Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
2.6.35.7-TOI-plus kernel patch & patchset
View unanswered posts
View posts from last 24 hours

 
Reply to topic    Gentoo Forums Forum Index Unsupported Software
View previous topic :: View next topic  
Author Message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Thu Sep 30, 2010 10:46 pm    Post subject: 2.6.35.7-TOI-plus kernel patch & patchset Reply with quote

Hi guys,

since I had some trouble with zen-sources getting tuxonice (hibernate) to work and wanted a kernel pretty close to vanilla with the possibility to add grsecurity with speed and stability in mind:

here is the result:

2.6.35.7_TOI-3.2-rc2_plus (this time I intentionally didn't tried to come up with an original name)

(this patchset applies to 2.6.35.7)

with the option to add grsecurity for 2.6.35.6 on top of it (it applies cleanly after reverting the fix from 2.6.35.6 to 2.6.35.7)

I only shortly tested grsecurity but it should work fine

0 core patchset (0_TOI_hardened-fix_custom-flags):

1_tuxonice-3.2-rc2-for-2.6.35.patch.bz2
2_Hardened_kernel_check_SSP_fix.patch
3_custom-cflags.patch

1 mm improvements (1_mm_vm):
0_PATCH\ 0_vmstat-update-zone-stat-threshold-when-onlining-a-cpu.patch
3_[PATCH]\ vmscan:\ prevent\ background\ aging\ of\ anon\ page\ in\ no\ swap\ system.txt

2 ext4 and general cfq improvements (2_fsync_cfq):

0_[PATCH]\ cfq:\ improve\ fsync\ performance\ for\ small\ files.txt
1_Re:\ [RFC]\ ext4:\ Don't\ send\ extra\ barrier\ during\ fsync\ if\ there\ are\ no\ dirty\ pages.txt
2_Re:\ [RFC]\ ext4:\ Don't\ send\ extra\ barrier\ during\ fsync\ if\ there\ are\ no\ dirty\ pages.txt

3 & 4 folders are left out due to trouble-making of those patches ;)

5 mm memcopy - using the memcopy from glibc, this speeds up btrfs and other filesystems (5_mm_memcopy):

[PATCH\ 1_3]\ lib:\ introduce\ some\ memory\ copy\ macros\ and\ functions.txt
[PATCH\ 2_3]\ lib:\ improve\ the\ performance\ of\ memcpy\ and\ memmove\ of\ the\ general\ version.txt

6 nofail variants of kmalloc kcalloc and kzalloc (6_kmalloc kcalloc kzalloc_nofail):

[patch\ v2\ 1_5]\ mm:\ add\ nofail\ variants\ of\ kmalloc\ kcalloc\ and\ kzalloc.txt
[patch\ v2\ 2_5]\ mm:\ add\ nofail\ variant\ of\ kmem_cache_zalloc.txt
[patch\ v2\ 3_5]\ fs:\ add\ nofail\ variant\ of\ alloc_buffer_head.txt
[patch\ v2\ 4_5]\ btrfs:\ add\ nofail\ variant\ of\ set_extent_dirty.txt

7-9 are left out due to same reasons like 3 & 4 ;)

10 fixes for memove - there's an alternative version in the subfolder which failed which rejects - feel free to revert the applied one and try out the other ... (10_mm_memmove):

[PATCH\ V3]\ [X86]\ Fix\ potential\ issue\ on\ memmove.txt

11 is left out due to same reasons like 3 & 4, ... ;)

12 remove annoying hang check error-message, especially useful for use with reiserfs & zfs or other filesystems suffering from slow transfers sometimes (12_hang_check long i_o):

[PATCH]\ block:\ Prevent\ hang_check\ firing\ during\ long\ I_O.txt

13 ck-patchset and its parts which are not dependent on BFS (13_ck-patchset):

01_mm-make_swappiness_really_mean_it.patch
02_mm-enable_swaptoken_only_when_swap_full.patch
03_mm-drop_swap_cache_aggressively.patch
04_mm-kswapd_inherit_prio-1.patch
05_mm-idleprio_prio-1.patch
06_mm-lru_cache_add_lru_tail.patch

