*** MohammadAG has quit IRC | 00:07 | |
*** xnt14 has quit IRC | 00:07 | |
*** _rd has joined #maemo-ssu | 00:07 | |
DocScrutinizer05 | there's "an app for that" (optification) | 00:35 |
---|---|---|
DocScrutinizer05 | comes per default with each maemo5 | 00:35 |
DocScrutinizer05 | each semi-*recent* maemo5 | 00:35 |
DocScrutinizer05 | zeq: I just sumbled across osso-backup-restore-locale, nfc what it does | 00:38 |
zeq | ? | 00:39 |
zeq | DocScrutinizer05: I'm trying to convince buildpackage to build the full locale cache | 00:40 |
zeq | it does build a locale*.deb by default, but it contains very little. | 00:40 |
kerio | would that fix the stupid thing where apt talks to me in english but it expects italian words at prompts? | 00:41 |
zeq | kerio: that's possible, I've no idea what's contained in the binary blob locale pkg | 00:42 |
zeq | as Doc mentioned earlier we might have some things missing. It's entirely possible there will be other things currently which currently are missing appear. | 00:43 |
zeq | From what I can gather recent debian libc builds have a locales-all target in addition to locale which by default just builds a package for the (build-)system current locale, or so it seems. | 00:45 |
kerio | zeq: would this end up in cssu-devel? | 00:46 |
zeq | There *must* be a way of cleaning all this up | 00:46 |
kerio | or is it cssu-extras material? | 00:46 |
zeq | kerio: don't know where it will end up. | 00:46 |
zeq | I'd quite like to have locale packages to install support for different lanaguages rather all or crypt symbols! | 00:47 |
zeq | cryptic* | 00:47 |
DocScrutinizer05 | kerio: maemo-optify-auto-opt.sh maemo-optify-firstboot.sh maemo-optify-make-mountbind.sh | 00:48 |
*** _rd has quit IRC | 00:50 | |
DocScrutinizer05 | zeq: I juggled with all that shite some 2 years ago. My conclusion been it's not possible to do anything about it. IIRC the 'normal' locale dirs don't contain anything, it's all in that 23MB closed blob locale-cache | 00:51 |
DocScrutinizer05 | I tried to get rid of zc, ru, whatnot else | 00:51 |
DocScrutinizer05 | no way | 00:51 |
DocScrutinizer05 | there's a 23MB blob in your / which you need like 2% of its content but you can't do anything and cant do without it | 00:52 |
DocScrutinizer05 | ask timeless, he published his own improved translations - or was it crashanddie? | 00:53 |
kerio | is mmcblk0 faster than emmc? | 00:54 |
kerio | er | 00:54 |
kerio | is mtd faster than mmcblk0 | 00:54 |
zeq | DocScrutinizer05: I don't give up too easily ;) | 00:55 |
DocScrutinizer05 | kerio: allegedly yes - slightly | 00:58 |
kerio | why can't we install everything there? | 00:58 |
DocScrutinizer05 | because it's only 200MB? | 00:59 |
kerio | er | 00:59 |
kerio | i meant on the emmc | 00:59 |
kerio | i can't believe there's no way to make a kernel that can use a rootfs on the emmc | 00:59 |
zeq | One good thing about mtd, in theory it should be possible to use XIP, no idea if it is used, or whether it's even possible in this case. I've not looked into it for years. | 01:01 |
zeq | XIP == eXecute In Place | 01:01 |
kerio | ooooh, that's neat | 01:01 |
zeq | same applies for data, it can be directly mapped on the storage | 01:02 |
zeq | this doesn't mean to say it's an overall win, or possible in this case though :) | 01:03 |
kerio | hmm, i wonder why the swap on mmc is used, given that there's plenty of space on the sd swap and it has a higher priority | 01:04 |
DocScrutinizer05 | hah? | 01:05 |
kerio | *emmc | 01:05 |
kerio | DocScrutinizer05: swap on mmcblk0p3, priority 0, used 16mb, swap on mmcblk1p2, priority 10, used 16mb | 01:06 |
DocScrutinizer05 | zeq: I'm not sure about XIP. I think even ROMBL loads mtd xloader/NOLO to ram | 01:07 |
DocScrutinizer05 | zeq: don't you mix that with NOR flash? | 01:08 |
zeq | There's no meaningful shortage of RAM on the device, and it's going to be a lot quicker than XIP, so that could be the reason. XIP is a cool technology though. | 01:08 |
kerio | i suppose the kernel is doing this because it's faster | 01:08 |
zeq | DocScrutinizer05: I think you're right | 01:08 |
kerio | but fucking hell, i would expect priorities to be respected | 01:08 |
zeq | As I said, it's been a while since I looked into it | 01:08 |
zeq | 5 years | 01:09 |
zeq | + | 01:09 |
DocScrutinizer05 | NOR can XIP, NAND not though, afaik | 01:09 |
zeq | Sounds right | 01:09 |
kerio | i wonder if compcache is actually worth it | 01:11 |
zeq | kerio: probably depends on workload. | 01:12 |
DocScrutinizer05 | nothing can replace more ram, except even MORE ram | 01:13 |
kerio | DocScrutinizer05: yeah but since you still haven't found a way to get more ram, we do what we can with what we have | 01:14 |
DocScrutinizer05 | though... since my laptop with 8GB RAM I actually think sometimes you can have *too much* RAM | 01:14 |
kerio | nonsense, you can't have too much RAM | 01:15 |
DocScrutinizer05 | I *had* too much RAM... for that 32bit OS | 01:15 |
kerio | DocScrutinizer05: what's your opinion on swap on uSD? | 01:16 |
kerio | too slow compared to emmc? | 01:16 |
DocScrutinizer05 | nasty | 01:16 |
DocScrutinizer05 | only for hardcore freaks | 01:16 |
kerio | ignore the backcover removal peril potential | 01:16 |
kerio | hm | 01:16 |
kerio | i tried it a bit | 01:16 |
kerio | compcache+swap on SD | 01:17 |
kerio | but... i'm noticing slowdowns on fucking SOLITAIRE | 01:17 |
kerio | wait, why "only for hardcore freaks"? | 01:17 |
DocScrutinizer05 | you already mentioned one of the main reasons | 01:18 |
kerio | maybe it's just compcache! | 01:18 |
* zeq is off to bed | 01:18 | |
* DocScrutinizer05 meant backcover | 01:18 | |
kerio | enabling both would be like striping, and it would increase speed! and stuff! | 01:18 |
kerio | oh, i see | 01:18 |
kerio | well, there's the Estel_ route of gluing a magnet on the sensor | 01:18 |
kerio | :D | 01:18 |
zeq | kerio: if you want to stripe you need to have both swaps at the same priority | 01:19 |
kerio | i could try that | 01:19 |
kerio | i wonder if it'll be faster than just emmc | 01:19 |
DocScrutinizer05 | well, I'd rather start messing with sys/class/gpio/* rather than doing that | 01:19 |
kerio | considering that emmc is still a lot faster than uSD | 01:19 |
zeq | anyway, goodnight guys | 01:20 |
kerio | night | 01:20 |
DocScrutinizer05 | n8 | 01:20 |
Estel_ | kerio, mtd isn't faster than emmc, sadly | 01:20 |
Estel_ | due to compression | 01:20 |
Estel_ | it is technically, but compression kills it | 01:20 |
kerio | yeah but ubifs is the shit | 01:20 |
Estel_ | freemangordon did extensive tests on that | 01:20 |
Estel_ | ubifs is ok, but nokia decided to enable transparent compression | 01:20 |
kerio | did someone try a lighter fs? | 01:20 |
Estel_ | ubifs without comrpession would be OK | 01:21 |
kerio | hm, can it be disabled? | 01:21 |
Estel_ | well, nokia did it, because before Pr 1.2 or P1.1 whole system ended up i nrootfs | 01:21 |
Estel_ | they hoped for no need for optification | 01:21 |
kerio | yeah right | 01:21 |
Estel_ | Pr 1.1 or 1.2 (can't remember) killed this ideas | 01:21 |
Estel_ | but compression remained | 01:21 |
Estel_ | well, considering that thumb makes 70-100 MB fre eon rotfs for average user... | 01:21 |
Estel_ | and that moving things to opt for increased performance actually makes sense for some things... | 01:22 |
Estel_ | disabling compression is an option to be considered | 01:22 |
kerio | i don't actually know how ubifs works | 01:22 |
Estel_ | it wil lfor sure be investigated more after thumb | 01:22 |
Estel_ | it just works ;l0 | 01:22 |
DocScrutinizer05 | mkfs.ubi /dev/mtd4 | 01:22 |
Estel_ | BTw, as for patches | 01:22 |
kerio | no, i mean how to make a ubifs volume | 01:22 |
kerio | DocScrutinizer05: i thought it had weird stuff wrt volumes | 01:22 |
kerio | it makes a device in /dev/ | 01:22 |
Estel_ | no idea if it will land on cssu-testing or cssu-devel or cssu-whatever, but for sure will land in cssu-thumb fork :) | 01:22 |
Estel_ | (zeq patches) | 01:22 |
kerio | can the transparent compression be transparently disabled? | 01:23 |
Estel_ | never tried, no idea, but disabling it and restoring from backupmenu should work | 01:23 |
Estel_ | honestly, you could do everything from backupmenu | 01:23 |
Estel_ | i.e. creating filesystem ubifs without compression and extracting rootfs | 01:23 |
kerio | oh right, backupmenu does a rm -rf and then a copy | 01:23 |
Estel_ | but BEFORE doing so, you should optify what is worth to be optified | 01:23 |
kerio | it doesn't remake the fs, for root | 01:23 |
kerio | isn't stuff optified by default anyway? | 01:23 |
Estel_ | to be 100% sure that without compression files will fit into rootfs | 01:23 |
Estel_ | but | 01:24 |
kerio | also i only have 55mb of free space | 01:24 |
kerio | :( | 01:24 |
DocScrutinizer05 | dafaq, mkfs.whateveryoulike /dev7mtd4 | 01:24 |
kerio | although ubifs disk space is weird as shit | 01:24 |
kerio | because of the compression | 01:24 |
DocScrutinizer05 | dev/mtd4 even | 01:24 |
Estel_ | kerio, ubifs is for now only sensible filesystem to use wioth rootfs | 01:24 |
kerio | DocScrutinizer05: explain /dev/ubi0, /dev/ubi0_0 and /dev/ubi_ctrl | 01:24 |
Estel_ | and few closed bits expect it there during boot | 01:24 |
Estel_ | so don't try with other filesystems | 01:24 |
kerio | Estel_: do they expect it to be the transparent flavour? | 01:24 |
Estel_ | probably not. | 01:25 |
Estel_ | they shouldn't care | 01:25 |
kerio | >you could do everything from backupmenu | 01:25 |
kerio | is backupmenu like zombo.com? | 01:25 |
Estel_ | well, 256 nAND mtd would be superfast for swapp, but would also live quite short | 01:25 |
Estel_ | kerio, you can enter root console in backuypmenu | 01:25 |
Estel_ | reformat mtd with ubifs without compression | 01:26 |
Estel_ | and then restore rootfs backup, one done before, AFTER optification | 01:26 |
Estel_ | BTw, check your rootfs, I have 70 MB free without any effort | 01:26 |
Estel_ | i.e. doing nothing special to have it this way | 01:26 |
DocScrutinizer05 | mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 mtd4 | 01:26 |
Estel_ | DocScrutinizer, why (trendy word) trolling? | 01:27 |
kerio | Estel_: used space? | 01:27 |
Estel_ | kerio, free space | 01:27 |
Estel_ | 70 mb free | 01:27 |
kerio | no, i mean | 01:27 |
kerio | what's your used? | 01:27 |
DocScrutinizer05 | Estel_: seventeen *exponent* shouldn't% | 01:27 |
Estel_ | 256-70? | 01:27 |
kerio | haha | 01:27 |
kerio | size 228M, used 169M, avail 55M | 01:28 |
kerio | somehow i think it'll fit | 01:28 |
kerio | :D | 01:28 |
Estel_ | should be 59 free | 01:28 |
Estel_ | but estimations | 01:28 |
Estel_ | some things like free are estimations in compressed | 01:28 |
kerio | yeah | 01:28 |
DocScrutinizer05 | meh, lamers | 01:28 |
Estel_ | and I would be carefull with "should fit", as some files have awesome compression ratio in ubifs | 01:29 |
Estel_ | for example 10 MB file becomes 100 kb (and it's not /dev/zero derived ;) ) | 01:29 |
Estel_ | kerio, easy experiment | 01:29 |
kerio | Estel_: i think 169 are true 169MB | 01:29 |
Estel_ | take Your rootfs backup tar | 01:29 |
kerio | hold on, i'll DU root | 01:29 |
kerio | oh right | 01:29 |
Estel_ | and unpack it | 01:29 |
kerio | i also have that | 01:29 |
Estel_ | on normal fs | 01:29 |
Estel_ | check space used | 01:29 |
Estel_ | job done | 01:29 |
* DocScrutinizer05 checks actual size of mtd4, checks what fs to format it with, and which speedtests to run on top of that fs | 01:30 | |
Estel_ | speedtest kills nand quite quickly | 01:30 |
Estel_ | wouldn't stress it, if others did it already | 01:30 |
kerio | DocScrutinizer05: apparently to properly use ubifs you need ubiattach and all sorts of shit | 01:30 |
Estel_ | kerio, google for ubifs, maybe there is a flag to just disable compression ;) | 01:31 |
Estel_ | but I doubt it | 01:31 |
Estel_ | but with thumb and speed advantage due to upgraded compiler, priority now is for compiling whatever possible and applicable with thumb | 01:31 |
DocScrutinizer05 | kerio: rtfm | 01:31 |
Estel_ | i.e. 3rd party programs, too, from extras and friends | 01:32 |
kerio | just -x none in mkfs-ubifs | 01:32 |
kerio | mkfs.ubifs | 01:32 |
DocScrutinizer05 | Estel_: I kill and torture my mtd4 as long as I like | 01:32 |
Estel_ | kerio, might want to check actual flags used for current ubifs on N900 | 01:32 |
kerio | DocScrutinizer05: you monster | 01:32 |
DocScrutinizer05 | it's made for that | 01:32 |
Estel_ | and then, use same flags, EXCEPT compression | 01:33 |
DocScrutinizer05 | or do YOU use an initrd on your N900? | 01:33 |
Estel_ | to re-create it | 01:33 |
kerio | i don't even know where they are | 01:33 |
Estel_ | mount | 01:33 |
Estel_ | ;) | 01:33 |
Estel_ | mount as root | 01:33 |
kerio | oic | 01:33 |
kerio | rw,bulk_read,no_chk_data_crc | 01:34 |
*** Estel_ is now known as sfdsdgfsd | 01:34 | |
*** sfdsdgfsd is now known as Estel_ | 01:34 | |
*** Estel_ is now known as asdasdas | 01:34 | |
DocScrutinizer05 | mtd4 is *unused* empty space! (well h-e-n copies all your p0rn from your attached memsticks to it, but otherwise it's never touched) | 01:35 |
*** asdasdas is now known as Estel_ | 01:35 | |
kerio | DocScrutinizer05: can't we use it? | 01:35 |
kerio | for... idk, rootfs | 01:35 |
kerio | it's like 200000 somethings | 01:36 |
DocScrutinizer05 | mkfs.ubi mtd4 | 01:36 |
kerio | i meant for rootfs | 01:36 |
kerio | does mtd have a partition table or something | 01:36 |
DocScrutinizer05 | incredible how LOUD I have to shout until you start to notice I have something to say | 01:37 |
* kerio hugs DocScrutinizer05 and apologizes | 01:38 | |
kerio | i meant merging the partitions or something | 01:38 |
DocScrutinizer05 | flasher and NOLO have partitioning hardcoded | 01:38 |
kerio | sizes too? .s | 01:38 |
kerio | :s | 01:38 |
kerio | shame | 01:38 |
kerio | but ok | 01:38 |
kerio | why the fuck is there a "rootfs.jffs2" in my COMBINED | 01:39 |
DocScrutinizer05 | ubi been jffs2 | 01:40 |
DocScrutinizer05 | for sentimental reasons the names stayed | 01:40 |
kerio | i see | 01:40 |
DocScrutinizer05 | or rather, for "flasher got it hardcoded, as does NOLO" reasons | 01:41 |
kerio | am i a bad enough dude to try this uncompressed rootfs thing? | 01:44 |
DocScrutinizer05 | I guess they used maemo4 and installed it to first n900 prototypes. then each crew started from that and wrote improved versions of their own little gem, so it would still work in m4 environment. One group made a better rootfs but called it rootfs.jffs2 so none of the other bits breaks. another crew made a NOLO that could also handle USB charging, but otherwise did everything like N8x0 NOLO | 01:45 |
DocScrutinizer05 | and that's where initrd/mtd4 comes from | 01:46 |
DocScrutinizer05 | though it *never* been used for maemo5 | 01:46 |
kerio | meh, i'm not a bad enough dude to try it | 01:50 |
kerio | although apparently there's a way to disable compression for single files | 01:50 |
kerio | i wonder how flasher makes the FS | 01:51 |
Estel_ | it's still not sure if nolo cares for sizes, though | 01:52 |
Estel_ | flasher remakes them for sure | 01:52 |
kerio | DocScrutinizer05: anyway, mkfs.ubifs /dev/mtd4 won't work at all | 01:53 |
Estel_ | but having working device and no need for flasdhing it via flasher - at least closed one - maybe you could alter size, just keep layout | 01:53 |
Estel_ | jsut a wild guess | 01:53 |
kerio | ubifs can be mkfsed on an ubi volume | 01:53 |
kerio | which is a layer between mtd and ubifs | 01:53 |
Estel_ | so create ubi volume first ;) but I would try with changing size of whole aprtition layout on NAND, just keep layout itself (adjusted sizes) | 01:53 |
kerio | i don't think there even *is* a partition | 01:54 |
Estel_ | werll, don't expect that DocScrutinizer will hug and apologize You in return ;) | 01:54 |
kerio | *partition layout | 01:54 |
Estel_ | it must be, as You have this mtd4 | 01:54 |
Estel_ | unused | 01:54 |
Estel_ | and our nice compressed ubifs used | 01:54 |
kerio | yeah but it's something passed to the kernel at boot time | 01:54 |
Estel_ | doesn't matter, it doesn't reformat it on every boot. Just use linux trools for handling ubifs and mtd | 01:55 |
Estel_ | changte sizes | 01:55 |
Estel_ | recreate ubifs | 01:55 |
Estel_ | extract rootfs | 01:55 |
kerio | no, i mean | 01:55 |
kerio | it's something that gets passed to the kernel on every boot | 01:55 |
Estel_ | s/trools/tools/ | 01:55 |
Estel_ | lol | 01:55 |
kerio | or it's compiled into the kernel | 01:55 |
kerio | it's not something that you can change easily | 01:55 |
Estel_ | and there is a chance that it doesn't care about sizes :) I hope | 01:55 |
kerio | ooh | 01:58 |
kerio | even easier | 01:58 |
kerio | like, a lot easier | 01:59 |
kerio | In UBIFS it is possible to enable or disable compression individually for each inode by setting or cleaning their compression flag. Note, the compression flag of directories is inherited, which means that when files and sub-directories are created, they inherit the compression flag of the parent directory. | 01:59 |
kerio | i don't know if i'm a bad enough dude to disable compression on / | 02:04 |
kerio | Estel_: wanna try? | 02:04 |
Estel_ | i'm dist-upgrading Ed now on device | 02:05 |
Estel_ | jsut unpack your rootfs basckup on normal filesystem | 02:05 |
Estel_ | and check it's sizes | 02:05 |
kerio | yeah, it's probably faster | 02:05 |
Estel_ | if it will fit, disable compression | 02:06 |
kerio | 313M | 02:06 |
kerio | :s | 02:06 |
Estel_ | if not, optifity something, until it will | 02:06 |
kerio | how big is the nand? | 02:06 |
kerio | er | 02:06 |
kerio | yeah, the nand | 02:06 |
Estel_ | see biggest files, and optify them | 02:06 |
kerio | aww, 256mb | 02:06 |
kerio | shame | 02:06 |
Estel_ | 228 in fact | 02:07 |
Estel_ | probably due to unused mtd4 | 02:07 |
kerio | hehe | 02:07 |
Estel_ | nand is 256 indeed | 02:07 |
kerio | nah, initfs is 0x200000 bytes | 02:07 |
kerio | just 2mb | 02:07 |
Estel_ | 200kb? | 02:07 |
kerio | oh, maybe | 02:07 |
kerio | no, 2mb | 02:07 |
kerio | says /proc/mtd | 02:08 |
Estel_ | so something else sits there :) not unexpected, like bits for nolo? | 02:08 |
Estel_ | well | 02:08 |
kerio | it's just wasted i think | 02:08 |
Estel_ | optify, until You will fit there :) | 02:08 |
Estel_ | in 228, not 2 mb | 02:08 |
kerio | nah, i'll pass | 02:08 |
Estel_ | kerio, Yuou use kp51r1, yep? | 02:08 |
kerio | yeah | 02:08 |
Estel_ | when you use Pali's charging module, does the diode keep green one like during emergency charging, despite, in fact, charging normally? | 02:09 |
Estel_ | it annoys me, that it stays even with screen on | 02:09 |
Estel_ | and I have own pattern for charging ;) | 02:09 |
Estel_ | I wonder if this is not implemented yet, or a bug | 02:09 |
kerio | \_o_/ i still use bme | 02:09 |
Estel_ | test is easy and safe | 02:09 |
Estel_ | as root | 02:09 |
kerio | i like my battery indicator | 02:09 |
Estel_ | stop bme | 02:09 |
Estel_ | sure, but during hostmode You doj't use bme | 02:09 |
kerio | of course not | 02:09 |
Estel_ | also, Pali's drivber, just like charge.sh by shadowJk, allow to charge my uber battery with higher current ;) | 02:10 |
kerio | hehe | 02:10 |
Estel_ | so, if You could | 02:10 |
Estel_ | stop bme | 02:10 |
kerio | oh shit it's 1am | 02:10 |
Estel_ | modprobe bq2415x_charger | 02:10 |
Estel_ | and connect charger | 02:10 |
Estel_ | then tell me about diode | 02:10 |
Estel_ | after finishing, rmod or modprobe -r bq2415x_charger | 02:11 |
Estel_ | and start bme | 02:11 |
