Official thread: "zen-sources" - Part II

Message

Post by **asturm** » Wed Feb 27, 2008 6:19 pm

Waninkoko wrote:Known bug.

Do you know which kernel option causes this?

Waninkoko · Post by **Waninkoko** » Wed Feb 27, 2008 9:40 pm

genstorm: It's caused by swap prefetch. But now it's fixed.

unK: yeah, happening here too. We think this happens if swap prefetch is enabled.

unK · Post by **unK** » Wed Feb 27, 2008 10:13 pm

Waninkoko wrote:unK: yeah, happening here too. We think this happens if swap prefetch is enabled.

Does it affect both master and master-devel?

Btw, it is possible to switch from latest revision to e.g v2.6.24-zen2?

Post by **asturm** » Wed Feb 27, 2008 10:27 pm

thx Waninkoko, trying zen3 out now!

kernelOfTruth · Post by **kernelOfTruth** » Wed Feb 27, 2008 10:53 pm

unK wrote:
Waninkoko wrote:unK: yeah, happening here too. We think this happens if swap prefetch is enabled.
Does it affect both master and master-devel?

Btw, it is possible to switch from latest revision to e.g v2.6.24-zen2?

*sigh*

Code: Select all

git checkout v2.6.24-zen2

should (hopefully) do the trick

unK · Post by **unK** » Wed Feb 27, 2008 11:01 pm

Thank you, I tried some combinations with git checkout, but apparently I didn't try the simplest one ;p

EnigmaticSeraph · Post by **EnigmaticSeraph** » Wed Feb 27, 2008 11:36 pm

@kernelOfTruth

BTW, I haven't forgotten about the config problem. ^^; It seems to be a bug in the config system. When I have the time, I'll investigate further, patch, and push upstream. For now, I'm really sorry that it fails for you. :/ The best workaround is to use one of the graphical configs or to throw your options in to the .config in a text editor.

Post by **asturm** » Thu Feb 28, 2008 1:14 am

Waninkoko wrote:genstorm: It's caused by swap prefetch. But now it's fixed.

positive, it works now.

Kwark · Post by **Kwark** » Thu Feb 28, 2008 9:22 am

My 2.6.24-zen2 does not poweroff properly. It just hangs on the very last part where it normally should poweroff the machine.
This is my .config. I have already tried booting with acpi=force but that didn't change anything.

Any ideas? Rebooting works fine.

loeb-it · Post by **loeb-it** » Thu Feb 28, 2008 10:40 am

Kwark wrote:My 2.6.24-zen2 does not poweroff properly. It just hangs on the very last part where it normally should poweroff the machine.
This is my .config. I have already tried booting with acpi=force but that didn't change anything.

Any ideas? Rebooting works fine.

Remove CONFIG_PREEMPT_RCU=y.

Post by **asturm** » Thu Feb 28, 2008 1:14 pm

I've reverted back to codename vipernicus.

zen3 doesn't hang my system anymore, but floods dmesg:

Code: Select all

klauncher[7118]: segfault at c9 rip 2b10e5b906bd rsp 7fffc682bce0 error 4
BUG: scheduling while atomic: X/5769/0x10000002
Pid: 5769, comm: X Tainted: P      D 2.6.24-zen3 #1

Call Trace:
 [<ffffffff8046e4a4>] schedule+0xd1/0x616
 [<ffffffff80298983>] __fput+0x181/0x1c1
 [<ffffffff80231999>] __cond_resched+0x1c/0x44
 [<ffffffff8046ec73>] cond_resched+0x2e/0x39
 [<ffffffff80237af0>] put_files_struct+0x6b/0xc5
 [<ffffffff80238e14>] do_exit+0x245/0x743
 [<ffffffff80368e4e>] do_unblank_screen+0xd/0x121
 [<ffffffff80221a44>] do_page_fault+0x592/0x65a
 [<ffffffff802a4e67>] __pollwait+0x0/0xe1
 [<ffffffff8022ebaa>] default_wake_function+0x0/0xe
 [<ffffffff8022ebaa>] default_wake_function+0x0/0xe
 [<ffffffff8022d62e>] __dequeue_entity+0x1c/0x32
 [<ffffffff8022d65c>] set_next_entity+0x18/0x3a
 [<ffffffff804706d9>] error_exit+0x0/0x51
 [<ffffffff80345e47>] clear_page_c+0x7/0x10
 [<ffffffff80271992>] get_page_from_freelist+0x49c/0x635
 [<ffffffff80271d8b>] __alloc_pages+0xa7/0x340
 [<ffffffff802743d5>] __pagevec_lru_add_active+0xd8/0xe9
 [<ffffffff8027c81a>] handle_mm_fault+0x24d/0x745
 [<ffffffff8022177a>] do_page_fault+0x2c8/0x65a
 [<ffffffff80213867>] read_hpet+0xf/0x12
 [<ffffffff8024c8c2>] getnstimeofday+0x2f/0x84
 [<ffffffff804706d9>] error_exit+0x0/0x51

Code: Select all

CONFIG_CLASSIC_RCU=y
# CONFIG_PREEMPT_RCU is not set
# CONFIG_RCU_TRACE is not set

mroconnor · Post by **mroconnor** » Thu Feb 28, 2008 2:54 pm

After recloning my repo yesterday I compiled 2.6.25-rc3-zen0-03138-g562d6ce and proceeded to emerge ati-drivers-8.433. Kernel compiled fine(they always do) but ati-drivers died with this message:

Code: Select all

make[1]: Entering directory `/usr/src/zen-sources'
  CC [M]  /var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.o
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c: In function 'firegl_put_user_ptr':
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:1620: warning: cast from pointer to integer of different size
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:1620: warning: cast from pointer to integer of different size
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:1620: warning: cast from pointer to integer of different size
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:1620: warning: cast from pointer to integer of different size
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c: In function 'KCL_PosixSecurityCapGetEffectiveVector':
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:1922: error: implicit declaration of function 'cap_t'
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c: In function 'KCL_PosixSecurityCapSetEffectiveVector':
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:1930: error: lvalue required as left operand of assignment
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c: In function '__ke_unregister_ioctl32_conversion':
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:3047: warning: 'return' with a value, in function returning void
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c: In function 'KAS_SlabCache_Initialize':
/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.c:5544: warning: cast from pointer to integer of different size
make[2]: *** [/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x/firegl_public.o] Error 1
make[1]: *** [_module_/var/tmp/portage/x11-drivers/ati-drivers-8.433/work/common/lib/modules/fglrx/build_mod/2.6.x] Error 2
make[1]: Leaving directory `/usr/src/zen-sources'
make: *** [kmod_build] Error 2
 * 
 * ERROR: x11-drivers/ati-drivers-8.433 failed.
 * Call stack:
 *               ebuild.sh, line   49:  Called src_compile
 *             environment, line 3268:  Called linux-mod_src_compile
 *             environment, line 2473:  Called die
 * The specific snippet of code:
 *               emake HOSTCC="$(tc-getBUILD_CC)" CC="$(get-KERNEL_CC)" LDFLAGS="$(get_abi_LDFLAGS)" ${BUILD_FIXES} ${BUILD_PARAMS} ${BUILD_TARGETS} || die "Unable to make ${BUILD_FIXES} ${BUILD_PARAMS} ${BUILD_TARGETS}.";
 *  The die message:
 *   Unable to make  GCC_VER_MAJ=4 KVER=2.6.25-rc3-zen0-03138-g562d6ce KDIR=/usr/src/linux kmod_build.
 * 
 * If you need support, post the topmost build error, and the call stack if relevant.
 * A complete build log is located at '/var/log/portage/x11-drivers:ati-drivers-8.433:20080228-144433.log'.
 * The ebuild environment file is located at '/var/tmp/portage/x11-drivers/ati-drivers-8.433/temp/environment'.
 *

I am using Classic RCU, I learned from last time. Did I miss something in .25 that changed radically?

Thanks in advance for any tips.

progresso

mroconnor · Post by **mroconnor** » Thu Feb 28, 2008 3:27 pm

Hmmm found this on phoronix.com. http://www.phoronix.com/forums/showthread.php?t=7631. Of course it doesnt offer a solution.

ilikenwf · Post by **ilikenwf** » Fri Feb 29, 2008 5:02 pm

I had to remove the first function from the C1E patch to get it to work, as well as disable aufs, unionfs, chunkfs, madwifi, and a ralink driver, but 2.6.25-rc3-zen0 compiles and runs cleanly. With it, I am able to boot my HP laptop dv9205us without using noapic or nolapic, or both, which is a monumental achievement. It is also wonderful that my wireless card is native now.

RMH, Waninkoko, devs, YOU ROCK!

dodo1122 · Post by **dodo1122** » Sat Mar 01, 2008 12:32 am

swap_prefetch bug in master-devel is being worked on (by tomorrow it should be completely fixed). There is also new alsa from mm in master-devel.
Also, i just pushed btrfs 0.13 to master.

dodo

are · Post by **are** » Sat Mar 01, 2008 6:40 pm

Waninkoko wrote:
are wrote:
genstorm wrote:That's what I got with every 2.6.24-zen version so far and why I'm still running 2.6.23-kamikaze5, because I didn't have the nerves and time to bother about the problem. It seems to me though that it only occured after building external modules against the kernel, so it could possibly be nvidia-related.
Waninkoko, you did an amazing work, but i have to agree: 2.6.23-kamikaze5 is much more mature especially for my notebook than zen is. It uses 2 watts less power, reboots/halts stable and hibernate/suspend works more stable too. Indeed I use this last kamikaze with updatet alsa and iwl-wifi.

Nevertheless i will test every new release! Thank you very much for your efforts!
Best regards
are
Try disabling preempt rcu.

Hi Waninkoko, i've tried this but it did not change anything. Latest kernel, that halts and hibernates properly, was 2.6.24-rc7-zen2 (with preempt rcu). Since then i can't power off the laptop, neither with or without preempt rcu. I have an completely intel based laptop.

Thanks for your efforts! and best regards
are

kernelOfTruth · Post by **kernelOfTruth** » Sat Mar 01, 2008 7:12 pm

anyone interested in an pax-patch for 2.6.24-zen4 ?

kernelOfTruth · Post by **kernelOfTruth** » Sat Mar 01, 2008 10:02 pm

kernelOfTruth wrote:anyone interested in an pax-patch for 2.6.24-zen4 ?

update:

sorry, no patch

, the patch I created was 140 MB big

so here's the tarball with grsecurity's pax-patch (http://www.grsecurity.com/test/pax-linu ... st31.patch) [February 29 2008 12:15:15],
zen-sources' state should be from commit: 04d280d4e981b4a2b3a14eae36aa7a0796566163

speed meets security Redux

see: http://forums.gentoo.org/viewtopic-t-66 ... ight-.html

kernelOfTruth · Post by **kernelOfTruth** » Sat Mar 01, 2008 10:10 pm

example kernel-config for amd64, zen-sources 2.6.24-zen4 + pax

moved to http://forums.gentoo.org/viewtopic-p-49 ... ml#4920047

gimpel · Post by **gimpel** » Sat Mar 01, 2008 11:08 pm

Errr, finger hurts due to excessive scrolling

Please use some nopaste service like http://phpfi.com for such things.

kernelOfTruth · Post by **kernelOfTruth** » Sat Mar 01, 2008 11:51 pm

gimpel wrote: Please use some nopaste service like http://phpfi.com for such things.

that paste-service looks good, I like it, I think I'll use it in the future

gimpel wrote: Errr, finger hurts due to excessive scrolling

wut ?

solved

, thanks

Gregoire · Post by **Gregoire** » Sun Mar 02, 2008 12:44 pm

What's the C1E's patch ?

Is there a way from linux to set the CPU ratio without speedstep activated as in windows ?

(I ask for core duo 2 CPU)

ilikenwf · Post by **ilikenwf** » Mon Mar 03, 2008 10:51 pm

Gregoire wrote:What's the C1E's patch ?

Is there a way from linux to set the CPU ratio without speedstep activated as in windows ?

(I ask for core duo 2 CPU)

The C1E patch disables a function in the AMD chips that causes lapic to incorrectly report that it is broken and disabled. It doesn't effect Intel users.

soundcheck · Post by **soundcheck** » Thu Mar 06, 2008 9:28 am

Hi folks.

Perhaps somebody can sched some light on below..

Can somebody explain what the difference between 2.6.24-zen4 and 2.6.24-zen-rt3.

Are these different revisions? What for? How can I get access to zen-rt3?

From what I recall Ingo offers a "Complete Preemption" mode and threaded SOFT-/HardIRQs.
Would be nice to have these in standard ZEN, if not already included. If these are somehow redundant
to other ZEN specific parameters forget that comment.

THX

kernelOfTruth · Post by **kernelOfTruth** » Thu Mar 06, 2008 5:47 pm

kls wrote:Hi folks.

Perhaps somebody can sched some light on below..

Can somebody explain what the difference between 2.6.24-zen4 and 2.6.24-zen-rt3.

Are these different revisions? What for? How can I get access to zen-rt3?

From what I recall Ingo offers a "Complete Preemption" mode and threaded SOFT-/HardIRQs.
Would be nice to have these in standard ZEN, if not already included. If these are somehow redundant
to other ZEN specific parameters forget that comment.

THX

Hi kls,

zen-rt is the branch with your mentioned features (soft- & hardirq)

if you're already cloned zen-sources checking out 2.6.24-zen-rt3 should suffice

Cheers