kleine merkseite fuer mich, werde bald darueber bloggen

the hp elitebook 8530w annoys the hell out of me:

  • problems with lvm+xfs+cryptfs:
    • many crashes
    • very slow io, probably a io-wait bug
  • nvidia driver has lots LOTS of quirks with resuming after pm-suspend ranging from backlight issues to fatal crashes with lots of file loss

today i decided to give the pm-suspend - resume crash another try. first i wanted to install a 'fedora core 13' live image. using the internal cdrom: 'DVD/CD REWRITABLE DRIVE', ModelNo: AD-7561S but guess what: it didn't work.

i've noticed latelty that this internal hp drive is broken as it can't burn CDR nor DVDs anymore. all images i was burning were corrupt. i've tried to burn a windows 7 dvd (this is legal, see the MSDN AA license). burning the first image using k3b worked but the dvd was corrupt and did not boot. so i burned another one, this time k3b brought up an error.

a friend of mine gave me his win7 copy. so the same drive was able to boot from that image. i've installed win7 two times. why did i install win7 at all? i wanted to check if my laptop is totally broken or if it's just the sloppy nvidia driver.

using win7 i downloaded a few demos using steam and i was able to play 3 different games for quite a while without any issue. i enabled the win7 suspend/hibernat equivalent to pm-suspend (which leaved the ram powered) and it worked even while the game was running. great news: expect the cdrom everything was working fine and it's the crappy linux system.

so i decided to use several linux live cds in order to find one which WAS actually able to perform a pm-suspend and resume cycle without a crash. as i heared that fc13 has a nice nouveau integration with kms right from scratch i'd like to give it a try.

now how to install such a thing? the dvd was laying right next to the internal drive and i decided to insert it. but booting from the dvd failed. i tried to change various bios values with no effect. so next i demounted the sata dvd drive from my desktop pc (before this i tested that the fc13-dvd would boot there, with great success).

it's quite funny that the pm-suspend worked on my desktop pc (also using a nvidia card and nouveau/kms) without any issue from the live dvd!

next i connected the sata drive to the esata port on my 'hp elitebook 8530w' but it could NOT boot any linux live-dvd i had hat hand, that was: fc13,fc12 and suse linux. suse linux was booting two times but somehow stalled after loading the kernel with NO message in the logs. win7 dvd bootet without any problems? why?!

next i put the external dvd drive into (or better onto) the 'ICY BOX IB-351StUS-B USB & eSATA 3.5" Enclosure' (in short icybox). using the icybox still only win7 was able to boot all other 3 dvds failed. so i asked myself if the linux images were broken?! i attached the icybox to my desktopmachine, also using usb. guess what: i could boot fc13 without any issue after enabling USB-DVD booting in the bios.

finally i put the internal harddrive into a external usb harddrive case similar to the icybox but for 2,5" drives. booting from the USB-DVD drive (the icybox) i then installed the fc13 live-image on the usb-hdd. a crazy setup if you'd ask me!

now i plugged the usb-hdd to my 'hp elitebook 8530w' and i could boot from it? don't ask me why THIS works, but it did. i did that with other linux systems in the past. to sad that does not work with windows.... but that's another story.

using the usb-hdd pm-suspend failed but not with a crash on resume as i'd expected it but with the fc13 background image and NO controls. so i just shut down the computer by issuing a 5sec powerbutton press.

next i plugged the external usb-hdd drive into the laptop, using a SATA setup (in the bios). trying pm-suspend and resume worked fine! this is great news. what is also great news is the fact, that the fan seems to be very low and there is no heat. that means that either the power-management works for the internal nvidia gpu OR that there is simply no heat?! i don't care as it works now!

so the next big quest will to adapt this setup to my gentoo installation!

note from fc13

too bad they did not provide /proc/config.gz! but i found it in /boot/config-2.6.34.7-56.fc13.i686, see http://dune2.dyndns.org/www-share/config-2.6.34.7-56.fc13.i686

so some benchmarks:

  • normal desktop usage both cpu and gpu usage low with good performance in firefox
  • watching Patent_Absurdity_HD-3540kbit.ogv did work with 30-50% cpu usage on BOTH cores, still no gpu heat! about 37-40W when film was playing zoomed, 30W when not playing!

that is great news!

  • a bug on firefox scrolling! which will draw the last buffer to the wrong place leaving the screen corrupted all over the place, a alt+tab fixes it as it redraws all pixels then, no partial drawing.
  • xrandr did not work that well yet
  • screen brightness issue after suspend/resume cycle leaving the screen backlight at 10 out of 100, i could not fix this by using [1] echo 100 > /proc/acpi/video/DGFX/LCD/brightness

[1] https://invalidmagic.wordpress.com/2010/06/30/no-more-crashes-after-resuming-from-pm-suspend/

  • the screen is too dark, have to check that. with ambient-sensors, where is my flashlight?! i used the ultra-bright n900 led light and that triggered the ambient-light sensor to set the screen to full brightness
somehow using echo 100 > /proc/acpi/video/DGFX/LCD/brightness worked now?! have to check that later...
WARNING: no pm-suspend was used YET, after pm-suspend it's simply VERY VERY DARK
to fix that, currently i reboot ... *G*

/var/log/Xorg.0.log

[ 23.309] (II) Loading /usr/lib/xorg/modules/drivers/nouveau_drv.so [ 23.322] (II) Module nouveau: vendor="X.Org Foundation" [ 23.322] compiled for 1.8.2, module version = 0.0.16 [ 23.322] Module class: X.Org Video Driver [ 23.322] ABI class: X.Org Video Driver, version 7.0

X.Org X Server 1.8.2 Release Date: 2010-07-01

Linux localhost.localdomain 2.6.34.7-56.fc13.i686 #1 SMP Wed Sep 15 03:33:58 UTC 2010 i686 i686 i386 GNU/Linux

/var/log/messages

Sep 19 17:09:16 ebooK kernel: [ 1036.685968] ata1: exception Emask 0x10 SAct 0x0 SErr 0x10000 action 0xf Sep 19 17:09:16 ebooK kernel: [ 1036.685977] ata1: SError: { PHYRdyChg } Sep 19 17:09:16 ebooK kernel: [ 1036.685989] ata1: hard resetting link Sep 19 17:09:16 ebooK kernel: [ 1037.408077] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

snd intel issues

00:1b.0 Audio device: Intel Corporation 82801I (ICH9 Family) HD Audio Controller (rev 03)

Sep 22 21:09:30 ebooK pulseaudio[4773]: alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write! Sep 22 21:09:30 ebooK pulseaudio[4773]: alsa-sink.c: Most likely this is a bug in the ALSA driver 'snd_hda_intel'. Please report this issue to the ALSA developers. Sep 22 21:09:30 ebooK pulseaudio[4773]: alsa-sink.c: We were woken up with POLLOUT set -- however a subsequent snd_pcm_avail() returned 0 or another value < min_avail.

update von alsa-utils hat das problem behoben, aber /etc/init.d/alsasound entlaed und laed jetzt auch keine modules mehre, denn alle snd_* sind noch da nach /etc/init.d/alsasound stop

====================================================================
 sata errors, frequently seen
====================================================================

Sep 22 21:09:33 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 22 21:09:52 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001

Sep 25 18:17:20 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:17:42 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:18:20 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:18:39 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:18:55 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:18:58 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:20:01 ebooK cron[12856]: (root) CMD (test -x /usr/sbin/run-crons && /usr/sbin/run-crons ) Sep 25 18:20:30 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:22:05 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:22:27 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:22:46 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:23:24 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:23:40 ebooK logger: ACPI event unhandled: battery PNP0C0A:00 00000080 00000001 Sep 25 18:24:56 ebooK kernel: [ 2989.757123] ata1: exception Emask 0x10 SAct 0x0 SErr 0x10000 action 0xf Sep 25 18:24:56 ebooK kernel: [ 2989.757134] ata1: SError: { PHYRdyChg } Sep 25 18:24:56 ebooK kernel: [ 2989.757146] ata1: hard resetting link Sep 25 18:24:57 ebooK kernel: [ 2990.480057] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300) Sep 25 18:24:57 ebooK kernel: [ 2990.482352] ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out Sep 25 18:24:57 ebooK kernel: [ 2990.485490] ata1.00: ACPI cmd ef/10:03:00:00:00:a0 (SET FEATURES) filtered out Sep 25 18:24:57 ebooK kernel: [ 2990.485935] ata1.00: configured for UDMA/133 Sep 25 18:24:57 ebooK kernel: [ 2990.485945] ata1: EH complete Sep 25 18:24:57 ebooK kernel: [ 2990.486235] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA Sep 25 18:24:58 ebooK kernel: [ 2990.821821] PM: Adding info for No Bus:vcs63 Sep 25 18:24:58 ebooK kernel: [ 2990.821846] PM: Adding info for No Bus:vcsa63

maybe related to the nvidia module? http://ubuntuforums.org/archive/index.php/t-1037819.html kann ich nicht nachpruefen, nv sowie vesa tun nicht richtig:

- nv: man kann damit arbeiten, beim resume blackscreen of death
- vesa: total defekt, bunte pixel zufaellig ueber dem bildschirm


/etc/init.d/alsasound restart

Freezing user space processes ... Freezing of tasks failed after 20.00 seconds (1 tasks refusing to freeze): rmmod D ffff88003e095fc0 0 1988 1891 0x00800004 ffff880028bd1ed8 0000000000000082 dead000000200200 ffff880028bd1fd8 0000000000013b00 ffff880028bd1fd8 ffff88003e095c00 0000000000013b00 0000000000013b00 0000000000013b00 0000000000013b00 ffff88003e095c00 Call Trace: [<ffffffff810592c0>] ? wakeme_after_rcu+0x0/0x14 [<ffffffff81070af2>] sys_delete_module+0x1b5/0x266 [<ffffffff814a1819>] ? do_page_fault+0x29e/0x2cd [<ffffffff81002db2>] system_call_fastpath+0x16/0x1b

siehe auch n900 bild vom crash, dort ist in stack-trace

nachricht an mark & mein log

news: verwendeter kernel: 2.6.34-gentoo-r6

-grafik 256.53 (nvidia proprietary)

tut soweit alles voll gut, aber meine festplatte macht jetzt spacken, wenn die im ahci mode ist (im ide mode ist alles kein problem)
im ahci mode kommt dann der fehler:
   Sep 25 18:24:56 ebooK kernel: [ 2989.757123] ata1: exception Emask 0x10 SAct
   0x0 SErr 0x10000 action 0xf
   Sep 25 18:24:56 ebooK kernel: [ 2989.757134] ata1: SError: { PHYRdyChg }
   Sep 25 18:24:56 ebooK kernel: [ 2989.757146] ata1: hard resetting link
   Sep 25 18:24:57 ebooK kernel: [ 2990.480057] ata1: SATA link up 3.0 Gbps
   (SStatus 123 SControl 300)
   Sep 25 18:24:57 ebooK kernel: [ 2990.482352] ata1.00: ACPI cmd
   ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
   Sep 25 18:24:57 ebooK kernel: [ 2990.485490] ata1.00: ACPI cmd
   ef/10:03:00:00:00:a0 (SET FEATURES) filtered out
   Sep 25 18:24:57 ebooK kernel: [ 2990.485935] ata1.00: configured for UDMA/133
   Sep 25 18:24:57 ebooK kernel: [ 2990.485945] ata1: EH complete
   Sep 25 18:24:57 ebooK kernel: [ 2990.486235] sd 0:0:0:0: [sda] Write cache:
   enabled, read cache: enabled, doesn't support DPO or FUA

- sound

- /etc/init.d/alsasound entlaed die module garnicht mehr
- manuelles entladen und neuladen der soundkarte tut, solange man bisher noch kein sound gespielt hat
- der kernel crashed wenn man sound spielt und dann die module entlaed (nachdem man mplayer wieder geschlossen hat)
  ev. tritt dieser fehler nur nach einem suspen/resume cycle auf, ungetestet

- wlan

- nach mehrfachen module unloading/reloading kommt es zu einem kernel-crash, stacktrace vorhanden (n900 foto)

- acpi

- ich hab die acpi keys volup/voldown, fn+f3, battery-inducator und fn+f4 (keine ahnung was das macht) hinbekommen
- mute macht weiterhin kein acpi event (bei System.map-2.6.34-gentoo-r1 war das alles etwas anderst)

- vt virtualisierung scheint mit meinem 64bit system schoen zu funktionieren, hab das mit virtualbox getestet


ich habe deine kernel config noch nicht probiert, das werde ich jetzt mal machen

Powered by MediaWiki