On Sat, 13 Mar 2010, VDR User wrote:

[pid 9629] 1268511516.215224 ioctl(18, 0x400c4150, 0xb43631a4) = 0

^^^^ here xine feeds new samples to the ALSA driver (0x400c4150 == SNDRV_PCM_IOCTL_WRITEI_FRAMES)

[pid 9629] 1268511516.215275 ioctl(18, 0x80044121, 0xb43632cc) = 0 [pid 9627] 1268511520.700109 ioctl(10, 0xc0104652, 0xb5443f00) = 0 [pid 9629] 1268511520.700340 ioctl(18, 0x80044121, 0xb43632cc) = -1 EPIPE (Broken pipe)

^^^^ next iteration (0x80044121 == SNDRV_PCM_IOCTL_DELAY) - see time difference - 4.485 seconds from last WRITEI

It seems like a xine bug or the system scheduler bug or any kernel bug affecting the system scheduler. The ALSA subsystem is not reponsible for this behaviour - the ring buffer in your case is for about 0.372 seconds of samples. This limit must not be crossed for the click-free playback.

Ask the xine developers for more help or try another player.

Jaroslav

----- Jaroslav Kysela perex@perex.cz Linux Kernel Sound Maintainer ALSA Project, Red Hat, Inc.