?18:57
wirrfirst test (using mass-storage! and kp52) gives approx 4.7MB/s18:57
wirrShadowJK, yep18:57
wirrway better than with default params18:57
ShadowJKwhat params did you change?18:59
ShadowJKUsing nr-requests tweak too?19:00
* ShadowJK is at work and not supposed to irc, so I might reply with great latency19:00
*** pvanhoof has joined #maemo19:01
*** tank-man has joined #maemo19:04
ShadowJKwould be very interesting to be able to read SD specs and see what they've changed over time19:04
wirrShadowJK, i've set nr_requests to 8192 using swappolube adjusted the partition layout to start with the first partition at 4megs (fat-optimized area from 4-12meg) and aligned all following partitions to 12meg boundaries (starting at 4, according to flashbench mailing list)19:04
wirradjusted fat & ext4 filesystems to page size19:05
*** Rantwolf has joined #maemo19:05
*** geaaru has quit IRC19:06
wirrShadowJK, i'm wondering if we should change default scheduler from cfq to noop (noop is ideal for ssd, but deadline, which is not available, would be better for sd, i think)19:07
*** pvanhoof has quit IRC19:08
ShadowJKOn N800 I only have noop, and keep wishing for cfq ;)19:13
wirrShadowJK, is that so?19:13
ShadowJKWell when I was researching this in maemo4 era last time (maemo5/n900 was such a huge improvement out of the box that I havent bothered much in that area), google searches found lots of eeepc users experimenting with this. Most users who tried noop came back saying "it's awesome" right after booting, then 15 minutes later started complaining about random apps or entire system mysterious stalling for tens of seconds at a time.. then they went back to cfq,19:16
ShadowJKwhich they deemed was "consistently fast, uh, consistently slow, but consistent"19:16
SkryI did test bfq some time ago, seemed to be better that cfq but I didn't benchmark it so might have been placebo19:17
ShadowJKThe fundamental issue is still that they're all flash-unaware.. some operations on flash will be essentially "free" while others will have a huge cost19:20
ShadowJKand all the schedulers more or less assume a harddrive, and cost is more or less a linear measure of how far away from last accessed block a potential disk access is19:21
wirrShadowJK, but I think there's some development going on in that area, i think it's called 'fios'19:22
*** M4rtinK has joined #maemo19:23
ShadowJKDocScrutinizer05 / SpeedEvil ; do you know if sd cards have internal oscillators or if they rely on host provided CLK? UHS cards running at really low speeds in SDHC hosts could be due to that too..19:24
Skryiirc bfq has at least some ssd optimizations too19:25
ShadowJKSSDs, flash, and HD are all 3 distinct19:26
ShadowJKssd optimization might hurt flash more than HD optimization19:26
DocScrutinizer05ShadowJK: sorry, no info here19:30
ShadowJKAfaict, the omap hsmmc hw's data timeout is specified in clock cycles, if we could drop hsmmc bus speed, the timeout would become longer.. but if card runs off of hsmmc clock, it'd run correspondingly slower too and it wouldn't change the situation :/19:31
*** shanttu has quit IRC19:31
* ShadowJK wonders if mmc clockspeed is related to cpu speed too, on N8x0 it was19:31
freemangordonon n900 too19:32
freemangordonthat's why cpu is locked 2 min 500 with USB connected19:32
Apic2 min 500?19:32
freemangordon@ min 500 MHz, sorry, typo19:33
Apicic, now I get it19:33
ShadowJKfreemangordon; can you tell what speeds mmc runs at with cpu @ 250, 500, 550, 600, 600+?19:34
freemangordonShadowJK: though not to VDD1 OPP but to VDD2 OPP19:34
freemangordonShadowJK: ^^^19:34
keriohold on, so setting the cpu to 500 or 600 would speed up the I/O? neat19:35
freemangordoncan't remember what were VDD2 freqs19:35
*** pablocastellanos has joined #maemo19:36
freemangordonkerio: no. you should tweak VDD2 frequency, CPU @ 500 or 600 does not make much of a difference19:36
freemangordonShadowJK: can't remember at what speed mmc is clocked, but it is on the top limit19:38
wirrfreemangordon, pali: would one of you be willing to compile an mmc module for power52 with timeout set to 7 or 13?19:39
SpeedEvilShadowJK: from memory, external clock19:40
freemangordonwirr: iirc it is set to 13 or 14 right now19:40
ShadowJKYeah it's 14 in powerkernel19:40
wirrfreemangordon, 14...19:41
freemangordonand that's the highes allowed value afaik19:41
wirrfreemangordon, yeah but strangely the card works fine on stock pr1.319:41
freemangordonI see19:41
* ShadowJK wonders why the number 15 is reserved19:41
freemangordonare you sure you tested it long enough? o stock that is.19:42
wirrfull 32gig read and write19:42
*** radic has joined #maemo19:44
ShadowJKry hit with bonnie++ (in repos, devel probably or tools)19:44
