2012-04-10 00:22 cladamw has joined #qi-hardware
2012-04-10 01:04 antgreen has joined #qi-hardware
2012-04-10 01:09 xiangfu has joined #qi-hardware
2012-04-10 01:24 Openfree` has joined #qi-hardware
2012-04-10 01:42 cladamwa has joined #qi-hardware
2012-04-10 02:46 xiangfu has joined #qi-hardware
2012-04-10 02:51 <qi-bot> [commit] Werner Almesberger: m1/xbrd/: reduced CB from 25 mm to 22 mm, reducing overall length to 82.08 mm (master) http://qi-hw.com/p/wernermisc/5bf0b30
2012-04-10 02:56 xian9fu has joined #qi-hardware
2012-04-10 02:59 orson_zhai has joined #qi-hardware
2012-04-10 04:06 AwAyla has joined #qi-hardware
2012-04-10 04:56 <xian9fu> kristianpaul, Hi I tried milkyminer some days ago. 
2012-04-10 04:57 <xian9fu> this one: http://fidelio.qi-hardware.com/~xiangfu/build-milkymist/milkyminer-firmware-20120401-0837/
2012-04-10 04:57 <xian9fu> I flash the soc.fpg by 'm1nor soc.fpg' then
2012-04-10 04:58 <xian9fu> it's only worked once. spend 5 minutes for get the result.  then what even I send to milkyminer. nothing return.
2012-04-10 04:58 <xian9fu> do you think we should build the 'soc' branch?
2012-04-10 05:18 jekhor has joined #qi-hardware
2012-04-10 05:24 cladamw has joined #qi-hardware
2012-04-10 06:23 <kyak> xian9fu: do you have an idea why this error happens from time to time (ERROR: "irq_set_chip_and_handler_name" [drivers/ieee802154/spi_atusb.ko] undefined!)?
2012-04-10 06:23 <kyak> i mean, during the build, which has failed
2012-04-10 06:24 <kyak> xian9fu: also, what's the difference between ~xiangfu/build-nanonote/ and ~xiangfu/building/Nanonote/Ben/ ?
2012-04-10 06:25 <xian9fu> when build was success it will move to ~xiangfu/building/Nanonote/Ben/ 
2012-04-10 06:25 <kyak> it seems like the minimal build always fails
2012-04-10 06:25 <xian9fu> sorry. all success will move to ~xiangfu/build-nanonote/
2012-04-10 06:25 <kyak> perhaps we miss something from config-minimal
2012-04-10 06:26 <xian9fu> kyak, we should not build atusb.ko under config.minimal. I think I already disable that.
2012-04-10 06:27 <kyak> ok.. is it building in a loop or you start it manually?
2012-04-10 06:29 <xian9fu> building in a loop
2012-04-10 06:29 <xian9fu> http://fidelio.qi-hardware.com/~xiangfu/bin/qi-hardware-build.sh and http://fidelio.qi-hardware.com/~xiangfu/bin/compile.sh
2012-04-10 06:30 antgreen` has joined #qi-hardware
2012-04-10 06:32 <kyak> cool, so it started the full build now
2012-04-10 06:32 <xian9fu> yes. 
2012-04-10 06:34 <xian9fu> I think the at_usb.h miss the irq.h
2012-04-10 06:37 <xian9fu> s/at_usb.h/at_usb.c
2012-04-10 06:37 <qi-bot> xian9fu meant: "I think the at_usb.c miss the irq.h"
2012-04-10 06:37 <xian9fu> spi_atusb.c
2012-04-10 06:41 * xian9fu testing the spi_atusb now... (by add irq.h)
2012-04-10 06:41 <kyak> looks like this.. wonder why it works in full image
2012-04-10 06:42 <xian9fu> kyak, full image don't build spi_atusb
2012-04-10 06:42 <xian9fu> disabled under config.full
2012-04-10 06:42 <kyak> ah!
2012-04-10 06:52 DocScrutinizer has joined #qi-hardware
2012-04-10 07:03 GeorgeH_ has joined #qi-hardware
2012-04-10 07:17 cladamw has joined #qi-hardware
2012-04-10 07:25 orson_zhai has joined #qi-hardware
2012-04-10 07:42 cladamw has joined #qi-hardware
2012-04-10 07:54 <qi-bot> [commit] Xiangfu: nanonote: config.minimal: disable spi_atusb since we don't have usb host (master) http://qi-hw.com/p/openwrt-packages/e702c87
2012-04-10 07:54 <xian9fu> kyak, I just disable the atusb under minimal config. I cannot find out the root cause faster. disable it first. when I work on WPAN patches. will try to make it working. 
2012-04-10 07:55 <kyak> xian9fu: atusb doesn't make sense for Ben anyway, right?
2012-04-10 07:55 <xian9fu> right. we don't ahve usb host. 
2012-04-10 08:06 xiangfu has joined #qi-hardware
2012-04-10 09:52 jivs_ has joined #qi-hardware
2012-04-10 09:54 orson_zhai has quit [#qi-hardware]
2012-04-10 10:12 <wpwrak> kyak: you'd have to implement USB host first :)
2012-04-10 10:12 <wpwrak> now, that's marketing: (in german, sorry) http://www.heise.de/newsticker/meldung/Raspberry-Pi-erhaelt-CE-Kennzeichnung-1517687.html
2012-04-10 10:13 <wpwrak> perhaps we should make a press release whenever Milkymist creator and chief architect Sebastien had a good morning dump ? ;-)
2012-04-10 10:19 <lekernel> chief architect has better plans, like turning the patch editor into something a bit like the first parts of this: http://vimeo.com/36579366
2012-04-10 10:25 <wolfspraul> here's hoping that migen gives an upside in several directions
2012-04-10 10:27 <lekernel> well this will come after migen. migen is about making the SoC more featureful, faster and more portable.
2012-04-10 10:30 <wpwrak> lekernel: (immediate feedback) sounds nice
2012-04-10 10:45 zrafa_ has joined #qi-hardware
2012-04-10 11:29 wej has joined #qi-hardware
2012-04-10 12:00 <viric> So, there is a working linux on milkymist SoC already?
2012-04-10 12:00 <lekernel> uclinux
2012-04-10 12:00 <viric> ah ok, no mmu
2012-04-10 12:01 <lekernel> Fallenou is working on a MMU
2012-04-10 12:01 <viric> ok
2012-04-10 12:01 <lekernel> you can, too. that's the beauty of MM SoC.
2012-04-10 12:01 <viric> :) clear
2012-04-10 12:01 <viric> If you want to be well served, make the bed yourself.
2012-04-10 12:06 jekhor has joined #qi-hardware
2012-04-10 12:32 <wolfspraul> viric: we can get you interested in Milkymist? :-)
2012-04-10 12:38 <viric> I'm interested in seeing some information about its software performance and power usage.
2012-04-10 12:38 <wolfspraul> right now it runs an RTEMS realtime/minikernel, and a specific app on top of that for the video synthesis
2012-04-10 12:38 <viric> aha
2012-04-10 12:38 <wolfspraul> power consumption is about 5W total
2012-04-10 12:39 <viric> with how much ram?
2012-04-10 12:39 <wolfspraul> 128 megabytes ddr1
2012-04-10 12:39 <wolfspraul> I don't think the ram's power consumption matters much at all btw
2012-04-10 12:39 <wolfspraul> but I have no exact number right now
2012-04-10 12:39 <viric> ok
2012-04-10 12:40 <viric> Could you publish a run of "openssl speed" in it?
2012-04-10 12:40 <wolfspraul> sorry don't understand. what exactly do you want to execute?
2012-04-10 12:40 <viric> "openssl speed"
2012-04-10 12:40 <viric> I imagine you can run openssl there. :)
2012-04-10 12:41 <viric> I don't know much about benchmarking... I use 'openssl speed' everywhere, when I want to compare performances.
2012-04-10 12:42 <wolfspraul> actually I don't think openssl is ported to m1 today
2012-04-10 12:42 <viric> it should be ansic code
2012-04-10 12:42 <wolfspraul> sure someone could do it
2012-04-10 12:43 <viric> ok
2012-04-10 12:43 <viric> well, I mean it's a benchmark quite broad and easy to build :)
2012-04-10 12:43 <wolfspraul> it's not very 'meaningful' for m1, but I 100% agree we should first publish any benchmark, so people have a starting point for their thinking
2012-04-10 12:44 <wolfspraul> you could then proceed to a hardware accelerated openssl speed benchmark :-)
2012-04-10 12:48 Ayla has joined #qi-hardware
2012-04-10 12:56 LunaVorax has joined #qi-hardware
2012-04-10 13:00 <wolfspraul> I'm trying to build a mobile & rechargeable battery pack for an m1 setup
2012-04-10 13:00 <wolfspraul> so basically I need 12V and 5V out of it
2012-04-10 13:03 <wolfspraul> I found nice little battery packs like this one http://www.iphonemili.com/products_detail/&productId=8f5df92b-03a0-441f-91bb-d444f512a769&comp_stats=comp-FrontProducts_list01-1311553918454.html
2012-04-10 13:03 <wolfspraul> so that works well, but I would like to use standard batteries, something more flexible
2012-04-10 13:04 <wolfspraul> I was hoping to find teeny tiny dc-dc converters that just allow me to put some wide voltage in, and get a stable 12V or 5V DC out, but cannot find them
2012-04-10 13:06 <wolfspraul> they have a lot of nice dc-dc chips, someone should make breakout boards for them :-)
2012-04-10 13:07 woakas has joined #qi-hardware
2012-04-10 13:13 <wpwrak> wolfspraul: you may be able to find some development boards. even for such simple things, they often make some. not sure how cost-effective they are, though
2012-04-10 13:14 <wolfspraul> well there are suppliers of battery packs, like that mili one
2012-04-10 13:14 <wolfspraul> but it's amazing how expensive they are, and of course no standard batteries
2012-04-10 13:15 <wolfspraul> they tend to use charging ics from TI like bqxxxx
2012-04-10 13:15 <wolfspraul> but in general I find the situation a little unflexible
2012-04-10 13:17 <wpwrak> that pack looks very stylish. maybe that explains the price :)
2012-04-10 13:17 <roh> what unit is that? 99 foo?
2012-04-10 13:17 <roh> us$?
2012-04-10 13:17 <wolfspraul> yes
2012-04-10 13:18 <roh> well... there are cheaper ones for sure... but will they do what you want and have a case?
2012-04-10 13:18 <wolfspraul> case not needed
2012-04-10 13:18 <roh> there is no standard in user serviceable lithium cells without packaging and security circuit, so they get integrated like in that pack
2012-04-10 13:19 <wolfspraul> how about every normal rechargeable aa battery?
2012-04-10 13:19 <roh> how many do you need? i think if its only for prototyping, just buy that thing
2012-04-10 13:19 <wolfspraul> sure, already done
2012-04-10 13:19 <wolfspraul> but don't tell me I shouldn't dig deeper :-)
2012-04-10 13:19 <roh> aa etc sucks. doesnt allow for lithium based cells
2012-04-10 13:24 <wolfspraul> why couldn't you fit a lithium-ion polymer into an aa case?
2012-04-10 13:25 <roh> wrong voltage, bad ratio of weight/case/power
2012-04-10 13:25 <wolfspraul> I didn't pay attention to battery chemicals yet indeed
2012-04-10 13:25 wej has joined #qi-hardware
2012-04-10 13:25 <roh> li based stuff has between 2.9 and 4.4V per cell, depending on chemistry and charge state
2012-04-10 13:26 <roh> mostly its between 3.9 and 4.2 or 3.0 and 3.4V... the first for li-ion and li-po based stuff, the latter for li-fe and similars
2012-04-10 13:28 <roh> and you need a balancing circuit for keeping proper care which has one line to every cell when used in series. parallel is doable, but needs low manufacturing tolerances in cell chemistry. so its much easier to just use one of the typical used 'plastic bag' packges and add standard industry charge control and safety chips (or outsource that)
2012-04-10 13:29 <wolfspraul> that's what those mili guys are doing (and similar vendors)
2012-04-10 13:29 <roh> sure. they buy cells and package them with some circuit i bet is very similar to the appnotes of the chips used
2012-04-10 13:30 <roh> or the reference design
2012-04-10 13:30 <wolfspraul> sure
2012-04-10 13:31 <roh> sparkfun and similars, some arduino boards and stuff watterott has also uses simple li-chargecontrollers and dc-dc switchers. not that complicated to use/design with. mostly its just some caps, some inductivity and a fast diode in addition to the chip
2012-04-10 13:32 <roh> but they use 'prepackged' cells, mostly single ones, but also small packs which then include the security and balancing chips
2012-04-10 13:33 <wolfspraul> interesting, ok. that's helpful!
2012-04-10 13:33 <roh> very similar (basically identical) to what one gets for cellphones, just with a hard shell there
2012-04-10 13:34 <roh> i can name you a book if you like to read up on more details on battery chemistry
2012-04-10 13:34 <wolfspraul> please, always
2012-04-10 13:34 <roh> 'moderne akkumulatoren richtig einsetzen' von andreas jossen und wolfgang weydanz im reichardt verlag
2012-04-10 13:35 <wolfspraul> thanks!
2012-04-10 13:35 <wolfspraul> will take a little time to get it and read, but it's queued :-)
2012-04-10 13:35 <roh> ;)
2012-04-10 13:36 <roh> its worth having when one has some detail question.. and rare indeed
2012-04-10 13:36 <roh> we got a copy in our raumfahrtagentur-library
2012-04-10 13:38 LunaVorax has joined #qi-hardware
2012-04-10 14:25 wej has joined #qi-hardware
2012-04-10 14:56 GNUtoo has joined #qi-hardware
2012-04-10 14:56 wej has joined #qi-hardware
2012-04-10 15:19 GNUtoo-desktop has joined #qi-hardware
2012-04-10 15:34 rejon has joined #qi-hardware
2012-04-10 15:50 emeb has joined #qi-hardware
2012-04-10 15:52 kilae has joined #qi-hardware
2012-04-10 16:35 <kyak> jow_laptop: it just caught my eye that OpenWrt-Toolchain-* doesn't have "for-Linux-x86_64/i686" in its name any more (unlike SDK). What does it mean?
2012-04-10 16:38 <jow_laptop> nothing, it was just forgotten I suppose
2012-04-10 16:39 <kyak> ah, so it's a bug then :)
2012-04-10 16:40 <jow_laptop> likely
2012-04-10 16:40 dptech has joined #qi-hardware
2012-04-10 16:44 dptech has joined #qi-hardware
2012-04-10 16:56 GNUtoo-desktop has joined #qi-hardware
2012-04-10 17:11 jurting has joined #qi-hardware
2012-04-10 17:28 jekhor has joined #qi-hardware
2012-04-10 17:55 <mth> kyak: I can reproduce the gmenu2x problem now
2012-04-10 18:01 LunaVorax has joined #qi-hardware
2012-04-10 18:02 <LunaVorax> Hello everyone!
2012-04-10 18:02 <kyak> mth: using Ben?
2012-04-10 18:04 <mth> no, using Dingoo
2012-04-10 18:06 <kyak> so you couldn't reproduce it before?
2012-04-10 18:11 <viric> Do you think linux 2.4 could run in the ben using less resources than 2.6?
2012-04-10 18:11 <viric> or 3
2012-04-10 18:12 <viric> or the configuration granularity still allows to achieve similar results, in relation to the features provided?
2012-04-10 18:14 <mth> kyak: no, but I was probably testing a binary from the wrong location then, so an older version instead of the one I just compiled
2012-04-10 18:14 <kyak> viric: i think you would spend more time porting 2.4 to Ben
2012-04-10 18:16 <viric> well I agree
2012-04-10 18:16 <viric> I just thought if in general linux is becoming too big.
2012-04-10 18:16 <viric> or the compile time configuration puts all in a good place.
2012-04-10 18:17 <Ayla> viric: it's becoming big because it have more drivers
2012-04-10 18:18 <kyak> viric: we could measure performance between 2.6 and 3.2.. probably you will not notice any changes, but who knows?
2012-04-10 18:18 <viric> but drivers can be disabled at build time
2012-04-10 18:19 <viric> Maybe linux is forgetting a bit the uniprocessor systems nowadays .)
2012-04-10 18:19 <viric> :)
2012-04-10 18:19 <viric> I always thought linux 2.2 was the fastest since 2.0 ;)
2012-04-10 18:19 <viric> and 2.4 the slowest.
2012-04-10 18:20 <viric> but I can't prove it :)
2012-04-10 18:20 <kyak> you are old! :) i barely touched 2.4
2012-04-10 18:20 <viric> I started at 2.0.34 :)
2012-04-10 18:20 <kyak> oooold! :)
2012-04-10 18:20 <viric> haha
2012-04-10 18:21 <viric> RH 5.2 it was. 5 is a big number.
2012-04-10 18:21 LunaVorax_ has joined #qi-hardware
2012-04-10 18:30 <zrafa_> kyak: viric is as old like me I guess :)
2012-04-10 18:32 <viric> :)
2012-04-10 18:32 <viric> We went to the magazine store to get updates ;)
2012-04-10 18:34 <viric> "Mama, can you bring me the latest kernel, at the time you buy your yellow press magazines? Thank you"
2012-04-10 18:36 <qi-bot> [commit] Maarten ter Huurne: Fixed GCC warning about initialization order. (master) http://qi-hw.com/p/gmenu2x/92d221a
2012-04-10 18:36 <qi-bot> [commit] Maarten ter Huurne: Don't overwrite link action provided to constructor. (master) http://qi-hw.com/p/gmenu2x/ed8b0c3
2012-04-10 18:36 <mth> kyak: that second commit should fix it
2012-04-10 18:37 <mth> viric: compile time configs help a lot, for example a lot of locking ops are reduced to nothing if you have 1 CPU and no kernel preemption
2012-04-10 18:37 <mth> but I do see the OpenDingux kernel growing over time and we're not writing that much code
2012-04-10 18:38 <mth> still, it's probably less effort to optimize 3.x than to backport to 2.4
2012-04-10 18:38 wej has joined #qi-hardware
2012-04-10 18:38 <viric> mth: that would mean using a non-SMP kernel, right?
2012-04-10 18:38 <viric> mth: or it's dynamically set?
2012-04-10 18:38 <mth> yes
2012-04-10 18:39 <mth> you can make a non-SMP kernel by config
2012-04-10 18:39 <viric> I know
2012-04-10 18:39 <viric> but there is the "nosmp" commandline, or "cpu=1", I can't recall... it isn't equivalent enough?
2012-04-10 18:39 <viric> I remember in 2.2 that preemption never helped me.
2012-04-10 18:39 <mth> I don't think so: checking that option might be slower than just doing the lock
2012-04-10 18:39 <jow_laptop> kyak: since a toolchain is not bound to a particular target it might also make sense to replace BOARD with ARCH
2012-04-10 18:40 <jow_laptop> kyak: so it reads OpenWrt-Toolchain-mipsel-for-Linux-x86_64-gcc-4.6-linaro_uClibc-0.9.33.tar.bz2
2012-04-10 18:40 <viric> mth: I should try in my nonsmp systems.
2012-04-10 18:40 <viric> In fact I don't own any SMP.
2012-04-10 18:40 <mth> my desktop is a Core 2 quad-core
2012-04-10 18:40 <mth> it's very using when doing big compiles
2012-04-10 18:40 <mth> *useful
2012-04-10 18:40 <viric> :)
2012-04-10 18:40 <viric> I don't doubt it
2012-04-10 18:40 <jow_laptop> kyak: and maybe replace -for- with -on-
2012-04-10 18:41 <viric> the 7h of battery of mine are very useful too ;)
2012-04-10 18:41 <viric> mth: 'perf' won't show me anything about the locks, right?
2012-04-10 18:41 <mth> it's connected to the wall socket and power outages are rare here, fortunately
2012-04-10 18:41 <jow_laptop> kyak: and prepend the -for- before the ARCH (toolchain-for-mipsel-on-linux-x86...)
2012-04-10 18:41 <viric> :)
2012-04-10 18:42 <mth> I've got a laptop with an i7 and that battery lasts about 7h as well, but not when doing make -j8 ;)
2012-04-10 18:42 abushcrafterfor1 has joined #qi-hardware
2012-04-10 18:42 <viric> mth: ah very good
2012-04-10 18:42 <mth> I don't know 'perf'
2012-04-10 18:42 <viric> I'll try about the preemption and all that...
2012-04-10 18:43 <mth> kernel preemption will likely make things slower, not faster
2012-04-10 18:43 <mth> it only helps if drivers are doing too much work in big chunks
2012-04-10 18:43 <viric> aha
2012-04-10 18:44 <viric> userspace preemption will still be there, right?
2012-04-10 18:44 <mth> yes, that's always there
2012-04-10 18:44 <mth> the kernel preemption is about whether a syscall or kernel thread can be preempted
2012-04-10 18:44 <viric> clera
2012-04-10 18:45 <mth> I guess it could protect against a user doing a denial of service attack by calling lots of expensive syscalls in a row
2012-04-10 18:45 <mth> but that's not very likely on single-user systems
2012-04-10 18:45 <viric> aha
2012-04-10 18:45 <viric> and what do you think of NO_HZ?
2012-04-10 18:46 <mth> it's good, but it's not perfect yet
2012-04-10 18:46 <mth> it only turns off the timer ticks when the CPU is idle, we recently found out
2012-04-10 18:46 <viric> and how is it bad?
2012-04-10 18:46 <viric> as for performance, may it be bad?
2012-04-10 18:47 <mth> it's not bad, it's just not always good enough
2012-04-10 18:47 <mth> in the sense that if you're running for example mplayer, the CPU won't be idle much
2012-04-10 18:47 <mth> but since there is only 1 process that wants to run, you could omit the timer ticks
2012-04-10 18:47 <viric> mth: PREEMPT_NONE/PREEMPT_VOLUNTARY/PREEMPT
2012-04-10 18:47 <mth> but the current kernel does not do that
2012-04-10 18:48 <viric> ah clear.
2012-04-10 18:48 <mth> someone is working on a patch to change that, but last time I checked it was still very experimental
2012-04-10 18:48 <viric> I imagine it's some kind of variable rate scheduler
2012-04-10 18:49 <mth> in OpenDingux we originally switched to HZ=1000 + NO_HZ, but that increased the overhead significantly
2012-04-10 18:49 <mth> so now we're using HZ=250 + NO_HZ
2012-04-10 18:49 <mth> the advantage is that sleep accuracy is now 4 ms instead of 10 ms
2012-04-10 18:49 <viric> ok
2012-04-10 18:49 <mth> a lot of emulators want 60 Hz output, so 10 ms is over half a frame,  not fine grained enough
2012-04-10 18:50 <viric> hm
2012-04-10 18:50 <viric> Ok, I'll go 250 for my desktop too
2012-04-10 18:51 <mth> hmm, we're using PREEMPT_VOLUNTARY at the moment, maybe we should try PREEMPT_NONE instead
2012-04-10 18:52 <mth> although the description of PREEMPT_VOLUNTARY says it only preempts at selected places, so maybe that's not so often
2012-04-10 18:54 <Ayla> PREEMPT_NONE means a much higher latency
2012-04-10 18:56 <viric> but it's the *kernel* preemption, no?
2012-04-10 18:56 <viric> latency in what?
2012-04-10 18:56 LunaVorax has joined #qi-hardware
2012-04-10 18:57 <mth> yes, but kernel preemption includes syscalls, so it's not fully independent of user space
2012-04-10 18:57 <mth> Ayla: according to the docs or measured on OD?
2012-04-10 18:57 <viric> but if the syscalls wait...
2012-04-10 18:57 <viric> they go away of the scheduler
2012-04-10 18:57 <viric> I don't know. I'll try PREEMPT_NONE in my desktop.
2012-04-10 18:57 <Ayla> mth: I did test it on OD, I honestly didn't see the difference
2012-04-10 18:58 <mth> if a syscall cannot get a semaphore for example, the process will be put in a wait state indeed
2012-04-10 18:58 <Ayla> but on OD, we rarely have more than one program running
2012-04-10 18:58 <mth> but with PREEMPT_VOLUNTARY even if the semaphore is available, other processes will get a chance at that point
2012-04-10 18:59 <mth> Ayla: and was there a different in throughput? (gpmark)
2012-04-10 18:59 <Ayla> I don't remember the details, that was a long time ago
2012-04-10 18:59 <Ayla> when we switched to 1000Hz
2012-04-10 18:59 <viric> what is gpmark?
2012-04-10 19:00 <Ayla> it's a program we use to benchmark the video driver
2012-04-10 19:00 <viric> ok
2012-04-10 19:01 <viric> Do you know any piece that, given a result of 'lsmod', would tell me the kernel options to enable? :)
2012-04-10 19:03 <viric> localmodconfig
2012-04-10 19:03 <viric> incredible
2012-04-10 19:03 <Ayla> make localmodconfig
2012-04-10 19:06 <viric> and it still keeps them as modules.
