All of lore.kernel.org
 help / color / mirror / Atom feed
From: Lee Revell <rlrevell@joe-job.com>
To: Remon Sijrier <remon@vt.shuis.tudelft.nl>
Cc: linux-kernel <linux-kernel@vger.kernel.org>
Subject: Re: Losing interrupts
Date: Fri, 16 Jul 2004 17:00:48 -0400	[thread overview]
Message-ID: <1090011648.30657.9.camel@mindpipe> (raw)
In-Reply-To: <200407161116.38493.remon@vt.shuis.tudelft.nl>

On Fri, 2004-07-16 at 05:16, Remon Sijrier wrote:
> Hello,
> 
> First, thanks a lot for the work done and still done for tackling the latency 
> issues in the kernel.
> 
> I'm interested in this area, and want to do some testing as well.
> Lee Revel, could you please sent me the changes you made for measuring 
> interrupt times to me?
> 

It's pretty sloppy, but it is just debugging code.  This patch is for
emu10k1, you will have to adapt this to your driver (ice1712). 
Basically you just add a couple of fields to the card record to track
the time the last interrupt occurred, and the time elapsed between the
previous two interrupts.  Then, if you launch a program like JACK that
should cause the soundcard interrupts to happen at a fixed interval, you
can use some heuristics to figure out whether you have lost an
interrupt.

Also the fscking tabs below have been mangled to spaces, I am not sure
if this is caused by copying from xterm, or pasting into Evolution, but
I am not too happy about it.

Index: include/emu10k1.h
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/include/emu10k1.h,v
retrieving revision 1.43
diff -u -r1.43 emu10k1.h
--- include/emu10k1.h   1 Jul 2004 09:22:16 -0000       1.43
+++ include/emu10k1.h   16 Jul 2004 20:54:16 -0000
@@ -1003,6 +1003,9 @@
        emu10k1_midi_t midi2; /* for audigy */
 
        unsigned int efx_voices_mask[2];
+
+       cycles_t last_interrupt_time;
+       int last_delay;
 };
 
 int snd_emu10k1_create(snd_card_t * card,
Index: pci/emu10k1/emupcm.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/pci/emu10k1/emupcm.c,v
retrieving revision 1.29
diff -u -r1.29 emupcm.c
--- pci/emu10k1/emupcm.c        1 Jul 2004 09:22:16 -0000       1.29
+++ pci/emu10k1/emupcm.c        16 Jul 2004 20:51:02 -0000
@@ -37,6 +37,26 @@
 static void snd_emu10k1_pcm_interrupt(emu10k1_t *emu, emu10k1_voice_t *voice)
 {
        emu10k1_pcm_t *epcm;
+       cycles_t then;
+       cycles_t now;
+       int delay;
+       int jitter;
+       
+       then = emu->last_interrupt_time;
+       now = get_cycles ();
+       emu->last_interrupt_time = now;
+       
+       delay = now - then;
+       jitter = abs( delay - emu->last_delay );
+
+       if (jitter > 50000) {
+           printk("IRQ: delay =  %i cycles, jitter = %i\n", delay,  jitter);
+       }
+        if (jitter * 2 > emu->last_delay ) {
+           printk("IRQ: delay =  %i cycles, jitter = %i - missed an interrupt?\n", delay,  jitter);
+       }
+           
+       emu->last_delay = delay;
 
        if ((epcm = voice->epcm) == NULL)
                return;
@@ -816,6 +836,8 @@
        mix->send_volume[1][0] = mix->send_volume[2][1] = 255;
        mix->attn[0] = mix->attn[1] = mix->attn[2] = 0xffff;
        mix->epcm = epcm;
+       emu->last_interrupt_time = 0;
+       emu->last_delay = 0;
        snd_emu10k1_pcm_mixer_notify(emu, substream->number, 1);
        return 0;
 }

Lee


  reply	other threads:[~2004-07-16 21:00 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-07-16  9:16 Losing interrupts Remon Sijrier
2004-07-16 21:00 ` Lee Revell [this message]
2004-07-17 13:29   ` Remon Sijrier
  -- strict thread matches above, loose matches on Subject: below --
2004-07-14 22:19 Lee Revell
2004-07-15 21:36 ` Lee Revell
2004-07-15 21:04   ` Andrew Morton
2004-07-16  3:24 ` Lee Revell

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1090011648.30657.9.camel@mindpipe \
    --to=rlrevell@joe-job.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=remon@vt.shuis.tudelft.nl \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.