Official thread: "zen-sources" - Part II

[mmike]

yeah! no hardlocks anymore (until now)
thanks Waninkoko

no more hardlocks while booting, but only if i disable NO_HZ and HI_RES_TIMERS

[termite]

ALSA problems here. For example, when restarting alsa:

* Storing ALSA Mixer Levels ...
alsactl: get_control:209: Cannot read control info '2,0,0,Master Playback Volume
,0': No such file or directory
* Error saving levels. [ !! ]
* Killing processes using ALSA ... [ ok ]
* Unloading ALSA modules ... [ ok ]
* Loading ALSA modules ...
* Could not detect custom ALSA settings. Loading all detected alsa drivers.
* Loading: snd_hda_intel ... [ ok ]
* Loading: snd-seq-oss ... [ ok ]
* Loading: snd-pcm-oss ... [ ok ]
* Restoring Mixer Levels ...
alsactl: set_control:983: warning: numid mismatch (9/11) for control #9
alsactl: set_control:985: warning: iface mismatch (2/2) for control #9
alsactl: set_control:987: warning: device mismatch (0/0) for control #9
alsactl: set_control:989: warning: subdevice mismatch (0/0) for control #9
alsactl: set_control:991: warning: name mismatch (Off-hook Switch/Off-hook Switc h) for control #9
alsactl: set_control:993: warning: index mismatch (0/0) for control #9
alsactl: set_control:991: warning: name mismatch (Caller ID Switch/Master Playba ck Switch) for control #10
alsactl: set_control:993: warning: index mismatch (0/0) for control #10
alsactl: set_control:995: failed to obtain info for control #10 (Operation not p ermitted)
* Errors while restoring defaults, ignoring [ ok ]

[Vakulinchuk]

@rmh3093: Sorted out my problem - turns out my usr dir was messed up (i.e. I was an idiot) - thanks for the help!

[forceflow2]

My memstick issue seems to have cropped up again on my latest pull of master-devel for some reason (Mentioned on page 8 of original thread)
It's a little different this time around though.

CC drivers/memstick/core/memstick.o
CC drivers/memstick/core/mspro_block.o
LD drivers/memstick/core/built-in.o
drivers/memstick/core/mspro_block.o: In function `memstick_priv':
mspro_block.c:(.text+0x0): multiple definition of `memstick_priv'
drivers/memstick/core/memstick.o:memstick.c:(.text+0x0): first defined here
drivers/memstick/core/mspro_block.o: In function `memstick_get_drvdata':
mspro_block.c:(.text+0x6): multiple definition of `memstick_get_drvdata'
drivers/memstick/core/memstick.o:memstick.c:(.text+0x6): first defined here
drivers/memstick/core/mspro_block.o: In function `memstick_set_drvdata':
mspro_block.c:(.text+0xd): multiple definition of `memstick_set_drvdata'
drivers/memstick/core/memstick.o:memstick.c:(.text+0xd): first defined here
make[3]: *** [drivers/memstick/core/built-in.o] Error 1
make[2]: *** [drivers/memstick/core] Error 2
make[1]: *** [drivers/memstick] Error 2
make: *** [drivers] Error 2

[Waninkoko]

I think you need to compile memstick support as modules.

Edited: Fixed.

[gerardo]

My wireless stopped working since upgrade from zen-sources-2.6.24_rc6-r0 to 2.6.24_rc7-r2
I'm using a Belkin PCI-card with the Ralink rt61pci module of the kernel.

I have exactly all the same options in 2.6.24_rc7-r2 as in 2.6.24_rc6-r0
(except for fb_condecor which was disabled before)

Here's dmesg (mac address is hidden here for safety):

Code: Select all

wlan0: Initial auth_alg=0
wlan0: authenticate with AP 00:12:bf:xx:xx:xx
wlan0: RX authentication from 00:12:bf:xx:xx:xx (alg=0 transaction=2 status=0)
wlan0: authenticated
wlan0: associate with AP 00:12:bf:xx:xx:xx
wlan0: associate with AP 00:12:bf:xx:xx:xx
wlan0: associate with AP 00:12:bf:xx:xx:xx
wlan0: association with AP 00:12:bf:xx:xx:xx timed out

iwconfig:

Code: Select all

wmaster0  no wireless extensions.

wlan0     IEEE 802.11g  ESSID:"myessid"  
          Mode:Managed  Frequency:2.417 GHz  Access Point: 00:12:BF:xx:xx:xx   
          Tx-Power=27 dBm   
          Retry min limit:7   RTS thr:off   Fragment thr=2352 B   
          Encryption key:off
          Power Management:off
          Link Quality:0  Signal level:0  Noise level:0
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:0   Missed beacon:0

ifconfig:

Code: Select all

wlan0     Link encap:Ethernet  HWaddr 00:0E:2E:xx:xx:xx  
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)

wmaster0  Link encap:UNSPEC  HWaddr 00-0E-2E-xx-xx-xx-8E-D6-00-00-00-00-00-00-00-00  
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)

/etc/conf.d/net:

Code: Select all

dns_domain_lo="MSHOME"

modules=( "wpa_supplicant" "dhcpcd" )
wpa_supplicant_wlan0="-Dwext"
config_wlan0=( "dhcp" )
routes_wlan0=( "dhcp" )

# cat /etc/wpa_supplicant/wpa_supplicant.conf

Code: Select all

ctrl_interface=/var/run/wpa_supplicant
ctrl_interface_group=0
ap_scan=1

network={
   ssid="myessid"
   proto=WPA
   key_mgmt=WPA-PSK
   pairwise=TKIP
   psk=myhexkey
   priority=2
 }

network={
	key_mgmt=NONE
	priority=-9999999
}

What has changed?

[forkboy]

I know this ain't gonna help but I've the same problem

[Kwark]

With the old -kamikaze and new -zen sources I have the following problem.
My machine is a dual Xeon, dual core setup, 3 GHz processors.

The second processor is detected as having 2.4GHz with with kernel series. It's detected alright with vanilla and gentoo sources.

Code: Select all

dmesg messages (good)
[    0.060804] CPU0: Intel P4/Xeon Extended MCE MSRs (24) available
[    0.263516] CPU0: Intel(R) Pentium(R) D CPU 3.00GHz stepping 05
[    0.000999] CPU1: Intel P4/Xeon Extended MCE MSRs (24) available
[    0.335436] CPU1: Intel(R) Pentium(R) D CPU 3.00GHz stepping 05

/proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 15
model           : 6
model name      : Intel(R) Pentium(R) D CPU 3.00GHz
stepping        : 5
cpu MHz         : 3000.000
cache size      : 2048 KB

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 15
model           : 6
model name      : Intel(R) Pentium(R) D CPU 3.00GHz
stepping        : 5
cpu MHz         : 2400.000

Any idea what may be causing this?
http://pastebin.com/m2763d309 for my .config

Sorry to be quoting myself, but I'm rather stuck. I don't mind testing different options, but I don't know where to start.

What may be the cause of the second processor having a lower clockspeed than the first one? Both should be 3GHz.

[RobertDavid]

What may be the cause of the second processor having a lower clockspeed than the first one? Both should be 3GHz.

Do you use some speedsteping method (dont know how it is with Xenon), I know that even if you use speedstep than the frequencies are lower both. but check that (/sys/devices/system/cpu/..), because speedsteped processor show its frequency in /proc/cpuinfo (my amd X2 does)

And do you test some performance loss? Becouse I dont think it is possible to run 2-core cpu on different freq. It will be only some mistake in /proc/cpuinfo.

[Kwark]

Yes, I do use cpu frequency scaling, however the performance governor is used, so the frequency should be set to the maximum.

[Waninkoko]

It happens with vanilla 2.6.24-rc7?

[Kwark]

It happens with vanilla 2.6.24-rc7?

No, it does not happen with any -vanilla (or -gentoo) at all.

[mroconnor]

for what ever reason since moving to zen-sources-2.6.24_rc7-r1 from the layman overlay, I can no longer modprobe or load the fglrx module. I am using ati-drivers-8.433 from portage. I'm I missing a post that gives the magic answer or might I be over looking something? Works fine in zen-sources-2.6.24_rc6-r0 and I haven't changed my .config.

Thanks in advance, cheers!

[kernelOfTruth]

for what ever reason since moving to zen-sources-2.6.24_rc7-r1 from the layman overlay, I can no longer modprobe or load the fglrx module. I am using ati-drivers-8.433 from portage. I'm I missing a post that gives the magic answer or might I be over looking something? Works fine in zen-sources-2.6.24_rc6-r0 and I haven't changed my .config.

Thanks in advance, cheers!

probably that one:

http://gentoo-wiki.com/HOWTO_ATI_Driver ... els_failed

[buddabrod]

for what ever reason since moving to zen-sources-2.6.24_rc7-r1 from the layman overlay, I can no longer modprobe or load the fglrx module. I am using ati-drivers-8.433 from portage. I'm I missing a post that gives the magic answer or might I be over looking something? Works fine in zen-sources-2.6.24_rc6-r0 and I haven't changed my .config.

Thanks in advance, cheers!

probably that one:

http://gentoo-wiki.com/HOWTO_ATI_Driver ... els_failed

in .rc7-zen* is the same preemptive stuff as in -rc6.

@progresso: is there any output? did you recompile the driver against the new kernel?

[Dottout]

latest rev doesnt compile

Code: Select all

  LD      .tmp_vmlinux1
kernel/built-in.o: In function `getnstimeofday':
(.text+0x23ffc): undefined reference to `__umoddi3'
kernel/built-in.o: In function `getnstimeofday':
(.text+0x2401c): undefined reference to `__udivdi3'
kernel/built-in.o: In function `do_gettimeofday':
(.text+0x2412e): undefined reference to `__udivdi3'
kernel/built-in.o: In function `do_gettimeofday':
(.text+0x2414c): undefined reference to `__umoddi3'
kernel/built-in.o: In function `timekeeping_resume':
timekeeping.c:(.text+0x243ec): undefined reference to `__umoddi3'
timekeeping.c:(.text+0x2440c): undefined reference to `__udivdi3'
kernel/built-in.o: In function `update_wall_time':
(.text+0x24716): undefined reference to `__umoddi3'
kernel/built-in.o: In function `update_wall_time':
(.text+0x24736): undefined reference to `__udivdi3'
kernel/built-in.o: In function `update_wall_time':
(.text+0x24ba6): undefined reference to `__umoddi3'
kernel/built-in.o: In function `update_wall_time':
(.text+0x24bc6): undefined reference to `__udivdi3'
make: *** [.tmp_vmlinux1] Error 1

edit: solved, it was my fault

[forceflow2]

Code: Select all

wlan0: WMM queue=1 aci=2 acm=0 aifs=2 cWmin=7 cWmax=15 burst=30
wlan0: WMM queue=0 aci=3 acm=0 aifs=2 cWmin=3 cWmax=7 burst=15
WARNING: at net/mac80211/rx.c:1702 __ieee80211_rx_handle_packet()
Pid: 0, comm: swapper Not tainted 2.6.24-rc7-zen2 #14
 [<c0365154>] __ieee80211_rx_handle_packet+0x96/0x819
 [<c0364fe2>] ieee80211_rx_irqsafe+0x2e/0x51
 [<f8aeaf44>] iwl3945_handle_data_packet+0xed/0xfb [iwl3945]
 [<c0365c1c>] ieee80211_rx_reorder_ampdu+0x122/0x12d
 [<c03665e6>] __ieee80211_rx+0x382/0x3d1
 [<c0132a6d>] ktime_get+0xf/0x2b
 [<c0358f63>] ieee80211_tasklet_handler+0x53/0xc9
 [<f8ae93bd>] iwl3945_irq_tasklet+0x3e9/0x55c [iwl3945]
 [<c0124646>] tasklet_action+0x48/0x8c
 [<c0124205>] __do_softirq+0x6c/0xcf
 [<c012429a>] do_softirq+0x32/0x36
 [<c01244ec>] irq_exit+0x38/0x6b
 [<c0106508>] do_IRQ+0x5c/0x70
 [<c0259089>] acpi_hw_register_read+0xd8/0x12a
 [<c01047fb>] common_interrupt+0x23/0x28
 [<c011007b>] acpi_map_lsapic+0x2d/0xda
 [<c026c136>] acpi_idle_enter_bm+0x23a/0x2aa
 [<c02ed302>] cpuidle_idle_call+0x53/0x75
 [<c02ed2af>] cpuidle_idle_call+0x0/0x75
 [<c0102579>] cpu_idle+0x96/0xb7
 [<c04608b7>] start_kernel+0x2bd/0x2c5
 [<c04600e0>] unknown_bootoption+0x0/0x196
 =======================

This pops up when I associate with an AP. Things still work, it's just annoying to see it there

[mroconnor]

I thought I did. But now that I am a zen kernel freak, maybe I got sloppy and missed something.

I'll check through my .config and make sure I actually did rebuild the drivers tonight and report back.

Thanks.

[mroconnor]

I did notice this warning when re-emerging ati-drivers against zen-sources-2.6.24_rc7-r1:

Code: Select all

WARNING: //lib/modules/2.6.24-rc7-zen1/video/fglrx.ko needs unknown symbol flush_tlb_page 

Is that from removing debug sysmbols?

I found this post http://www.cs.ucla.edu/~okada/Memo/Fedo ... fglrx.html, in the section he says:

(Reference: http://d.hatena.ne.jp/viver/20061116 and http://pastebin.ca/350588 )
I borrowed flush_tlb_page() function from /usr/src/kernels/2.6.20-2925.13.fc7xen-x86_64/include/asm-x86_64/tlbflush.h
because I have "undefined symbol" error when linking with kernel by modprobe.

What did I do to my poor kernel?

On another note I seem to be hitting swap more since moving to zen-sources-2.6.24_rc6-r0. Would that have anything to do with Zen tunables such as Gaming etc?

Cheers.

[Waninkoko]

On another note I seem to be hitting swap more since moving to zen-sources-2.6.24_rc6-r0. Would that have anything to do with Zen tunables such as Gaming etc?

Yes.

[mroconnor]

I guess it really doesn't matter, ram usage has gone up and the swap has gone up BUT I think that performance hasn't been touched and I think it has improved(no benchmarks to really quantify it).

I am using the Gaming tunable, tickless(don't think that would impact anything) deadline and a timer freq of 100. Anybody have succes with anything different on a Core2 laptop w/ 1GB ram?

BTW - I really like the new menuconfig layout w/ Zen tunables as a separate item. Excellent work.

Cheers!

[kEiNsTeiN]

hi, sorry for not wanting to read a gazillion pages

what is the difference between xen and zen sources? some patches appear to work for both (nvidia stuff). AFAI understand, xen-sources is patched for being inside a Xen VM, and zen-sources are simply a very popular kernel patchset, and successor of kamikaze...

okay, and my second question: nvidia-drivers fail to compile. if you could look into this?

servus (as I use to say in germany )

[hoacker]

okay, and my second question: nvidia-drivers fail to compile. if you could look into this?

Here's another thread about nvidia not compiling with .24 (vanilla in this case):http://forums.gentoo.org/viewtopic-p-47 ... ml#4728548, but we don't have a solution, yet. Could the missing "arch/i386" be a problem as stated in the linked comment?

[kEiNsTeiN]

I dont think that's related as my emerge goes further than theirs...

oh, btw, try "make clean". There are some scripts missing. I copied them from gentoo-sources and they work. Is this intentional?

[kernelOfTruth]

okay, and my second question: nvidia-drivers fail to compile. if you could look into this?

Here's another thread about nvidia not compiling with .24 (vanilla in this case):http://forums.gentoo.org/viewtopic-p-47 ... ml#4728548, but we don't have a solution, yet. Could the missing "arch/i386" be a problem as stated in the linked comment?

try:
1) sabayon overlay -> nvidia-drivers
2) paludis
3) chmod -R go+rX . (inside /usr/src/linux/), then emerge nvidia-drivers
4) ???