Hello everybody.
Today (errrrrr, yesterday, since it is now well past midnight and I am
still pulling my hair off!), UPS brought me 10.0. With eager
anticipation I started upgrading 9.3 to 10.0 on my Presario 3240 laptop
(2.2 GHz AMD64 based, nVidia chipset, nVidia GeForce4 440, Go with
64MB dedicated video memory). Now I regret it ever doing this upgrade.
Below I describe some of the many problems I am facing after the upgrade
and then provide what I think are pertinent syslogs.
--What happened to alsaplayer? It seems it has disappeared, even from
the repositories!
--Often the cursor does not "grab" when it is supposed to (left-clicking
on sliders or window frames for resizing or when attempting to "block"
text to be pasted by left-clicking and dragging --I use KDE).
--The internet connection unexpectedly freezes after about 10 minutes of
uptime (it did not do it just before the upgrade and after the upgrade I
tested it with a 32-bit system; the ethernet cable and the cable modem
seem to work fine with the 32-bit system --running 32-bit SuSE 9.3).
This happens even if there is virtually no internet activity. See the
pertinent syslogs below. Could this be a kernel problem? 'uname -r'
responds '2.6.13-15-default', the stock SuSE kernel.
--Shutdown takes forever (maybe over a minute) to commence.
--Lost ticks are reported in the syslogs, even booting with the
'no_timer_check' kernel directive. (And no, I need ACPI because this is
a laptop!) See the pertinent syslogs below.
--DMA and IRQ errors also get registered in the syslogs, as you can see
below.
--The upgrade trashed all my "politically incorrect" media players! (I
remember that before SuSE was aquired by Novel, it offered, on the
installation CD/DVD, rpms for hard-to-install media players. Those were
the days!) with some effort, I can tolerate "politically correct"
libraries, such as 'xine-lib', if I know what I have to delete and
install from other sources. But in my case it insisted on replacing
perfectly good libraries, claiming inconsistencies. When the upgrade
was done, kaffeine, xine, and MPlayer, which worked reasonably well
under 9.3 (with the exception of some --not all-- .wmv files) had been
virtually eviscerated. None could play DVDs, or mpegs, or whatever
else! None! I uninstalled 'xine-lib', reinstalled 'libxine1',
uninstalled and then re-installed kaffeine, and made sure that
'w32codec' was installed, but the end result was the same. (It could be
that some packages are 32-bit and some others 64.) At least, if the
user has "politically incorrect" libraries and executables installed,
the upgrade procedure should make an attempt to leave them alone!!!!
Some libraries, such as 'libdvdcss' are not even available in the
apt-get repositories and one has to go on a treasure hunt to find them
--if he is able to at all. Hear me out, SuSE/Novell! I cannot spend
all my waking hours and cut into my sleep hunting for dependencies to
get my computer to do the things I want. I am seriously thinking about
switching to Mandriva or some other distribution which has all the
software I want ready to go. If this had been a friend's computer, most
probably I WOULD have gone with Mandriva. (Where is the ease of use,
advertised on the box my10.0 came in? Novell has been quoted in the
newsmedia about its desire to pernetrate the desktop. How are they going
to attract people away from Windows with problems like these? Sure, I am
no guru, but I am no novice either.)
--Why does 'lspci' report my video card as '01:00.0 VGA compatible
controller: nVidia Corporation NV17 [GeForce4 420 Go 32M] (rev a3)',
while I know it is a *64MB* GeForce *440* Go?
Now some snippets of the syslogs (obtained via 'tail -n 100 -f
/var/sys/log/messages'):
Internet connection hangs:
Oct 25 00:04:53 presario gconfd (nonsense-7105): GConf server is not in
use, shutting down.
Oct 25 00:04:53 presario gconfd (nonsense-7105): Exiting
Oct 25 00:06:25 presario kernel: NETDEV WATCHDOG: eth0: transmit timed out
Oct 25 00:06:25 presario kernel: eth0: Transmit timeout, status 0c 0055
c07f media 10.
Oct 25 00:06:25 presario kernel: eth0: Tx queue start entry 5131 dirty
entry 5127.
