**** BEGIN LOGGING AT Thu May 31 02:59:58 2012
May 31 03:35:34 <pabs3> samgee: please change the wiki page to point at the new site qtmoko.sf.net
May 31 03:42:45 <CIA-17> freesmartphone.org: 03morphis 07cornucopia * rfb8ce7ea9b4d 10/fsodeviced/src/plugins/kernel26_rfkill/Makefile.am: fsodeviced: kernel26_rfkill: $(AM_VALAFLAGS) should be used in kernel26_rfkill_la_VALAFLAGS
May 31 03:53:35 <CIA-17> freesmartphone.org: 03morphis 070.11 * rff0f3bed88b2 10cornucopia/fsodeviced/src/plugins/kernel26_rfkill/Makefile.am: fsodeviced: kernel26_rfkill: $(AM_VALAFLAGS) should be used in kernel26_rfkill_la_VALAFLAGS
May 31 03:53:35 <CIA-17> freesmartphone.org: 03morphis 070.11 * r4364968276e5 10cornucopia/fsodeviced/ (ChangeLog configure.ac): fsodeviced: release version 0.11.4
May 31 05:36:09 <CIA-17> SHR: 03Martin.Jansa 07meta-smartphone * r8ebcd2d6a42d 10/meta-fso/recipes-freesmartphone/cornucopia/ (fsodeviced_0.11.3.bb fsodeviced_0.11.4.bb): fsodeviced: upgrade to 0.11.4
May 31 06:10:59 <CIA-17> SHR: 03shr-devel 07buildhistory * r7ed06716cf1a 10/packages/ (453 files in 453 dirs): packages: Build 201205310411 of shr 20120531 for machine nokia900 on opmbuild
May 31 06:33:55 <DocScrutinizer51> PaulFertser: ping
May 31 06:37:59 <CIA-17> SHR: 03shr-devel 07buildhistory * rb312285698ae 10/packages/armv4t-oe-linux-gnueabi/ (51 files in 51 dirs): packages: Build 201205310812 of shr 20120531 for machine om-gta02 on opmbuild
May 31 06:51:20 <JaMa|Off> morphis: is this expected to be missing libfsonetwork in fsonetworkd? http://pastebin.ca/2156426
May 31 06:56:29 <DocScrutinizer51> JaMa|Off: what calls to kernel provided functions will involve overhead penalty for copying data buffers which carry the payload? (e.g. write() )
May 31 06:58:15 <DocScrutinizer51> JaMa|Off: a bit of a fuzzy question - background is I need a way to test run a kernel module (here caif) at max speed, with minimum latency between return of one callto datasend() and next call to same function
May 31 06:58:53 <JaMa|Off> I see, but don't know
May 31 06:59:00 <DocScrutinizer51> :-)
May 31 06:59:05 <DocScrutinizer51> thanks anyway
May 31 06:59:21 <JaMa|Off> heh, yw anytime :)
May 31 07:00:18 <DocScrutinizer51> I'd guess a lot of those basic 'interface' functions (like write() ) invole quite some memcpy()
May 31 07:00:25 <DocScrutinizer51> involve*
May 31 07:01:09 <DocScrutinizer51> while ideally the userland would just pass a pointer to a buffer, and kernel land would directly read from this userland buffer
May 31 07:01:52 <JaMa|Off> and cannot you generate "random" data to send in that kernel module? if it's only about testing max speed
May 31 07:02:17 <DocScrutinizer51> yep, that's what I'm considering to do right now
May 31 07:02:46 <JaMa|Off> and then maybe compare with writing from user space to see if it's limited by those memcpy calls or not
May 31 07:02:51 <DocScrutinizer51> basically run the test loop in a test kernel module
May 31 07:05:41 <CIA-17> SHR: 03shr-devel 07buildhistory * r183ef57781f0 10/packages/armv7a-vfp-neon-oe-linux-gnueabi/fsodeviced/ (45 files in 45 dirs): packages: Build 201205310841 of shr 20120531 for machine nokia900 on opmbuild
May 31 07:05:42 <DocScrutinizer51> the pita is my colleagues haven't updated docs for CAIF: http://lxr.free-electrons.com/source/Documentation/networking/caif/Linux-CAIF.txt
May 31 07:06:43 <DocScrutinizer51> for example it still mentions a Kernel API and caif_kernel.h that got binned long ago
May 31 07:19:41 <DocScrutinizer51> aah, wasn't it gnutoo who asked what's HSI? http://lxr.free-electrons.com/source/drivers/net/caif/caif_hsi.c
May 31 07:22:22 <lindi-> I've done some zero-copy disk to network tests with splice()
May 31 07:23:39 <DocScrutinizer51> lindi-: man 2 splice ?
May 31 07:24:23 <DocScrutinizer51> :nod:
May 31 07:26:26 <lindi-> DocScrutinizer51: yes, I found some bugs and sent a few patches
May 31 07:26:59 <lindi-> DocScrutinizer51:   https://bugzilla.kernel.org/show_bug.cgi?id=24832
May 31 07:27:57 <DocScrutinizer51> >> The initial implementation of this flag  was  buggy:  therefore starting  in  Linux  2.6.21  it  is a no-op (but is still permitted in a splice() call); in the future, a correct implementation may be restored.
May 31 07:28:00 <DocScrutinizer51> ;-P
May 31 07:29:09 <jake42> JaMa|Off: i've put up a log of fsogsmd on http://shr-project.org/trac/ticket/2012#comment:3
May 31 07:31:04 <JaMa|Off> jake42: thanks.. I guess it's that timeout.. ping morphis about if if you see him here
May 31 07:31:26 <jake42> tnx, will do :)
May 31 07:33:06 <JaMa|Off> jake42: probably related to this commit http://git.freesmartphone.org/?p=cornucopia.git;a=commit;h=d6ce0b8bdd48fd07dcf851184a50967d0826fbdf
May 31 07:37:37 <jake42> thanks
May 31 07:38:37 <JaMa|Off> I thought we had this configurable from fsogsmd.conf but doesn't like it
May 31 07:40:41 <JaMa|Off> jake42: ah not this commit, this isn't in 0.11
May 31 07:46:50 <DocScrutinizer51> lindi-: >>  (e.g. run netcat -lp1234 -u > /dev/zero there)  << dev/zero??
May 31 07:47:22 <CIA-17> SHR: 03shr-devel 07buildhistory * r7f83df7a602e 10/packages/om_gta04-oe-linux-gnueabi/shr-version/ (latest shr-version/latest): packages: Build 201205310915 of shr 20120531 for machine om-gta04 on opmbuild
May 31 07:50:55 <DocScrutinizer51> well, does anybody see instantly (and can share) how a test-caif.ko would interface to caif.ko?
May 31 07:51:33 <DocScrutinizer51> http://lxr.free-electrons.com/source/include/net/caif/caif_dev.h seems about socks only
May 31 07:52:14 <DocScrutinizer51> I think I don't want to use socks for test-caif.ko interface to caif.ko
May 31 07:54:57 <DocScrutinizer51> http://lxr.free-electrons.com/source/Documentation/networking/caif/Linux-CAIF.txt  >line47 Using the Kernel API< suggests nice  caif_transmit(&caif_dev, skb); and  .receive_cb = my_receive,
May 31 08:18:26 <JaMa|Off> morphis_: is this expected to be missing libfsonetwork in fsonetworkd? http://pastebin.ca/2156426
May 31 08:18:40 <JaMa|Off> morphis_: and please check http://shr-project.org/trac/ticket/2012
May 31 08:26:08 <CIA-17> SHR: 03shr-devel 07buildhistory * rc2be92f623f4 10/packages/crespo-oe-linux-gnueabi/shr-version/ (latest shr-version/latest): packages: Build 201205310956 of shr 20120531 for machine crespo on opmbuild
May 31 08:26:18 <CIA-17> SHR: 03shr-devel 07buildhistory * rf3d041f4fb7f 10/images/crespo/eglibc/chroot-image/ (5 files): images: Build 201205310956 of shr 20120531 for machine crespo on opmbuild
May 31 08:30:03 <morphis_> JaMa|Off: yes, the libfsonetwork was removed (it was empty)
May 31 08:31:39 <JaMa|Off> ok fine
May 31 08:36:18 <morphis_> hm CME ERROR: 100 is a unknown error
May 31 08:42:24 <morphis_> JaMa|Off: you can reproduce this?
