**** BEGIN LOGGING AT Sat Feb 06 02:59:58 2016
Feb 06 05:49:09 <capitanocrunch> DocScrutinizer51:  strange,conky says i have 938MB free on /home
Feb 06 05:52:26 <capitanocrunch> so why maemo-optify is complaining about not enough space at least 200MB needed
Feb 06 06:06:32 <Vajb_> capitanocrunch: just a thought, but maybe it tries to move something big and it needs 200mb more
Feb 06 06:43:19 <capitanocrunch> it seems noting big to move in /etc/maemo-optify-auto-opt.conf
Feb 06 07:18:27 <Maxdamantus> Just avoid optification and put everything on eMMC in the first place.
Feb 06 09:42:03 <freemangordon> Pali: whi is 550 MHz disabled?
Feb 06 09:42:29 <Pali> freemangordon: do not remember now...
Feb 06 09:42:44 <freemangordon> Pali: http://pastebin.com/sVFgySwx
Feb 06 09:43:03 <freemangordon> also, what about that latency, is it the same on n900 as on generic omap?
Feb 06 09:44:01 <Pali> we have some table in comment in board-rx51*.c files
Feb 06 09:44:12 <Pali> which contains these informations
Feb 06 09:45:06 <freemangordon> Pali: alos, you should enable    CONFIG_CPUFREQ_DT, disable CONFIG_ARM_OMAP2PLUS_CPUFREQ and make CONFIG_THERMAL=y. Please, enable  CONFIG_KERNEL_MODE_NEON while at it
Feb 06 09:45:10 <freemangordon> *also
Feb 06 09:45:23 <freemangordon> ok, will look at it
Feb 06 09:46:25 <Pali> why is CONFIG_THERMAL=y needed? is not CONFIG_THERMAL=m enough?
Feb 06 09:46:26 <freemangordon> Pali: that table is for cpuidle C-states, not cpufreq
Feb 06 09:46:52 <Pali> then you need to dig table from kernel-power
Feb 06 09:46:53 <freemangordon> Pali: "# if CPU_THERMAL is on and THERMAL=m, CPUFREQ_DT cannot be =y"
Feb 06 09:46:57 <Pali> ok
Feb 06 09:47:10 <Pali> rx51 has different tables as generic omap
Feb 06 09:47:25 <Pali> we have some patches for legecy board code in linux-n900 tree
Feb 06 09:47:49 <freemangordon> saw no latency there, but will look agan
Feb 06 09:48:49 <Pali> fc858246d6e5735efb64016e4a453face1228c11 and df5c1e60a5213755d31631b7c12bf33bc1a6bb70
Feb 06 09:48:58 <Pali> thats alls what is in linux-n900 tree
Feb 06 09:50:10 <freemangordon> yep. is that latency used at all in older kernels?
Feb 06 09:51:23 <Pali> this was is in linux-n900 tree for a long time
Feb 06 09:51:34 <Pali> but do not know if it matches data from kernel-power
Feb 06 09:51:56 <freemangordon> where is that tada in KP, I can't find it
Feb 06 09:52:12 <freemangordon> I am not talking about C-states :)
Feb 06 09:54:43 <freemangordon> Pali: also, please merge my onenand patch, until we have a better fix from tony
Feb 06 09:54:55 <Pali> ok
Feb 06 09:54:56 <freemangordon> very bad things happen without it :)
Feb 06 10:01:31 <Pali> freemangordon: are we really able to use CONFIG_ARM_OMAP2PLUS_CPUFREQ?
Feb 06 10:01:49 <Pali> CONFIG_CPUFREQ_DT and not CONFIG_ARM_OMAP2PLUS_CPUFREQ?
Feb 06 10:12:56 <freemangordon> Pali: at least in qemu
Feb 06 10:13:13 <freemangordon> Pali: see omap2plus_defconfig
Feb 06 10:13:37 <Pali> ok, it is there enabled....
Feb 06 10:13:37 <freemangordon> it has CONFIG_CPUFREQ_DT=y and CONFIG_ARM_OMAP2PLUS_CPUFREQ disabled
Feb 06 10:14:13 <freemangordon> Pali: do you know what is the diference?
Feb 06 10:14:34 <freemangordon> as I suspect ONFIG_ARM_OMAP2PLUS_CPUFREQ cannot be used with DT
Feb 06 10:16:38 <Pali> you are probably, rigth, see arch/arm/mach-omap2/pm.c
Feb 06 10:16:47 <Pali> function omap_init_cpufreq
Feb 06 10:17:19 <freemangordon> :nod:
Feb 06 10:17:52 <Pali> patches pushed to linux-n900 tree
Feb 06 10:18:25 <freemangordon> Pali: also, why we have g_nokia and ko built-in?
Feb 06 10:18:34 <freemangordon> *co
Feb 06 10:18:49 <Pali> nfs usb root without initfs
Feb 06 10:19:01 <freemangordon> hmm
Feb 06 10:19:02 <Pali> it provides usb networking
Feb 06 10:19:11 <freemangordon> I know about usb net
Feb 06 10:19:24 <Pali> anyway, g_nokia now contains also removable mass storage subgadget
Feb 06 10:19:27 <freemangordon> the point is that with those built-in kernel is too big for qemu
Feb 06 10:19:42 <freemangordon> at least for the scripts you gave me
Feb 06 10:19:50 <Pali> use qflasher -v -m rx51 -p k=4094,i=2
Feb 06 10:20:06 <Pali> it will increase size of kernel partition and decrease initfs
Feb 06 10:20:10 <freemangordon> ok
Feb 06 10:26:14 <freemangordon> Pali: also, please put that bqxxx name patch
Feb 06 10:34:16 <ceene> Pali: which tag are you developing on now? 4.5-rc1?
Feb 06 10:34:31 <Pali> freemangordon: bq patch pushed
Feb 06 10:34:45 <Pali> ceene: v4.5-rc1-n900
Feb 06 10:34:53 <freemangordon> Pali: thanks
Feb 06 10:35:18 <freemangordon> Pali: any ide why "omap2_set_init_voltage: unable to find boot up OPP for vdd_core"
Feb 06 10:35:27 <Pali> no idea
Feb 06 10:44:43 * jonwil has just about figured out all the steps necessary to update the root certificate store in CSSU
Feb 06 10:45:35 <jonwil> Just need to figure out how to change maemo-security-certman so that things will do what they are supposed to do on Scratchbox (where /etc is a symlink)
Feb 06 10:58:22 <freemangordon> Pali: cpufreq_dt seems to work on the device too :)
Feb 06 10:58:31 <Pali> great
Feb 06 10:58:59 <freemangordon> at least there is stats with some values, going to check the current
Feb 06 11:00:31 <freemangordon> POWER_SUPPLY_CURRENT_NOW=92463
Feb 06 11:00:52 <freemangordon> POWER_SUPPLY_CURRENT_NOW=91749
Feb 06 11:01:13 <freemangordon> I guess that's fine, given that deep idle states does not work, still
Feb 06 11:05:03 <freemangordon> Pali: there is something weird for SR nvalues, all of them are different
Feb 06 11:14:49 <Wizzup> jonwil: good news
Feb 06 11:20:27 <freemangordon> Pali: http://pastebin.com/nYGWAAwA :)
Feb 06 11:39:30 <ceene> still no mmc :(
Feb 06 11:39:36 <ceene> [    6.122467] omap_hsmmc 4809c000.mmc: omap_device: omap_device_enable() called from invalid state 1
Feb 06 11:39:39 <ceene> [    6.217590] omap_hsmmc 480b4000.mmc: omap_device: omap_device_enable() called from invalid state 1
Feb 06 11:39:42 <ceene> [    8.263732] mmc0: card never left busy state
Feb 06 11:39:44 <ceene> [    8.288024] mmc0: error -110 whilst initialising SD card
Feb 06 11:41:11 <Maxdamantus> Maybe you shouldn't've used it for swap.
Feb 06 11:41:27 <ceene> this sd card works fine on kernel power
Feb 06 11:41:32 <ceene> but doesn't on pali's kernel
Feb 06 11:41:59 <Maxdamantus> Oh, it's just an SD card then?
Feb 06 11:42:16 <ceene> well, i haven't tested any other sd card
Feb 06 11:46:18 <ceene> ah, you thought i was talking about internal eMMC
Feb 06 11:49:09 <ceene> i wanted to place there some rootfs, but since the kernel doesn't see it, i just can't :/
Feb 06 11:50:11 <Maxdamantus> Does it see the eMMCL
Feb 06 11:50:14 <Maxdamantus> s/L$/?/
Feb 06 11:52:50 <ceene> i can't see it now either
Feb 06 11:52:57 <ceene> i think i could with 4.4-rc7
Feb 06 11:53:23 <ceene> uhm, no i couldn't either
Feb 06 11:53:40 <ceene> what i do see is mtd devices
Feb 06 11:54:30 <Maxdamantus> which are fairly independent from the eMMC/SD card.
Feb 06 11:54:33 <ceene> yep
Feb 06 11:55:23 <ceene> i hadn't thought about emmc as i didn't want to use it for now
Feb 06 11:59:47 <freemangordon> ceene: you've pulled the latest from github?
Feb 06 12:00:42 <freemangordon> ceene: also, make sure to power-off before booting mainline kernel
Feb 06 12:00:54 <ceene> v4.5-rc1-n900, just an hour ago or so
Feb 06 12:01:01 <freemangordon> pulll again
Feb 06 12:01:15 <freemangordon> Pali pushed some stuff
Feb 06 12:01:18 <ceene> ok
Feb 06 12:01:36 <freemangordon> ceene: eMMC works fine here
Feb 06 12:02:06 <ceene> is there any module i have to load or something like that?
Feb 06 12:02:20 <ceene> i'm not booting into maemo, just to a small custom rootfs
Feb 06 12:02:42 <freemangordon> ceene: I boot maemo, so don't know
Feb 06 12:03:09 <freemangordon> lemme check
Feb 06 12:03:56 <freemangordon> ceene: no, all the sd/mmc stuff seems to be built-in
Feb 06 12:04:12 <freemangordon> ceene: but, it is very important to power-off befor booting
Feb 06 12:04:19 <ceene> i even removed battery
Feb 06 12:04:28 <freemangordon> otherwise some i2c bus get stuck
Feb 06 12:04:53 <ceene> i can try to boot my other n900
Feb 06 12:05:01 <freemangordon> shouldn;t make difference
Feb 06 12:05:13 <ceene> well, it works on yours
Feb 06 12:05:17 <ceene> why doesn't on mine?
Feb 06 12:05:29 <freemangordon> ceene: did you do make mrproper recently?
Feb 06 12:05:51 <ceene> i'm rebuilding totally from scratch right now
Feb 06 12:06:12 <freemangordon> ceene: branch v4.5-rc1-n900?
Feb 06 12:06:16 <ceene> yes
Feb 06 12:06:47 <freemangordon> ceene: add http://pastebin.com/sVFgySwx as well, otherwise cpufreq won;t work
Feb 06 12:06:50 <freemangordon> afaik
Feb 06 12:07:03 <ceene> ok, i don't mind much right now about that
Feb 06 12:07:12 <freemangordon> :)
Feb 06 12:07:25 <freemangordon> ceene: do you want me to share my kernel?
Feb 06 12:08:00 <ceene> sure, let's give it a try
Feb 06 12:08:00 <freemangordon> to see if it will have eMMC working?
Feb 06 12:08:08 <freemangordon> ok, just a minute
Feb 06 12:09:32 <freemangordon> ceene: http://46.249.74.23/linux/
Feb 06 12:11:18 <ceene> downloading, thanks
Feb 06 12:18:19 <ceene> uhm, i need to tweak my rootfs to load your kernel
Feb 06 12:18:32 <ceene> do you mind if we continue at some other time?
Feb 06 12:18:32 <freemangordon> hmm?
Feb 06 12:18:36 <freemangordon> sure
Feb 06 12:18:39 <ceene> i'm late to eat lunch with the in-laws
Feb 06 12:18:40 <ceene> :&
Feb 06 12:18:41 <ceene> :/
Feb 06 12:19:14 <ceene> it has booted your kernel but is complaining about /dev/null not existing
Feb 06 12:19:34 <freemangordon> but does mmc work?
Feb 06 12:19:41 <ceene> don't know
Feb 06 12:19:44 <ceene> can't do anything at all
Feb 06 12:19:45 <ceene> :)
Feb 06 12:19:45 <freemangordon> ok
Feb 06 12:19:53 <ceene> since there are no dev entries, nothing really works
Feb 06 12:19:59 <freemangordon> ok
Feb 06 12:20:03 <ceene> just console output complaining about that around 100 times per second
Feb 06 12:20:08 <freemangordon> :)
Feb 06 12:20:41 <ceene> i may be able to tinker a bit this afternoon, but don't know for sure
Feb 06 12:20:51 <ceene> thanks, cya later!
Feb 06 12:33:33 <Maxdamantus> mount -t devtmpfs dev /dev
Feb 06 13:16:37 <joga> hmm.. is it possible to use n900 as a microsd card reader over usb, ie. expose the card if I insert one in to the slot and plug it in?
Feb 06 13:17:57 <joga> I need to write an image to a microsd but can't find my card adapters *anywhere* and the laptops that have readers only take in a typical size sd card :) 
Feb 06 13:18:50 <freemangordon> joga: yes, you can
Feb 06 13:19:29 <joga> are there any caveats, I remember I've had crashes when trying to write too fast to mydocs
Feb 06 13:19:42 <freemangordon> none I am aware of
Feb 06 13:19:55 <joga> guess I could try it..
Feb 06 13:29:51 <jonwil> bah, I cant seem to figure out the behaviour of glibc when you call chdir(blah) where blah is a symlink to something else
Feb 06 14:21:46 <DocScrutinizer05> jonwil: >>  ELOOP  Too many symbolic links were encountered in resolving path. << I can't deduce anything special about symlinks from this error
Feb 06 14:22:42 <jonwil> I ran a test and it looks like yes, in the environment I have with the glibc I have etc, chdir does resolve symlinks
Feb 06 14:24:49 <DocScrutinizer05> yeah sure, there's also >>Depending on the file system, other errors can be returned.<<
Feb 06 14:25:12 <DocScrutinizer05> see FAT
Feb 06 14:25:54 <DocScrutinizer05> well, FAT is an inappropriate example since you won't even _find_ any symlinks on it
Feb 06 14:26:40 <DocScrutinizer05> so unlike trying to create a symlink, for chdir() symlinks on FAT are a non-issue
Feb 06 14:27:33 <DocScrutinizer05> I recently wondered if a symlink on ext* can point to a VFAT file. I guess "why not"
Feb 06 14:44:22 <DocScrutinizer05> POSIX seems identical
Feb 06 14:44:56 <DocScrutinizer05> (man 3p chdir)
Feb 06 14:46:26 <DocScrutinizer05> what I would love is a option to ls that colors path parts that are actually symlinks
Feb 06 14:47:46 <DocScrutinizer05> well, I made lstree which does something similar
Feb 06 14:48:28 <DocScrutinizer05> http://paste.opensuse.org/51236423
Feb 06 14:49:10 <DocScrutinizer05> http://paste.opensuse.org/82985592
Feb 06 16:09:53 <jonwil> ok, so somehow I gotta figure out just what absolute_pathname in https://github.com/community-ssu/maemo-security-certman/blob/master/lib/security/maemosec_common.cpp is really doing and how to make it do what we want (return paths where the symlinks haven't been dereferenced) all without breaking anything else
Feb 06 16:27:45 <DocScrutinizer05> hmm, I wonder why absolute_pathname should _not_ dereference symlinks, after all this is what it's all about
Feb 06 16:28:25 <DocScrutinizer05> almost, it also changes relative pathname to /-anchored absolute pathname (as the name implies)
Feb 06 16:37:18 <DocScrutinizer05> I'm much more concerned about https://github.com/community-ssu/maemo-security-certman/blob/master/lib/security/maemosec_common.cpp#L220
Feb 06 16:44:51 <DocScrutinizer05> process_name(string& to_this)  returns absolute_pathname() in to_this only if the process is not /usr/bin/maemo-launcher. AIUI otherwise it does return the program **TEXT** (the binary!!) of the <appname>.launch which is launched by maemo-launcher 
Feb 06 16:47:27 <DocScrutinizer05> sprintf(exe_name, "/proc/%ld/cmdline", (long)my_pid);  fd = open(exe_name, O_RDONLY);  if (fd >= 0) {ssize_t len = read(fd, buf, sizeof(buf));	if (len) {to_this = buf;
Feb 06 16:48:21 <DocScrutinizer05> aah oops, sorry. no this should work
Feb 06 16:49:20 <DocScrutinizer05> I had one mental dereference too much ;-)
Feb 06 18:21:52 <ceene> freemangordon: i've been testing different things and it seems it was the poweroff thing you told me about
Feb 06 18:22:07 <ceene> i now see eMMC and external SD card
Feb 06 18:23:15 <ceene> if i reboot and reload the same kernel, they won't show
Feb 06 18:23:53 <ceene> i can poweroff from pali's kernel, boot that kernel again, and then they will appear again
Feb 06 19:57:54 <Maxdamantus> 02:29:51 < jonwil> bah, I cant seem to figure out the behaviour of glibc when you call chdir(blah) where blah is a symlink to something else
Feb 06 19:58:13 <Maxdamantus> chdir in glibc pretty much just directly invokes the chdir system call, which handles them.
Feb 06 20:04:17 <Maxdamantus> This is fucking horrible.
Feb 06 20:04:27 <Maxdamantus> tgtname = (char*)malloc(PATH_MAX);
Feb 06 20:04:34 <Maxdamantus> That memory is never freed.
Feb 06 20:04:55 <Maxdamantus> How is this function actually used?
Feb 06 20:06:13 <Maxdamantus> Unless it's meant to be used with something like boehmgc, but that's horrible (and I highly doubt that's the case).
Feb 06 20:09:53 <Maxdamantus> Well, it's freed in one possible code path.
Feb 06 20:13:30 <Maxdamantus> /proc/%ld/cmdline is the arguments passed to the current program btw, not the link to the binary (that's /proc/%ld/exe)
Feb 06 22:14:19 <fishbulb> hi, I'm getting messages not delivered with yappari
Feb 06 22:14:26 <fishbulb> and I think I might have missed a call
Feb 06 22:14:31 <fishbulb> the 3g modem is on all the time
Feb 06 22:14:46 <fishbulb> is there a way I can test that somehow? some diagnostic tests or whatever?
Feb 06 22:15:42 <Sicelo009N> test what?
Feb 06 22:15:54 <fishbulb> I have no idea. the modem?
Feb 06 22:16:05 <fishbulb> gprs/3g
Feb 06 22:16:11 <fishbulb> that modem
Feb 06 22:16:14 <Sicelo009N> ping?
Feb 06 22:16:40 <Sicelo009N> open other websites, etc.
Feb 06 22:17:20 <fishbulb> yeah that all works
Feb 06 22:17:31 <fishbulb> maybe yappari isn't working I dunno.
Feb 06 22:17:53 <fishbulb> though it does.. pretty much. it's only skipping messages from one chick on an iphone
Feb 06 22:18:14 <Sicelo009N> known problem
Feb 06 22:45:41 <fishbulb> shit
Feb 06 22:45:43 <fishbulb> it is?
Feb 06 22:45:54 <fishbulb> these are pretty important messages to me
Feb 06 22:46:03 <fishbulb> that's just an ongoing problem?
Feb 06 22:50:41 <Sicelo009N> you may have to buy an iphone or android phone then
Feb 06 22:51:37 <Sicelo009N> it's a pretty 'big' problem ... we donm5 currently have full textsecure on N900 ... that's the cause of your problem
Feb 06 22:51:45 <Sicelo009N> *don't
Feb 06 22:53:33 <fishbulb> what is textsecure
Feb 06 22:54:18 <fishbulb> I have no money for a friggen android
Feb 06 22:54:20 <ceene> encryption protocol of whatsapp
Feb 06 22:54:22 <fishbulb> I have a tablet...
Feb 06 22:54:32 <fishbulb> yeah but these aren't encrypted messages
Feb 06 22:54:41 <ceene> they are
Feb 06 22:55:15 <ceene> there are android phones for like 60 bucks
Feb 06 22:55:37 <ceene> second hand shitty ones should cost a fraction of that
Feb 06 22:56:15 <fishbulb> what are?
Feb 06 22:56:27 <fishbulb> you know they are encrypted for sure?
Feb 06 22:57:01 <fishbulb> dude I have this phone partly because I'm broke
Feb 06 22:57:22 <ceene> yes, they are encrypted
Feb 06 22:57:33 <fishbulb> how come half of them work then
Feb 06 22:57:47 <fishbulb> it's just one contact that doesn't
Feb 06 23:04:26 <ceene> things that happen when a protocol is kept secret and undocumented
Feb 06 23:04:37 <ceene> sometimes thing work, sometimes don't
Feb 06 23:19:16 <fishbulb> secret?
Feb 06 23:19:19 <fishbulb> how'd you get this far then?
Feb 06 23:20:07 <fishbulb> it's not like I'm complaining either this is totally awesome, but then one person freaks out if their messages don't go through, because it's like... one of the reasons I can still keep this phone
Feb 07 00:10:29 <fishbulb> I like the improvements too
Feb 07 00:10:44 <fishbulb> the images go to the image directory with the other photos now
Feb 07 00:13:56 <bencoh> ceene: have you looked at the other opensource implems?
Feb 07 00:15:12 <bencoh> (yowsup for example)
Feb 07 00:16:26 <bencoh> (and whatsapp-purple)
Feb 07 00:21:39 <bencoh> ah, looks like whatsapp-purple has the same issue
Feb 07 00:53:53 <lopx> long time
Feb 07 00:53:55 <lopx> hi
Feb 07 00:54:17 <lopx> anyone on? 
Feb 07 01:04:52 <lopx> i know someone is always on :) 
Feb 07 01:26:52 <Muumipeikko> hello?
**** ENDING LOGGING AT Sun Feb 07 02:59:58 2016