Oct 25 00:06:25 presario kernel: eth0: Tx descriptor 0 is 0008a047.
Oct 25 00:06:25 presario kernel: eth0: Tx descriptor 1 is 0008a063.
Oct 25 00:06:25 presario kernel: eth0: Tx descriptor 2 is 0008a047.
Oct 25 00:06:25 presario kernel: eth0: Tx descriptor 3 is 0008a047.
(queue head)
Oct 25 00:06:25 presario kernel: eth0: link up, 100Mbps, full-duplex,
lpa 0x41E1
Oct 25 00:06:37 presario kernel: NETDEV WATCHDOG: eth0: transmit timed out
Oct 25 00:06:37 presario kernel: eth0: Transmit timeout, status 0c 0055
c07f media 10.
Oct 25 00:06:37 presario kernel: eth0: Tx queue start entry 4 dirty
entry 0.
Oct 25 00:06:37 presario kernel: eth0: Tx descriptor 0 is 0008a047.
(queue head)
Oct 25 00:06:37 presario kernel: eth0: Tx descriptor 1 is 0008a047.
Oct 25 00:06:37 presario kernel: eth0: Tx descriptor 2 is 0008a047.
Oct 25 00:06:37 presario kernel: eth0: Tx descriptor 3 is 0008a047.
Oct 25 00:06:37 presario kernel: eth0: link up, 100Mbps, full-duplex,
lpa 0x41E1
[This pattern repeats itself.]
Sound, DMA, interrupt, and timing problems (disregard the errors about
skype segfaulting; it was happening under 9.3, too):
Oct 25 00:14:17 presario kernel: hda: dma_timer_expiry: dma status == 0x24
Oct 25 00:14:27 presario kernel: hda: DMA interrupt recovery
Oct 25 00:14:27 presario kernel: hda: lost interrupt
Oct 25 00:14:50 presario hp: unable to open /var/run/hpiod.port: No such
file or directory: prnt/hpijs/hplip_api.c 75
Oct 25 00:15:23 presario kernel: hda: dma_timer_expiry: dma status == 0x24
Oct 25 00:15:33 presario kernel: hda: DMA interrupt recovery
Oct 25 00:15:33 presario kernel: hda: lost interrupt
Oct 25 00:17:04 presario kernel: Losing some ticks... checking if CPU
frequency changed.
Oct 25 00:20:40 presario kernel: ALSA sound/core/pcm_lib.c:2126:
playback write error (DMA or IRQ trouble?)
Oct 25 00:20:50 presario kernel: ALSA sound/core/pcm_lib.c:2429: capture
read error (DMA or IRQ trouble?)
Oct 25 00:20:59 presario kernel: skype[6459]: segfault at
0000000057318920 rip 0000000057318920 rsp 00000000567e742c error 14
Oct 25 00:21:09 presario kernel: ALSA sound/core/pcm_native.c:1451:
playback drain error (DMA or IRQ trouble?)
Oct 25 00:21:15 presario su: (to root) nonsense on /dev/pts/1
Oct 25 00:21:42 presario kernel: ALSA sound/core/pcm_lib.c:2126:
playback write error (DMA or IRQ trouble?)
Oct 25 00:21:52 presario kernel: ALSA sound/core/pcm_lib.c:2126:
playback write error (DMA or IRQ trouble?)
Oct 25 00:21:58 presario kernel: skype[6508]: segfault at
00000000572f1920 rip 00000000572f1920 rsp 00000000567e742c error 14
Oct 25 00:22:08 presario kernel: ALSA sound/core/pcm_native.c:1451:
playback drain error (DMA or IRQ trouble?)
presario:~ #
Here is output of 'vmstat 5'; I am used to seeing *five* times as many
interrupts under the same conditions (idling system):
procs -----------memory---------- ---swap-- -----io---- --system--
----cpu----
r b swpd free buff cache si so bi bo in cs us sy
id wa
0 0 0 32140 150092 360188 0 0 109 31 300 377 7 1
88 4
0 0 0 32156 150100 360188 0 0 0 16 255 284 0 0
100 0
0 0 0 32156 150100 360188 0 0 0 6 253 276 0 0
100 0
CF