2012-04-10 19:06 <viric> hm interesting
2012-04-10 19:07 LunaVorax has joined #qi-hardware
2012-04-10 19:07 kuribas has joined #qi-hardware
2012-04-10 19:26 LunaVorax has joined #qi-hardware
2012-04-10 19:44 GNUtoo has joined #qi-hardware
2012-04-10 19:45 LunaVorax has joined #qi-hardware
2012-04-10 20:55 <qi-bot> [commit] Maarten ter Huurne: SettingsDialog: Code layout cleanup. (master) http://qi-hw.com/p/gmenu2x/fd642ff
2012-04-10 20:55 <qi-bot> [commit] Maarten ter Huurne: Surface cleanup. (master) http://qi-hw.com/p/gmenu2x/b248aaf
2012-04-10 20:55 <qi-bot> [commit] Maarten ter Huurne: Touchscreen: Avoid constructing an SDL_Rect for is-inside tests. (master) http://qi-hw.com/p/gmenu2x/58d6077
2012-04-10 20:55 <qi-bot> [commit] Maarten ter Huurne: BrowseDialog: Code layout cleanup. (master) http://qi-hw.com/p/gmenu2x/46f2edb
2012-04-10 20:55 <qi-bot> [commit] Maarten ter Huurne: Explicitly convert 32-bit integers to 16-bit. (master) http://qi-hw.com/p/gmenu2x/8f57afc
2012-04-10 20:55 <Ayla> whohooo!
2012-04-10 20:59 <mth> we're back at 0 warnings :)
2012-04-10 21:01 <wpwrak> mth: so how do you know that the mechanism that generates warnings isn't broken ? :)
2012-04-10 21:02 <mth> I'm hoping GCC has unit tests for that
2012-04-10 21:03 <wpwrak> heh :)
2012-04-10 21:46 <whitequark> viric: openssl is in openwrt, isn't? and I'm pretty sure it is a requirement for some pretty basic apps
2012-04-10 21:47 <Ayla> mth: with binutils 2.20.1, uClibc 0.9.31 (so, old versions) and GCC 4.7.0 *without* LTO, the libdl bug is still present
2012-04-10 21:48 <mth> ok, interesting
2012-04-10 21:50 <viric> whitequark: I hoped so, but we talked about the milkymist
2012-04-10 21:50 <viric> does the milkymist run owrt?
2012-04-10 22:01 <whitequark> ah. no idea.
2012-04-10 22:01 <whitequark> I misread your message then
2012-04-10 22:11 <viric> :)
2012-04-10 22:11 <viric> hm using flashplayer... makes my cs/s go to 4000...
2012-04-10 22:11 <viric> ha, 1000 came from firefox.
2012-04-10 22:12 <viric> vimprobable2 the same
2012-04-10 22:13 <viric> mth: if a software makes the cpu go to 4000/5000 cs/s....
2012-04-10 22:13 <viric> I better blame the sw, no?
2012-04-10 22:24 <mth> cs/s?
2012-04-10 22:26 <Ayla> context switch per second, I guess
2012-04-10 22:26 <viric> yes
2012-04-10 22:27 <mth> I don't know what a normal amout is, but 4000 does sound very high
2012-04-10 22:27 <mth> s/amout/amount/
2012-04-10 22:27 <qi-bot> mth meant: "I don't know what a normal amount is, but 4000 does sound very high"
2012-04-10 22:27 <viric> I've no idea why this happens. maybe some preemption
2012-04-10 22:27 <Ayla> it should be equal to HZ, if the program never sleeps
2012-04-10 22:27 <mth> flash player is the easiest way to drain your battery in any case
2012-04-10 22:27 <viric> it's firefox with its multiples threads
2012-04-10 22:28 <Ayla> a context switch will occur if the thread waits for some I/O to finish, is blocked on a mutex, or calls usleep()
2012-04-10 22:28 <viric> Ayla: this is HZ_250 + NO_HZ, PREEMPT_NONE
2012-04-10 22:28 <mth> sometimes I'm not doing anything CPU intensive and notice the laptop getting slightly warn... then it's probably a flash ad in a web page that's still open
2012-04-10 22:28 <Ayla> so the number of context switches is HZ minimum, but can be much higher
2012-04-10 22:28 <viric> Ayla: due to NO_HZ?
2012-04-10 22:29 <mth> with NO_HZ the number of switches can be lower as well
2012-04-10 22:29 <viric> ok
2012-04-10 22:29 <Ayla> viric: no, due to the fact that < Ayla> a context switch will occur if the thread waits for some I/O to finish, is blocked on a mutex, or calls usleep()
2012-04-10 22:29 <viric> yes, clear
2012-04-10 22:29 <mth> maybe flash uses a lot of threads that block on semaphores or something
2012-04-10 22:29 <Ayla> that's the case for Firefox at least
2012-04-10 22:29 <viric> yes
2012-04-10 22:29 <viric> then you get firefox+flash...
2012-04-10 22:29 <mth> or is a different thread in the same process the same context?
2012-04-10 22:29 <viric> which adds up
2012-04-10 22:30 <viric> mth: no, not the same context
2012-04-10 22:30 <mth> most browsers just take up a lot of memory, but not CPU if you're not interacting with them
2012-04-10 22:30 <viric> ha.
2012-04-10 22:30 <mth> while flash keeps consuming CPU even in the background
2012-04-10 22:30 <viric> there is javascript, flash, ...
2012-04-10 22:31 <mth> poorly written javascript can also be a problem, yes
2012-04-10 22:31 <viric> animated gifs
2012-04-10 22:32 <viric> in firefox, even downloading files takes a lot of cpu
2012-04-10 22:32 <Ayla> mth: a "context" is just the state of the CPU registers, so two threads of the same process don't have the same context
2012-04-10 22:32 <mth> animated gifs don't have to be updated unless the animation is currently visible
2012-04-10 22:32 <viric> "dont have to"... :)
2012-04-10 22:33 <mth> yeah, theory and practice are often not the same
2012-04-10 22:33 <viric> with elinks, my cs/s average goes ~ 100/s :)
2012-04-10 22:34 <viric> mplayer playing a flash video, 500/s
2012-04-10 22:35 <viric> all looks like too much for me
2012-04-10 22:35 <viric> it only gets reasonable when doing nothing nothing.
2012-04-10 22:35 <Ayla> why should it be "too much"?
2012-04-10 22:36 <Ayla> getting a lower number of context switches won't make your system faster
2012-04-10 22:37 <mth> well, a context switch is overhead, so it does help to reduce the count, the question is when it's significant and when not
2012-04-10 22:38 <mth> speaking of context switches, we should add register saving for MXU
2012-04-10 22:39 lekernel has joined #qi-hardware
2012-04-10 22:39 <Ayla> good idea
2012-04-10 22:40 <mth> with a check whether MXU is enabled for the current context: no need to save/restore it if it's not in use
2012-04-10 22:41 <Ayla> it'd be useful to add the MXU instructions to binutils too
2012-04-10 22:51 dvdk has joined #qi-hardware
2012-04-10 22:52 <dvdk> could somebody please remove the spambot user "Bailey Helton" who spammed some qi-hw wiki pages last week?  can his changes be reverted automatically?
2012-04-10 22:55 <dvdk> kristianpaul^ wolfspraul^
2012-04-10 23:16 <Ayla> mth: do you know a practical way to make the MXU instructions visible?
2012-04-10 23:17 <Ayla> I mean something like a lib
2012-04-10 23:17 <mth> visible in what way?
2012-04-10 23:19 <Ayla> so that they could be used by several apps or libs, like libpng or mplayer
2012-04-10 23:20 <wolfspraul> dvdk: a spammer went through unnoticed? wait I check
2012-04-10 23:20 <mth> I don't think we need a lib for that, just inline asm support
2012-04-10 23:20 <mth> or maybe intrinsics
2012-04-10 23:21 <dvdk> wolfspraul: ha, good to meet you here.  just wanted to drop a mail; sharims.cc/blog has problems with recent iceweasel webbrowser (i.e. debian's firefox)
2012-04-10 23:21 <mth> MXU will only be used in performance critical code and it's not easy to write code that is both very efficient and generic
2012-04-10 23:21 <wolfspraul> oops
2012-04-10 23:21 <wolfspraul> thanks!
2012-04-10 23:22 <wolfspraul> just blocked helton
2012-04-10 23:22 <wolfspraul> I think you reverted the edits already, no?
2012-04-10 23:23 <dvdk> wolfspraul: wrt shop website, see screenshot im mail
2012-04-10 23:24 <wolfspraul> yes will do, thanks a lot!
2012-04-10 23:26 <dvdk> wolfspraul: yes found out how to revert.
2012-04-10 23:26 <dvdk> wolfspraul: just sent mail.  hope you remember your GPG password :)
2012-04-10 23:27 <dvdk> think website problem is certificate-related.  CSS is accessed via https://?
2012-04-10 23:27 <wolfspraul> maybe the cert just expired? :-) I'll check
2012-04-10 23:28 <dvdk> wolfspraul: riht, on april 1st.  ha that was simple
2012-04-10 23:28 <dvdk> btw typo above it's sharism.cc/shop, not /blog