Linux Sound subsystem development
 help / color / mirror / Atom feed
* Re: hdd & sound
@ 1999-08-12  1:59 Jonathan Masters
  1999-08-12 16:22 ` Jonathan Masters
  1999-08-16 15:35 ` Jonathan Masters
  0 siblings, 2 replies; 3+ messages in thread
From: Jonathan Masters @ 1999-08-12  1:59 UTC (permalink / raw)
  To: linux-sound

[-- Attachment #1: Type: text/plain, Size: 4834 bytes --]

Roger Larsson wrote:

> Hi all,
>
> All of you that are having this problem. Could you reply to me with:
> a) type of ISA cards (as specific as possible, especially SCSI and
> sound)
>

My machine has a single ISA card which is a SB 64 AWE GOLD. The HDDs are running of
the motherboard controller. I have two hdds on one controller, two cd-roms (1
reader, 1 rewriter) on the other. This test was performed when ripping the Britney
Spears Sometimes single to create a wav file. cdrecord produces some interuption, I
used cdparanoia becausethat seems to be much worse so that you can seethe effects.
This also occurs when I write cd-roms, but I don't want to waste a couple of cds
unless you think it's useful in which case I will.



> b) verify that you do not run SCSI/EIDE in non bus master mode.
>

As far as I am aware they are in Bus master mode. The machine was built by myself
and is less than 1 year old (PII 400 Mhz 128Mb RAM, 256 MbSWAP (128Mb on each of
the two hdds). I did not adjust this setting so I assume that it is using Bus
mastering mode which is normal on all modern machines.


> c) attach a output from vmstat 2 > vmstat.out
>    (play sound, start this command, wait 5 secs, start disk activity,
>     wait 5 secs, stop vmstat with ^C
>

I already had sound playing, started the command, waited 5 secs, started ripping,
continued until ripping finished, waited a couple of secs (I think 5) and then
stopped vmstat. During the ripping there was a horrible "slow motiony slow down" of
the music, like when you hold a tape and physically slow it down as it moves inside
a tape player. Also, it often "cut's out" for several seconds at a time as it did
once or twice here.


>
>     This was checked on Wakkos machine, and it indicated that no bus
>     mastering / DMA was used.
>     Note: my SB16SCSI II does not have DMA enabled by default... and
>           ISA cards can not do bus mastering on their own.)
>

I assume you are refering to the sbs which have a hdd controller onboard? Mine are
on the motherboard separate from the sb64.

If you want any further info, just tell me what you need and you've got it
yesturday.


>
> [I will make a summary]
>
> PS.
>
> I do think that this is related (until proven otherwise)
>
> Check out
>   http://www.gardena.net/benno/linux/audio/
>
> But make sure that you are using Bus Master EIDE / dma transfers since
> it does not handle that yet !
>
> /RogerL
>
> Mike wrote:
> >
> > On Sat, 7 Aug 1999, Wakko Warner wrote:
> >
> > > I'm noticing this one one of my machines.  When writing to a hard drive, the
> > > sound drags.  The machine I'm on has piix4 ide and an isa scsi card.  I'm
> > > playing sounds from the drive on the scsi card (aha1510 I think is the scsi
> > > card).  I tried writing to both an ide and scsi drive, both drag the sound.
> > > Sound card is an SB16 pnp /w ide port (not disablable unfortunately).  I had
> > > a similar problem on an isa ide card playing sounds off of the hard drive
> > > (like 44k 16bit 2ch sound off the drive, sound card was an ess card).  My
> > > home box that this is happneing on is 2.2.7.  The one with the isa ide is no
> > > longer active, but it was running 2.0.36.
> > >
> > I see much the same thing under (extremely) heavy disk load, both on
> > onboard piix4 ide, and on an onboard aic7??? scsi.
> >
> > Things it isn't:
> >   hardware - too many people are seeing the same thing for that
> >   driver specific - it happens on both ide and scsi, and with both sb and
> >                     ess cards
> >   anything blindingly obvious
> >
> > Currently I suspect something evil either in sound_core.c or sound_timer.c
> > or (more likely) something scheduler related, but this is wild guessing.  Any
> > ideas?
> >
> > --
> > Mike <rickettm@ox.compsoc.net>
> >
> > Q:      What's yellow, and equivalent to the Axiom of Choice?
> > A:      Zorn's Lemon.
> >
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.rutgers.edu
> > Please read the FAQ at http://www.tux.org/lkml/
>
> --
>
> The Internet interprets Windows as damage,
>              and routes around it.
>
> Roger Larsson
> Skellefteå
> Sweden
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.rutgers.edu
> Please read the FAQ at http://www.tux.org/lkml/

--
Jonathan C. Masters                     (jonathan@oxlug.org)
                                        PGP: www.brookes.ac.uk/~95227860/KEY

           "Upon this rock I will build my church,
            and the gates of hell shall not prevail against it".

                 -- Matthew 16, 17-18



[-- Attachment #2: vmstat.out --]
[-- Type: text/plain, Size: 5040 bytes --]

   procs                      memory    swap          io     system         cpu
 r  b  w   swpd   free   buff  cache  si  so    bi    bo   in    cs  us  sy  id
 1  0  0  29248   2996  34524  23732   1   0    72    45  170   148  15   4  82
 1  0  0  29248   3044  34520  23684   0   0    38     1  880  1931   2   4  94
 1  0  0  29248   3180  34480  23596   0   0     0     7  816  1901   2   4  94
 2  0  0  29248   1832  34236  23584   0   0     0     4  821  1775   4   7  89
 2  0  0  29248   3136  32876  23484   0   0    39   281  784  1704   7  12  81
 0  0  0  29248   3168  32624  23480   0   0     0     0  842  1861  10   8  82
 1  0  0  29248   3140  32392  23532   0   0    38     0  799  1894   7  12  81
 2  0  1  29248   3708  32304  23492   0   0     0     0  746  1695  10  10  80
 3  0  1  29248   3580  32304  23492   0   0     0   642  767  1748  11   9  80
 2  0  1  29248   3368  32304  23568   0   0    38     0  756  1725  11  10  79
 2  0  1  29248   3112  32304  23568   0   0     0     0  768  1826   6  12  82
 1  0  1  29248   3772  32304  23260   0   0    38  1130  794  1841   8  11  81
 2  0  1  29248   3696  32304  23260   0   0     0     0  774  1860   9   9  82
 1  0  1  29248   3948  32304  23260   0   0     1   729  792  1765  10  10  80
 2  0  0  29248   3596  32304  23336   0   0    39     7  752  1679   7  10  83
 1  0  0  29248   3404  32304  23336   0   0     0     0  768  1819   7   6  87
 2  0  0  29248   3792  32304  23336   0   0     1  1023  791  1737   8   9  83
 2  0  0  29248   3600  32304  23412   0   0    38     0  769  1704  10   9  80
 3  0  1  29248   3244  32304  23412   0   0     0  1015  828  1771   9  12  79
 5  0  1  29248   3644  30276  23512 176   0   341   500  935  2358  35  18  47
 2  0  1  29248   3024  30020  23944   0   0   190   250  831  2017  34  18  47
   procs                      memory    swap          io     system         cpu
 r  b  w   swpd   free   buff  cache  si  so    bi    bo   in    cs  us  sy  id
 3  0  0  29248   3476  29900  23936   0   0     9   266  791  1754  13   6  81
 3  0  0  29248   3140  29900  24012   0   0    39     0  756  1700  11  11  78
 2  0  1  29248   3456  29780  24004   0   0     0  1347  821  1727  10  13  77
 3  0  0  29248   3264  29780  24004   0   0     0     0  762  1696  11   7  82
 2  0  0  29248   3596  29780  24080   0   0    38     0  758  1716  31  18  51
 1  0  1  29248   5564  29536  24152   0   0    45  1016  771  1734  15  10  75
 3  0  1  29248   5208  29536  24228   0   0    38     0  798  1958   6  11  82
 2  0  0  29248   5368  29536  24228   0   0     0   769  766  1788  10  12  78
 3  0  1  29248   5212  29536  24228   0   0     0     0  789  1885   9   8  83
 3  0  0  29248   5460  29536  24304   0   0    39     0  766  1869  10   8  82
 3  0  0  29248   5204  29536  24304   0   0     0  1016  773  1822  11   9  80
 3  0  1  29248   5464  29536  24304   0   0     0     0  772  1860   8  12  80
 2  0  1  29248   5056  29536  24380   0   0    38  1012  784  1808  10   9  80
 1  0  0  29248   5320  29536  24380   0   0     0     0  759  1863  10   7  83
 1  0  1  29248   4984  29536  24456   0   0    38  1014  833  1977  10  12  78
 2  0  0  29248   5308  29536  24456   0   0     0     0  769  1797  10  11  79
 3  0  1  29248   5120  29536  24456   0   0     0     0  760  1763   7  10  82
 3  0  1  29248   5300  29536  24532   0   0    38  1014  775  1722  11  13  77
 2  0  1  29248   4900  29740  24532   0   0     0     0  756  1717  10   9  80
 4  0  0  29248   4164  30672  24532   0   0     0  1015  768  1749  14  10  76
 2  0  1  29248   3076  31424  24608   0   0    39     0  758  1791   6  11  83
   procs                      memory    swap          io     system         cpu
 r  b  w   swpd   free   buff  cache  si  so    bi    bo   in    cs  us  sy  id
 3  0  1  29492   3356  32252  24052   0 122     0   781  894  2059  12  14  74
 5  0  0  29492   2860  33004  23512   0   0    52   515  784  2033  23  13  64
 4  0  0  29492   3196  33840  22512   0   0    26   500  762  1877   8  10  82
 3  0  1  29492   2852  33912  22500   0   0    38   653  765  1843  11  11  79
 2  0  1  29492   2936  34048  22476   0   0    39   500  834  1981  14  10  76
 3  0  1  29492   2728  34356  22024   0   0     0     0  750  1987   9  12  79
 4  0  1  29492   3052  34660  21632   0   0     0   517  774  1876  11  12  78
 3  0  1  29492   2528  35056  21420   0   0    38   500  766  1879  11   9  80
 2  0  0  29492   2800  35232  21232   0   0     0   266  768  1919  11  13  76
 2  0  1  29492   2872  35700  20928   0   0     0     0  750  1853   9  11  79
 2  0  1  29492   2896  35420  20884   0   0    38     0  753  1812  11  11  78
 2  0  0  29492   5076  35620  20740   0   0     0  1052  771  1804  13  11  76
 3  0  1  29492   4988  35624  20816   0   0    39     0  889  2107   2   3  94
 2  0  0  29492   4988  35624  20816   0   0     0  1012  822  1889   3   4  93
 1  0  0  29492   4988  35624  20816   0   0     0     0  836  1925   2   3  95

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: hdd & sound
  1999-08-12  1:59 hdd & sound Jonathan Masters
@ 1999-08-12 16:22 ` Jonathan Masters
  1999-08-16 15:35 ` Jonathan Masters
  1 sibling, 0 replies; 3+ messages in thread
From: Jonathan Masters @ 1999-08-12 16:22 UTC (permalink / raw)
  To: linux-sound

I noticed that too. I only have esd running on the sound card, but it
seems to do this with/without esd on /dev/dsp. Anyway, here is:

cat /proc/interupts
-------------

          CPU0
  0:    5059285          XT-PIC  timer
  1:      46017          XT-PIC  keyboard
  2:          0          XT-PIC  cascade
  4:    1320764          XT-PIC  serial
  5:   34087419          XT-PIC  soundblaster
 11:        187          XT-PIC  eth0
 12:    1547997          XT-PIC  PS/2 Mouse
 13:          1          XT-PIC  fpu
 14:     834840          XT-PIC  ide0
 15:      10788          XT-PIC  ide1
NMI:          0
ERR:          0

-------
killall esd
-------


cat /proc/interupts
-------------

           CPU0
  0:    5060624          XT-PIC  timer
  1:      46051          XT-PIC  keyboard
  2:          0          XT-PIC  cascade
  4:    1320769          XT-PIC  serial
  5:   34093966          XT-PIC  soundblaster
 11:        187          XT-PIC  eth0
 12:    1548207          XT-PIC  PS/2 Mouse
 13:          1          XT-PIC  fpu
 14:     834860          XT-PIC  ide0
 15:      10788          XT-PIC  ide1
NMI:          0
ERR:          0



and here is

cat /proc/sound:

[mastersj@periscope mastersj]$ cat /proc/sound
OSS/Free:3.8s2++-971130
Load type: Driver loaded as a module
Kernel: Linux periscope 2.2.11 #22 SMP Thu Aug 12 03:00:17 BST 1999 i686

Config options: 0

Installed drivers:

Card config:

Audio devices:
0: Sound Blaster 16 (4.16) (DUPLEX)

Synth devices:
0: AWE32-0.4.3 (RAM512k)

Midi devices:
0: Sound Blaster 16
1: AWE Midi Emu

Timers:
0: System clock

Mixers:
0: Sound Blaster


So esd makes no difference.
HOPE this helps.

Jon.


Roger Larsson wrote:

> Hi,
>
> The soundblaster!
>
> And you did not play any sound?
> During 5 s it gave 3460 interrupts! That is almost 700 interrupts / s
> (every 1.5 ms) What is going on?
>
> What sound system are you using?
> cat /proc/sound
>
> An idea is forming!
>
> Wakko could you also try (idle system):
>  cat /proc/interrupts > irq.proc
>  sleep 5
>  cat /proc/interrupts >> irq.proc     # note: append!
>
> /RogerL
>
> >            CPU0
> >   0:    3797062          XT-PIC  timer
> >   1:      41269          XT-PIC  keyboard
> >   2:          0          XT-PIC  cascade
> >   4:    1238719          XT-PIC  serial
> >   5:   25616073          XT-PIC  soundblaster
> >  11:        153          XT-PIC  eth0
> >  12:    1294293          XT-PIC  PS/2 Mouse
> >  13:          1          XT-PIC  fpu
> >  14:     340745          XT-PIC  ide0
> >  15:      10788          XT-PIC  ide1
> > NMI:          0
> > ERR:          0
> >            CPU0
> >   0:    3797565          XT-PIC  timer
> >   1:      41273          XT-PIC  keyboard
> >   2:          0          XT-PIC  cascade
> >   4:    1239076          XT-PIC  serial
> >   5:   25619533          XT-PIC  soundblaster
> >  11:        153          XT-PIC  eth0
> >  12:    1294866          XT-PIC  PS/2 Mouse
> >  13:          1          XT-PIC  fpu
> >  14:     340765          XT-PIC  ide0
> >  15:      10788          XT-PIC  ide1
> > NMI:          0
> > ERR:          0

--
Jonathan C. Masters                     (jonathan@oxlug.org)
                                        PGP: www.brookes.ac.uk/~95227860/KEY

           "Upon this rock I will build my church,
            and the gates of hell shall not prevail against it".

                 -- Matthew 16, 17-18

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: hdd & sound
  1999-08-12  1:59 hdd & sound Jonathan Masters
  1999-08-12 16:22 ` Jonathan Masters
@ 1999-08-16 15:35 ` Jonathan Masters
  1 sibling, 0 replies; 3+ messages in thread
From: Jonathan Masters @ 1999-08-16 15:35 UTC (permalink / raw)
  To: linux-sound

Exactly. Someone else agrees with me.
Oh, since you work for creative labs, any chance of getting DVD support from you
guys soon?
According to the guys at creative technical support (the one in Ireland that
everyone in the UK and Ireland goes through to), it would be best for me to "wait
for someone to reverse engineer our code than to wait for a driver" - that was very
honest now wasn't it? :) (probably, he'dbe fired very quickly if they knew). Any
official comment?

Jon.

Joe Cotellese wrote:

> I'd like to add my 2cents FWIW.
>
> 1) I don't believe it's hardware. I agree that too may people are seeing this
> with different hardware.
>
> 2) I'm not so sure it's specific to any one driver.  I've been working on the
> ES1371 driver for about a month now (this was also my first experience with
> Linux and sound) and immediately noticed the problems described in this thread.
> For me renicing didn't seem to do anything.
>
> I'm trying to play around with things in the driver to see if they make any
> difference.  If I come up with any useful information I'll post it.
>
> Regards,
> Joe C.
>
> /*******************************
> Joe Cotellese  Software Engineer
> Creative Labs, Malvern
> 610 647 3930(p)  610 647 8908(f)
> 215 868 0986(m) joec@ensoniq.com
> *******************************/
>
>
>   --
> Jonathan C. Masters                     (jonathan@oxlug.org)
>                                         PGP: www.brookes.ac.uk/~95227860/KEY
>
>            "Upon this rock I will build my church,
>             and the gates of hell shall not prevail against it".
>
>                  -- Matthew 16, 17-18
>

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~1999-08-16 15:35 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
1999-08-12  1:59 hdd & sound Jonathan Masters
1999-08-12 16:22 ` Jonathan Masters
1999-08-16 15:35 ` Jonathan Masters

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox