On Sun, 15 Feb 1998, Andrea Arcangeli wrote:
> I find out that the lock happen inside DMAbuf_getwrbuffer() called from
> audio_write(). The mp3 file I am playing is "112 kbit/s, 44100 Hz
> joint-stereo". Are you playing the same file format? The file is also 4/5
> Mbyte long...
Okay, I just tried it again and - again - no problem. I know now, that there
are 51682 Files on my harddisk, and mpeg-files of all sampling rates are
playing parallel to the find command (I emptied the disc cache reading a big
file from disc between the tests). I tried different harddisk activity and
some network traffic, too. This was tested on a dual pentium board with SMP
enabled.
But, during the tests I discovered the following problem: the following
command makes the system unuseable (64 Megabytes available, about 52 MB
free)
python:~>ll /dev/urandom /dev/null
crw-rw-rw- 1 root root 1, 3 Jan 1 1980 /dev/null
crw-r--r-- 1 root root 1, 9 Dec 17 1995 /dev/urandom
python:~>id
uid=3015(carsten) gid=101(knwu)
python:~>dd if=/dev/urandom of=/dev/null bs=43000000 count=2
The process 'dd' grows slowly, is unkillable and eats all available cpu time
(load avg. exactly 2.00, ps means state "Running", but no chance to kill or
stop it). This was done as ordinary user, not as root. Kernel was 2.1.86,
compiled for and using smp.
Regards
Carsten
-- Carsten Gross Internet: carsten@sol.wohnheim.uni-ulm.de Wohnheim Heilmeyersteige: Sebastian Kneipp Weg 6, 89075 Ulm
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu