[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-rt
Subject:    [rtl] mlockall broken ?
From:       j_clifton () aerosimulation ! com
Date:       1999-09-24 15:33:42
[Download RAW message or body]


I never got an answer to my previous query. I wrote a test program which mallocs
memory and does mlockall until mlockall fails at the arbitrary 50% of physical
memory limit. I use memset to dirty up the memory and then go into a spin loop
calling rand() to insure I stay in memory. The problem is mlockall claims that
the memory is resident, locked, not pagable yet I can load more copies of the
program than there is physical memory. Below is a snippet from top:

62 processes: 58 sleeping, 4 running, 0 zombie, 0 stopped
CPU states: 111.4% user, 11.9% system,  0.0% nice,  0.0% idle
Mem:   63136K av,  61880K used,   1256K free,  27072K shrd,  20504K buff
Swap:      0K av,      0K used,      0K free                 19872K cached

  PID USER     PRI  NI  SIZE  RSS SHARE STAT  LIB %CPU %MEM   TIME COMMAND
11096 root      12   0 32600  31M   632 R       0 39.1 51.6   0:09 sparemem
11098 root      10   0 32600  31M   632 R       0 35.7 51.6   0:05 sparemem
11105 root      10   0 32600  31M   632 R       0 34.0 51.6   0:04 sparemem
11118 jon        8   0   716  716   552 R       0 13.6  1.1   0:00 top

Another weirdness is that the memory test will make my application disappear
from the top display even though I have the application set to priority 98.


--- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail majordomo@rtlinux.cs.nmt.edu OR
echo "unsubscribe rtl <Your_email>" | mail majordomo@rtlinux.cs.nmt.edu
----
For more information on Real-Time Linux see:
http://www.rtlinux.org/~rtlinux/

[prev in list] [next in list] [prev in thread] [next in thread]