| View previous topic :: View next topic |
| Author |
Message |
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
 Posted: Sun Jun 26, 2005 4:49 pm Post subject: 2.6.12-nitro1 [__DEPRECATED__] |
 |
|
2.6.12-nitro1 "Broken Pipe"
READ THIS FIRST
Nitro-sources is an experimental kernel patchset which is NOT supported by the Gentoo devs. If you have problems with something, see if the problem persists with a kernel supported by Gentoo. If you still have the same problem with a kernel supported by Gentoo (like vanilla-sources), then file a bug report at h**p://bugs.gentoo.org . If you only had the problem with 2.6.12-nitro1, and not with a vanilla kernel, then bug us in this thread.
Best Kernel ive had since quite a long time.
DO NOT POST ABOUT GCC4 PROBLEMS, ATI/NVIDIA MODULES OR REISER4, UNLESS YOU
WANT OTHER USERS TO HELP YOU, BECAUSE I WONT, I ONLY CAN HELP WITH IN-KERNEL __AND__
NOT-BROKEN-BY-DESIGN STUFF
This kernel Patchset is unsupported on 64BIT Architectures, we can only try to help with compile-errors
on x86_64 and thelike
| Patchlist wrote: |
n0-patch-2.6.12-git4.diff
/* Newest git tree as of this writing */
n1-sk98lin_v8.16.2.3.diff
/* latest SK98LIN driver from SysKonnect */
n2-patch-2.6.12-ck2.diff
/* Best desktop interactivity ever */
n3-gregkh-all-2.6.12-nitro.diff
/* All patches from GREG as of post-git4 */
n4-Mortonize_nitro-aka-mm_patches.diff
/* patches from MM-tree, summary below this patchlist */
n5-reiser4_nitro.diff
/* all-in-1 reiser4 patch, seems newer than in mm1 */
n6-hfsplus_force_rw.diff
/* mount hfsplus partitions with force_rw option to force read-write mount,
* DANGEROUS! */
n7-i386-emu-rdtsc.diff
/* emulation of RDTSC instructionset on archs older than i686, select the arch
* and a menuconfig option appears */
n8-swsusp-2.1.9-nitro_diff.diff
/* SWSUSP2, __might be broken__, untested !!! */
n9-usbhid-readd-kconfig-r2.diff
/* chmp- through Kconfig */
n10-config-hz-2.6.12.diff
/* change internal kernel freq for x86, x86_64 and ppc */
n11-squashfs2.1-r2.diff
/* compact readonly filesystem */
n12-compile_fixes_1.diff
/* nitro-specific compile-fixes part1 */
n13-vesafb-tng-0.9-rc6-r2.diff
n14-fbsplash-0.9.2-r3.diff
/* Beautiful framebuffer AKA eyecandy */
n15-slabtune.diff
/* slab-tuning from M.BLIGH */
n16-it821x.diff
/* Newest IT821X controller driver */
n17-inotify-0.23.diff
/* inotify from robert love, 2.6.12 official */
n18-lufs2.diff
/* lufs from the pretty old cvs tree, although newer than the one floating
* around */
n19-daconfig-2.2.0-FROM_NO.diff
/* DaMouse's pretty Kconfig, props to NO-sources */
n20-cflags-selection-FROM_NO.diff
/* oh-so-ricy CFLAGS-selection, props to NO-sources */
n21-config-nr-tty-devices-FROM_NO.diff
/* configure the number of tty* entries in /dev, props to NO-sources */
n22-Win4Lin3-23_3_05-nitro.diff
/* NeTraverse'S Win4Lin, newest version, you need at least Win4Lin5 userspace
* subsystem */
n23-nitro-fb_logo.diff
/* the comeback of the nitro-logo for framebuffer */
n24-2.6.12.1-update.diff
/* updated to 2.6.12.1 */
n25-nitro_release.diff
/* TAG */
n26-compile_fixes_2.diff
/* nitro-specific compile-fixes part2 */
n27-adjust_nr_running.diff
/* slight change of nr_running calculation from Con, originally for mm */
n28-compile_fixes_3.diff
/* nitro-specific compile-fixes part3 */
|
This kernel contains the ipw2100 and ipw2200 patches!!
| MM patches wrote: |
add-check-to-proc-devices-read-routines.patch
add-do_proc_doulonglongvec_minmax-to-sysctl-functions.patch
agp-AGP-01-printk-levels.patch
as-limit-queue-depth-fix.patch
as-limit-queue-depth.patch
avoid-recursive-oopses.patch
char-tty_io-replace-schedule_timeout-with-msleep_interruptible.patch
con_consdev-bit-not-set-correctly-on-last-console.patch
drivers-net-wireless-ipw2100-use-the-dma_32bit_mask-constant.patch
drivers-net-wireless-ipw2200-use-the-dma_32bit_mask-constant.patch
export-file_ra_state_init-again.patch
ext3-reduce-allocate-with-reservation-lock-latencies-tidy.patch
ext3-reduce-allocate-with-reservation-lock-latencies.patch
fix-of-dcache-race-leading-to-busy-inodes-on-umount-fix.patch
fix-of-dcache-race-leading-to-busy-inodes-on-umount-tidy.patch
fix-of-dcache-race-leading-to-busy-inodes-on-umount.patch
fix-recursive-ipw2200-dependencies.patch
git-alsa-fix.patch
git-alsa-usbaudio-fix.patch
git-alsa.patch
git-netdev-chelsio.patch
git-netdev-janitor.patch
git-netdev-register-netdev.patch
git-netdev-smc91x-eeprom.patch
git-netdev-we18-ieee80211-wifi.patch
git-ntfs.patch
i386-never-block-forced-sigsegv.patch
improve-cd-dvd-packet-driver-write-performance.patch
ipw2100-assume-recent-kernel.patch
ipw2100-cleanup-debug-prints.patch
ipw2100-kill-dead-macros.patch
ipw2100-old-gcc-fix.patch
ipw2100-small-cleanups.patch
irqpoll.patch
mempool-only-init-waitqueue-in-slow-path.patch
no-acpi-build-fix.patch
o1-sb-list-traversing-on-syncs.patch
optimize-sys_times-for-a-single-thread-process-update-2.patch
optimize-sys_times-for-a-single-thread-process-update.patch
optimize-sys_times-for-a-single-thread-process.patch
preempt_count-is-int-remove-cast-and-dont-assign-to.patch
reiserfs-add-checking-of-journal_begin-return-value.patch
tty-output-lossage-fix.patch
vfs-bugfix-two-read_inode-calles-without.patch
x86-local-apic-fix.patch
x86_64-never-block-forced-sigsegv.patch
AND a FEW MORE.....
|
Special Notes:
Win4Lin users: you must enable networking support and sysctl support or unresolved symbols appear,
you must disable 1Gig Lowmem support, or it will not work even though it compiles.....
Reiser4 users: Reiser4 now depends on __selected__ CRYPTO_DEFLATE and __NOT__
selected 4K stacks.
SWSUSP2: Untested, need reports <UPDATE: seems to work with Swapwriter and lzf compression>
FBSplash + Vesafb-tng : also untested, try and tell us too.... <UPDATE: seems to work as well>
BrokenOut: Its rediffed, so these are the patches you can cleanly apply in the right order and thus get this nitro release,
I hope the compile fixes can be of any help for other maintainers, as it partially was a hard time to find out
what's wrong.... 
OK, now the LINKS:
Ebuild
Patch
Relnotes
Broken out (tar-bz2-ipped)
| Gentree wrote: |
Gentoo + Nitro = Gazzzzz!
|
And as usual:
HAVE FUN GODDAMMIT!!!!
_________________
Retired gentoo user
Last edited by Tiger683 on Thu Jun 30, 2005 4:37 pm; edited 5 times in total |
|
| Back to top |
|
 |
ejs76
Tux's lil' helper
Joined: 07 Feb 2004
Posts: 148
|
| Posted: Sun Jun 26, 2005 5:00 pm Post subject: |
|
|
THANKS MAN!!!  . I'm compiling it right now. And "having fun" I shall... |
|
| Back to top |
|
|
Jazz
Guru
Joined: 16 Nov 2003
Posts: 543
Location: Melbourne, Australia
|
| Posted: Sun Jun 26, 2005 5:20 pm Post subject: |
|
|
OMFG ! win4lin finally !!, awesome........
It was all worth a wait in da end
Will report soon !
