Messages in this thread | | | From | Steve VanDevender <> | Date | Wed, 6 Jan 1999 23:33:00 -0800 (PST) | Subject | phew, that's better |
| |
In terms of memory management, 2.2.0pre4 seemed to throw things back to the bad old days of 2.1.129-130 where on my 32M system, the system would have a tendency to have at least half and sometimes as much as 2/3 of memory in the "cached" category. 2.2.0pre4 seemed to perform better in that state than 2.1.129 did, but somehow that still looked wrong to me, and it was apparent that a couple of large applications (for me Netscape and XEmacs) would tend to throw each other heavily into swap when switching between them.
I just got 2.2.0pre5 and it seems to have improved dramatically, or at least undone the cache-happy behavior of 2.2.0pre4. Currently only about 1/4 to 1/3 of my memory is in the "cached" state, and not nearly as much is gratuitously tossed into swap so switching between two large applications doesn't induce such heavy swapping.
Now I'm trying to figure out why 2.2.0pre5 'free' says I have only 30752k memory available when 2.2.0pre4 said I had 30816k with an identical configuration.
Jan 3 13:43:08 localhost kernel: Linux version 2.2.0-pre4 (stevev@tzadkiel) (gcc version 2.7.2.3) #26 Sun Jan 3 02:24:03 PST 1999 Jan 3 13:43:08 localhost kernel: Memory: 30776k/32768k available (916k kernel code, 408k reserved, 628k data, 40k init)
[ here 'free' used to report total mem of 30816k, I swear ]
Jan 6 22:35:58 localhost kernel: Linux version 2.2.0-pre5 (stevev@tzadkiel) (gcc version 2.7.2.3) #27 Wed Jan 6 21:28:58 PST 1999 Jan 6 22:35:58 localhost kernel: Memory: 30776k/32768k available (916k kernel code, 408k reserved, 628k data, 40k init)
$ uname -a Linux tzadkiel 2.2.0-pre5 #27 Wed Jan 6 21:28:58 PST 1999 i486 unknown $ free total used free shared buffers cached Mem: 30752 29888 864 12384 840 9632 -/+ buffers/cache: 19416 11336 Swap: 99324 10900 88424
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |