Re: mmap() thing. (2.1.44-pre2)

Matthew Kirkwood (matthew.kirkwood@lmh.ox.ac.uk)
Thu, 26 Jun 1997 01:51:18 +0100 (BST)


My disks have just finished fscking after that mapsort thing killed
my machine :-(

For what it's worth, I can report that on my PC [1] I could
indeed reproduce this rather antisocial behaviour.

Running the program annouces that:
File "/var/tmp/mapfile_test" with 48761856 bytes will be created.

In itself this wouldn't be a problem, as I have 55Mb free, but the
thing soon starts to swap anyway.

procinfo (when _will_ we get a new top which works?) announces that
47Mb is being cached, which seems fair, but then, the machine starts
to eat up swap. Swap increases over a minute or so, hitting about
15Mb when the machine finally dies. VC switching is still OK, but
that's all I could get from the thing. (Forgot to try Alt-Scroll,
etc, but I'll recompile with the magic SysReq thing as well before
retrying.)

Matthew.

[1] P120, 64Mb EDO, everything mounted (including 64Mb swap)
on hda, s/w mostly RH4.1.

--
Matthew Kirkwood  |  Mail:  weejock@ferret.lmh.ox.ac.uk
LMH JCR,          |  Web:   http://www-jcr.lmh.ox.ac.uk/~weejock/
Oxford OX2 6QA,   |  PGP:   finger weejock@ferret.lmh.ox.ac.uk
England.          |