[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <66c81c16-737d-4ec6-b590-a6b1070b991b@sirena.org.uk>
Date: Thu, 13 Jul 2023 21:39:25 +0100
From: Mark Brown <broonie@...nel.org>
To: Takashi Iwai <tiwai@...e.de>
Cc: "NĂcolas F. R. A." Prado
<nfraprado@...labora.com>, Jaroslav Kysela <perex@...ex.cz>,
kernel@...labora.com,
AngeloGioacchino Del Regno
<angelogioacchino.delregno@...labora.com>,
Shuah Khan <shuah@...nel.org>, Takashi Iwai <tiwai@...e.com>,
alsa-devel@...a-project.org, linux-kernel@...r.kernel.org,
linux-kselftest@...r.kernel.org
Subject: Re: [PATCH 2/2] kselftest/alsa: pcm-test: Decrease stream duration
from 4 to 2 seconds
On Thu, Jul 13, 2023 at 10:47:43AM +0200, Takashi Iwai wrote:
> Mark Brown wrote:
> > # default.time3.0.0.0.PLAYBACK - 44.1kHz stereo large periods
> > # default.time3.0.0.0.PLAYBACK hw_params.RW_INTERLEAVED.S16_LE.44100.2.16383.131064 sw_params.131064
> > not ok 10 default.time3.0.0.0.PLAYBACK
> > # time mismatch: expected 2000ms got 2229
> > reliably (the actual time drifts by a few ms). The other boards I've
> > got coverage of seem fine, and I didn't check any broader CI yet.
> Interesting. With the current patch, we rather extended the margin in
> proportion; formerly 4 sec +/- 0.1s, now 2 sec +/- 0.1s. And it
> exceeded out of sudden.
Right.
> I guess this rather caught a problem of the driver itself.
Well, there's doubtless something driver/hardware related going on but
I'm not sure if it's a problem there or not. The results from the 4s
runtime were:
# default.time3.0.0.0.PLAYBACK - 44.1kHz stereo large periods
# default.time3.0.0.0.PLAYBACK hw_params.RW_INTERLEAVED.S16_LE.44100.2.16383.131064 sw_params.131064
ok 10 default.time3.0.0.0.PLAYBACK
so the same buffer sizes and so on, and the period size is only 10ms
unless I miscalculated which should be quite a long way off from the
100ms of margin we give ourselves. It does seem a little suspect that
it's the large periods test that's falling over though.
Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)
Powered by blists - more mailing lists