kerio | i'm gonna go to sleep actually :3 | 02:11 |
Estel_ | hm, can't do short test? Pity, but I won't stop You | 02:12 |
kerio | Estel_: diode stays on constantly | 02:13 |
kerio | probably because it's the best the kernel can do, it can't really know your mce patterns | 02:14 |
Estel_ | it can | 02:14 |
Estel_ | it know during hostmode and booston :) | 02:14 |
kerio | well of course it can, it's the fucking kernel | 02:14 |
Estel_ | I think it's jsut not implemented yet | 02:14 |
Estel_ | thanks a lot for confirming | 02:14 |
kerio | Estel_: no, it's usbmode.sh that does the cool light shows | 02:14 |
kerio | during hostmode | 02:14 |
Estel_ | right | 02:14 |
Estel_ | no problem to have something that does light shows during charging, in userspace | 02:15 |
Estel_ | after all, booston is kernel situation, too | 02:15 |
Estel_ | i.e. kernel does it, and something take lightshow | 02:15 |
Estel_ | in stock bme does it | 02:15 |
Estel_ | in oour case <whatever> may do it too ;) | 02:15 |
Estel_ | and this dioe is eatin 0.5 mA of my charging current! :P: | 02:16 |
kerio | haha | 02:16 |
kerio | anyway, wait for pali's userland bme replacement | 02:16 |
kerio | hm, i just realized that there would've been a problem with decompressing rootfs by recreating the fs | 02:19 |
kerio | the ubitools are on the rootfs | 02:19 |
*** jonwil has joined #maemo-ssu | 03:22 | |
*** arcean has quit IRC | 03:55 | |
*** nox- has quit IRC | 04:09 | |
*** LaoLang_cool has joined #maemo-ssu | 04:33 | |
*** jonwil has quit IRC | 04:58 | |
*** amiconn has quit IRC | 05:31 | |
*** amiconn_ has joined #maemo-ssu | 05:31 | |
*** amiconn_ is now known as amiconn | 05:31 | |
*** zeq1 has joined #maemo-ssu | 07:54 | |
*** _rd has joined #maemo-ssu | 08:41 | |
*** _rd has quit IRC | 08:50 | |
*** LaoLang_cool has quit IRC | 09:14 | |
*** jonwil has joined #maemo-ssu | 09:50 | |
*** Pali has joined #maemo-ssu | 09:52 | |
zeq1 | as long as the SDK stays as is, is there a good reason not to have a more recent glibc on device? | 10:22 |
*** Pali has quit IRC | 10:27 | |
zeq1 | freemangordon: I've held back putting the cssu0 build up until I know what's happening with the linux-header dependencies. | 10:30 |
*** Pali has joined #maemo-ssu | 10:30 | |
*** infobot has quit IRC | 10:32 | |
*** luf has joined #maemo-ssu | 11:14 | |
luf | cssu version fixed in curl, obexd and eds in git. | 11:14 |
kerio | yo freemangordon | 11:51 |
kerio | wtf is tinymail | 11:51 |
kerio | and do i have to reboot after upgrading it? | 11:51 |
Pali | library for modest | 11:54 |
kerio | ohai pali | 11:57 |
kerio | Pali: Estel_ was a bit confused by the values in the sysfs entries of your battery module | 12:00 |
kerio | what units are those? why 1800 for 950? | 12:03 |
kerio | or something | 12:03 |
Pali | kerio, all values are in standard units (mAh, mV, mA, ...) | 12:05 |
kerio | yeah but the charging current was 1800 for Estel_ | 12:06 |
Pali | cat registers and check them in datasheet | 12:06 |
Pali | also rmmod driver and use i2c get/set | 12:07 |
Pali | and ask Doc about this problem | 12:07 |
Pali | I ported charger shell script with info from datasheet | 12:07 |
Pali | check if same problem is i charger script | 12:08 |
kerio | i see | 12:08 |
*** luf has left #maemo-ssu | 12:15 | |
*** Pali has quit IRC | 12:45 | |
*** macmaN has quit IRC | 13:08 | |
*** macmaN has joined #maemo-ssu | 13:12 | |
*** andre__ has quit IRC | 13:43 | |
*** lizardo has joined #maemo-ssu | 13:54 | |
freemangordon | zeq1: ok | 13:58 |
zeq1 | freemangordon: I've confirmed again it is working fine with 4.2, trying with 4.6... | 13:59 |
freemangordon | 4.6? | 13:59 |
*** Pali has joined #maemo-ssu | 14:01 | |
zeq1 | linaro-4.6 | 14:01 |
zeq1 | rebooting now | 14:02 |
*** zeq1 has quit IRC | 14:06 | |
zeq | freemangordon: linaro-4.6 does exactly the same thing. I | 14:15 |
freemangordon | and it is? | 14:15 |
zeq | I can't believe the generated locale format is compiler dependent, so it must be something not building correctly on later compilers. | 14:16 |
freemangordon | well, lets stick to 4.2.1 for libc then | 14:17 |
zeq | That why I asked ^ whether we *really* need to stick to this old version for anything other than the SDK | 14:17 |
freemangordon | old version of? libc? | 14:17 |
zeq | yeah | 14:18 |
zeq | what stops us moving closer to Debian? | 14:18 |
freemangordon | well, i have no clue what will break if we do that | 14:18 |
freemangordon | really, having pselect working should be enough, even without newer compiler | 14:19 |
zeq | it's a separate thing for sure | 14:19 |
freemangordon | :nod: | 14:19 |
zeq | There must be security vulnerabilities in this version at least? | 14:20 |
freemangordon | sure, but we don't know about them | 14:20 |
zeq | (libc I mean) | 14:20 |
freemangordon | but it has nothing to do with the compiler | 14:20 |
freemangordon | s/but/and/ | 14:21 |
freemangordon | ~ping | 14:21 |
freemangordon | hmm | 14:21 |
zeq | no, but being able to get Maemo closer to a debian distro will pick these things up for free. | 14:21 |
freemangordon | but we don;t know what will break. though libc should be ABI compatible. | 14:21 |
zeq | there *shouldn't* be breakage (in theory) | 14:22 |
freemangordon | yeah | 14:22 |
zeq | but won't know without testing | 14:22 |
zeq | If I get my dev device I think I will make an experimental target and get as close as i can | 14:22 |
*** LaoLang_cool has joined #maemo-ssu | 14:22 | |
freemangordon | zeq: well, lets split that. | 14:22 |
freemangordon | 1. pselect with 4.2.1 | 14:23 |
freemangordon | 2. libc upgrade | 14:23 |
freemangordon | agree? | 14:23 |
zeq | absolutely! | 14:23 |
freemangordon | deal :) | 14:23 |
zeq | do we keep --without-fp in the glibc rebuild? | 14:24 |
zeq | I think we should probably if only there must have been some reason to have it...? | 14:24 |
freemangordon | hmm, did you find any rationale for that? | 14:24 |
zeq | no | 14:24 |
zeq | but there must be one | 14:24 |
zeq | surely | 14:24 |
freemangordon | glibc? or libc? | 14:24 |
zeq | it's in the armel.mk of the Maemo glibc build -> libc6 | 14:25 |
freemangordon | zeq: ^^^ | 14:25 |
zeq | it's actually eglibc afaik | 14:25 |
freemangordon | zeq: though it might be there by historical reasons | 14:25 |
zeq | could be | 14:25 |
zeq | support ARM CPUs without FP hw | 14:25 |
freemangordon | because AFAIK on N8X0 vfp is veeery slow | 14:26 |
zeq | or very slow FP indeed | 14:26 |
freemangordon | not sure how it is on A8, but should not be that bad | 14:26 |
