Dne 28. 07. 20 v 20:04 Pavel Hofman napsal(a):
Dne 28. 07. 20 v 19:04 Takashi Iwai napsal(a):
On Tue, 28 Jul 2020 18:46:00 +0200, Pavel Hofman wrote:
The debug around the event looks like this, suddenly meter->now drops down:
s16_update 1: meter->now 2567498, s16->old 2566593, size 905 s16_update 1: meter->now 2568401, s16->old 2567498, size 903 s16_update 1: meter->now 20786, s16->old 20786, size 0 s16_update 1: meter->now 1065, s16->old 20786, size -19721 s16_update 1: meter->now 24839, s16->old 24838, size 1 s16_update 1: meter->now 701, s16->old 24839, size -24138 s16_update 1: meter->now 1253162, s16->old 701, size 1252461 s16_update 1: meter->now 1255148, s16->old 1253162, size 1986
..........
s16_update 1: meter->now 11136, s16->old 10261, size 875 s16_update 1: meter->now 22525, s16->old 22524, size 1 s16_update 1: meter->now 963, s16->old 22525, size -21562 s16_update 1: meter->now 1270914, s16->old 963, size 1269951 s16_update 1: meter->now 1272917, s16->old 1270914, size 2003
I think the problem is that s16->old is not reset when status.appl_ptr is zeroed and starts running again. There is a call
static void s16_reset(snd_pcm_scope_t *scope) { snd_pcm_scope_s16_t *s16 = scope->private_data; snd_pcm_meter_t *meter = s16->pcm->private_data; s16->old = meter->now; }
but I do not know when this method is called and whether the meter->now is already assigned to the newly zeroed status.appl_ptr.
Would adding atomic_add(&meter->reset, 1) in snd_pcm_meter_reset() help?
Unfortunately not.
s16_reset is called correctly, setting s16->old = meter->now; But at that time meter->now is still 22751, setting s16->old to the same value.
s16_update 1: meter->now 22751, s16->old 22751, size 0
However, in the next update call meter->now comes from the freshly started application pointer:
s16_update 1: meter->now 839, s16->old 22751, size -21912
Of course this helps:
if (size < 0)
size += spcm->boundary;
if (size < 0) {
size = meter->now;
s16->old = 0;
}
But I understand this is not a solution because:
- it will not work at reaching spcm->boundary (after thousands of hours?)
- it will cause the same problem when the stream is rewound (which is
the problem now too) - size will equal to large meter->now (length from the beginning of the stream minus the rewound = large number).
IMHO there are two cases of the [application pointer + delay] drop compared to the previous run:
* stream start, rewinding => s16->old = meter->now; size =0, i.e. skipping the samples to show * wrapping at spcm->boundary => size += spcm->boundary, i.e. showing the wrapped samples
Optionally the second case could be handled just like the first case by resetting s16->old, assuming the boundary wrap occurs very infrequently.
Pavel.