*** radic_ has quit IRC19:45
*** tank-man has quit IRC19:51
*** teotwaki has quit IRC19:55
*** Woody14619 has joined #maemo19:58
*** Woody14619 has quit IRC19:58
*** Woody14619 has joined #maemo19:58
*** qwazix has joined #maemo20:01
*** toxaris has joined #maemo20:04
*** NIN101 has quit IRC20:05
*** valerius has quit IRC20:06
*** valerius has joined #maemo20:06
*** eMHa has quit IRC20:20
*** NIN101 has joined #maemo20:21
*** luf has joined #maemo20:21
*** sq-one has joined #maemo20:22
*** sasquatch has quit IRC20:25
*** sasquatch has joined #maemo20:26
*** Martix has joined #maemo20:27
*** Darkchaos has quit IRC20:38
*** zhxt has quit IRC20:39
*** sq-one has quit IRC20:39
*** sasquatch has quit IRC20:40
*** stardiviner has quit IRC20:40
*** luf has quit IRC20:49
*** retro|cz has quit IRC20:50
*** zhxt has joined #maemo20:51
*** luf has joined #maemo20:52
*** shamus has quit IRC20:53
*** shamus has joined #maemo20:53
*** eMHa has joined #maemo20:57
*** mvp_ has joined #maemo21:09
*** strohi has quit IRC21:10
*** strohalm has joined #maemo21:10
*** toxaris has quit IRC21:12
*** Sc0rpius has quit IRC21:12
*** valerius has quit IRC21:15
*** pvanhoof has joined #maemo21:20
*** Sc0rpius has joined #maemo21:20
*** pablocastellanos has quit IRC21:23
*** xmlich02 has quit IRC21:27
*** valerius has joined #maemo21:33
*** Theorbo has quit IRC21:33
*** rcg has joined #maemo21:36
*** nox- has joined #maemo21:37
*** flx_ has joined #maemo21:39
*** flux has quit IRC21:40
*** flx_ is now known as flux21:41
*** zhxt has quit IRC21:46
*** louisdk has joined #maemo21:49
*** louisdk has quit IRC21:49
*** Milhouse has quit IRC22:08
*** florian has quit IRC22:10
*** pvanhoof has quit IRC22:18
*** Corsac has quit IRC22:20
*** Corsac has joined #maemo22:20
*** Milhouse has joined #maemo22:24
*** florian has joined #maemo22:30
wirrShadowJK: I've run tests with bonnie++: Despite the geometry, fs and scheduler tweaks this resulted in i/o errors and a complete crash of the device. next i've replaced the omap_hsmmc.ko with the one from cssu-thumb. and voilĂ  now no more io errors and acceptable performance...22:31
wirrPali: perhaps you could compile new omap_hsmmc modules for power52 with timeout values of 7 and 13? i'd be happy to test.22:32
wirrif someone's interested in the bonnie output just tell me22:33
*** Martix has quit IRC22:33
*** Martix has joined #maemo22:33
*** Martix_ has joined #maemo22:35
*** Martix_ has quit IRC22:37
*** zogg_ has quit IRC22:38
*** ZogG_laptop has quit IRC22:38
*** ZogG_lap1op has joined #maemo22:39
*** zogg_ has joined #maemo22:39
*** Martix has quit IRC22:39
*** stardiviner has joined #maemo22:39
ShadowJKI forget, is cssu-thumb kernel different from powerkernel?22:42
*** _rd has joined #maemo22:42
ShadowJKwirr: did bonnie fail on stock?22:43
wirrhaven't tried yet22:43
wirrand yes, cssu-thumb and power52 modules differ22:43
ShadowJKwhat do you mean "7 and 13"?22:44
*** etrunko has quit IRC22:50
lufShadowJK: if I remember right, cssu-thumb kernel is power-kernel (but not non-release kp52)22:55
*** Pali has quit IRC22:57
ShadowJKso it probably has same hsmmc code as kp5122:57
wirrShadowJK, timeout values22:57
*** _rd has quit IRC22:59
*** _rd has joined #maemo22:59
lufShadowJK: the best man to ask is freemangordon22:59
lufShadowJK: or take a look into sources ;)22:59
ShadowJKwirr: btw where did you find out about the 12M thing?23:00
wirrShadowJK, flashbench, wait a sec23:00
*** lizardo has quit IRC23:01
wirrMine is the Sandisk Ultra SDSDU-032G-A11 "SU32G" (2012)23:01
wirrcan someone point me to the sources of cssu-thumb and power52?23:06
*** shamus has quit IRC23:09
*** shamus has joined #maemo23:10
*** thedead1440 has quit IRC23:11
*** anthony__ has quit IRC23:13
*** Martix has joined #maemo23:15
*** thedead1440 has joined #maemo23:16
* ShadowJK vaguely recalls there used to be a benchmark utility on that flashcardsurveypage :/23:16
*** thedead1440_ has joined #maemo23:17
*** brzys has quit IRC23:18
*** sq-one has joined #maemo23:18
*** murrayc has joined #maemo23:19
*** thedead1440 has quit IRC23:20
*** brzys has joined #maemo23:31
wirrShadowJK, Git - people/arnd/flashbench.git/summary;a=summary23:36
ShadowJKdid you run it?23:37
ShadowJK(it's probably destructive)23:37
wirr(only through mass storage)23:39
ShadowJKget this too23:40
ShadowJKI'll give you params23:41
ShadowJKOr here's the script I normally use:
ShadowJKObviously edit the device name so you dont overwrite harddrives, and ediit path to executable23:46
ShadowJKBasically goes through random writes with 4k ... 12M block size..23:46
*** shamus has quit IRC23:46
*** shamus has joined #maemo23:48
*** xes has joined #maemo23:48
*** croppa has quit IRC23:49
*** croppa has joined #maemo23:49
*** xes has left #maemo23:50
ShadowJKWarning: as it's a write test it will destroy existing data23:52
*** Theorbo has joined #maemo23:52

Generated by 2.15.1 by Marius Gedminas - find it at!