[Cooker] 2.4.17.5mdk: bug in snd (devfs=nomount)
Hi! Everytime I shutdown cooker, the snd used counter is increased by one instead of decreased by one. I think the problem has to do with alsactl or devfs. After booting the kernel (devfs=nomount due to the already several times reported bug in devfsd) the usage pointer of snd is already at 4, but alsa drivers are not loaded (This worked when I booted with the old devfsd after restarting alsa. I have installed the alsa modules after rebooting to 2.5.17.5mdk). I also tested rebooting with the MDK 8.1 devfsd and found out that booting works here (after compose.inc message there is a kernel oops caused by insmod on my desktop). Still alsa is not started, but when restarting alsa, the kernel modules are loaded. It does not work however, somehow the sound stuff is screwed up, f.e. no /dev/sound devices, cat /proc/sound/cards crashes etc. All this happens on my AMD-K62 desktop, on my laptop sound seems to work. Maybe a problem with my 3DNOW patch ? Regards, Reinhard -- Software-Engineer, Developer for Embedded Devices Project: HyperPen Tablet USB Driver for Linux GnuPG Public Key available on request msg51457/pgp0.pgp Description: PGP signature
Re: [Cooker] 2.4.17.5mdk: bug in snd (devfs=nomount)
On þÔ×, 2002-01-17 at 19:04, Reinhard Katzmann wrote:
Hi!
Everytime I shutdown cooker, the snd used counter is increased by one
instead of decreased by one. I think the problem has to do with
alsactl or devfs.
After booting the kernel (devfs=nomount due to the already
several times reported bug in devfsd) the usage pointer of snd
is already at 4, but alsa drivers are not loaded (This worked
when I booted with the old devfsd after restarting alsa. I
have installed the alsa modules after rebooting to 2.5.17.5mdk).
I also tested rebooting with the MDK 8.1 devfsd and found out
that booting works here (after compose.inc message there is a kernel
oops caused by insmod on my desktop). Still alsa is not started,
but when restarting alsa, the kernel modules are loaded. It does
not work however, somehow the sound stuff is screwed up, f.e.
no /dev/sound devices, cat /proc/sound/cards crashes etc.
All this happens on my AMD-K62 desktop, on my laptop sound seems
to work. Maybe a problem with my 3DNOW patch ?
Works here with devfs on almost up-to-date cooker (yesterday sync) with
ALSA and OSS. Not tried booting without devfs.
2.4.17-2mdk + patch (irrelevant in this case)
2.4.17-5mdk
PIII733, i815e chipset, SB Live! Player 1024
{pts/1}% cat /etc/modules.conf
alias usb-interface usb-uhci
#alias sound-slot-0 emu10k1
alias sound-card-0 snd-card-emu10k1
probeall scsi_hostadapter ppa ide-scsi
options scsi_mod scsihosts='ppa,ide-scsi'
{pts/1}% lsmod
Module Size Used byTainted: P
ppp_deflate42144 0 (autoclean)
bsd_comp4512 0 (autoclean)
ppp_async 6624 1 (autoclean)
ppp_generic19560 3 (autoclean) [ppp_deflate bsd_comp
ppp_async]
slhc5072 1 (autoclean) [ppp_generic]
binfmt_misc 6148 1
snd-synth-emu10k1 4384 0 (autoclean) (unused)
snd-synth-emux 25376 0 (autoclean) [snd-synth-emu10k1]
snd-seq-midi-emul 4848 0 (autoclean) [snd-synth-emux]
snd-seq-virmidi 8320 0 (autoclean) [snd-synth-emux]
snd-seq-midi3424 0 (autoclean) (unused)
snd-seq-oss25600 0 (unused)
snd-seq-midi-event 3184 0 [snd-seq-virmidi snd-seq-midi
snd-seq-oss]
snd-seq40880 0 [snd-synth-emux snd-seq-virmidi
snd-seq-midi snd-seq-oss snd-seq-midi-event]
snd-pcm-oss18816 0
snd-pcm-plugin 16080 0 [snd-pcm-oss]
snd-mixer-oss 4704 0 [snd-pcm-oss]
snd-card-emu10k12400 2
snd-emu10k122912 0 [snd-synth-emu10k1 snd-card-emu10k1]
lp 6432 0 (autoclean)
snd-pcm31648 0 [snd-pcm-oss snd-pcm-plugin
snd-emu10k1]
snd-timer 8576 0 [snd-seq snd-pcm]
snd-ac97-codec 24864 0 [snd-emu10k1]
snd-emux-mem1456 0 [snd-synth-emux snd-emu10k1]
snd-mixer 24328 2 [snd-mixer-oss snd-emu10k1
snd-ac97-codec]
snd-rawmidi10112 0 [snd-seq-midi snd-emu10k1]
snd-seq-device 3948 0 [snd-synth-emu10k1 snd-synth-emux
snd-seq-midi
snd-seq-oss snd-seq snd-card-emu10k1 snd-rawmidi]
snd34304 1 [snd-synth-emux snd-seq-virmidi
snd-seq-midi snd-seq-oss snd-seq-midi-event snd-seq snd-pcm-oss
snd-pcm-plugin snd-mixer-oss snd-card-emu10k1 snd-emu10k1 snd-pcm
snd-timer snd-ac97-codec snd-emux-mem snd-mixer snd-rawmidi
snd-seq-device]
soundcore 4068 9 [snd]
irtty 5920 2 (autoclean)
irda 92300 1 (autoclean) [irtty]
af_packet 12552 0 (autoclean)
keybdev 1888 0 (unused)
mousedev4160 0 (unused)
hid18432 0 (unused)
input 3584 0 [keybdev mousedev hid]
usb-uhci 21540 0 (unused)
usbcore50976 1 [hid usb-uhci]
nls_koi8-r 3872 2 (autoclean)
nls_cp866 3872 2 (autoclean)
vfat9852 2 (autoclean)
fat31928 0 (autoclean) [vfat]
supermount 57668 3 (autoclean)
sd_mod 11580 0 (autoclean) (unused)
w83781d17216 0
i2c-proc6464 0 [w83781d]
i2c-i8014516 0 (unused)
i2c-core 13600 0 [w83781d i2c-proc i2c-i801]
ide-scsi8096 0
scsi_mod 91560 2 [sd_mod ide-scsi]
parport_pc 21704 1 (autoclean)
parport23872 1 [lp parport_pc]
rtc 5880 0 (autoclean)
reiserfs 168672 2
-andrej
Re: [Cooker] 2.4.17.5mdk: bug in snd (devfs=nomount)
Hi! My Module list: Module Size Used byTainted: P af_packet 12808 0 (unused) lp 6176 0 parport_pc 21800 1 parport23456 1 [lp parport_pc] ipsec 135008 1 serial 55712 0 (autoclean) (unused) isa-pnp29416 0 (autoclean) [serial] usb-ohci 19328 0 (unused) 8139too13664 1 (autoclean) snd-timer 10464 1 (initializing) snd41888 5 (autoclean) [snd-timer] soundcore 3844 0 (autoclean) [snd] nls_iso8859-1 2880 1 (autoclean) jfs 129024 1 (autoclean) nls_iso8859-15 3392 1 (autoclean) nls_cp850 3616 1 (autoclean) vfat9916 1 (autoclean) fat31704 0 (autoclean) [vfat] xfs 476064 1 (autoclean) xfs_support 8080 0 (autoclean) [xfs] pagebuf23584 1 (autoclean) [xfs xfs_support] reiserfs 160800 4 (autoclean) ext3 60192 6 (autoclean) jbd38680 6 (autoclean) [ext3] sd_mod 11644 0 (autoclean) floppy 46432 0 (autoclean) md 57632 0 (unused) sr_mod 14744 0 (unused) cdrom 28096 0 [sr_mod] loop 10704 0 (unused) ide-floppy 13152 0 ide-scsi8160 0 usb-storage51804 0 (unused) usbcore50528 1 [usb-ohci usb-storage] tmscsim30208 0 (unused) scsi_mod 91320 4 [sd_mod sr_mod ide-scsi usb-storage tmscsim] rtc 6008 0 (autoclean) unix 15396 80 (autoclean) I just noted that snd-timer is in initializing status.. Seems to be because I got the following oops while loading, could be confirmed by cat /proc/asound/cards: Jan 17 21:35:49 linux kernel: invalid operand: Jan 17 21:35:49 linux kernel: CPU:0 Jan 17 21:35:49 linux kernel: EIP: 0010:[af_packet:__insmod_af_packet_O/lib/modules/2.4.17-1sm/kernel/net/pack+-99315/96] Tainted: P Jan 17 21:35:49 linux kernel: EIP:0010:[e1d0dc0d]Tainted: P Jan 17 21:35:49 linux kernel: EFLAGS: 00010212 Jan 17 21:35:49 linux kernel: eax: 0016 ebx: 1000 ecx: 0016 edx: Jan 17 21:35:49 linux kernel: esi: da373b34 edi: da4561c4 ebp: 1000 esp: d9f25f6c Jan 17 21:35:49 linux kernel: ds: 0018 es: 0018 ss: 0018 Jan 17 21:35:49 linux kernel: Process cat (pid: 5454, stackpage=d9f25000) Jan 17 21:35:49 linux kernel: Stack: da4561c4 ffea 1000 c013c755 da4561c4 0804e618 1000 Jan 17 21:35:49 linux kernel:da4561e4 0002 c012c1d6 0804f000 2000 3c4735a5 3c4735a5 d9f24000 Jan 17 21:35:49 linux kernel:0006 d9f24000 1000 0804e618 0003 c0106fd3 0003 0804e618 Jan 17 21:35:49 linux kernel: Call Trace: [sys_read+149/208] [sys_brk+198/256] [system_call+51/64] Jan 17 21:35:49 linux kernel: Call Trace: [c013c755] [c012c1d6] [c0106fd3] Jan 17 21:35:49 linux kernel: Jan 17 21:35:49 linux kernel: Code: 0f 47 d8 2b 47 20 39 d8 0f 4c d8 53 8b 47 20 03 06 50 ff 74 With kernel-2.4.16.2mdk I didn't have such problems at all, looking closer to the boot process I don't think it's a devfs related problem (the oops also happens with devfs=nomount). Anything what I could try to find the problem ? I strongly suppose it has something todo with 3DNOW, but how do I get sound if I don't even can install a self-made AMD-K62 kernel ? Until tomorrow I recompile my kernel to a normel i586 one which should work as well I hope and report again if there are trouble with that. (Recompiling lasts quite long even with a 500 Mhz AMD-K62 :-( ) Regards, Reinhard -- Software-Engineer, Developer for Embedded Devices Project: HyperPen Tablet USB Driver for Linux GnuPG Public Key available on request msg51483/pgp0.pgp Description: PGP signature
Re: [Cooker] 2.4.17.5mdk: bug in snd (devfs=nomount) - Solved!
Hi! The problem is solved. It seems I have used the wrong version of the alsa driver (installed those from the laptop, ie. the Duron drivers instead of the desktop i.e. K6-2 driver, stupid me). Thanks for all the help!! Reinhard -- Software-Engineer, Developer for Embedded Devices Project: HyperPen Tablet USB Driver for Linux GnuPG Public Key available on request msg51486/pgp0.pgp Description: PGP signature
