| Daeken | will there be any improvements in NTFS support in the 2.5.x kernel line? |
|---|---|
| fork | actually i would ask what are the biggest improvements regarding posix compliance. It seem to be important to port more software to linux. And to use linux as commercial platform at all |
| @mulix | what do you guys feel represents the biggest change for out-of-tree developers to cope with? |
| Daeken | sarnold: no problem ;) |
| ShawnWerk | team: when can we expect some updated documentation for 2.5? I'm nearly ripping my hair out ;-) |
| noph | well, i'm not that "kernel-deep" as the other people around here. the only problem i had with the kernels nowdays is when i put in a weird cd in my cd-drive and mount the cd. sometimes the kernel crashes :\ Will mount be different in any way in the new kernel-serie? |
| arashi | noph, it isn't |
| @mulix | davej, such a document would be very useful... |
| imbezol | would it be possible to have a quick introduction of the speakers.. what role they play in the kernel devlopment/community? |
| @mulix | davej, I'll be delighted to help with the fleshing out part. |
| PiotR | were can we read about the changed cli() /sti() stuff? |
| @rml | PiotR: Documentation/cli-and-sti.txt or somesuch in the source tree |
| ShawnWerk | davej wrote the Happy Meal(tm) :-) |
| andre | PiotR: sadly us who have been around a long time just do it |
| @mulix | davej, maybe he borrowed some of Alan's gnomes :-) |
| PiotR | ok, thanks |
| ShawnWerk | hi ata =) |
| PiotR | I'm learning 2.4 stuff, I hope the transition to 2.5 isn't too hard :) |
| ShawnWerk | PiotR: they changed alot ;) |
| will-h | Lots will change as ever |
| ShawnWerk | (*cough* PnP) |
| ShawnWerk | davej: I see, but im looking for a document on what was overhauled =) |
| Daeken | in linux, minor versions don't mean minor changes :P |
| ShawnWerk | Question to anyone: Is there a reason I can't register a AMD Northbridge with the PCI layer? It refuses to register. :( |
| ShawnWerk | davej: I see. |
| ShawnWerk | I find 2.5.52 blazingly fast with desktop response |
| @mulix | sarnold approved my off topic question, so: |
| @mulix | what do you consider the best technical book you ever read? |
| arashi | heh, maybe we could see something like CONFIG_THROUGHPUT for large machines, and CONFIG_RESPONSIVENESS for desktops ... |
| Lovechild | did I miss the lecture? |
| Daeken | mulix: The XML Black Book |
| @rml | GED |
| Daeken | Lovechild: not at all ;) |
| Lovechild | ahh... |
| Lovechild | hi rml |
| @rml | hi Lovechild |
| @gregkh | ShawnWerk: sorry :( |
| ShawnWerk | gregkh: It looks easier to use |
| Lovechild | rml: thanks for the interview |
| noph | how about userspace in the new kernel? will the new kernel-serie prior userpace or speed? |
| @rml | indeed |
| noph | (sorry for my bad english) |
| Daeken | another question... will supermount be included in the kernel itself this time? |
| ShawnWerk | trying to port Alan's AMD 76x PM to 2.5 is a nightmare :( |
| @mulix | Daeken, not likely, we're in "feature freeze" time. |
| Daeken | mulix: ahhh... ok, thanks :) |
| ShawnWerk | hence the need to register the Northbridge (if we need to) |
| ShawnWerk | I do notice though that SCSI is completely busted in 2.5.52 :-( |
| ShawnWerk | when I try to extract a tape w/ cpio i get corruption |
| ShawnWerk | ;-) |
| Lovechild | oh I want my software RAID to work with 2.5... but no... it says that root=/dev/md1 is invalid... |
| Lovechild | but that can't be right... since I'm on the very same raid setup now on 2.4 |
| ShawnWerk | davej: but how come 2.5 won't let me register the Northbridge? when compile kernel with -DDEBUG, the closes thing pci_scan_bus() returns is the 700d device but I need 700c :-( |
| ShawnWerk | unless this is a bug(?) |
| ShawnWerk | it registered the Southbridge though |
| ShawnWerk | ok |
| andre | eew that sounds line a nasty |
| ShawnWerk | davej: lemme see |
| bugfixer | What is the status of Hyper-Thread support in 2.5, and has any of that been back ported to 2.4? |
| ShawnWerk | davej: pci_match_bus(): DECISION ids->device 700c == pci_dev->device 700d? |
| ShawnWerk | pci_device |
| Lovechild | bugfixer: the O(1) scheduler is SMT aware |
| ShawnWerk | but pci_match_bus skips over |
| ShawnWerk | 00:00.0 Class 0600: 1022:700c (rev 11) <---------- Northbridge |
| ShawnWerk | something clearly is weird |
| andre | ShawnWerk: is that by name or by id ? |
| ShawnWerk | id |
| ShawnWerk | the 700d is the IDE |
| andre | this is amd |
| ShawnWerk | yes |
| andre | I had a typo for a while but it should have been fixed |
| ShawnWerk | this problem you mean? |
| andre | yes the exact |
| ShawnWerk | so it should be reporting the 700c |
| ShawnWerk | is this in 2.5.52? |
| mcp | good evening |
| Lovechild | hi mcp |
| mas | Hi, I wanted to know panel's idea about the MSFT's trusted computing environment that basically means (to me as a programmer): hardware won't run the software not signed by a trusted party of MSFT? In what ways it'll affect Linux communitiy and development efforts for it? |
| ShawnWerk | .51 |
| ShawnWerk | i have not tried this in .52 yet |
| andre | should not be, the was almost a year ago hen my brain slipped |
| ShawnWerk | hmm |
| ShawnWerk | here's what i logged (partly) |
| ShawnWerk | amd76x_pm: amd76x_pm hardware power management 0.1.0 |
| ShawnWerk | amd76x_pm: Looking for Northbridge... |
| ShawnWerk | bus pci: add driver amd76x_pm-nb |
| ShawnWerk | kobject amd76x_pm-nb: registering. parent: <NULL>, subsys: drivers |
| ShawnWerk | Inside driver_attach() |
| ShawnWerk | !bus->match not true continue on.. |
| ShawnWerk | !dev->driver keep going... |
| ShawnWerk | inside bus_match() |
| ShawnWerk | inside pci_bus_match() |
| mcp | .oO(I am wondering why #linux is +M) |
| ShawnWerk | pci_match_bus(): DECISION: ids->vendor 1022 == pci_dev->vendor 1022? |
| ShawnWerk | pci_match_bus(): DECISION ids->device 700c == pci_dev->device 700d? |
| ShawnWerk | pci_match_bus(): DECISION ids->subvendor ffffffff == pci_dev->subsystem_vendor |
| ShawnWerk | 0? |
| ShawnWerk | pci_match_bus(): DECISION ids->subdevice ffffffff == pci_dev->subsystem_device |
| ShawnWerk | 0? |
| ShawnWerk | bus_match(): match(): We didn't find a match. Failed |
| ShawnWerk | bus_match(): Returns -19 [-ENODEV] |
| ShawnWerk | Dec 14 23:20:37 unknown kernel: Driver_attach(): Returns 0 |
| ShawnWerk | done. |
| ShawnWerk | its NODEV but thats a lie =) |
| ShawnWerk | pci_dev should have reported 700c |
| ShawnWerk | but the southbridge: |
| ShawnWerk | pci_match_bus(): DECISION: ids->vendor 1022 == pci_dev->vendor 1022? |
| ShawnWerk | pci_match_bus(): DECISION ids->device 7443 == pci_dev->device 7443? |
| ShawnWerk | bus_match(): We matched! |
| ShawnWerk | amd76x_pm: Initializing southbridge Advanced Micro Devic AMD-768 [Opus] ACPI |
| ShawnWerk | bus_match(): Success! Attach it! |
| ShawnWerk | bound device '00:07.3' to driver 'amd76x_pm-sb' |
| ShawnWerk | sysfs_create_link: depth = 4, size = 33 |
| ShawnWerk | sysfs_create_link: path ='../../../../devices/pci0/00:07.3' |
| ShawnWerk | bus_match(): Returns 0 |
| ShawnWerk | driver_attach():bus_match() returns 1 |
| ShawnWerk | so im confused =) |
| mcp | rml: interactive_delta=2 and prio_bonus_ratio=10 are good |
| cdub_ | ShawnWerk: i think this level of debugging might best be done offline |
| ShawnWerk | it was done :) |
| ShawnWerk | but nobody replied in lkml ;/ |
| ShawnWerk | after 2 weeks |
| mas | andre: I believe that also means that Linux would have to lose awful lot of functionality to become a trusted party of RIAA/MPAA and MSFT and other big names, such as cdrom (or dvd) raw access and so on |
| ShawnWerk | and i have no clue to why the kernel is not seeing 700c |
| mcp | hahahaha |
| @davej | ShawnWerk: wait until its merged, and then hit bugzilla 8) |
| ShawnWerk | davej ;) good idea |
| ShawnWerk | then i can see why it didnt work for me (compare my code vs) |
| ShawnWerk | i've given up for now :) |
| ShawnWerk | I will keep trying out the code in each new 2.5 kernel release, to see if somehow something fixed it :) but im sure this code is ok. |
| ShawnWerk | </ot> ;) |
| andre | mas: dirty secret, the newst Mt-Rainier drives have a hidden raw mode :-) |
| mas | andre: w00t :-) |
| ShawnWerk | ata: are you working on the 2.5 ide? or only 2.4? |
| Lovechild | rik is so silent, I was looking forward to hearing about rmap in 2.5... damn I didn't miss it did I? |
| ShawnWerk | Lovechild: no :) |
| @sarnold | Lovechild: rik is currently consulting for a client, so his time isn't really his at the moment.. :-/ |
| Lovechild | sarnold: that's what I get for sleeping... stupid me :) |
| @sarnold | Lovechild: you didn't miss him.. riel missed us. :-/ there is a small chance he might be able to participate later on. |
| Lovechild | sarnold: or we might be able to convince him to give a small extra lecture on rmap as a christmas present :) |
| Lovechild | sarnold: at his own time of course |
| ShawnWerk | hmm |
| andre | ShawnWerk: both but now I send to Axboe and Cox as Torvalds and I are still mending the past |
| ShawnWerk | andre: good to here :-) |
| ShawnWerk | wouldn't want to loose you, I respect your work and appreciate it. |
| Phantom | Why isn't IPSec included in the Linux kernel yet (or is it inthe latest 2.5)? |
| mcp | Lovechild: Joachim did rmap15b for wolk4. And his impressions are that w/ rmap15b it is feelable slower than w/o. As I expected it. |
| mcp | Phantom: it is in 2.5 |
| Lovechild | mcp: so wolk4-realpre tonighT?? |
| Phantom | ups. That's great! |
| Phantom | thanks |
| mcp | Lovechild: no because that thing does not work for _me_. |
| ShawnWerk | er hear |
| mcp | Lovechild: I am thinking about to drop rmap again. |
| Lovechild | mcp: arrggh.. because wolk4-up3 is awefully choppy for me |
| ShawnWerk | do any of you need some testers? |
| mcp | Lovechild: system does to boot up |
| mcp | Lovechild: please what? |
| ShawnWerk | i've got some spare CPU cycles to give :) |
| ShawnWerk | two boxes |
| ShawnWerk | w/ serial console |
| ShawnWerk | being a kernel newbie, i'd like to help out as much as I can while learning in the process |
| Lovechild | mcp: it seems that 2.4.18 which with wolk wasn't so laggy in X, in 2.4.19/20 laggyness was really bad and the read latency thing only made it worse |
| mcp | Lovechild: therefore there is a low latency elevator think in it you know? |
| will-h | Is the ipsec support going to support tunnel interfaces? I've been using Tobias Ringstrom's ipsec_tunnel patch on 2.4 with a lot of success. |
| mcp | Lovechild: rl2 does not make things more worse, it speeds up throughput. For desktop while heavy i/o 2.4.19 and 2.4.20 etc. are not usable w/o the lowlat elevator. |
| Lovechild | mcp: I have it enabled... and by god it's lag hell... it's worse and q3 in 28.8 baud under heavy CPU load + disk IO |
| mcp | Lovechild: ok while you are complaining here, why you don't say anything on the mailinglist? Should I guess? should I read your mind? |
| Lovechild | rml: in theory who would we have to bribe to get it :) |
| * fork:#qc don't thinks he wont a better oops | |
| mcp | Lovechild: if you don't let us know you live with that asis. |
| fork | it good enough :) |
| @sarnold | fork :) |
| Lovechild | mcp: no no... I've been writting my semester project so I've had little time to test this... after I've tinker with the settings a bit I'll post to wolk.dev |
| arashi | how's async i/o going? i noticed it's in "before 2.6" status on Guillaume's list ... |
| ShawnWerk | I think we need a kernel developer appreciation day =) for all the hard/good work you people do. |
| ShawnWerk | and the help you give :) |
| mcp | Lovechild: also I am pretty sure you cannot game quake3 while doing heavy i/o and cpu load in a reasonable speed with _any_ kernel on a usual desktop mashine. |
| mcp | Lovechild: I can game quake3 while doing heavy disk i/o and copying gigabytes of data w/o any slowdown with the lowlat elevator hack. |
| Lovechild | mcp: hehe, I mean X is very laggy (compared to wolk3.8) when I compile stuff... |
| @mulix | what do you guys think about the variety of kernel trees we have lately? too many, not enough? |
| Lovechild | mulix: can we ever have to many kernel trees? |
| ShawnWerk | mulix: one more is coming ;) |
| arashi | ah, the time-honored viro stink test ^_^ |
| mcp | Lovechild: you may want to play with the scheduler knobs. |
| ShawnWerk | but not until 2.6 is released |
| @mulix | lovechild, of course, if they are not contributing to World Domination |
| ShawnWerk | i might as well ask this since some core people are here |
| Lovechild | mcp: after the lecture I'll tinker with it... |
| ShawnWerk | I'm planning with another person on creating a 2.6 to 2.7 kernel patch tree |
| @mulix | s/not contributing/hurting/ |
| ShawnWerk | collecting all good and some experimental patches for upcoming 2.7ish kernel so we can get a head-start when 2.6 stablizes |
| ShawnWerk | you think this is a good idea? |
| will-h | What are the changes with md and LVM in 2.5? Does anyone want to summarise? |
| @MJesus | please more slowsly (for traslators) |
| @MJesus | please more slowsly (for traslators) |
| @MJesus | please more slowsly (for traslators) |
| ShawnWerk | if we can get some experimental patches working well with a 2.6 tree, then when time comes for 2.7 branch things from the 2.6 experimental tree could go into 2.7 with Linus/core developers blessings. |
| ShawnWerk | does this sound good? |
| @raul | please, take pity of us 0:) |
| @mulix | sarnold, good addition. my answer is "it's a good goal to work for, but reaching it isn't important at all" |
| ShawnWerk | rml: thats what i want this kernel tree (eXtendable Linux Kernel -xlk) to do. |
| ShawnWerk | will do. |
| ShawnWerk | 2.6 experimental meaning -xlk based off of a 2.6 vanilla tree |
| Lovechild | on a more pratical note, is there's any plans to merge Reiser4 soon ? |
| @mulix | what do you guys think about mosix/openmosix? other Linuxkernel based clustering code? |
| @mulix | does it have any reason to be merged into mainline, sometime in the future? |
| ShawnWerk | I want reiser4 in -xlk to get it working for 2.6/2.7 inclusion |
| fscked | do you think the kind of improvements made within the kernel in this last development cycle will add too much complexity as to scare newbie kernel hackers off? Or on the other hand the new APIs make it simpler and more accessible? |
| Lovechild | rml: then again, it's a great feature... I wonder why it's in none of the trees for 2.5 I saw it in -mcp but that hasn't been updated for a long time |
| Mave | which journalled file system from 2.6 is recommended ? |
| * jaytee:#qc whispers..... XFS | |
| * Lovechild:#qc hits jaytee... Reiser4 !!! | |
| jaytee | :-) |
| ShawnWerk | rml: it's not about the changes the scares me, it's the lack of documentation. I understand that everyone is busy making changes to 2.5 but we're in freeze now no? does this include API freezes? |
| Mave | :) |
| RaiST | In the future will be a general fix, correction phase for general fixing and for code optimizing? |
| ShawnWerk | davej: ok, so code freeze isn't happening yet |
| ShawnWerk | davej: thanks :-) |
| fscked | thanks for picking up the question, rml, and thanks for all the replies |
| RaiST | sorry for my question's problem. |
| RaiST | In the future, will be a general code fixing and optimizing phase for the kernel? |
| ShawnWerk | do any of you developers have good tips on doing kernel debugging with 2.5? |
| ShawnWerk | rml: I'm using 2.5.52 for my desktop @ home and it hasn't crashed (excluding the module mess) :-) so judging on the stability im having I see 2.6 coming soon for at least me :-) |
| ShawnWerk | w/ KDE |
| ShawnWerk | so im pushing the kernel to its limits with compatability between things |
| @rml | ShawnWerk: yes, 2.5 is very stable |
| @rml | I recently switched my personal workstation to it |
| ShawnWerk | :-) |
| ShawnWerk | ive been using 2.5 for a month now |
| ShawnWerk | 2.5.26+ |
| ShawnWerk | about that time |
| ShawnWerk | I'd like to see lm_sensors in 2.5. How come it didn't get accepted? It's working great in 2.5.52 right now for me. (with the 2.5.49 patches fixed up) |
| radical | whats the best way for a newbie to test/try out/fiddle with 2.5.x - vmware or something? |
| RaiST | In the future, will be a general code fixing and optimizing phase for the kernel? |
| @sarnold | RaiST: tends to happen in "feature freeze" and "code freeze" stages |
| ShawnWerk | rml: I've had SCSI issues though on my older P233MMX box so no SCSI on the workstation. |
| ShawnWerk | I shall report those tonight on bugzilla |
| RaiST | thanks. |
| ShawnWerk | gregkh: SCSI indeed =) |
| ShawnWerk | then again, I could just be extremely lucky to have 2.5.x working at all ;-) (grins at zwane) |
| @gregkh | ShawnWerk: probably because no one has submitted lm_sensors for inclusion. |
| ShawnWerk | gregkh: interesting |
| @gregkh | will-h: did that make sense? |
| will-h | thanks gregkh |
| will-h | So we will have a number of options for lvm/etc type functionality? |
| ShawnWerk | gregkh: I have seen many changes from lm_sensors (i2c) changes added to 2.5.51/.52 |
| ShawnWerk | since the 2.5.49 patch broke a little due to some of its i2c code added in |
| @gregkh | ShawnWerk: good. |
| @gregkh | will-h: yes we will. |
| will-h | Thanks |
| * arashi:#qc watches everyone say "no, modules are broken" ;/ | |
| ShawnWerk | hehehehe |
| Zadeh | What do lk devs think of intel compiler support? |
| ShawnWerk | module mess ;) |
| ShawnWerk | i really wish they held off on the module maddness |
| arashi | modules are okay if you don't do fancy autoload stuff, but that's just my experience :) |
| ShawnWerk | but module changes came so late |
| arashi | yeah |
| ShawnWerk | had it gone in eariler 2.5 i think we wouldn't have such a problem |
| arashi | but it'd been talked about forever ... ah well |
| ShawnWerk | hmm |
| @gregkh | Zadeh: if the intel compiler can build the kernel, great! |
| ShawnWerk | thanks ata |
| Zadeh | gregkh: think many people will actually use it? |
| Mave | will gcc 3.2 "officailly" supported soon ? |
| radical | has the networking code changed a lot from 2.4.18/19 to 2.5.x? |
| @gregkh | Zadeh: don't know, if the compiler was free, more might :) |
| Mave | heh pasted with my typo :) |
| radical | davej: what do you mean by "poisoning" ? |
| ShawnWerk | gcc 3.3 just branched ;) oh yay |
| * ShawnWerk:#qc waits for the compile errors to pile up | |
| Mave | rml: ok thanks. i use 3.2 too :) |
| Zadeh | gregkh: I've been using it on 2.5 (non commercial ver :)) and really only notice that it compiles faster.. |
| @rml | Mave: now that the major distributors ship it, it is pretty common |
| @rml | it does some good optimizations too |
| Mave | rml: yep |
| @rml | very sloooow though |
| ShawnWerk | we need PCH in gcc :-( |
| Zadeh | yup :/ |
| @gregkh | Zadeh: but does it make better executable? Last I looked in i386 it didn't, but ia64 it did. |
| @rml | our build farms at MontaVista are taking measurably longer, I am told |
| radical | why would you get an oops when freeing a memory with illegal instructions? (ignore if its too stupid!) |
| ShawnWerk | im hoping some people put pressure on the gcc people to merge PCH into 3.4/3.5 |
| ShawnWerk | so we can really speed things up |
| cdub_ | radical, later if it's used |
| Zadeh | gregkh: Not really that I can see.. I suppose intel's primary interest is with IA64 support. |