BYe,
Jazz
_________________
In 2010, M$ Windows will be a quantum processing emulation layer for a 128-bit mod of a 64-bit hack of a 32-bit patch to a 16-bit GUI for an 8-bit operating system written for a 4-bit processor from a 2-bit company that can't stand 1 bit of competition. |
|
| Back to top |
|
|
SchAmane
Tux's lil' helper
Joined: 11 Sep 2004
Posts: 107
|
| Posted: Sun Jun 26, 2005 5:21 pm Post subject: |
|
|
| Going to try it |
|
| Back to top |
|
|
tagwar
Tux's lil' helper
Joined: 11 Aug 2004
Posts: 147
Location: Karlsruhe
|
| Posted: Sun Jun 26, 2005 8:11 pm Post subject: great!! |
|
|
fbsplash + vesafb-tng are working just nicely.
I'm going to try swsusp2 now and report back.
[EDIT]: swsusp2 works fine using the swapwriter and lzf compression here.
Thanks for another great patchset!! |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Sun Jun 26, 2005 8:19 pm Post subject: |
|
|
Thanks man, every feedback is welcome 
_________________
Retired gentoo user |
|
| Back to top |
|
|
priestjim
Tux's lil' helper
Joined: 09 Jan 2005
Posts: 100
Location: Never never land
|
| Posted: Sun Jun 26, 2005 8:57 pm Post subject: |
|
|
Very good patchset! Nice work really. Been waiting for it for a long time...
_________________
My soul is my antenna, I am the instrument + the guitar is my amplifier.
-- Ottmar Liebert |
|
| Back to top |
|
|
SchAmane
Tux's lil' helper
Joined: 11 Sep 2004
Posts: 107
|
| Posted: Sun Jun 26, 2005 9:04 pm Post subject: |
|
|
Feedback:
Win4Lin is selectable even if 1Gblowmem is selected.
For reiser4 you need CRYPTO_DEFLATE option to set =y and not zlib... |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Sun Jun 26, 2005 9:11 pm Post subject: |
|
|
Selectable is not the problem, and i turned the compile error into the compile warning in the code,
but the problem is that it __WILL_NOT__WORK__, you have been warned....
anyways, thanks for great feedback y'all !!!!!
cheers,
T
_________________
Retired gentoo user |
|
| Back to top |
|
|
Zephyrus
Apprentice
Joined: 01 Sep 2004
Posts: 204
|
| Posted: Sun Jun 26, 2005 10:22 pm Post subject: |
|
|
Thanks for the new release but on my Amd64 system i get some errors. 
It seems that the kernel has some problems during the initialization of the PCI subsystem:
1]the nvidia ethernet driver reports an error in "pci_enable_dev_failed"
2]the nvidia ide driver reports that "the device is unavailable due to resource conflicts"
Maybe there are other errors but i don't have logs and this is all i can see.
The boot process ends with a panic due to the fact that the kernel cannot open the root device (most probably because of the previous errors). |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Sun Jun 26, 2005 10:31 pm Post subject: |
|
|
Sorry, thats something i cannot check before a release, 32-bit here.
_________________
Retired gentoo user |
|
| Back to top |
|
|
SchAmane
Tux's lil' helper
Joined: 11 Sep 2004
Posts: 107
|
| Posted: Sun Jun 26, 2005 10:38 pm Post subject: |
|
|
3 Hours playing ut2004 on new nitro and all works fine
Best kernel ever. Latency of whole system is very high. |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Sun Jun 26, 2005 10:45 pm Post subject: |
|
|
You mean latency is very low, dont you?
high latency = bad.
hmm, apparently we have a good 32 bit release here.
i need more AMD64 feedback though before i make the second release, probably only updates and bugfixing.
_________________
Retired gentoo user |
|
| Back to top |
|
|
Zephyrus
Apprentice
Joined: 01 Sep 2004
Posts: 204
|
| Posted: Sun Jun 26, 2005 11:10 pm Post subject: |
|
|
| Tiger683 wrote: |
[...]
i need more AMD64 feedback though before i make the second release, probably only updates and bugfixing. |
I get exactly the same errors with 2.6.12-mm1 therefore this is not a "nitro-only" kernel behaviour.
I checked lkml.org but it seems that no-one has this kind of problem.
I don't know...maybe it is something strictly linked to my system or maybe i'm doing something wrong with the kernel configuration |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Sun Jun 26, 2005 11:25 pm Post subject: |
|
|
| Quote: |
I get exactly the same errors with 2.6.12-mm1 therefore this is not a "nitro-only" kernel behaviour.
I checked lkml.org but it seems that no-one has this kind of problem.
I don't know...maybe it is something strictly linked to my system or maybe i'm doing something wrong with the kernel configuration
|
Thanks! Its important to know you also have this problem with mm-sources.
Thanks for your effort,
Do you by any chance use ext3 ?
[EDIT]
What chipset do you have on your MoBo?
[/EDIT]
[EDIT2]
Zephyrus: I can not find any mm-related patch in nitro that could lead to such behaviour... if you could verify occurence of
the bug with the syme .config in vanilly kernel, then you should post a bug to lkml.org or even better first try changing your .config (probably
add something rather than remove)
[/EDIT]
_________________
Retired gentoo user |
|
| Back to top |
|
|
SchAmane
Tux's lil' helper
Joined: 11 Sep 2004
Posts: 107
|
| Posted: Mon Jun 27, 2005 12:25 am Post subject: |
|
|
| Tiger683 wrote: | | You mean latency is very low, dont you? |
Ops, yes, its low. Just love 
Mash racie  |
|
| Back to top |
|
|
lavacano
Apprentice
Joined: 29 May 2005
Posts: 190
Location: Poulsbo, WA
|
| Posted: Mon Jun 27, 2005 12:28 am Post subject: |
|
|
make modules_install
WARNING: /lib/modules/2.6.12-nitro1/alsa-driver/acore/snd.ko needs unknown symbol devfs_mk_dir
WARNING: /lib/modules/2.6.12-nitro1/alsa-driver/acore/snd.ko needs unknown symbol devfs_mk_cdev
WARNING: /lib/modules/2.6.12-nitro1/alsa-driver/acore/snd.ko needs unknown symbol devfs_remove
dmesg
[ 1693.527579] snd: Unknown symbol devfs_mk_dir
[ 1693.527614] snd: Unknown symbol devfs_mk_cdev
[ 1693.527907] snd: Unknown symbol devfs_remove
emerge alsa-driver
/var/tmp/portage/alsa-driver-1.0.9b/work/alsa-driver-1.0.9b/acore/sound.c: In function `snd_register_device':
/var/tmp/portage/alsa-driver-1.0.9b/work/alsa-driver-1.0.9b/acore/sound.c:241: warning: implicit declaration of function `devfs_mk_cdev'
/var/tmp/portage/alsa-driver-1.0.9b/work/alsa-driver-1.0.9b/acore/sound.c: In function `snd_unregister_device':
/var/tmp/portage/alsa-driver-1.0.9b/work/alsa-driver-1.0.9b/acore/sound.c:279: warning: implicit declaration of function `devfs_remove'
/var/tmp/portage/alsa-driver-1.0.9b/work/alsa-driver-1.0.9b/acore/sound.c: In function `alsa_sound_init':
/var/tmp/portage/alsa-driver-1.0.9b/work/alsa-driver-1.0.9b/acore/sound.c:361: warning: implicit declaration of function `devfs_mk_dir'
_________________
Sincerely,
Chadwick Ferguson |
|
| Back to top |
|
|
Legoguy
Apprentice
Joined: 22 Dec 2003
Posts: 166
Location: Edmonton, Alberta, Canada
|
| Posted: Mon Jun 27, 2005 5:57 am Post subject: |
|
|
That looks like something to do with devfs... you should have that disabled in the kernel by now and use udev 
[edit] nevermind, happens here too... we have a problem here..
with LIRC too:
| Code: | /var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:54:35: linux/devfs_fs_kernel.h: No such file or directory
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c: In function `cleanup':
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:132: warning: implicit declaration of function `devfs_remove'
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:134: warning: implicit declaration of function `class_simple_device_remove'
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c: In function `lirc_register_plugin':
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:374: warning: implicit declaration of function `devfs_mk_cdev'
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:378: warning: implicit declaration of function `class_simple_device_add'
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c: In function `lirc_dev_init':
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:847: warning: implicit declaration of function `class_simple_create'
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:847: warning: assignment makes pointer from integer without a cast
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c: In function `cleanup_module':
/var/tmp/portage/lirc-0.7.1/work/lirc-0.7.1/drivers/lirc_dev/lirc_dev.c:887: warning: implicit declaration of function `class_simple_destroy' |
looks like a missing header file
I'm going to try copying the file (include/linux/devfs_fs_kernel.h) from vanilla and see if that works...
[edit2] Yep, that works... just "cp /usr/src/linux-2.6.12-gentoo-r1/include/linux/devfs* /usr/src/linux-2.6.12-nitro1/include/linux" and it solves the problems. |
|
| Back to top |
|
|
b3cks
Veteran
Joined: 23 Mar 2004
Posts: 1481
Location: Bremen (GER)
|
| Posted: Mon Jun 27, 2005 6:38 am Post subject: |
|
|
W00t... Yeees, I can't wait until todays eve to test this kernel!
Anyone testet ATI 3D/drm yet?
@Tiger: Thx a lot!
If you want another mirror for the patcheset, tell me.
_________________
I am /root and if you see me laughing you better have a backup. |
|
| Back to top |
|
|
lavacano
Apprentice
Joined: 29 May 2005
Posts: 190
Location: Poulsbo, WA
|
| Posted: Mon Jun 27, 2005 7:54 am Post subject: |
|
|
yeah devfs was disabled long ago in favor of udev here. I just used the .config from nitro-2.6.12-rc4 so its b0rked migrated back to rc4. Oh I was also wondering : any posibility of kexec in the future? thanks, Chadwick
_________________
Sincerely,
Chadwick Ferguson |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Mon Jun 27, 2005 8:11 am Post subject: |
|
|
Guys: welcome to reality !
DEVFS is being removed beginning with 2.6.12+ kernels, so blame the kernel devs, or the driver devs
for not keeping up with kernel development.....
in short: i will not reintroduce __deprecated__ and __removed__ stuff back into the kernel...
PS: lavacano: why in the hell do you emerge ALSA????? its in the kernel....
cheers
T
_________________
Retired gentoo user |
|
| Back to top |
|
|
andrew_wiggin
n00b
Joined: 21 Apr 2005
Posts: 59
Location: Israel
|
| Posted: Mon Jun 27, 2005 8:30 am Post subject: |
|
|
| looks great compiling now |
|
| Back to top |
|
|
Thetargos
Apprentice
Joined: 28 May 2003
Posts: 155
Location: Mexico City
|
| Posted: Mon Jun 27, 2005 8:58 am Post subject: |
|
|
| Same here. ALSA even from the official page tarball, gives a warning, which I believe is the culprit for it giving unresolved symbols upon driver loading. I have checked and there is NO option whatsoever of DevFS, however there are other pseudo filesystem options (like proc, pts, shm, etc), however when building ALSA it complains about devfs... I've cat'ed the .config file, and there doesn't seem to be a configuration option for CONFIG_DEVFS_FS, not "pounded out" nor present at all. Why, then is ALSA complaining about it, and possibly compiling with support for devfs, when there is indeed no support whatsoever for it compiled in the kernel either module of built-in? (unless, of course it would be silently being compiled as built-in, but then again, there would not be a complaint from ALSA about it, now would it?). |
|
| Back to top |
|
|
Thetargos
Apprentice
Joined: 28 May 2003
Posts: 155
Location: Mexico City
|
| Posted: Mon Jun 27, 2005 8:59 am Post subject: |
|
|
| Tiger683 wrote: | Guys: welcome to reality !
DEVFS is being removed beginning with 2.6.12+ kernels, so blame the kernel devs, or the driver devs
for not keeping up with kernel development.....
in short: i will not reintroduce __deprecated__ and __removed__ stuff back into the kernel...
PS: lavacano: why in the hell do you emerge ALSA????? its in the kernel....
cheers
T |
Well, I usually install ALSA CVS or their latest stable release, whch not always is included in either vanilla or some patches, unless nitro1 includes 1.0.9b, or some CVS stuff in it, I would compile ALSA externally.
EDIT: I'm rebuilding with ALSA support, just for the heck of it. |
|
| Back to top |
|
|
Tiger683
Veteran
Joined: 08 Jan 2005
Posts: 1347
Location: Heffner's House
|
| Posted: Mon Jun 27, 2005 9:07 am Post subject: |
|
|
It has the git snapshot from mm, which should be pretty recent,
as i said, compile it in the kernel tree, and preferably as builtin....
cheers
T
PS: All user-comments will be included in the next release, so for now it would be:
1) the two missing devfs headers, just for compatibility sake
2) LIRC-CVS
3) I will look around for the most recent alsa snapshot
4) Genetic anticipatory
5) mmap fragmentation patches from mm
To Be Continued....
Anyone interested in svgalib-helper included in the kernel via Kconfig ?
expect nitro2 after release of -ck3
_________________
Retired gentoo user
Last edited by Tiger683 on Mon Jun 27, 2005 9:22 am; edited 2 times in total |
|
| Back to top |
|
|
|
Unsupported Software
