| 00:06.45 | *** join/#debianppc s^3 (n=s^3@tor/session/external/x-a303e689b9607266) |
| 00:37.49 | *** join/#debianppc karim (n=karim@ip-7.net-82-216-143.rev.numericable.fr) |
| 01:18.13 | *** join/#debianppc _rene__ (n=rene@dslb-084-056-069-127.pools.arcor-ip.net) |
| 02:33.11 | *** join/#debianppc xbaez (n=xbaez@dsl-201-128-167-184.prod-infinitum.com.mx) |
| 02:35.16 | *** join/#debianppc sk1p (i=alex@p548D531A.dip.t-dialin.net) |
| 02:41.38 | *** join/#debianppc yipstar (n=yipstar@cpe-68-173-28-208.nyc.res.rr.com) |
| 02:45.12 | *** join/#debianppc yipstar_ (n=yipstar@cpe-68-173-28-208.nyc.res.rr.com) |
| 03:23.43 | *** join/#debianppc peterrooney (n=peter@S0106001310a01647.gv.shawcable.net) |
| 03:45.18 | *** join/#debianppc wishmaster (n=wishmast@p54863CEE.dip0.t-ipconnect.de) |
| 03:46.38 | wishmaster | I have an irritating question. |
| 03:46.40 | wishmaster | Is there somebody awake? |
| 03:47.09 | wishmaster | Why does mplayer not find the libx264.so.47 ? |
| 03:47.16 | wishmaster | Where can I add the path? |
| 03:47.22 | tcsetattr | awake, yes. irritated, well let's wait and see |
| 03:47.35 | tcsetattr | at runtime or compile time? |
| 03:47.38 | wishmaster | runtime |
| 03:48.13 | tcsetattr | where is it, /usr/local/lib? |
| 03:48.16 | wishmaster | yes |
| 03:48.38 | tcsetattr | put that in /etc/ld.so.conf and run ldconfig |
| 03:48.51 | wishmaster | aha, looking |
| 03:49.42 | wishmaster | that's it |
| 03:49.47 | wishmaster | thank you very much. |
| 03:53.43 | wishmaster | Now I can play DVDs like a world champion. |
| 04:41.17 | *** join/#debianppc MasterTom (n=MajorTom@p54863CEE.dip0.t-ipconnect.de) |
| 06:45.42 | *** join/#debianppc marculix (n=marculix@gw.ptr-80-238-227-85.customer.ch.netstream.com) |
| 06:45.57 | *** join/#debianppc ravenbird (i=ravenbir@unaffiliated/simonrvn/bot/ravenbird) |
| 06:49.45 | *** join/#debianppc marculix (n=marculix@gw.ptr-80-238-227-85.customer.ch.netstream.com) |
| 06:58.46 | *** join/#debianppc simonrvn (i=simon@unaffiliated/simonrvn) |
| 07:12.23 | *** join/#debianppc angel2 (n=ds9@ALyon-102-1-1-148.w80-15.abo.wanadoo.fr) |
| 07:24.56 | *** join/#debianppc tailok (n=tailok@p54BF7883.dip.t-dialin.net) |
| 07:26.54 | *** join/#debianppc s^3 (n=s^3@tor/session/external/x-af5a68d47daa0334) |
| 08:24.53 | *** join/#debianppc reDFog (n=redfog@tor/session/external/x-1fba638db9e09552) |
| 08:31.01 | *** join/#debianppc dolphy (n=dolphy@polar.es6.egwn.net) |
| 09:33.39 | *** join/#debianppc Q-FUNK (n=q-funk@dsl-hkigw7-fe12f900-46.dhcp.inet.fi) |
| 09:41.26 | *** join/#debianppc johill (n=johannes@p5487D6F2.dip.t-dialin.net) |
| 09:42.22 | Corsac | hmhm, rc5 is out and i didn't even booted my -rc4 |
| 09:42.23 | Corsac | :D |
| 09:43.28 | johill | heh |
| 09:44.21 | Corsac | <PROTECTED> |
| 09:46.50 | Corsac | but 2.6.17-rc aren't compatible with mipv6 kernel :/ |
| 09:49.00 | johill | mipv6? |
| 09:49.17 | johill | hah, you easily beat my uptime, I always kill my box with experimental snd-aoa modules ;) |
| 09:49.33 | johill | so that I have to reboot to get rid of them |
| 09:52.50 | Corsac | mobile ipv6 |
| 09:53.02 | Corsac | my internship is about that, so I wanted to do some tests |
| 09:53.13 | Corsac | but if I don't have bcm43xx on it, it's not very fun |
| 09:53.14 | Corsac | :) |
| 09:53.31 | Corsac | i didn't found a really good reason to reboot |
| 09:53.45 | Corsac | even the OOM error yesterday didn't force me to reboot :) |
| 09:54.07 | johill | hehe |
| 09:54.17 | johill | you should develop more kernel modules ;) |
| 09:54.22 | johill | that'll force you to reboot more often |
| 09:54.27 | Corsac | eheh |
| 09:54.31 | Corsac | i will |
| 09:54.45 | Corsac | during my internship we'll code some mipv6 linux & bsd modules |
| 09:58.27 | johill | :) |
| 10:02.07 | johill | hoping that some will make it into 2.6.18 |
| 10:02.09 | *** join/#debianppc benzea (n=benzea@p5487D6F2.dip.t-dialin.net) |
| 10:02.41 | Corsac | eheh |
| 10:03.23 | Corsac | mivp6 kernel is a patched 2.6.16 but there is no extraversion so maybe it's a 2.6.16.0 |
| 10:03.30 | Corsac | NAME=Sliding Snow Leopard |
| 10:03.38 | Corsac | and they don't provide patches, it seems |
| 10:03.42 | Corsac | only for the stable release |
| 10:03.55 | johill | that's .0 then |
| 10:03.58 | johill | just diff it against that :) |
| 10:05.34 | Corsac | yeah but I don't think patching the 2.6.17-rc5 kernel with that diff will success |
| 10:05.58 | johill | no idea how invasive mipv6 is |
| 10:12.18 | Corsac | hmhm, diff -ur isn't sufficient ? |
| 10:12.52 | johill | should |
| 10:13.30 | Corsac | Only in mipv6-20060525: README.IPsec |
| 10:13.30 | Corsac | Only in linux-2.6.16: REPORTING-BUGS |
| 10:13.35 | Corsac | ??? |
| 10:13.53 | johill | so? |
| 10:13.59 | johill | maybe they accidentally removed it |
| 10:14.09 | Corsac | yes but shouldn't I get a real patch ? |
| 10:14.34 | johill | only with -N |
| 10:14.47 | Corsac | ah thanks |
| 10:19.23 | Corsac | johill: when using my hd extensively (large compile or untaring or something), i find my powerbook quite iresponsive |
| 10:19.28 | Corsac | do you have such behavior ? |
| 10:19.36 | johill | depends on the filesystem |
| 10:19.50 | Corsac | ext3 |
| 10:19.51 | johill | that's some locking issue -- you don't have preempt I guess |
| 10:19.56 | Corsac | hmhm |
| 10:20.55 | Corsac | not here, CONFIG_PREEMPT_VOLUNTARY=y |
| 10:21.01 | Corsac | -not here, |
| 11:33.02 | *** join/#debianppc marculix (n=marculix@gw.ptr-80-238-227-85.customer.ch.netstream.com) |
| 12:06.22 | *** join/#debianppc karim (n=karim@ip-7.net-82-216-143.rev.numericable.fr) |
| 12:07.54 | karim | [|`\^@ |
| 12:09.04 | pahartik | karim: what is that all about? |
| 12:09.48 | karim | keyborad test |
| 12:09.50 | karim | error |
| 12:09.58 | karim | the keymap is fine now |
| 12:19.10 | pahartik | karim: well fixed |
| 12:19.23 | karim | yep |
| 12:19.42 | karim | I am using a G4 sawtooth |
| 12:19.55 | karim | I am not sure if the soundchip is supported by alsa now |
| 12:20.12 | karim | I use oss module, but it creates an horrible whistle from the internal speaker |
| 12:20.20 | karim | I need to mute record to not have that |
| 12:47.15 | karim | I try to build a simple thing like joe, I do "apt-build install joe", and it fails with checking for C compiler default output file name... configure: error: C compiler cannot create executables |
| 12:48.21 | Q-FUNK | karim: johill's WIP module might work |
| 12:48.55 | johill | what's a sawtooth? |
| 12:49.03 | johill | I doubt snd-aoa supports it |
| 12:49.03 | karim | johill, some fish |
| 12:49.18 | johill | G4 sawtooth? |
| 12:49.22 | karim | yes |
| 12:49.39 | karim | it's the yosemite folower |
| 12:49.40 | johill | try |
| 12:49.45 | Q-FUNK | doesn't snd-aoa support pretty much all of the G4 and G5? |
| 12:49.51 | johill | ind /proc/device-tree/ -name layout-id | xargs hexdump -e '1/4 "0x%x\n"' |
| 12:49.54 | johill | eh |
| 12:49.55 | johill | find /proc/device-tree/ -name layout-id | xargs hexdump -e '1/4 "0x%x\n"' |
| 12:50.17 | johill | Q-FUNK: don't think so, the list is on http://johannes.sipsolutions.net/Projects/snd-aoa/ |
| 12:50.45 | Q-FUNK | johill: is thta line supposed to return anything? |
| 12:50.59 | johill | Q-FUNK: yes, if snd-aoa can possibly support your machine it should print out something |
| 12:51.09 | Q-FUNK | ah |
| 12:51.16 | johill | like 0x46 |
| 12:51.20 | johill | the layout-id |
| 12:51.26 | Q-FUNK | is it merged in ALSa upstream yet? |
| 12:51.34 | johill | no, there are a few small things to do before that |
| 12:51.35 | Q-FUNK | it would allow for a wider range of testers |
| 12:51.43 | johill | like headphone detection |
| 12:51.51 | johill | karim: does that line return anything? |
| 12:52.18 | karim | no |
| 12:52.29 | johill | then snd-aoa won't support your machine (yet) |
| 12:53.09 | karim | ha |
| 12:53.14 | Q-FUNK | the device-tree has a lot of interesting stuff, but nothing for sound, afaik |
| 12:53.25 | johill | Q-FUNK: on newer machines all the sound stuff is there too |
| 12:53.31 | johill | which snd-aoa currently relies on |
| 12:53.36 | Q-FUNK | ah |
| 12:53.39 | johill | well, not all of the really very new stuff |
| 12:53.46 | Q-FUNK | is that an I2C shortcut? ;) |
| 12:53.49 | johill | the last powerbooks had more than mine, and it works on mine, but ... |
| 12:53.58 | johill | yeah there's some i2c stuff on the very latest models |
| 12:54.04 | johill | it doesn't need that |
| 12:54.14 | johill | but it needs the i2s stuff and the layout-id to know what to do |
| 12:55.31 | Q-FUNK | snd_powermac calls i2c_core |
| 12:55.47 | johill | but in a very stupid way, mind you |
| 12:56.11 | johill | i2c is a bit stupid anyway, you can't really detect if there's something on it or not |
| 12:56.47 | johill | (snd-powermac relies on the i2c bus being there, see the recent bug reports about it segfaulting when i2c-powermac was renamed) |
| 12:57.57 | karim | my error is checking for C compiler default output file name... configure: error: C compiler cannot create executables |
| 12:58.44 | johill | maybe you don't even have one installed? |
| 12:59.49 | pahartik | karim: have you got "libc6-dev", "gcc" and "g++" installed, if you prefer building packages from sources instead of using existing binary packages? |
| 13:00.39 | karim | pahartik, I have all that, I even did cross compilation once |
| 13:00.46 | karim | with distcc |
| 13:05.51 | Q-FUNK | karim: install "build-essentials" |
| 13:06.10 | karim | it's already installed |
| 13:06.23 | Q-FUNK | then it should work |
| 13:06.32 | karim | the build works with dpkg-buildpackage but apt-build fails |
| 13:06.53 | Q-FUNK | what about with debuild? |
| 13:06.59 | Q-FUNK | or pbuilder? |
| 13:08.21 | karim | don't know for this one |
| 13:09.00 | pahartik | karim: what one really needs to execute is "debian/rules build" and "fakeroot debian/rules binary" :) |
| 13:09.35 | karim | ah ? |
| 13:11.15 | pahartik | karim: "debian/rules" is package-specific executable script within package source folder that one can retrieve with "apt-get source" |
| 13:11.35 | Q-FUNK | yes, but that won't solve depenencies & such |
| 13:14.39 | karim | something must be missing |
| 13:16.44 | karim | it always fail here : checking for C compiler default output file name... configure: error: C compiler cannot create executables after 10 lines of configure |
| 13:17.14 | pahartik | Q-FUNK: oh yes, but I often edit "debian/rules" to make it have options to link with libraries and other things I prefer on "./configure" stage... otherwise binary package would be just as good... |
| 13:17.42 | Q-FUNK | pahartik: do you practice safe macros? :) |
| 13:18.25 | pahartik | Q-FUNK: I do not know what that is... :) |
| 13:19.00 | Q-FUNK | pahartik: there's a debian-safe way of altering those ./configure options |
| 13:21.12 | pahartik | Q-FUNK: examples I can think of right away... disabling some options on "libxine1", disabling terrible GUI on "mplayer" |
| 13:23.44 | pahartik | Q-FUNK: what would be safer than editing "debian/rules"? |
| 13:24.33 | Q-FUNK | it IS done in debian/rules, but there are debian-specific ways of passing configure options via there |
| 13:25.17 | pahartik | Q-FUNK: I have been naughty with custom packages? |
| 13:29.30 | Q-FUNK | could be :) |
| 13:29.48 | Q-FUNK | you can look at my packages for examples of how it's done |
| 13:35.05 | *** join/#debianppc tailok (n=tailok@p54BF7883.dip.t-dialin.net) |
| 13:38.42 | *** join/#debianppc dpro (n=x@chello080109121047.8.15.vie.surfer.at) |
| 13:41.05 | karim | when you do /bin/uname -p -i what message do you have ? |
| 13:41.13 | karim | I think that's the problem |
| 13:41.17 | karim | it says unkown for me |
| 13:42.00 | karim | ok I have the same message on K7 ... |
| 13:42.40 | Corsac | corsac@hidalgo: uname -a [~] |
| 13:42.44 | Corsac | Linux hidalgo 2.6.17-rc2 #4 Fri Apr 28 16:04:48 CEST 2006 ppc GNU/Linux |
| 13:42.46 | Corsac | corsac@hidalgo: uname -p -i [~] |
| 13:42.49 | Corsac | unknown unknown |
| 13:55.15 | pahartik | karim: "unknown unknown" |
| 13:57.22 | karim | may I ask you to try apt-buil for txt editor joe ? |
| 13:58.59 | pahartik | karim: yes, though I prefer "emacs" and "nano"... |
| 13:59.46 | pahartik | karim: "bash: apt-build: command not found" :) |
| 13:59.56 | karim | joe is the text editor they told us to use at the unic on this crap unix terminals |
| 14:00.07 | karim | you need to install it :D |
| 14:29.34 | karim | hi |
| 14:29.54 | karim | fot g4 the -mtune option should be what ? |
| 14:29.56 | karim | -mtune=type-cpu |
| 14:30.18 | karim | I am not sure it it's what is in /proc/cpu |
| 14:36.43 | *** join/#debianppc iBalo (n=user@dslb-088-072-212-138.pools.arcor-ip.net) |
| 15:06.30 | dpro | karim: I use -mcpu 7450 -maltivec |
| 15:06.49 | karim | dpro, with apt-build |
| 15:06.50 | karim | ? |
| 15:07.08 | karim | dpro, when you do cat /proc/cpuinfo, what's your cputype ? |
| 15:07.18 | karim | mine is 7400 |
| 15:07.23 | dpro | karim: as in export CFLAGS="..." ; export CXXFLAGS="..." |
| 15:07.28 | karim | I was wondering if there was more than one g4 version |
| 15:07.42 | karim | dpro, what is it ? |
| 15:07.52 | dpro | mine is 7455 (as reported by /proc/cpuinfo) |
| 15:07.58 | karim | dpro, ok |
| 15:08.12 | karim | where can I find the standard CFLAGS things ? |
| 15:08.50 | dpro | karim: autoconf / configure honors those environment vars scons doesn't |
| 15:09.23 | karim | dpro, ok so I don't need to set that in apt-build.conf |
| 15:09.38 | karim | I added -maltivec, but I am not sure if it's not implied |
| 15:15.32 | dpro | karim: I am really neither a gcc nor a apt-build guru, I think I mostly use this out of superstition (I never actually profiled code checking the difference) |
| 15:16.10 | dpro | karim: what are you compiling anyway ? |
| 15:18.01 | *** join/#debianppc Q-FUNK (n=q-funk@dsl-hkigw7-fe12f900-46.dhcp.inet.fi) |
| 15:18.59 | karim | dpro, just having fun |
| 15:19.10 | karim | I am about to build world |
| 15:19.13 | karim | just for fun |
| 15:28.21 | dpro | hehe ... a strange sense of humour I'm tempted to say ;) |
| 15:35.30 | *** join/#debianppc tcsetattr (i=pacman@dialup58.clss.net) |
| 16:35.22 | *** join/#debianppc simonrvn (i=simon@unaffiliated/simonrvn) |
| 16:48.15 | *** join/#debianppc MajorTom (n=MajorTom@p54863CEE.dip0.t-ipconnect.de) |
| 16:50.24 | Q-FUNK | <PROTECTED> |
| 16:57.04 | *** join/#debianppc dreamind (i=ENETDOWN@fw-ext.konaktiva.tu-darmstadt.de) |
| 16:58.04 | Bor_Ed | good one |
| 16:58.45 | karim | I was googling and heard about apt-fu |
| 16:59.00 | karim | seems it's like apt-build, but I can't find a trace of apt-fu |
| 16:59.12 | karim | seems it died or something while it was promising |
| 16:59.18 | Bor_Ed | I'd gues it's like apt-get moo |
| 17:01.38 | karim | moo ? |
| 17:01.47 | Bor_Ed | moo |
| 17:01.58 | karim | you mean meuh |
| 17:02.03 | Bor_Ed | nope |
| 17:02.16 | karim | apt-fu is supposed to be like emerge |
| 17:02.24 | karim | moo is the call of the cow |
| 17:02.31 | karim | in french it's meuh |
| 17:02.33 | karim | lol |
| 17:02.33 | Bor_Ed | pretty much the same :) |
| 17:03.02 | Bor_Ed | emerge port on debian... sheesh |
| 17:03.14 | karim | why not |
| 17:03.26 | karim | I hate gentoo because there is no config tools |
| 17:03.26 | Bor_Ed | apt-build is enough for me |
| 17:03.48 | karim | well I don't like how it handle world upgrades |
| 17:03.53 | Bor_Ed | apt-get source -b package or similar, haven't used it in ages |
| 17:04.08 | *** join/#debianppc s^3 (n=s^3@tor/session/external/x-4f7a65d0e44dbd90) |
| 17:04.13 | karim | apt-src does that well |
| 17:04.40 | karim | but I don't see how to give compile options |
| 17:05.02 | Bor_Ed | I don't recall the package name... but the options are configurable on apt options |
| 17:05.18 | Bor_Ed | like the CFLAGS and so on leet tuning options |
| 17:05.27 | karim | great |
| 17:06.23 | karim | I guess apt-src can be used much like apt-build then, by using it in a loop on all packages for the world |
| 17:06.32 | karim | I hope to speed up gnome a bit |
| 17:06.41 | Bor_Ed | I doubt |
| 17:07.03 | karim | well I installed gentoo and felt it was a bit faster |
| 17:07.24 | karim | maybe placebo |
| 17:07.49 | Bor_Ed | gentoofs might be faster than potato, but on recent releases I have pretty heavy doubts on the optimization diff on performance |
| 17:08.43 | karim | potato ? |
| 17:09.09 | Bor_Ed | debian 2.2 |
| 17:09.34 | karim | on dumb thing from apt-build is it seems he also rebuild package with arch "all" |
| 17:09.50 | karim | it shouldn't take time anyway |
| 17:11.06 | Bor_Ed | you can probably opt it to build just for the current arch too |
| 17:11.35 | karim | that must be done through the dpkg --get-selection command |
| 17:12.13 | Bor_Ed | no idea... I really haven't hacked in apt setups that much |
| 17:12.35 | Bor_Ed | I know build-dep and source... those are pretty much enough :) |
| 17:14.10 | karim | by the way since one source package can provide more than one binary .deb |
| 17:14.41 | karim | if the list is build uon insstalled packages I am not sure how it will behave |
| 17:35.38 | Corsac | johill: is there a "memtestppc" (like memtest86) to check ram ? |
| 17:35.55 | Corsac | (and is there still a 1GB limit with bcm43xx ?) |
| 17:36.28 | johill | no, there isn't afaik |
| 17:38.03 | simonrvn | there was one. dunno why it got pulled. called memtest |
| 17:42.21 | johill | yeah, it's not arch specific though, it just tries to allocate as much mem as possible |
| 17:42.29 | johill | memtester |
| 18:04.18 | Q-FUNK | http://q-funk.blogspot.com/2006/05/brightness-control-on-imac-g3-under.html |
| 18:25.36 | karim | when some package a specific -dev files to be built that are not *all* acrichitecture |
| 18:26.05 | karim | if the normal package is build with headers issued of different optimisation, will it work ? |
| 18:26.35 | karim | or should the -dev package depend of the arch also be compiled with the specific options ? |
| 18:35.15 | *** join/#debianppc MelkorD (n=MorgothD@eha202.internetdsl.tpnet.pl) |
| 18:45.19 | johill | the -dev package usually also ships .a files for static linking which are obviously arch-dependent |
| 18:54.45 | simonrvn | johill: ok |
| 18:55.33 | johill | Q-FUNK: you probably want hansmi's recent brightness infrastructure changes for that |
| 18:56.01 | Q-FUNK | johill: could be |
| 18:56.16 | johill | if it needs to be in the kernel, that is |
| 18:56.25 | Q-FUNK | most brigthness/contrast gizmos so far only seem to support powerbooks |
| 18:56.31 | johill | maybe you can just bit-bang the busses from userespace |
| 18:56.40 | johill | he also supports the apple cinema usb displays now |
| 18:56.46 | Q-FUNK | maybe, but I wouldn't know how |
| 18:57.01 | Q-FUNK | now would i risk it |
| 18:57.05 | johill | me neither |
| 18:57.05 | Q-FUNK | nor |
| 18:57.21 | simonrvn | having a backside cache on a cpu is nice. the box is faster. ;) |
| 18:57.23 | Q-FUNK | all i recall was Ben saying that this would be child's play |
| 18:57.39 | simonrvn | Q-FUNK: the h/w upgrade happened and works, finally |
| 18:59.02 | Q-FUNK | nice :) |
| 18:59.37 | simonrvn | the bogomips in /proc/cpuinfo is buggered though, it shows 41-ish |
| 18:59.54 | Corsac | the bogomips are only bogus |
| 19:00.06 | johill | simonrvn: that's because ppc no longer uses a loop for measurements |
| 19:00.31 | simonrvn | oh, i see. it's running some version of 2.6.16 |
| 19:00.54 | johill | yeah, sounds about right, .16 should have had that |
| 19:01.07 | simonrvn | ok. nothing to worry about then |
| 19:01.18 | johill | yeah |
| 19:01.47 | simonrvn | also i noticed the lack of a sysctl l2cr interface |
| 19:02.14 | johill | l2cr? |
| 19:02.17 | simonrvn | which i was using to set some CPU options |
| 19:02.25 | simonrvn | l2 cache control |
| 19:02.32 | johill | oh. no idea |
| 19:02.48 | simonrvn | /proc/sys/kernel/l2cr or somewhere |
| 19:03.19 | simonrvn | yeh, there |
| 19:04.31 | simonrvn | exactly. got removed |
| 19:08.32 | karim | hi |
| 19:08.48 | simonrvn | tansi |
| 19:08.55 | karim | why are some package in *-dev not for architecture *all* ? |
| 19:09.14 | karim | there is something I don't get |
| 19:10.25 | johill | karim: I told you |
| 19:10.25 | johill | <johill> the -dev package usually also ships .a files for static linking which are obviously arch-dependent |
| 19:10.27 | simonrvn | what he said |
| 19:10.35 | karim | johill, well you told the room, but not to me in particular :) I missed that message |
| 19:10.55 | johill | heh ok |
| 19:11.06 | karim | johill, what makes it arch independent ? |
| 19:11.20 | karim | I am not sure what is static linking I mean |
| 19:11.28 | karim | -in |
| 19:11.53 | johill | well, using cc's -static flag to get a binary that doesn't have dynamic library dependencies |
| 19:12.39 | karim | ok but if it's static what does it mean ? It means that .a file will be kind of included into the final binary ? |
| 19:12.52 | johill | sort of |
| 19:13.31 | karim | ok, so if I want to rebuild a package with some cpu optimisation, will it work ? |
| 19:13.49 | karim | I mean if the static lib is not rebuild with those optimisations |
| 19:14.07 | simonrvn | it's not an optimisation |
| 19:14.16 | karim | simonrvn, ? |
| 19:15.18 | karim | simonrvn, mm ? |
| 19:16.43 | karim | so if I rebuild a package with -maltivec and -mtune=7400, and that the depencies resolution installs a -dev package with static libs, will it work ? |
| 19:17.05 | karim | the static lib beeing already built |
| 19:17.22 | johill | the lib won't be rebuilt, but yes, it'll work |
| 19:17.32 | johill | though, usually, you won't even be building a static binary |
| 19:17.35 | johill | just linking against the dynamic one |
| 19:18.50 | karim | johill, but if it builds against static ones that are prebuilt, I will loose some optimisation then. |
| 19:19.01 | karim | I guess |
| 19:19.38 | johill | yes, but have you actually measured the performance increase? it's probably not much more over the time you'll be using the software than the time you spent today trying to make it work ;) |
| 19:20.17 | simonrvn | no kidding ;) |
| 19:21.12 | karim | johill, don't try to reason me |
| 19:22.48 | karim | well about the static thing the thing that precupied most was the risk it would not work at all |
| 19:33.53 | simonrvn | i'm so glad i went through the pain of getting this beige g3 set up. it's zippier than the power tower pro with a similar cpu |
| 19:35.04 | simonrvn | i'm booting off of an atto/53c875 scsi card. does it nicely. the other drives "live" on an adaptec 2940 (my u-131 isn't liking the beige for some reason, or it's buggered) |
| 19:40.57 | *** join/#debianppc s^3 (n=s^3@tor/session/external/x-bf3a1ad33e3d76a1) |
| 19:54.37 | *** join/#debianppc dolphy (n=dolphy@218.Red-80-38-145.staticIP.rima-tde.net) |
| 20:19.55 | *** join/#debianppc dahkneth (n=dahkneth@81-236-225-223-no51.tbcn.telia.com) |
| 20:43.07 | *** join/#debianppc tcsetattr (i=pacman@dialup20.clss.net) |
| 21:18.24 | *** join/#debianppc deadcat (n=deadcat@adsl-69-110-235-170.dsl.pltn13.pacbell.net) |
| 21:18.55 | deadcat | i am having trouble booting up my 2.6.16.18 kernel. complaining about the root= line, but my old 2.6.14 and the debian .16 works fine. any clue? |
| 21:21.20 | *** join/#debianppc tmbinc (i=XXX@e176158103.adsl.alicedsl.de) |
| 21:36.10 | *** join/#debianppc deadcat (n=deadcat@adsl-69-110-235-170.dsl.pltn13.pacbell.net) |
| 21:36.15 | deadcat | VFS: cannot open root device "hda3" or unknow-block(0,0) |
| 21:36.19 | deadcat | A64: the 2.6.14 worked fine. an i am pretty much using the same .config on the 2.6.16.18 dont know what up |
| 21:36.29 | deadcat | plearse append a correct "root=" boot option. \n kernel panic - not syncing: VFS: unable to mount root fs on uknown-block(0,0) |
| 21:36.52 | johill | deadcat: probably forgot about the initrd or something |
| 21:37.09 | deadcat | johill: i never include initrd in my kernels though. |
| 21:37.26 | johill | heh, well, dunno then |
| 21:37.34 | johill | lots of config options changed |
| 21:37.40 | deadcat | damn |
| 21:37.43 | johill | maybe for some reason the relevant stuff got kicked out |
| 21:37.53 | johill | 2.6.14->2.6.16 is a job for a very careful make oldconfig |
| 21:37.56 | deadcat | maybe i should build it with initrd support |
| 21:38.05 | deadcat | damn |
| 21:41.28 | deadcat | johill: if i enable initrd, what else do i need to enable? |
| 21:41.53 | johill | if you don't use an initrd you don't need to enable that |
| 21:41.58 | johill | better check if IDE is built-in etc |
| 21:42.39 | deadcat | k. i never use initrd. but i cant figure out why this kernel is failing on me. )= |
| 21:47.02 | *** join/#debianppc reDFog (n=redfog@tor/session/external/x-a341b19769427970) |
| 22:17.35 | *** join/#debianppc MajorTom (n=wishmast@p54863CEE.dip0.t-ipconnect.de) |
| 22:33.56 | *** join/#debianppc s^3 (n=s^3@tor/session/external/x-55b333e69c15bbe5) |
| 22:58.13 | deadcat | fuck. goddamn kernel pacnic with the freakin VFS: unable to mount root fs on uknown-block(0,0) AGAIN!!! |
| 23:15.28 | *** join/#debianppc TBBle (i=identdsu@202.55.154.106) |
| 23:54.05 | *** join/#debianppc tcsetattr (i=pacman@dialup90.clss.net) |