14 fix some scheduler-problems after resuming from suspend-to-ram / hibernate (14_corei7_resume):

[patch\ 1_3]\ sched:\ init\ rt_avg\ stat\ whenever\ rq\ comes\ online.txt
[patch\ 2_3]\ sched:\ fix\ minimum\ power\ returned\ by\ update_cpu_power.txt

(patch 3 is in 17_power\ consumption_sched)

15 make the scheduler more immune against CPU hogs (15_[PATCH 0_3][RFC] Improve load balancing when tasks have large weight differential):

[PATCH\ 1_3]\ sched:\ set\ group_imb\ only\ a\ task\ can\ be\ pulled\ from\ the\ busiest\ cpu.txt
[PATCH\ 2_3]\ sched:\ drop\ group_capacity\ to\ 1\ only\ if\ remote\ group\ has\ no\ running\ tasks.txt
[PATCH\ 3_3]\ sched:\ do\ not\ consider\ SCHED_IDLE\ tasks\ to\ be\ cache\ hot.txt

16 vmscan - this one should fix lags desktop responsiveness problems mostly on 64bit systems / in general (16_vmscan-stuff):

[PATCH]\ vmscan:\ raise\ the\ bar\ to\ PAGEOUT_IO_SYNC\ stalls.txt

17 lower power consumption and fix issues with nohz on 2.6.35:

1_sched:\ Fix\ string\ comparison\ in\ proc_sched_features.txt
2_sched:\ Change\ nohz\ idle\ load\ balancing\ logic\ to\ push\ model.txt
3_sched:\ Avoid\ side-effect\ of\ tickless\ idle\ on\ update_cpu_load.txt
4_sched:\ Fix\ nohz\ balance\ kick.txt

<-- those patches are from the zen-kernel stable git-repository: kudos to the zen devs !
you guys saved me some headaches ;)

in the folder upstream\ patches_used\ zen-versions\ instead
are the upstream / unmodified versions

behind the folder
5_core\ i7\ resume_after_Change\ nohz\ idle\ load\ balancing\ logic\ to\ push\ model

[patch\ 3_3\]\ sched\:\ move\ sched_avg_update\(\)\ to\ update_cpu_load_updated.txt

fixes another issue for core i7 cpus after resume from suspend-to-ram/hibernate (the 3rd patch to the series from 14)


18 for your convenience: grsecurity patch to apply if needed

grsecurity-2.2.0-2.6.35.6-201009281623.patch

simply revert:

-static struct irq_chip en_percpu_chip __read_mostly = {
+static struct irq_chip xen_percpu_chip __read_mostly = {

in drivers/xen/events.c

and the grsecurity-patch should apply without rejects and work fine ...

enjoy ! :)


http://depositfiles.com/files/rq75f5t2q (for the patch)

(
sha256sum wrote:
42ce9027566bd4ccd9a63f04c92f7c4f20e22f11a5ed11738982c3cfb243ceeb 2.6.35.7_TOI-3.2-rc2_plus.7z
)

http://depositfiles.com/files/gsu3p5d2v (for the broken-out patches - they're raw so you might have to fix rejects)

sha256sum wrote:
(5087c6c08dffc3028a64109fe04af4cfe5a0c9417ce6812802d17369b87ff9b1 2.6.35.7_TOI-3.2-rc2_plus_broken-out.7z)



P.S.:

if someone can point me to another upload site like omploader I'd gladly upload the archives there
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D


Last edited by kernelOfTruth on Mon Oct 04, 2010 3:21 pm; edited 2 times in total
Back to top
View user's profile Send private message
Rion
Guru
Guru


Joined: 12 Oct 2006
Posts: 377
Location: Ekaterinburg, Russia

PostPosted: Fri Oct 01, 2010 2:57 am    Post subject: Reply with quote

gotta try it, maybe i'll be more lucky than with tuxonice-sources =)
tuxonice in tuxonice-sources are broken for me (hangs on hibernate).
_________________
~amd64
rion-overlay
Back to top
View user's profile Send private message
broch
Tux's lil' helper
Tux's lil' helper


Joined: 09 Jul 2005
Posts: 94

PostPosted: Sat Oct 02, 2010 4:32 pm    Post subject: Reply with quote

I don't have any issues with toi but tried your patch out of curiosity.
I have got following error during compilation:

CC block/blk-ioc.o
CC block/blk-map.o
CC block/blk-exec.o
block/blk-exec.c: In function ‘blk_execute_rq’:
block/blk-exec.c:101:15: error: ‘sysctl_hung_task_timeout_secs’ undeclared (first use in this function)
block/blk-exec.c:101:15: note: each undeclared identifier is reported only once for each function it appears in
make[1]: *** [block/blk-exec.o] Error 1
make: *** [block] Error 2


thanks
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Sat Oct 02, 2010 6:12 pm    Post subject: Reply with quote

broch wrote:
I don't have any issues with toi but tried your patch out of curiosity.
I have got following error during compilation:

CC block/blk-ioc.o
CC block/blk-map.o
CC block/blk-exec.o
block/blk-exec.c: In function ‘blk_execute_rq’:
block/blk-exec.c:101:15: error: ‘sysctl_hung_task_timeout_secs’ undeclared (first use in this function)
block/blk-exec.c:101:15: note: each undeclared identifier is reported only once for each function it appears in
make[1]: *** [block/blk-exec.o] Error 1
make: *** [block] Error 2


thanks


I currently have no idea (and am very busy so I can't come up with a real fix for it):

actually this shouldn't be necessary - but:

reverting:

[PATCH]\ block:\ Prevent\ hang_check\ firing\ during\ long\ I_O.txt

should fix it

thanks for trying out :)
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Sun Oct 03, 2010 3:17 pm    Post subject: Reply with quote

if you're interest in improving your desktop interactivity further try:

[PATCH 0/2] Reduce mmap_sem hold times during file backed page faults
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Mon Oct 04, 2010 1:41 pm    Post subject: Reply with quote

there's an important fix for the xfs-users out there:

[patch] xfs: properly account for reclaimed inodes

<-- the 3rd hunk won't apply cleanly and I currently can't figure out how to fix that ...

... so leave that one out for now and wait until upstream has ported it to 2.6.35 (if applicable), they mentioned porting it to 2.6.34 so it must be really important

edit:

the fix could be:
http://marc.info/?l=linux-kernel&m=128618776007499&w=2
Quote:
diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c
index 6845db9..3314f2a 100644
--- a/fs/xfs/xfs_iget.c
+++ b/fs/xfs/xfs_iget.c
@@ -499,6 +499,7 @@ xfs_ireclaim(
write_lock(&pag->pag_ici_lock);
if (!radix_tree_delete(&pag->pag_ici_root, agino))
ASSERT(0);
+ pag->pag_ici_reclaimable--;
write_unlock(&pag->pag_ici_lock);
xfs_perag_put(pag);


still waiting for approval ...
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Tue Oct 05, 2010 1:25 am    Post subject: Reply with quote

kernelOfTruth wrote:
broch wrote:
I don't have any issues with toi but tried your patch out of curiosity.
I have got following error during compilation:

CC block/blk-ioc.o
CC block/blk-map.o
CC block/blk-exec.o
block/blk-exec.c: In function ‘blk_execute_rq’:
block/blk-exec.c:101:15: error: ‘sysctl_hung_task_timeout_secs’ undeclared (first use in this function)
block/blk-exec.c:101:15: note: each undeclared identifier is reported only once for each function it appears in
make[1]: *** [block/blk-exec.o] Error 1
make: *** [block] Error 2


thanks


I currently have no idea (and am very busy so I can't come up with a real fix for it):

actually this shouldn't be necessary - but:

reverting:

[PATCH]\ block:\ Prevent\ hang_check\ firing\ during\ long\ I_O.txt

should fix it

thanks for trying out :)


setting:

Quote:
Slow work debugging through debugfs (SLOW_WORK_DEBUG) [N/y/?] n


might also fix it ...
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Wed Oct 27, 2010 11:31 pm    Post subject: Reply with quote

new patchset-revision out:

https://forums.gentoo.org/viewtopic-p-6468142.html#6468142 <--
2.6.36_TOI-plus kernel patch & patchset
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Unsupported Software All times are GMT
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum