diff mbox series

[FFmpeg-devel] oss: account for sample size when computing timestamp

Message ID 20220601090615.52719-1-mhjacobson@me.com
State Accepted
Commit b3e261bab381f43ab5f842725d30479d511d1111
Headers show
Series [FFmpeg-devel] oss: account for sample size when computing timestamp | expand

Checks

Context Check Description
andriy/make_x86 success Make finished
andriy/make_fate_x86 success Make fate finished
andriy/make_armv7_RPi4 success Make finished
andriy/make_fate_armv7_RPi4 success Make fate finished

Commit Message

Matt Jacobson June 1, 2022, 9:06 a.m. UTC
This is my first patch to ffmpeg; please let me know if I've made any process 
errors.  Thanks!

---

Don't assume each sample is one byte in size.  Doing so results in wrong and
occasionally non-monotonically-increasing timestamps.

Fix nearby cosmetic typo.

Signed-off-by: Matt Jacobson <mhjacobson@me.com>
---
 libavdevice/oss.c     | 4 +++-
 libavdevice/oss.h     | 1 +
 libavdevice/oss_dec.c | 2 +-
 3 files changed, 5 insertions(+), 2 deletions(-)

Comments

Matt Jacobson June 15, 2022, 8:08 p.m. UTC | #1
Hi,

> On Jun 1, 2022, at 5:06 AM, Matt Jacobson <mhjacobson@me.com> wrote:
> 
> Don't assume each sample is one byte in size.  Doing so results in wrong and
> occasionally non-monotonically-increasing timestamps.
> 
> Fix nearby cosmetic typo.

Friendly ping on this patch.  Original mail:

<https://ffmpeg.org/pipermail/ffmpeg-devel/2022-June/297066.html>

Please let me know if any changes are desired.  If not, could someone please 
help me commit it?  (Is there a process to request access?)

Thank you,

Matt
Marton Balint June 19, 2022, 8:08 p.m. UTC | #2
On Wed, 15 Jun 2022, Matt Jacobson wrote:

> Hi,
>
>> On Jun 1, 2022, at 5:06 AM, Matt Jacobson <mhjacobson@me.com> wrote:
>>
>> Don't assume each sample is one byte in size.  Doing so results in wrong and
>> occasionally non-monotonically-increasing timestamps.
>>
>> Fix nearby cosmetic typo.
>
> Friendly ping on this patch.  Original mail:
>
> <https://ffmpeg.org/pipermail/ffmpeg-devel/2022-June/297066.html>
>
> Please let me know if any changes are desired.  If not, could someone please
> help me commit it?  (Is there a process to request access?)

Thanks, will apply.

Regards,
Marton
diff mbox series

Patch

diff --git a/libavdevice/oss.c b/libavdevice/oss.c
index eddc2ddf1a..b042f58875 100644
--- a/libavdevice/oss.c
+++ b/libavdevice/oss.c
@@ -102,9 +102,11 @@  int ff_oss_audio_open(AVFormatContext *s1, int is_output,
     switch(tmp) {
     case AFMT_S16_LE:
         s->codec_id = AV_CODEC_ID_PCM_S16LE;
+        s->sample_size = 2;
         break;
     case AFMT_S16_BE:
         s->codec_id = AV_CODEC_ID_PCM_S16BE;
+        s->sample_size = 2;
         break;
     default:
         av_log(s1, AV_LOG_ERROR, "Soundcard does not support 16 bit sample format\n");
@@ -112,7 +114,7 @@  int ff_oss_audio_open(AVFormatContext *s1, int is_output,
         return AVERROR(EIO);
     }
     err=ioctl(audio_fd, SNDCTL_DSP_SETFMT, &tmp);
-    CHECK_IOCTL_ERROR(SNDCTL_DSP_SETFMTS)
+    CHECK_IOCTL_ERROR(SNDCTL_DSP_SETFMT)
 
     tmp = (s->channels == 2);
     err = ioctl(audio_fd, SNDCTL_DSP_STEREO, &tmp);
diff --git a/libavdevice/oss.h b/libavdevice/oss.h
index 66d1a34cf6..f1da2b1bec 100644
--- a/libavdevice/oss.h
+++ b/libavdevice/oss.h
@@ -30,6 +30,7 @@  typedef struct OSSAudioData {
     AVClass *class;
     int fd;
     int sample_rate;
+    int sample_size; /* in bytes ! */
     int channels;
     int frame_size; /* in bytes ! */
     enum AVCodecID codec_id;
diff --git a/libavdevice/oss_dec.c b/libavdevice/oss_dec.c
index d3dbe77cf9..2cdc4324e8 100644
--- a/libavdevice/oss_dec.c
+++ b/libavdevice/oss_dec.c
@@ -91,7 +91,7 @@  static int audio_read_packet(AVFormatContext *s1, AVPacket *pkt)
         bdelay += abufi.bytes;
     }
     /* subtract time represented by the number of bytes in the audio fifo */
-    cur_time -= (bdelay * 1000000LL) / (s->sample_rate * s->channels);
+    cur_time -= (bdelay * 1000000LL) / (s->sample_rate * s->sample_size * s->channels);
 
     /* convert to wanted units */
     pkt->pts = cur_time;