zeq | might be worth benching with and without? | 14:26 |
freemangordon | hmm, can we do some kind of benchmark? | 14:26 |
freemangordon | :D | 14:26 |
freemangordon | yeah, do you know any? | 14:27 |
zeq | there must be a libm benchmark.... | 14:27 |
zeq | I'll google | 14:27 |
freemangordon | ok | 14:27 |
*** infobot has joined #maemo-ssu | 14:30 | |
*** ChanServ sets mode: +v infobot | 14:30 | |
freemangordon | thanks doc | 14:31 |
DocScrutinizer05 | yw | 14:31 |
zeq | freemangordon: nbench should be ok i think | 14:33 |
freemangordon | ok. will you do that? | 14:33 |
DocScrutinizer05 | ~+uptime | 14:33 |
infobot | - Uptime for purl - | 14:33 |
infobot | Now: 3m 17s running infobot 1.5.4 (SVN) -- linux | 14:33 |
infobot | 1: 59d 8h 41m 19s running infobot 1.5.4 (SVN) -- linux, ended Sun Nov 14 18:39:57 2010 | 14:33 |
infobot | 2: 57d 3h 9m 23s running infobot 1.5.4 (SVN) -- linux, ended Fri Jun 26 20:39:27 2009 | 14:33 |
infobot | 3: 36d 20h 47m 14s running infobot 1.5.4 (SVN) -- linux, ended Tue Aug 4 17:38:59 2009 | 14:33 |
freemangordon | zeq: hmm, stupid question | 14:33 |
DocScrutinizer05 | ~+chaninfo | 14:33 |
infobot | 124/124 channels, 5033 users, 3618 unique: #debian/1182, #kde/353, #maemo/284, #gsoc/237, #meego/214, #asterisk/188, #harmattan/188, #oe/151, #wowuidev/138, #wowace/114, #webos-internals/111, #openmoko/95, #wowhead/88, #arm-netbook/86, #openmoko-cdevel/73, #uclibc/71, #n9/70, #utah/69, #bzflag/61, #edev/61, #htc-linux/59, #tomcat/56, #qi-hardware/54, #curseforge/51, #maemo-ssu/50, #asterisk-dev/45, #diablofans/43, #elinux/42, #flow3/41, ... | 14:33 |
zeq | what is? :) | 14:33 |
freemangordon | it is only you who can do it right now :D | 14:34 |
zeq | :D | 14:34 |
freemangordon | nbech with/out VFP | 14:34 |
zeq | true | 14:34 |
freemangordon | nbench | 14:34 |
*** mickname has quit IRC | 14:36 | |
DocScrutinizer05 | whenever somebody notices infobot being unresponsive or even not joined, please ping me about it | 14:37 |
DocScrutinizer05 | *sometimes* I can do sth about it | 14:37 |
freemangordon | :) | 14:37 |
DocScrutinizer05 | ~+status | 14:39 |
infobot | Since Wed Aug 1 11:30:17 2012, there have been 0 modifications, 2 questions, 0 dunnos, 0 morons and 3 commands. I have been awake for 8m 43s this session, and currently reference 118668 factoids. I'm using about 18484 kB of memory. With 0 active forks. Process time user/system 3.08/0.15 child 0/0 | 14:39 |
DocScrutinizer05 | thanks tim riker we got purl aka infobot aka ibot aka apt | 14:44 |
infobot | DocScrutinizer05: de rien | 14:44 |
*** gri has quit IRC | 14:44 | |
freemangordon | WTF? I am running my device without OC for several hour without noticing it. hmmm, not sure if it is good or bad thing :) | 14:45 |
freemangordon | hours even | 14:45 |
DocScrutinizer05 | definitely a good thing, it proves OC is placebo | 14:45 |
freemangordon | no, it proves thumb2 is really good thing :P | 14:46 |
DocScrutinizer05 | hey, probably that too | 14:46 |
freemangordon | ;) | 14:46 |
DocScrutinizer05 | and then it demonstrates OC << thumb | 14:46 |
freemangordon | but of course | 14:46 |
freemangordon | that is for sure, no CPU can replace free RAM | 14:47 |
DocScrutinizer05 | I'm still amazed about that reported speed gain of thumb | 14:47 |
freemangordon | DocScrutinizer05: you won;t believe it | 14:47 |
DocScrutinizer05 | I'm already pondering to experience it myself | 14:47 |
freemangordon | but I think it is a combination of less code and newr compiler | 14:47 |
DocScrutinizer05 | :nod: | 14:49 |
DocScrutinizer05 | since I just come to think of it: you usually compile bug proving code with -O0 | 14:49 |
freemangordon | yeah | 14:50 |
freemangordon | though not necesserily | 14:50 |
DocScrutinizer05 | yeah, for bugs in compiler optimization you use whatever is needed ;-P | 14:51 |
* jonwil is still no closer to figuring out the mysteries of the universe (or rather, the mysteries of osso-wlan-security) :( | 14:51 | |
DocScrutinizer05 | for demonstrating any other bugs elsewhere you either use -O0 or assembler ;-) | 14:52 |
DocScrutinizer05 | generally it's a good idea to look at assembler listing anyway | 14:53 |
DocScrutinizer05 | to make sure the binary does what you planned it shall do | 14:53 |
DocScrutinizer05 | for bug testing code | 14:53 |
freemangordon | :nod: | 14:53 |
freemangordon | Pali: ping | 14:54 |
Pali | pong | 14:56 |
freemangordon | Pali: could you confirm tinymail bug fixed? | 14:56 |
freemangordon | Pali: in -thumb bersion | 14:57 |
zeq | freemangordon: just for fun (and because it was ready to go) here is linaro-4.6 compiled libc6 --with-fp: http://pastebin.com/mfATnev5 compare to sulu's nbench results for Easy Debian: http://talk.maemo.org/showthread.php?t=34550&page=263 | 14:58 |
*** arcean has joined #maemo-ssu | 14:59 | |
freemangordon | Pali: also, could you look at https://gitorious.org/community-ssu/tinymail/blobs/1955bad9fa612e7313ad753726e8cda68f14d0a8/libtinymail-camel/camel-lite/camel/providers/imap/camel-imap-store.c#line2717 | 14:59 |
freemangordon | I still have a feeeling the code is buggy | 15:00 |
freemangordon | zeq: ok | 15:00 |
zeq | should add, that's thumb too | 15:00 |
Pali | freemangordon, fixed | 15:00 |
zeq | much faster than the ED results | 15:00 |
DocScrutinizer05 | now who's going to prepare a generic "backup" containing all the "good stuff" *apps* we like to offer to our beloved customers? | 15:01 |
freemangordon | zeq: yeah | 15:02 |
DocScrutinizer05 | idea: download the "backup" somewhere, and restore, picking which apps to restore/install and which not | 15:02 |
freemangordon | Pali: does not matter, please take a look at do-while cycle, i think if there is only one "item" to add, the list will contain 2 elements | 15:03 |
freemangordon | i.e. item-> next will point to item | 15:03 |
freemangordon | and that will cause a memory corruption. sorry, I am sick these days and cannot think :( | 15:04 |
freemangordon | zeq: yeah, deffinitely better. I am curious how it will be with 4.2.1 | 15:06 |
zeq | I'm rebuilding to make sure I have everything set correctly | 15:07 |
freemangordon | ok | 15:07 |
freemangordon | zeq: could you look at the tinymail code above | 15:07 |
freemangordon | seems Pali is too busy now | 15:08 |
zeq | ok | 15:08 |
zeq | the while loop? | 15:08 |
freemangordon | yep | 15:09 |
freemangordon | what will happen if we have only one element added to the list | 15:09 |
zeq | I'm not sure what the data looks like...? | 15:11 |
freemangordon | does not matter | 15:11 |
freemangordon | my point is that if we add only one element (i.e. item), then its "next" member will point to itself | 15:12 |
freemangordon | i.e. item->next==item | 15:12 |
zeq | that's true | 15:12 |
freemangordon | thats a nasty bug, look at the code that frees the list | 15:13 |
zeq | I take it, it's possible to have only one item be added | 15:14 |
zeq | ? | 15:14 |
freemangordon | yes | 15:14 |
freemangordon | https://gitorious.org/community-ssu/tinymail/blobs/1955bad9fa612e7313ad753726e8cda68f14d0a8/libtinymail-camel/camel-lite/camel/providers/imap/camel-imap-store.c#line2620 | 15:15 |
freemangordon | zeq: ^^^ | 15:15 |
freemangordon | the heap will be raped :D | 15:15 |
zeq | ouch | 15:15 |
zeq | This has hit people? | 15:16 |
freemangordon | well, heap corruption. who knows :D | 15:17 |
zeq | looks like it just needs an extra check somewhere | 15:17 |
freemangordon | tail->next = item; should be "else tail->next = item;" | 15:18 |
freemangordon | aiui | 15:18 |
freemangordon | correct? | 15:18 |
zeq | yes, otherwise the tail=item is redundant anyway. must have been intended | 15:19 |
freemangordon | that comes from my last commit, it was even uglier before that | 15:20 |
zeq | oh | 15:20 |
DocScrutinizer05 | (item-> next will point to item) classical single linked list, empty :-) | 15:20 |
freemangordon | not empty unfortunately, with one element pointing to iself | 15:21 |
freemangordon | *itself | 15:21 |
DocScrutinizer05 | for a double linked list both item->next and item->prev point to item when empty | 15:21 |
freemangordon | DocScrutinizer05: look at the code that frees the list, it is not done like that | 15:22 |
DocScrutinizer05 | each linked list consists of a root elem that's always 'virtual' and never holds real data | 15:22 |
freemangordon | ^^^ | 15:22 |
DocScrutinizer05 | then the implementation is brainfsckd ;-D | 15:22 |
freemangordon | :nod: | 15:22 |
freemangordon | this is how it was: | 15:23 |
freemangordon | https://gitorious.org/community-ssu/tinymail/commit/1955bad9fa612e7313ad753726e8cda68f14d0a8 | 15:23 |
freemangordon | somehow that was working with gcc 4.2.1 | 15:23 |
freemangordon | newer compilers resulted in NULL being returned | 15:24 |
freemangordon | insted of pointer to first element | 15:24 |
zeq | hmmm | 15:24 |
freemangordon | ? | 15:25 |
zeq | items = NULL; | 15:25 |
zeq | tail = (struct imap_status_item *) &items; | 15:25 |
zeq | ? | 15:25 |
freemangordon | yes :D | 15:25 |
zeq | wtf | 15:25 |
freemangordon | brainfcked code, though I can understand what was the intention | 15:25 |
zeq | maybe it meant if items != NULL? | 15:26 |
freemangordon | first time tail will hold the address of items | 15:26 |
freemangordon | and it will be assigned the first element address | 15:26 |
freemangordon | but gcc 4.6.2 and 4.7.2 cannot deal with that | 15:27 |
zeq | probably because it's broken ;) | 15:27 |
freemangordon | totally hackish and unreadable code | 15:28 |
freemangordon | yes | 15:28 |
freemangordon | anyway, goping to fix it | 15:28 |
freemangordon | *going | 15:28 |
zeq | ok :) | 15:28 |
zeq | I'll get back to rebuild libc6 | 15:28 |
zeq | handy thing I can just static link nbench to the SDK libc6 :) | 15:32 |
*** Pali has quit IRC | 15:33 | |
kerio | DocScrutinizer05: try thumb! it's great! | 15:34 |
* DocScrutinizer05 sucks his thumb | 15:34 | |
kerio | and then provide a service to upgrade the n900's ram! | 15:34 |
* kerio doesn't know if it's even theoretically possible | 15:34 | |
DocScrutinizer05 | sure, I'll ship the .deb tomorrow | 15:34 |
DocScrutinizer05 | in theory we could add "ultrafast" RAM_as_uSD | 15:35 |
kerio | how? | 15:35 |
kerio | do they even exist? | 15:35 |
DocScrutinizer05 | adapter | 15:35 |
DocScrutinizer05 | anyway, telco waiting for me, cya l8r | 15:36 |
kerio | http://www.hwtools.net/ExtenderBoard/TFEX.html | 15:36 |
zeq | bye doc | 15:36 |
kerio | looks small | 15:37 |
zeq | freemangordon: can't we just say the non-sdk linux-kernel-*-headers "Provides: linux-kernel-headers" ? | 15:39 |
kerio | Estel_: pls add a slot for DDR3 ram in your replacement aluminium cases | 15:40 |
zeq | otherwise we need "... | ...power...| ...cssu..." | 15:40 |
zeq | in the libc6 control | 15:41 |
freemangordon | zeq: i think that is provided, lemme check | 15:42 |
freemangordon | no, it is not :( | 15:43 |
zeq | bug eh? | 15:43 |
freemangordon | zeq: but what is the problem to build it with SDK headers? | 15:43 |
freemangordon | wel,, not sdk, but maemo-extras | 15:44 |
zeq | you end up without any pselect() :P | 15:44 |
zeq | at all! | 15:44 |
freemangordon | why? is it something in headers? | 15:44 |
zeq | yes, the headers hold the syscall tables | 15:44 |
freemangordon | can't recall what your patch is doing | 15:44 |
freemangordon | aah, ok | 15:44 |
freemangordon | hmm, it will take a while | 15:45 |
zeq | for some reason, if you tell glibc to use a later min kernel, it assumes pselect() et al is provided without checking and silently builds libc without it entirely if it isn't in the syscall list. | 15:46 |
freemangordon | hehe | 15:46 |
zeq | rather unfortunate :) | 15:46 |
freemangordon | zeq: lemme finish with tinymail, we'll need Pali too | 15:46 |
zeq | ok | 15:46 |
amiconn | freemangordon, zeq: Afaik the "fp" in --without-fp means old arm hardfp, which no modern arm cpu supports. So this option makes sense. Using vfp otoh is a good idea (if not already in use) | 15:52 |
amiconn | Afaik binaries compiled with hardfp do run on arm cores without hardfp support, but very slowly because hardfp will be emulated through exceptions (of course the emulation code needs to be present) | 15:54 |
freemangordon | amiconn: ok, sounds sane | 15:55 |
amiconn | Our arm cortex-a8 should support vfp, as it definitely does support neon, and iirc these two features are closely related | 15:57 |
freemangordon | vfp is supported for sure | 15:57 |
freemangordon | amiconn: why "old", there is some hand written assembly code? | 16:00 |
*** mase76 has joined #maemo-ssu | 16:01 | |
amiconn | I mean FPA, which is not binary compatible with VFP | 16:04 |
freemangordon | ok | 16:04 |
amiconn | And what relation is there between "old" and "assembly"? | 16:04 |
zeq | if it's FPA support mode, why isn't it called such ??? | 16:04 |
zeq | EABI dropped support for FPA afaik | 16:05 |
zeq | how can the switch do *anything* | 16:05 |
zeq | ? | 16:05 |
amiconn | Hmm, check the gcc --help output? | 16:06 |
zeq | amiconn: it's a glibc build switch | 16:06 |
freemangordon | amiconn: it was a question, i.e "is there some hand written assembly code?" | 16:06 |
amiconn | Code for what? | 16:07 |
zeq | freemangordon: I suspect there is since I get exactly the same nbench result with ARM gcc-4.2 as with THUMB linaro-4.6 compiled libcs | 16:07 |
freemangordon | zeq: what ./configure --help gives about that switch? | 16:07 |
zeq | freemangordon: about to check that ;) | 16:07 |
freemangordon | amiconn: scratch that :) | 16:07 |
amiconn | ok | 16:08 |
zeq | --with-fp if using floating-point hardware [default=yes] | 16:08 |
freemangordon | nice. now, WTF is that disabled? | 16:09 |
freemangordon | zeq: who is the maintainer? | 16:09 |
amiconn | Hmm, not very descriptive. What type of floating point hardware? | 16:09 |
zeq | *any* probalby | 16:09 |
amiconn | I guess a quick check would tell | 16:09 |
zeq | probably | 16:09 |
amiconn | (compile using --with-fp and check with objdump) | 16:10 |
zeq | amiconn: I'm in the process of compiling with-fp and without-fp for benchmarking purposes :) | 16:10 |
zeq | it is possible that it's a win to use softfloat in libm, who knows? | 16:11 |
freemangordon | zeq: but your device will be busted if libc does a SIGILL | 16:11 |
zeq | freemangordon: that's okay because I'm just installing it in the SDK and building nbench with -static :) | 16:11 |
freemangordon | aah, ok :) | 16:11 |
zeq | Juha Kallioinen <juha.kallioinen@nokia.com> <- maintainer | 16:15 |
zeq | I guess that means unmaintained? | 16:15 |
*** lizardo has quit IRC | 16:16 | |
freemangordon | zeq: well, unmaintained, but we can ask him why | 16:17 |
freemangordon | if he can recall | 16:17 |
freemangordon | zeq: but on the other hand if it works we can skip that | 16:17 |
*** lizardo has joined #maemo-ssu | 16:18 | |
zeq | it definitely works, it's on my device ;) | 16:18 |
freemangordon | zeq: built with no --without-fp? | 16:18 |
zeq | yeah, it was an accident | 16:19 |
freemangordon | hehe, a brave one :D | 16:19 |
zeq | I backed up /lib before playing with all this | 16:19 |
zeq | so I could always use LD_LIBRARY_PATH | 16:19 |
freemangordon | thats in case you can boot | 16:20 |
freemangordon | and I am not sure you can with broken libc6 | 16:20 |
zeq | other than the messed up locales, it's easy enough to test before rebooting :) | 16:20 |
zeq | the locales only break after reboot | 16:20 |
zeq | when I tested the other day, and thought it was working, I hadn't fully realised | 16:21 |
freemangordon | zeq: hmm, isn't dpkg-configure all needed? | 16:21 |
freemangordon | or something like that, to rebuild the locales | 16:21 |
zeq | ? | 16:22 |
freemangordon | yeah, lemme check | 16:22 |
zeq | ok | 16:22 |
freemangordon | dpkg-reconfigure locales | 16:22 |
zeq | interesting | 16:23 |
freemangordon | not sure that works on maemo though | 16:23 |
freemangordon | hmm, no dpkg-reconfigure | 16:23 |
zeq | nope | 16:24 |
zeq | another good reason to get closer to debian | 16:24 |
freemangordon | zeq: BTW how are the locales broken? | 16:25 |
zeq | well, if using (apparently) any gcc other than 4.2 on reboot the UI translations are missing | 16:25 |
zeq | just symbolic labels appear for all the UI elements | 16:26 |
zeq | makes some things very hard to use! | 16:27 |
*** luf has joined #maemo-ssu | 16:27 | |
freemangordon | missing as in? there is nothing in /usr/whatever_the_locale_are or gettext fails? | 16:27 |
zeq | remember my what's that button do question the other day! | 16:27 |
freemangordon | yes | 16:27 |
zeq | the locales are provided by posix-locales pkg | 16:27 |
zeq | in /usr/lib/locales | 16:28 |
zeq | or is it /usr/lib/locale | 16:28 |
zeq | anyway it's just a huge bin only pkg containing the generated locales | 16:28 |
zeq | (from nokia) | 16:28 |
freemangordon | locale-archive? | 16:29 |
*** Pali has joined #maemo-ssu | 16:29 | |
zeq | yes | 16:29 |
freemangordon | a huge 20MB file? | 16:29 |
zeq | on rootfs | 16:29 |
zeq | :) | 16:30 |
kerio | :s | 16:30 |
kerio | can it be optified? | 16:30 |
freemangordon | so, the file is there, but cannot be read? | 16:30 |
zeq | for some *unknown* reason | 16:30 |
kerio | also Pali: can kp51 mount from the emmc before requiring anything in /usr/? | 16:30 |
zeq | I don't see an error, is there a log I can check? | 16:31 |
zeq | obviously it's not going to be in dmesg | 16:31 |
freemangordon | syslog? | 16:31 |
Pali | and ubifs is rootfs | 16:31 |
zeq | silly question, but where is syslog kept? | 16:31 |
freemangordon | "/var/log/syslog" | 16:32 |
zeq | honestly? | 16:32 |
zeq | hehe | 16:32 |
freemangordon | but you should install sysklogd | 16:32 |
zeq | ah | 16:32 |
zeq | maybe that's why I didn't know | 16:32 |
freemangordon | and reboot after that, so everyone is aware there is syslog daemon | 16:32 |
zeq | ok, that explains it | 16:32 |
zeq | I don't currently have a /var/log | 16:33 |
zeq | installing now | 16:33 |
kerio | can /usr/ be moved away from the rootfs? | 16:33 |
zeq | Well, I'll give it another go later, see if there are any errors | 16:34 |
kerio | /etc/mtab is linked /proc/mounts, and unmount tries to change /etc/mtab which causes warnings. | 16:34 |
kerio | fukken messybox again? | 16:34 |
kerio | ...wait, why does my sbin/preinit have something like a menu to choose between flash and mmc boot? :o | 16:35 |
kerio | oh, r&d mode | 16:36 |
freemangordon | Pali: I pushed another fix to tinymail. | 16:38 |
kerio | what the hell, the boot process requires /usr/sbin/chroot | 16:39 |
kerio | i really can't follow what happens in preinit | 16:39 |
Pali | freemangordon, ok | 16:39 |
zeq | I'm thinking nbench isn't a good way to benchmark libm | 16:42 |
freemangordon | why? | 16:42 |
zeq | I'm getting the same results, within ~1% | 16:43 |
zeq | even less than 1% actually | 16:44 |
freemangordon | hmm, strange, with or without FP | 16:45 |
freemangordon | ? | 16:45 |
zeq | between every test | 16:45 |
freemangordon | aaah | 16:45 |
zeq | with-fp without-fp and linaro-4.6 | 16:45 |
zeq | diff says the bins differ though | 16:45 |
zeq | what's a good fp instruction to look for in ARM disassembly? | 16:46 |
zeq | I can't even think what the register symbols are | 16:47 |
freemangordon | well, check sin/cos, whatever trigonometric function, it should be done with one instruction AFAIK | 16:47 |
zeq | good idea :) | 16:47 |
zeq | I'll look in the debug lib | 16:48 |
freemangordon | or even better fsin/fcos | 16:48 |
zeq | vcvt.f64.f32 d7, s16 | 16:50 |
zeq | ? | 16:50 |
zeq | fp? | 16:50 |
freemangordon | yep | 16:50 |
zeq | perhaps without-fp is ignored? | 16:51 |
freemangordon | could be | 16:51 |
zeq | shouldn't be there anyway though | 16:51 |
zeq | and it must be producing different code ... | 16:52 |
zeq | otherwise the binaries should have matched | 16:52 |
zeq | actually the libm bins do match | 16:54 |
zeq | for some reason the static nbench bin didn't | 16:55 |
kerio | Pali: what modules are needed to access the emmc? | 17:00 |
Pali | omap hsmmc | 17:01 |
Pali | + deps | 17:01 |
kerio | are they big? | 17:01 |
Pali | I do not know | 17:02 |
kerio | hm, omap_hsmmc.ko is 21k | 17:03 |
*** DocScrutinizer has quit IRC | 17:06 | |
*** DocScrutinizer has joined #maemo-ssu | 17:06 | |
*** DocScrutinizer05 has quit IRC | 17:06 | |
*** DocScrutinizer05 has joined #maemo-ssu | 17:06 | |
*** LinuxCode has joined #maemo-ssu | 17:09 | |
*** LaoLang_cool has quit IRC | 17:13 | |
*** LaoLang_cool has joined #maemo-ssu | 17:32 | |
*** LaoLang_cool has left #maemo-ssu | 17:42 | |
*** Estel_ has quit IRC | 18:28 | |
*** Estel_ has joined #maemo-ssu | 18:37 | |
*** Pali has quit IRC | 18:47 | |
*** ivgalvez has joined #maemo-ssu | 18:54 | |
*** Pali has joined #maemo-ssu | 18:58 | |
*** NIN101 has joined #maemo-ssu | 19:01 | |
*** jonwil has quit IRC | 19:13 | |
*** Pali has quit IRC | 19:20 | |
*** Pali has joined #maemo-ssu | 19:23 | |
*** ivgalvez has quit IRC | 19:28 | |
*** Pali has quit IRC | 19:37 | |
*** Pali has joined #maemo-ssu | 19:38 | |
*** mase76 has quit IRC | 20:05 | |
*** mase76 has joined #maemo-ssu | 20:27 | |
*** mase76 has quit IRC | 20:27 | |
zeq | I've (re-) found a really good reason to upgrade glibc. | 20:38 |
zeq | C99 inline semantics | 20:38 |
zeq | as discussed on tmo: | 20:38 |
zeq | http://talk.maemo.org/archive/index.php/t-81861.html | 20:38 |
kerio | <freemangordon> afaik THUMB is broken on the N900's SoC, and the workaround's overhead > benefit of THUMB | 20:51 |
kerio | LOL | 20:51 |
kerio | i guess that programmers really *are* awful at guessing where's the slowness | 20:53 |
*** luf has left #maemo-ssu | 20:53 | |
zeq | You don't know until you try ;) | 20:57 |
*** Estel_ has quit IRC | 21:14 | |
*** Estel_ has joined #maemo-ssu | 21:14 | |
*** taziff has quit IRC | 21:49 | |
*** taziff has joined #maemo-ssu | 21:49 | |
DocScrutinizer51 | well, producing decent working code is craftmanship | 21:55 |
DocScrutinizer51 | producing decent working *readable* code is an art | 21:56 |
DocScrutinizer51 | producing decent working readable *optimized* code that results in an optimauzed fast system, this is the dark science of magic spells | 21:57 |
*** arcean_ has joined #maemo-ssu | 22:00 | |
*** arcean has quit IRC | 22:03 | |
*** xnt14 has joined #maemo-ssu | 22:11 | |
*** xnt14 has quit IRC | 22:12 | |
*** xnt14 has joined #maemo-ssu | 22:13 | |
*** xnt14 has quit IRC | 22:17 | |
*** xnt14 has joined #maemo-ssu | 22:17 | |
*** MohammadAG has joined #maemo-ssu | 22:25 | |
*** xnt14 has quit IRC | 22:25 | |
*** xnt14 has joined #maemo-ssu | 22:26 | |
*** xnt14 is now known as xnt14_ | 22:29 | |
*** xnt14_ is now known as xnt14i | 22:29 | |
*** xnt14i is now known as xnt14 | 22:29 | |
*** xnt14 is now known as _xnt14 | 22:30 | |
*** Pali has quit IRC | 22:32 | |
*** _xnt14 has joined #maemo-ssu | 22:39 | |
*** arcean_ is now known as arcean | 22:41 | |
amiconn | If it's both working and fast, why does it need to be readable? :P | 22:46 |
DocScrutinizer05 | that's one of the magic words nobody outside the magic circle knows the meaning of: maintainability | 22:55 |
DocScrutinizer05 | one of the foundation concepts of our black magic | 22:55 |
DocScrutinizer05 | it's about heraitage and honour of your ancestors | 22:56 |
DocScrutinizer05 | transfer of sekrit knowledge to the next genration, and stuff like that | 22:57 |
DocScrutinizer05 | so basically nobody outside the magic circle needs or wants to know about | 22:57 |
DocScrutinizer05 | but occasionally you witness one of our black magic circle dispute with a charlatan at the IRC market, and then you can hear the old wise man shout something like "man 3p kill" and the charlatan falls to the ground hiurt deadly while the guru just turns away and walks outa the scene - then you witnessed the powers of the true dark science of magic spells | 23:08 |
*** _rd has joined #maemo-ssu | 23:09 | |
kerio | hmm, what's 3p? | 23:09 |
DocScrutinizer05 | pshhhhh that's sekrit! | 23:09 |
kerio | oh, posix 2001 | 23:10 |
kerio | or something weird like that | 23:10 |
*** infobot has quit IRC | 23:15 | |
amiconn | Black magicians are able to deal with code that looks unreadable to normal programmers ;) | 23:19 |
kerio | DocScrutinizer05: anyway... | 23:22 |
kerio | who fixes the bugs in the decent working readable optimized code that results in an optimauzed (sic!) fast system? | 23:22 |
*** taziff has quit IRC | 23:30 | |
*** taziff has joined #maemo-ssu | 23:30 | |
DocScrutinizer05 | kerio: obviously such code is bugfree per definitionem | 23:31 |
kerio | ooh, tinymail updates again | 23:32 |
kerio | i don't even use modest | 23:33 |
DocScrutinizer05 | ( optimauzed (sic!) ) demonstrate that you tought your cat a better typing style ;-P | 23:33 |
amiconn | lol | 23:34 |
kerio | optimauzed is yours | 23:34 |
kerio | and "tought", really | 23:34 |
DocScrutinizer05 | curse KDE4 bullshit, do you think it's possible to switch the spellchecker to en_EN in konversation's input field permanently? | 23:36 |
DocScrutinizer05 | mauze is what my cat types when he suddenly feels hungry | 23:37 |
*** _rd has quit IRC | 23:38 | |
kerio | oh, btw | 23:38 |
kerio | who recompiles the decent working readable optimized code that results in an optimized system as thumb2? | 23:39 |
DocScrutinizer05 | then he runs out and comes back with some mize | 23:39 |
DocScrutinizer05 | I told him to avoid thoise friggin warez sitez | 23:40 |
DocScrutinizer05 | recompile is gentoo | 23:41 |
DocScrutinizer05 | gentoo even nears to other OS | 23:42 |
Estel_ | DocScrutinizer, Pali to9ld me to contact You, about strange behavior of sysfs entries in bq2415x_charger module (sic!) | 23:46 |
Estel_ | as method of gathering them is, reportedly, derived from your work | 23:46 |
Estel_ | mind excuse wood faeries for a while, and look at it?;) | 23:46 |
Estel_ | and before You ask, I was also surprised that you're one to help with module written by Pali (despite methods used) | 23:47 |
*** NIN101 has quit IRC | 23:51 | |
*** lizardo has quit IRC | 23:56 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!