| c0nan | what is pch? |
| ShawnWerk | precompiled headers |
| c0nan | oh ok |
| ShawnWerk | and other goodies |
| @gregkh | Zadeh: agreed. |
| radical | cdub_: hmm okk.. so its kinda like "bookmarking" that mem location.. |
| @davej | ShawnWerk: ISTR reading on gcc-patches that someone was pushing it for 3.4 |
| ShawnWerk | davej: WooHoo! |
| ShawnWerk | that is great |
| radical | davej: aah okk.. got it. thanx |
| ShawnWerk | I hope to start compiling 2.5.5(3) with gcc 3.3 as a test base |
| ShawnWerk | and fix some compile bugs w/ some patches |
| cdub_ | radical: maybe not bookmarking, but marking it as poisoned ;-) |
| Zadeh | Apple did the PCH stuff and its being fed back into gcc by RH |
| @davej | ShawnWerk: 3.3 (or head cvs as it was when I last tried it) spat out a *lot* of warnings. |
| ShawnWerk | tomorrow :) (that is if .53 is out) |
| radical | cdub_ : ya.. got it! :) |
| @davej | there are fixes for lots of them in -dj, which I really should push to Linus, but theres more important bits.. |
| ShawnWerk | davej: i'll try to fix some of those and submit. |
| ShawnWerk | oh :) |
| ShawnWerk | then i won't :-) |
| @davej | you see the same errors with older gcc and -W |
| radical | cdub_: im working on the networking doc of lkdp(linux kernel doc-n project) and using 2.4.18/19 for it. has the networking code undergone major changes in 2.5.x? |
| ShawnWerk | sarnold: good question :-) I'd like documents :) |
| RaiST | what is the current stage of tux webserver? |
| radical | what exactly is the O(1) scheduler? |
| ShawnWerk | davej: is the new threading in 2.5 or in a secondary tree? |
| ShawnWerk | I know he has patches w/ glibc 2.3.1 and userland libs to use it (?) or am i just muddled up ;) |
| radical | cdub_: but no design/other major changes? normal code cleanups/optimisations.. etc? |
| ShawnWerk | rml: ahh so i need to find out how to enable NPTL w/ glibc 2.3.1 |
| @rml | ./configure --enable-nptl --kernel=2.5.52 or something like that |
| radical | rml: hmm ya thats what i was wondering as to "how" it does it in constant time.. |
| ShawnWerk | ah :) I will do that thanks rml. |
| ShawnWerk | has the DRM code been updated in 2.5.52? I've had serious ATI Radeon lockups with system, syslog reports AGP ART errors |
| radical | cdub_: cool! so shouldnt be much of a problem updating the doc for 2.6, whenever it is out! |
| ShawnWerk | X-cvs loads but either the screen is garbled, wrong resolution or the monitor refuses to sync and i hard lock. If i console /ssh in I have full access to the machine however. |
| radical | rml: guess i'll have to look at the code, is it there only in 2.5.x ? |
| @rml | radical: its in 2.4-ac and some other patches, too |
| ShawnWerk | davej :-) great thanks. |
| radical | cdub_: okk.. cool |
| radical | cool :) |
| ShawnWerk | davej: I will get the error for you and submit a bugzilla. |
| ShawnWerk | i might have submitted that on lkml |
| radical | cdub_: okk.. noted that down! |
| ShawnWerk | drm] Initialized radeon 1.7.0 20020828 on minor 0 |
| ShawnWerk | [drm:radeon_do_init_cp] *ERROR* PCI GART not yet supported for Radeon! |
| ShawnWerk | [drm:radeon_ati_pcigart_cleanup] *ERROR* no scatter/gather memory! |
| ShawnWerk | [drm:radeon_do_cleanup_cp] *ERROR* failed to cleanup PCI GART! |
| ShawnWerk | i dont have a PCI card though ;-) |
| @davej | ah, pci gart. outside my world. |
| @davej | erk |
| ShawnWerk | heh |
| @davej | thats one for the drm folks. |
| @sarnold | leave it to ShawnWerk to come up with nasty error messages... |
| radical | what would you guys recommend as a testbed for a newbie with access to only one machine! umm vmware? (i use uml right now) |
| ShawnWerk | sarnold ;-) |
| c0nan | uml is good |
| ShawnWerk | davej: I will submit a newer report with 2.5.52 tonight to the drm people. |
| ShawnWerk | sarnold: I like to scatter myself everywhere :) |
| Mave | uhm how about bochs ? :) |
| ShawnWerk | ah |
| radical | havent tried bochs.. |
| ShawnWerk | with the new kbuild qt frontend, it seems broken :( |
| jmgv | uml is a bad name. That remenber Unified Modeling Language :-) |
| c0nan | yeah jmgv i agree |
| c0nan | :) |
| ShawnWerk | menuconfig works fine though |
| radical | sarnold: hmm ok.. |
| ShawnWerk | cdub: I really hope we get some sort of kernel dumping, even though i have serial now it would be good for other people w/o serial and such |
| cdub_ | heh, just a joke. sarnold said philosophical... ;-) |
| ShawnWerk | ;-) |
| ShawnWerk | I do think it's a good thing to share between the Linux and BSD people |
| ShawnWerk | we both benefit |
| c0nan | hmm any plans on role based access control mechanisms? |
| ShawnWerk | how will We address so-called 'Patent' concerns in current and future kernel development? |
| c0nan | What is the roadmap on that generic access control mech.? |
| nfms | Linux is now taking advantage of the faster syscall stuff on P4 chips (SYSNTER et al). Are there any other OS's out there already taking advantage of these features? |
| radical | in include/asm-ia64/unistd.h, the syscall numbers start from 1024.. why is that? on i386 ,sparc etc this is not the case.. |
| c0nan | beer break in #linux? :)) |
| Zadeh | Q: Are there any major parts that seem to have 'stabilized' in 2.5.x to a point where their designs are pretty much going to stay relatively the same for quite a long time? |
| radical | sarnold: where can i find links to the logs of the other umeet talks? |
| ShawnWerk | thanks for your time kernel maintainer/developers, time to go :-) |
| @sarnold | radical: http://umeet.uninet.edu/umeet2002/english/des.eng.html |
| radical | sarnold: thanx |
| c0nan | who's idea was this umeet. That's my first time, and it is GREAT! thank you all! |
| RaiST | Last days in kml, there was a topic about kernel bug database with bugzilla. what do you think of this? |
| debissued | this is my first time as well ;) |
| mulix | rml, did you read "Unix Internals", by Uresh Vahalia (sp)? |
| mulix | I got it at work on Thursday, but didn't read it yet. |
| mulix | cdub_, I've got most of the Stevens books :-) |
| cdub_ | ah, good ;-) |
| RaiST | For a newbie, is it good for starting to work with latest kernel or starting with historical ones? |
| mulix | did any of you read UTLK, 2nd edition? (not out yet) |
| RaiST | thanx |
| radical | but how does a newbie find a "problem" for which he needs to change the code?? |
| mulix | sarnold, YES! :-) |
| RaiST | radical: same thing for me but i decided to find simple but very extreme things. |
| c0nan | sarnold: good philosophic question last one :) |
| radical | RaiST: like what? |
| c0nan | i think if you got paid for something that you love to do is the best one right? |
| RaiST | for example you can start working with tux webserver thing |
| mulix | radical, use it daily, find something that doesn't work or that you think could work better, or just do it. |
| RaiST | it will help you understand kernel structure and topic is very clear to a userspace programmer |
| mulix | if everything works perfectly, just reimplement something from scratch :-) |
| RaiST | am i wrong? |
| radical | hmm i guess... cool |
| RaiST | clap clap clap :) |
| radical | thank you all for this talk!!! :) |
| RaiST | i know that there will be a log for #linux. |
| RaiST | will be one for #qc? |
| fernand0 | yes raist |
| RaiST | ok |
| fernand0 | always is being logged |
| fernand0 | ops |
| fernand0 | sorry if i looked too much as the FBI |
| fernand0 | 0:) |
| radical | RaiST: there is.. http://umeet.uninet.edu/umeet2002/english/des.eng.html, i guess it'll be updated soon for todays talk |
| c0nan | fernand0 :) |
| RaiST | :) |
| fernand0 | /always/all/ |
| RaiST | <fernand0> always is being logged |
| RaiST | <fernand0> ops |
| RaiST | <fernand0> sorry if i looked too much as the FBI |
| RaiST | :) |