| === asac_ is now known as asac | ||
| dholbach | good morning | 07:15 |
|---|---|---|
| === cjwatson_ is now known as cjwatson | ||
| === doko_ is now known as doko | ||
| === cjwatson_ is now known as cjwatson | ||
| repete | !seen kwii | 12:08 |
| ubotu | Sorry, I don't know anything about seen kwii - try searching on http://ubotu.ubuntu-nl.org/factoids.cgi | 12:08 |
| === repete is now known as repete_bbl | ||
| === repete_bbl is now known as repete | ||
| === cjwatson_ is now known as cjwatson | ||
| === \sh_away is now known as \sh | ||
| Mithrandir | are there anybody around who would mind if we had the meeting in person? | 17:02 |
| Mithrandir | given that it tends to be Canonical and Intel people contributing and most of those are already present in the room. | 17:03 |
| Mithrandir | we'll be happy to have it here if there are anybody external, though. | 17:03 |
| alek_desk | Mithrandir: hi | 17:03 |
| === robr_ is now known as robr | ||
| alek_desk | robr: hi | 17:04 |
| Mithrandir | alek_desk: yo | 17:04 |
| alek_desk | Mithrandir, hoyo. Then I could sleep again. if no IRC meeting here :) | 17:06 |
| Mithrandir | alek_desk: you could come into the office? :-P | 17:06 |
| alek_desk | Mithrandir, :) go to office at 1AM in PRC is not very good idea. | 17:07 |
| Mithrandir | alek_desk: oh, you're in China. Sorry. :-) | 17:07 |
| Mithrandir | I thought you were local to Portland, for some reason. | 17:08 |
| GrueMaster | Awe, where's your sense of adventure? | 17:08 |
| alek_desk | Mithrandir, haha. | 17:08 |
| Mithrandir | alek_desk: we have all the names of your kernel people, for access to kernel.ubuntu.com? If so, we'll take the rest offline here in Oregon | 17:09 |
| robr | alek_desk: hi | 17:09 |
| alek_desk | Mithrandir, yes. I think so. | 17:09 |
| Mithrandir | alek_desk: ok, goodie. See you around then. :-) | 17:10 |
| alek_desk | robr, about the booting performance issue. | 17:10 |
| alek_desk | Mithrandir, see you. | 17:10 |
| alek_desk | Martin got the "real" data comparison, do you see the mail ? | 17:11 |
| Jay-laptop | alek_desk: robr is busy checking his emails next to me :-) | 17:11 |
| alek_desk | robr, we think for Ubuntu booting counting, our data included BIOS time originally. | 17:12 |
| alek_desk | Jay-laptop, thanks :) | 17:12 |
| alek_desk | robr, while for RF's data, it is pure OS booting time | 17:12 |
| alek_desk | so actually, our image boot faster than RF image. | 17:13 |
| robr | alek_desk: what do you mean? | 17:13 |
| alek_desk | Jay-laptop, do you see the patch feng sent out | 17:14 |
| alek_desk | robr, it was said RF booting only need 51 seconds and Moblin image booting need 61 seconds in a mail sent by Ubuntu people. | 17:15 |
| robr | alek_desk: i saw martin's email | 17:15 |
| alek_desk | robr, actually the RF data did not count BIOS time. | 17:16 |
| Jay-laptop | alek_desk: Not yet, did he send it to ubuntu kernel-team? | 17:16 |
| alek_desk | Jay-laptop, no. | 17:16 |
| robr | alek_desk: the timings that Martin included, by whom and where was the timings taken? | 17:16 |
| alek_desk | robr, Martin tested it on a CRB C0 board. | 17:17 |
| robr | alek_desk: we think the numbers that were sent out were on a different BIOS and customer hw | 17:19 |
| GrueMaster | I'm in the lab now with D0 CB and Bios 67. My timing for Moblin 20080114 is 61 seconds (less 18 seconds for bios). | 17:20 |
| GrueMaster | My tests on RF were ~65 seconds. | 17:21 |
| GrueMaster | This is from power on. | 17:21 |
| alek_desk | robr, GrueMaster, I think we should only count pure OS booting time for both. | 17:21 |
| alek_desk | robr, BTW, we got a EFI bios from tinao team, it only costs 3 seconds. | 17:22 |
| Mithrandir | I don't think we want EFI if we can avoid it. According to what I've heard from mjg59, suspend and such can become interesting then | 17:23 |
| alek_desk | Mithrandir, the EFI team here want to push that though :( | 17:25 |
| Mithrandir | also, 3s is quite a bit | 17:26 |
| Mithrandir | alek_desk: do you mean the total is 3s, or it is 18s+3s? | 17:28 |
| robr | alek_desk: i agree we need to be talking about the same thing and breakout the numbers ... we can't ignore bios time, so we need to measure it, all measurements need to be broken out | 17:29 |
| robr | Mithrandir: what about EFI makes suspend interesting? | 17:30 |
| Mithrandir | I don't remember the details, sorry. It might be something else than suspend | 17:32 |
| Mithrandir | usplash will break, at least, unless EFI maps the video bios at the same address as the legacy address. | 17:32 |
| StevenK | Which I suspect is unlikely | 17:33 |
| alek_desk | Mithrandir, according to our test, usplash works. | 17:33 |
| Mithrandir | alek_desk: hum, ok. | 17:33 |
| Mithrandir | I'm not even pretending to know much about EFI; I just know that if mjg59 tells me it might well break something, it might well break something. :-) | 17:34 |
| robr | there are other reasons to not like EFI -- i should give the disclaimer that i used to work in the EFI group at Intel | 17:34 |
| alek_desk | :) | 17:34 |
| suihkulokki | linuxbios! | 17:35 |
| Mithrandir | alek_desk: could you respond to my question about whether it's 3s total or 3s more (with EFI)? | 17:35 |
| alek_desk | Mithrandir, oh, sorry, it is total time. | 17:35 |
| alek_desk | from power on to loading kernel by Elilo | 17:36 |
| robr | alek_desk: that's 3s for EFI and booting the kernel and launching X Windows? | 17:36 |
| alek_desk | robr, from press power button to Elilo appears | 17:36 |
| robr | alek_desk: ahh, you mean 3 seconds to load ELILO | 17:37 |
| Mithrandir | robr: slightly optimistic, that. :-) | 17:37 |
| alek_desk | robr, the BIOS initialization is very fast | 17:38 |
| alek_desk | And X window startup is really time consuming. | 17:39 |
| robr | alek_desk: i wouldn't say 3 seconds is fast, it's just fast for a PC BIOS | 17:41 |
| alek_desk | robr, it is much faster then current CRB bios | 17:42 |
| robr | anyway it's a little pointless to be talking about which flavor of BIOS is used or not used -- it will be the customer that decides one way or another | 17:42 |
| alek_desk | robr, yes. we know. | 17:42 |
| robr | we can only reduce the OS boot time and that is what we should focus on | 17:42 |
| alek_desk | robr, yes. But currently have no idea how to make kernel boot faster or to let X startup faster. :( | 17:47 |
| robr | alek_desk: one thing that will help with booting the kernel faster is to compile all the required drivers into the kernel instead of loadable modules | 17:48 |
| robr | alek_desk: i believe most drivers are compiled in on the moblin kernel, but not all | 17:49 |
| robr | alek_desk: but i think we need to have brainstorming meeting with Ubuntu engineers and rusty and see if we can get a bin list of things to try to reduce boot speed | 17:50 |
| alek_desk | robr, yes. especially for GUI stuff, since almost half of time is eaten by them. | 17:51 |
| alek_desk | robr, there is another issue about HDA audio | 17:53 |
| alek_desk | there are 50+ extra nobody care IRQs trigger during HDA working | 17:53 |
| robr | GrueMaster: any ideas on alek_desk 's comment about HDA? | 17:54 |
| alek_desk | Martin already filed a launchpad report to track that. | 17:54 |
| GrueMaster | Oops, missed it. | 17:55 |
| GrueMaster | I was away fending off another fire. | 17:55 |
| GrueMaster | I'm going to do some quick tests of alsa 1.0.16 today. I think it will help, as there have been some substantial changes to the core code. | 17:56 |
| === mawhalen_ is now known as mawhalen | ||
| === mawhalen_ is now known as mawhalen | ||
| === mawhalen_ is now known as mawhalen | ||
| bspencer | bfiller, looking at the log at: https://code.launchpad.net/~ubuntu-mobile/hildon-desktop/ubuntu | 19:24 |
| HappyCamp_laptop | bspencer: smagoun rustyl Mithrandir robr: So are we going to have a hardy sync up? Figure out what still needs to be done, etc... | 19:40 |
| bspencer | HappyCamp_laptop, good to do. When? We are in the middle of finding out what that is. | 19:42 |
| HappyCamp_laptop | bspencer: I would think sooner the better, so people know what needs to be done. | 19:42 |
| === mawhalen_ is now known as mawhalen | ||
| Mithrandir | HappyCamp_laptop: we had the small bits that were still hanging from Monday discussed earlier. Read https://intel.wiki.canonical.com/Mobile/UbuntuMobileSprintJanuary2008/MoblinReviewNotes and strike through what's done. | 20:13 |
| === mawhalen_ is now known as mawhalen | ||
| === mawhalen_ is now known as mawhalen | ||
| === \sh is now known as \sh_away | ||
| === mawhalen_ is now known as mawhalen | ||
| smagoun | inuka_desk: http://people.ubuntu.com/~bryce/Testing/libdrm/Hardy/libdrm2_2.3.0-4ubuntu4~804um1_lpia.deb | 22:47 |
| smagoun | http://people.ubuntu.com/~bryce/Testing/libdrm/Hardy/libdrm2-dbg_2.3.0-4ubuntu4~804um1_lpia.deb | 22:47 |
| smagoun | http://people.ubuntu.com/~bryce/Testing/libdrm/Hardy/libdrm-dev_2.3.0-4ubuntu4~804um1_lpia.deb | 22:47 |
| smagoun | http://people.ubuntu.com/~bryce/Testing/xserver-xorg-video-psb/Hardy/xserver-xorg-video-psb_0.2.6-0ubuntu1~804um1_i386.deb | 22:47 |
| smagoun | http://people.ubuntu.com/~bryce/Testing/xserver-xorg-video-psb/Hardy/xserver-xorg-video-psb-dbg_0.2.6-0ubuntu1~804um1_i386.deb | 22:47 |
| bspencer | bfiller, I have patches against hm-statusbar-container | 22:57 |
| bspencer | which doesn't exist in Hildon 2.0 | 22:57 |
| bspencer | it is for embedding the statusbar | 22:57 |
| bspencer | do you want to add these in stages, or just a single patch with a fully updated hm-statusbar container ? | 22:58 |
| bfiller | bspencer: yup, I'm creating a patch right now that adds it | 22:58 |
| bspencer | hm... then my patch which also adds it needs to be changed :) | 22:58 |
| bfiller | bspencer: I wonder if it might be easier for you to wait till I'm done | 22:58 |
| bspencer | yep | 22:58 |
| bfiller | bspencer: I'm almost there | 22:58 |
| bfiller | bspencer: probably 1 more hour | 22:58 |
| GrueMaster | smagoun: Do you already have the Beta6 kernel module intigrated? These drivers are very version dependent on the kernel module. | 23:00 |
| bspencer | bfiller, np | 23:00 |
| bspencer | in fact all the changes I have touch that file (dynamic screen size, statusbar length fix) | 23:00 |
| Mithrandir | ToddBrandt: were you part of the moblin integration discussion monday? | 23:00 |
| smagoun | GrueMaster: for gutsy or hardy? | 23:01 |
| GrueMaster | Both. If you plan on integrating the beta6 video drivers for X, you need the corresponding kernel module. | 23:02 |
| ToddBrandt | Mithrandir: no I missed that one | 23:02 |
| GrueMaster | I just saw your post with the links. | 23:02 |
| Mithrandir | ToddBrandt: ok, you have an action item to get the matchbox keyboard manager things into the PPA (and upstream). | 23:02 |
| Mithrandir | ToddBrandt: I'm not going to ask how it's going, then. :-) | 23:03 |
| smagoun | GrueMaster: Inuka has a LUM with the beta 6 driver, and we're trying that alongside bryce's updates to the x server | 23:03 |
| ToddBrandt | Mithrandir: who gave me the AR? Was it Mauri? | 23:03 |
| Mithrandir | ToddBrandt: I think she was in the meeting, yes. | 23:04 |
| GrueMaster | ok, just checking. I remember that Gutsy had the Beta3 X drivers and the alpha kernel module for a while. | 23:04 |
| smagoun | GrueMaster: yup, it's a bit of a mess :) | 23:05 |
| GrueMaster | It happens. | 23:06 |
| sabotage | kyleN: whered yall go? | 23:26 |
| === robr_ is now known as robr | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!