On Thu, 28 Oct 2004 17:33:50 +0100 (WEST) "Rui Nuno Capela" wrote: > OK. Here are my early consolidated results. Feel free to comment. > > 2.6.9 RT-U3 RT-V0.4.3 > --------- --------- --------- > XRUN Rate . . . . . . . . . . . 424 8 4 /hour > Delay Rate (>spare time) . . . 496 0 0 /hour > Delay Rate (>1000 usecs) . . . 940 8 4 /hour > Maximum Delay . . . . . . . . . 6904 921 721 usecs > Maximum Process Cycle . . . . . 1449 1469 1590 usecs > Average DSP CPU Load . . . . . 38 39 40 % > Average Context-Switch Rate . . 7480 8929 9726 /sec > > Note: all tests were carried out running jackd -v -dalsa -dhw:0 -r44100 > -p128 -n2 -S -P, loaded with 9 (nine) fluidsynth instances, on a > P4@2.533Ghz laptop, against the onboard sound device (snd-ali5451). > > On the RT kernels only, the following optimizations were issued: > > chrt --pid --fifo 90 2 (pidof "ksoftirqd/0" = 2) > chrt --pid --fifo 60 `pidof "IRQ 5"` (snd-ali5451) > > Take care. hi, i tried a V0.5.2 with PREEMPT_REALTIME and all debugging off (config attached). I cannot reproduce your results. I have experienced around 30 xruns in 10 minutes. And big ones, too (> 5ms). I don't know exactly what kind of load triggers them. Here's a bit of qjackctl message window (btw: jackd was idle, no clients connected, except for qjackctl): **** alsa_pcm: xrun of at least 0.594 msecs **** alsa_pcm: xrun of at least 0.037 msecs 16:20:44.832 XRUN callback (2). 16:20:44.840 XRUN callback (3). **** alsa_pcm: xrun of at least 0.024 msecs **** alsa_pcm: xrun of at least 3.682 msecs 16:22:59.834 XRUN callback (4). 16:22:59.839 XRUN callback (5). **** alsa_pcm: xrun of at least 0.016 msecs **** alsa_pcm: xrun of at least 3.460 msecs 16:23:03.513 XRUN callback (6). **** alsa_pcm: xrun of at least 2.028 msecs 16:23:03.869 XRUN callback (7). **** alsa_pcm: xrun of at least 7.061 msecs **** alsa_pcm: xrun of at least 4.510 msecs 16:23:03.894 XRUN callback (8). 16:23:03.895 XRUN callback (9). **** alsa_pcm: xrun of at least 0.428 msecs **** alsa_pcm: xrun of at least 0.157 msecs 16:23:54.546 XRUN callback (10). 16:23:54.547 XRUN callback (11). 16:23:54.549 XRUN callback (12). **** alsa_pcm: xrun of at least 1.507 msecs 16:25:49.194 XRUN callback (13). subgraph starting at qjackctl-1013 timed out (subgraph_wait_fd=14, status = 0, state = Triggered) **** alsa_pcm: xrun of at least 1.534 msecs 16:28:23.530 XRUN callback (14). **** alsa_pcm: xrun of at least 0.222 msecs **** alsa_pcm: xrun of at least 2.674 msecs **** alsa_pcm: xrun of at least 3.904 msecs 16:28:26.790 XRUN callback (15). 16:28:26.794 XRUN callback (16). 16:28:26.795 XRUN callback (17). **** alsa_pcm: xrun of at least 0.701 msecs Here's some setup info: ksoftirqd/0: mango:~# chrt -p 2 pid 2's current scheduling policy: SCHED_FIFO pid 2's current scheduling priority: 99 snd-cs46xx: mango:~# chrt -p `pidof "IRQ 3"` pid 118's current scheduling policy: SCHED_FIFO pid 118's current scheduling priority: 90 mango:~# pidof jackd 1014 mango:~# chrt -p 1014 pid 1014's current scheduling policy: SCHED_OTHER pid 1014's current scheduling priority: 0 mango:~# chrt -p 1015 pid 1015's current scheduling policy: SCHED_OTHER pid 1015's current scheduling priority: 0 mango:~# chrt -p 1016 pid 1016's current scheduling policy: SCHED_FIFO pid 1016's current scheduling priority: 70 mango:~# chrt -p 1017 pid 1017's current scheduling policy: SCHED_FIFO pid 1017's current scheduling priority: 60 Hmm, it seems i haven't disabled all debugging. This is from dmesg: BUG: atomic counter underflow at: [] dump_stack+0x1e/0x20 (20) [] qdisc_destroy+0xd9/0xe0 (28) [] dev_shutdown+0x36/0xb0 (28) [] unregister_netdevice+0x129/0x2b0 (40) [] unregister_netdev+0x19/0x40 (16) [] ppp_shutdown_interface+0x64/0xd0 [ppp_generic] (36) [] ppp_release+0x86/0x90 [ppp_generic] (16) [] __fput+0x14f/0x170 (36) [] filp_close+0x57/0x80 (28) [] sys_close+0x70/0x90 (32) [] syscall_call+0x7/0xb (-8124) Here's some context from syslog: Oct 29 15:04:01 mango pppd[324]: LCP: timeout sending Config-Requests Oct 29 15:04:01 mango pppd[324]: Connection terminated. Oct 29 15:04:01 mango kernel: BUG: atomic counter underflow at: Oct 29 15:04:01 mango kernel: [valid_stack_ptr+46/96] dump_stack+0x1e/0x20 (20) Oct 29 15:04:01 mango kernel: [unregister_netdevice+553/672] qdisc_destroy+0xd9/0xe0 (28) Oct 29 15:04:01 mango kernel: [ethtool_op_get_tso+6/32] dev_shutdown+0x36/0xb0 (28) Oct 29 15:04:01 mango kernel: [__sock_create+185/704] unregister_netdevice+0x129/0x2b0 (40) Oct 29 15:04:01 mango kernel: [blk_wait_queue_drained+25/304] unregister_netdev+0x19/0x40 (16) Oct 29 15:04:01 mango kernel: [pg0+810783028/1069765632] ppp_shutdown_interface+0x64/0xd0 [ppp_generic] (36) Oct 29 15:04:01 mango kernel: [pg0+810766502/1069765632] ppp_release+0x86/0x90 [ppp_generic] (16) Oct 29 15:04:01 mango kernel: [shmem_writepage+463/480] __fput+0x14f/0x170 (36) Oct 29 15:04:01 mango kernel: [sys_swapon+1591/2080] filp_close+0x57/0x80 (28) Oct 29 15:04:01 mango kernel: [sys_swapon+1744/2080] sys_close+0x70/0x90 (32) Oct 29 15:04:01 mango kernel: [do_notify_resume+15/76] syscall_call+0x7/0xb (-8124) Oct 29 15:04:10 mango pppoe[326]: Timeout waiting for PADS packets Oct 29 15:04:31 mango pppd[324]: Serial connection established. Oct 29 15:04:31 mango pppd[324]: Using interface ppp0 Oct 29 15:04:31 mango pppd[324]: Connect: ppp0 <--> /dev/pts/0 Oct 29 15:04:47 mango pppoe[822]: PADS: Service-Name: '' Oct 29 15:04:47 mango pppoe[822]: PPP session is 2133 Oct 29 15:04:47 mango pppd[324]: CHAP authentication succeeded Oct 29 15:04:48 mango pppd[324]: Cannot determine ethernet address for proxy ARP Oct 29 15:04:48 mango pppd[324]: local IP address 213.23.197.161 Oct 29 15:04:48 mango pppd[324]: remote IP address 145.253.4.148 Oct 29 15:04:48 mango pppd[324]: primary DNS address 195.50.140.250 Oct 29 15:04:48 mango pppd[324]: secondary DNS address 145.253.2.11 flo