diff mbox

snd_interval_refine_first/last: exclude value only if also excluded before

Message ID 1530543186-20962-1-git-send-email-twischer@de.adit-jv.com (mailing list archive)
State New, archived
Headers show

Commit Message

Timo Wischer July 2, 2018, 2:53 p.m. UTC
From: Timo Wischer <twischer@de.adit-jv.com>

Without this commit the following intervals [x y), (x y) were be
replaced to (y-1 y) by snd_interval_refine_last(). This was also done if
y-1 is part of the previous interval.
With this changes it will be replaced with [y-1 y) in case of y-1 is
part of the previous interval. A similar behavior will be used for
snd_interval_refine_first().

This solves the issue reported here:
https://bugzilla.opensuse.org/show_bug.cgi?id=1033179
and work arounded with commit
e736715 ("pcm: dmix: Disable var_periodsize as default").

I am able to reproduce the issue with a simplified aplay use case using
the following configuration:
pcm_slave.adr3_tdm_8ch {
    pcm {
        type hw
        card "Loopback"
        device 0
    }
    rate 48000
    period_size 128
    buffer_size 1024
    channels 2
}

pcm.dshare_Playback_3 {
    type dmix
    ipc_key 600
    ipc_perm 0660
    ipc_gid audio
    var_periodsize true
    slave adr3_tdm_8ch
}

pcm.AdevAcousticoutSpeech {
    type rate
    slave.pcm dshare_Playback_3
    slave.rate 48000
}

$ modprobe snd_aloop
$ aplay -v --period-size=352 -c2 -fS16_LE -r22500 -DAdevAcousticoutSpeech /dev/urandom
...
Rule 9 (0xffff91d1f230): PERIODS=(0 2) -> NONE BUFFER_SIZE=480 PERIOD_SIZE=[240 240]
refine_soft 'AdevAcousticoutSpeech' (end--22)
...
aplay: ../../alsa-utils-1.1.5/aplay/aplay.c:1390: set_params: Assertion `err >= 0' failed.
Aborted by signal Aborted...

The following stack trace shows where the -EINVAL will be thrown:
__snd_pcm_hw_params_set_period_size_near()
snd1_pcm_hw_param_set_near()
snd1_pcm_hw_param_set_last()
snd1_pcm_hw_refine_slave()
snd1_pcm_hw_refine_soft()
snd_pcm_hw_rule_div()
snd1_interval_refine()

This issue exists due to PERIODS does not include 2
Rule 9 (0xffff91d1f230): PERIODS=(0 9) -> (0 2) BUFFER_SIZE=[120 480]
PERIOD_SIZE=(240 241)
because of an invalid integer inverval of PERIOD_SIZE of (240 241).
This interval is set by snd_interval_refine_last().

Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
---
 src/pcm/interval.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

Comments

Takashi Iwai July 3, 2018, 10:48 a.m. UTC | #1
On Mon, 02 Jul 2018 16:53:06 +0200,
<twischer@de.adit-jv.com> wrote:
> 
> From: Timo Wischer <twischer@de.adit-jv.com>
> 
> Without this commit the following intervals [x y), (x y) were be
> replaced to (y-1 y) by snd_interval_refine_last(). This was also done if
> y-1 is part of the previous interval.
> With this changes it will be replaced with [y-1 y) in case of y-1 is
> part of the previous interval. A similar behavior will be used for
> snd_interval_refine_first().
> 
> This solves the issue reported here:
> https://bugzilla.opensuse.org/show_bug.cgi?id=1033179
> and work arounded with commit
> e736715 ("pcm: dmix: Disable var_periodsize as default").
> 
> I am able to reproduce the issue with a simplified aplay use case using
> the following configuration:
> pcm_slave.adr3_tdm_8ch {
>     pcm {
>         type hw
>         card "Loopback"
>         device 0
>     }
>     rate 48000
>     period_size 128
>     buffer_size 1024
>     channels 2
> }
> 
> pcm.dshare_Playback_3 {
>     type dmix
>     ipc_key 600
>     ipc_perm 0660
>     ipc_gid audio
>     var_periodsize true
>     slave adr3_tdm_8ch
> }
> 
> pcm.AdevAcousticoutSpeech {
>     type rate
>     slave.pcm dshare_Playback_3
>     slave.rate 48000
> }
> 
> $ modprobe snd_aloop
> $ aplay -v --period-size=352 -c2 -fS16_LE -r22500 -DAdevAcousticoutSpeech /dev/urandom
> ...
> Rule 9 (0xffff91d1f230): PERIODS=(0 2) -> NONE BUFFER_SIZE=480 PERIOD_SIZE=[240 240]
> refine_soft 'AdevAcousticoutSpeech' (end--22)
> ...
> aplay: ../../alsa-utils-1.1.5/aplay/aplay.c:1390: set_params: Assertion `err >= 0' failed.
> Aborted by signal Aborted...
> 
> The following stack trace shows where the -EINVAL will be thrown:
> __snd_pcm_hw_params_set_period_size_near()
> snd1_pcm_hw_param_set_near()
> snd1_pcm_hw_param_set_last()
> snd1_pcm_hw_refine_slave()
> snd1_pcm_hw_refine_soft()
> snd_pcm_hw_rule_div()
> snd1_interval_refine()
> 
> This issue exists due to PERIODS does not include 2
> Rule 9 (0xffff91d1f230): PERIODS=(0 9) -> (0 2) BUFFER_SIZE=[120 480]
> PERIOD_SIZE=(240 241)
> because of an invalid integer inverval of PERIOD_SIZE of (240 241).
> This interval is set by snd_interval_refine_last().
> 
> Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>

Thanks for spotting this bug.  Applied now.

We need the similar fix in the kernel code, too...


Takashi
diff mbox

Patch

diff --git a/src/pcm/interval.c b/src/pcm/interval.c
index 95cedd5..ef4c2ed 100644
--- a/src/pcm/interval.c
+++ b/src/pcm/interval.c
@@ -200,27 +200,33 @@  int snd_interval_refine(snd_interval_t *i, const snd_interval_t *v)
 
 int snd_interval_refine_first(snd_interval_t *i)
 {
+	const unsigned int last_max = i->max;
+
 	if (snd_interval_empty(i))
 		return -ENOENT;
 	if (snd_interval_single(i))
 		return 0;
 	i->max = i->min;
-	i->openmax = i->openmin;
-	if (i->openmax)
+	if (i->openmin)
 		i->max++;
+	/* only exclude max value if also excluded before refine */
+	i->openmax = (i->openmax && i->max >= last_max);
 	return 1;
 }
 
 int snd_interval_refine_last(snd_interval_t *i)
 {
+	const unsigned int last_min = i->min;
+
 	if (snd_interval_empty(i))
 		return -ENOENT;
 	if (snd_interval_single(i))
 		return 0;
 	i->min = i->max;
-	i->openmin = i->openmax;
-	if (i->openmin)
+	if (i->openmax)
 		i->min--;
+	/* only exclude min value if also excluded before refine */
+	i->openmin = (i->openmin && i->min <= last_min);
 	return 1;
 }