May 31 08:47:45 <JaMa|Off> morphis: yes
May 31 08:54:16 <morphis> even after a reboot?
May 31 08:56:11 <JaMa|Off> I'll reboot tonight and try
May 31 08:58:16 <morphis> tonight?
May 31 09:02:06 <CIA-17> SHR: 03shr-devel 07buildhistory * r4a42bdea48ad 10/packages/palmpre-oe-linux-gnueabi/shr-version/ (latest shr-version/latest): packages: Build 201205311035 of shr 20120531 for machine palmpre on opmbuild
May 31 09:04:16 <JaMa|Off> morphis: I cannot reboot it remotely
May 31 09:04:37 <JaMa|Off> morphis: and I won't be home from work before 9pm
May 31 09:10:26 <morphis> ah ok
May 31 09:11:14 <radekp> CME ERROR: 100 during call?
May 31 09:12:25 <radekp> IIRC it's quite normal to get these errors during CLCC polling, we just ignore them and the CLCC returns good result next time (or after few msecs)
May 31 09:13:16 <radekp> http://lists.ofono.org/pipermail/ofono/2012-March/012525.html
May 31 09:14:57 <radekp> http://git.kernel.org/?p=network/ofono/ofono.git;a=commitdiff;h=ebac438d9325c34bfe22322c6234e25c86e125d7
May 31 09:18:06 <morphis> radekp: good to know
May 31 09:19:54 <radekp> morphis:  it seems that the gta04 modem has some intervals (~100ms) - when the call is ending - when it returns the error 100 instead of good results
May 31 09:34:25 <JaMa|Off> morphis: and any idea about those timeouts jake42 is seeing on gta02?
May 31 09:35:51 <morphis> not really
May 31 09:46:48 <CIA-17> SHR: 03shr-devel 07buildhistory * r0deb23b8774c 10/packages/palmpre2-oe-linux-gnueabi/shr-version/ (latest shr-version/latest): packages: Build 201205311111 of shr 20120531 for machine palmpre2 on opmbuild
May 31 09:47:26 <Q-Master> guys, can anyone help me with sysfs usage from userspace C code?
May 31 09:47:55 <JaMa|Off> morphis: ok, then I'll close 061 just with fsodeviced/frameworkd fix and we'll see if jake42 would be able to reproduce it with 061
May 31 09:49:22 <morphis> JaMa|Off: it would be great if jake42 can reproduce it
May 31 09:49:28 <morphis> otherwise it will be hard to track it down
May 31 09:56:21 <morphis> JaMa|Off: for your report with the CME ERROR 100 I have now a fix
May 31 09:56:23 <morphis> which just needs testing
May 31 10:00:11 <CIA-17> freesmartphone.org: 03morphis 07cornucopia * r29addebadc36 10/fsogsmd/src/lib/at/atcommandqueue.vala: fsogsmd: lib: atcommandqueue: don't write out any response in INFO log level
May 31 10:00:13 <CIA-17> freesmartphone.org: 03morphis 07cornucopia * r481ecca5e6df 10/fsogsmd/src/plugins/modem_option_gtm601/mediators.vala: 
May 31 10:00:13 <CIA-17> freesmartphone.org: fsogsmd: modem_option_gtm601: reimplement CallListCalls mediator to fix strange CLCC behavior
May 31 10:00:13 <CIA-17> freesmartphone.org: For reference see
May 31 10:00:13 <CIA-17> freesmartphone.org: - https://git.kernel.org/?p=network/ofono/ofono.git&a=commitdiff&h=ebac438d9325c34bfe22322c6234e25c86e125d7
May 31 10:00:13 <CIA-17> freesmartphone.org: - http://lists.ofono.org/pipermail/ofono/2012-March/012525.html
May 31 10:00:14 <CIA-17> freesmartphone.org: 03morphis 07cornucopia * r799aecf205d4 10/fsogsmd/ (3 files in 3 dirs): fsogsmd: lib: at: implement first bits of AT+CCFC command
May 31 10:09:16 <JaMa|Off> morphis: that's it ^ ?
May 31 10:09:41 <JaMa|Off> I'll upgrade to git version to test it later
May 31 10:14:43 <morphis> JaMa|Off: it should
May 31 10:14:52 <JaMa|Off> ok
May 31 10:18:13 <CIA-17> freesmartphone.org: 03morphis 07cornucopia * ra5e2fb679ed6 10/fsogsmd/src/plugins/modem_option_gtm601/mediators.vala: fsogsmd: modem_option_gtm601: don't intend to poll endless
May 31 10:18:34 <CIA-17> SHR: 03shr-devel 07buildhistory * r4da118bf5d93 10/packages/armv4t-oe-linux-gnueabi/opkg/ (11 files in 11 dirs): packages: Build 201205311148 of shr 20120531 for machine om-gta02 on opmbuild
May 31 10:18:38 <CIA-17> SHR: 03shr-devel 07buildhistory * rd4b7b3c5b12d 10/images/om_gta02/eglibc/shr-image/ (11 files): images: Build 201205311148 of shr 20120531 for machine om-gta02 on opmbuild
May 31 10:24:24 <JaMa|Off> morphis: what about aurora-images? should I still build them on shr buildhost?
May 31 10:24:30 <jake42> this?
May 31 10:24:30 <jake42> 10:47 < JaMa|Off> morphis: yes
May 31 10:24:30 <jake42> 10:54 < morphis> even after a reboot?
May 31 10:24:30 <jake42> 10:56 < JaMa|Off> I'll reboot tonight and try
May 31 10:24:30 <jake42> 10:58 < morphis> tonight?
May 31 10:24:32 <jake42> 11:02 < CIA-17> SHR: shr-devel buildhistory * r4a42bdea48ad /packages/palmpre-oe-linux-gnueabi/shr-version/ (latest shr-version/latest): packages: Build 201205311035 of shr 20120531 for machine palmpre on opmbuild
May 31 10:24:36 <jake42> 11:04 < JaMa|Off> morphis: I cannot reboot it remotely
May 31 10:24:39 <jake42> 11:04 < JaMa|Off> morphis: and I won't be home from work before 9pm
May 31 10:24:41 <jake42> 11:10 < morphis> ah ok
May 31 10:24:44 <jake42> 11:11 < radekp> CME ERROR: 100 during call?
May 31 10:24:46 <jake42> 11:12 < radekp> IIRC it's quite normal to get these errors during CLCC polling, we just ignore them and the CLCC returns good result next time (or after few msecs)
May 31 10:24:49 <jake42> 11:13 < radekp> http://lists.ofono.org/pipermail/ofono/2012-March/012525.html
May 31 10:24:53 <jake42> 11:14 < radekp> http://git.kernel.org/?p=network/ofono/ofono.git;a=commitdiff;h=ebac438d9325c34bfe22322c6234e25c86e125d7
May 31 10:24:56 <jake42> 11:18 < morphis> radekp: good to know
May 31 10:24:58 <jake42> 11:19 < radekp> morphis:  it seems that the gta04 modem has some intervals (~100ms) - when the call is ending - when it returns the error 100 instead of good results
May 31 10:25:01 <jake42>  [11:21] [jake42(+i)] [3:freenode/#openmoko-cdevel(+Cn)] [Act: 2,4,5,6,7]
May 31 10:25:04 <jake42> [#openmoko-cdeve]
May 31 10:25:07 <jake42> oops
May 31 10:25:09 <jake42> sry.. 
May 31 10:25:20 <morphis> JaMa|Off: no you don't need to build them
May 31 10:25:31 <jake42> puttys nice rightklick paste feature..
May 31 10:27:00 <JaMa|Off> jake42: it's configureable :) you can change it to middlebutton
May 31 10:28:05 <JaMa|Off> but then I usually forget to pres Ctrl+C after selecting some text.. and then paste something much bigger then expected :)
May 31 10:50:29 <CIA-17> SHR: 03shr-devel 07buildhistory * r138dfb42591a 10/packages/armv7a-vfp-neon-oe-linux-gnueabi/opkg/ (11 files in 11 dirs): packages: Build 201205311221 of shr 20120531 for machine nokia900 on opmbuild
May 31 10:50:30 <CIA-17> SHR: 03shr-devel 07buildhistory * rd2fc4e6ffc20 10/images/nokia900/eglibc/shr-image/ (11 files): images: Build 201205311221 of shr 20120531 for machine nokia900 on opmbuild
May 31 11:08:00 <jake42> JaMa|Off: I changed it to opening a menu as I don't need that feature..
May 31 11:20:13 <CIA-17> SHR: 03shr-devel 07buildhistory * rf504f0c6fe04 10/images/om_gta04/eglibc/shr-image/ (11 files): images: Build 201205311252 of shr 20120531 for machine om-gta04 on opmbuild
May 31 11:48:01 <CIA-17> SHR: 03shr-devel 07buildhistory * r15ac56773c13 10/images/crespo/eglibc/ (12 files in 2 dirs): images: Build 201205311322 of shr 20120531 for machine crespo on opmbuild
May 31 11:57:54 <JaMa|Off> jake42: 061 closed, please test when you find time
May 31 12:05:46 <CIA-17> SHR: 03Martin.Jansa 07shr-chroot * re2e611928fbd 10/ (1737 files in 46 dirs): system upgrade
May 31 12:08:54 <DocScrutinizer51> JaMa|Off: (paste om MMB) I tried this xmouse thingy last few days, but basically windoze isn't prepared for this sorta stuff 
May 31 12:09:24 <DocScrutinizer51> so basically it doesn't work nicely
May 31 12:10:27 <JaMa|Off> pitty, maybe windows9 :)
May 31 12:10:42 <DocScrutinizer51> the only semi-usable 'app' on my work PC is cygwin :-P
May 31 12:11:25 <DocScrutinizer51> I *always* forget to press ctrl-C
May 31 12:11:35 <JaMa|Off> true about cygwin
May 31 12:11:57 <DocScrutinizer51> windoze is such a shite
May 31 12:13:06 <DocScrutinizer51> re CAIF, it seems there's no function call API anymore, I need to use SOCKS even in kernel land
May 31 12:13:12 <DocScrutinizer51> :-(
May 31 12:13:54 <DocScrutinizer51> afk
May 31 12:32:03 <JaMa|Off> I was using putty to connect to cygwin sshd to have usable term until I've discovered mintty in cygwin
May 31 13:11:03 <ayaka> I wonder something about s-on and bootloader, does mobile use EFI now
May 31 15:42:28 <PaulFertser_> DocScrutinizer51: pong
May 31 15:52:17 <DocScrutinizer51> PaulFertser_: hi :-)
May 31 15:52:50 <DocScrutinizer51> I planned to discuss a few kernel development problems with you
May 31 15:54:26 <PaulFertser_> DocScrutinizer51: i'd be very glad if i'm able to help
May 31 15:54:27 <DocScrutinizer51> they ordered me to build a testframe for caif_hsi.c et rel, to run the shit at MAX POSSIBLE thruput,
May 31 15:54:52 <DocScrutinizer51> CAIF only offers SOCKS as API
May 31 15:55:27 <DocScrutinizer51> context switxching between kernel and userland for each pkg to send considdered already too much
May 31 15:55:33 <DocScrutinizer51> overhead
May 31 15:57:31 <DocScrutinizer51> so I'm already about to look into a yet-to-build kernel module caif-test that opens SOCK-DRAM to CAIF and spams it with test data to send over HSI to modem and via loopback there back to andridiot
May 31 15:57:57 <lindi-> andridiot?
May 31 15:58:08 <DocScrutinizer51> err android
May 31 15:58:23 <lindi-> oh, you switched to android?
May 31 15:58:33 <DocScrutinizer51> aka 'linux' (they call it linux here at work)
May 31 15:59:02 <DocScrutinizer51> nah, our customers are using android on OMAP
May 31 15:59:16 <DocScrutinizer51> and CAIF, to talk to our modem
May 31 15:59:22 <PaulFertser_> What's CAIF?
May 31 15:59:42 <DocScrutinizer51> check LXR, or google
May 31 15:59:57 <DocScrutinizer51> it's a muxing protocol, like phonet
May 31 16:00:14 <DocScrutinizer51> google for caif-hsi.c
May 31 16:00:43 <DocScrutinizer51> afk, another 60 min work, then beback
May 31 16:00:58 <PaulFertser_> Ok
May 31 16:01:00 <PaulFertser_> Got it
May 31 16:05:28 <DocScrutinizerST> PaulFertser_: http://lxr.free-electrons.com/source/net/caif/caif_socket.c
May 31 16:05:53 <DocScrutinizerST> http://lxr.free-electrons.com/source/Documentation/networking/caif/Linux-CAIF.txt?a=microblaze
May 31 16:07:45 <DocScrutinizerST> ^^^though terribly outdated
May 31 16:08:00 <DocScrutinizerST> you'll notice HSI isn'
May 31 16:08:02 <DocScrutinizerST> t
May 31 16:08:09 <DocScrutinizerST> mentioned at all
May 31 16:10:48 <DocScrutinizerST> http://www.google.com/search?q=caif&sitesearch=lxr.free-electrons.com%2Fsource
May 31 16:11:34 <DocScrutinizerST> http://lxr.free-electrons.com/source/include/net/caif/caif_hsi.h
May 31 16:12:18 <DocScrutinizerST> As with phonet, there are other transports as well: USB for example
May 31 16:32:07 <DocScrutinizerST> (actually incredible I may use this web-interface from my workstation @ work ;-P)
May 31 16:33:53 <PaulFertser_> Sorry, but i do not have any immediate suggestion. I will surely try to investigate it more this evening.
May 31 16:36:20 <DocScrutinizerST> well, I have no specific question, rather a few general points I want to discuss with you, e.g. overhead of using SOCK vs passing a buff* to a function (esp memcpy involved?)
May 31 17:30:55 <DocScrutinizer51> well, now I got time to chat, but no remaining charge in battery ;-)
May 31 18:06:58 <CIA-17> freesmartphone.org: 03morphis 07cornucopia * r7fee73d7d2ae 10/fsogsmd/ (4 files in 3 dirs): 
May 31 18:06:58 <CIA-17> freesmartphone.org: fsogsmd: lib: at: add implementation of an AT result parser
May 31 18:06:58 <CIA-17> freesmartphone.org: Implementation is fairly adopted from the ofono project and their GAtResultIter class. See
May 31 18:06:58 <CIA-17> freesmartphone.org: https://git.kernel.org/?p=network/ofono/ofono.git;a=blob;f=gatchat/gatresult.c;hb=HEAD
May 31 19:57:46 <nschle85> hello does anybody know how the ipk is called for the kernel sources ?
May 31 20:04:04 <PaulFertser_> nschle85: just copy the headers or even the whole build tree from your OE build 
May 31 20:04:11 <PaulFertser_> but i'm afk sleeping :(
May 31 20:04:43 <nschle85> PaulFertser_: ok see you tomorrow ?
May 31 20:07:06 <PaulFertser_> nschle85: sure
May 31 20:07:33 <nschle85> i copied already the headers but directory modules is missing
May 31 21:16:41 <prahal2> currently all audio clients needs to be tweaked to call into fso 
**** ENDING LOGGING AT Fri Jun 01 02:59:57 2012