Subject: interesting bug
To: None <>
List: current-users
Date: 01/17/1994 02:30:35

	well, this certainly is interesting.

	a while back, i reported an interesting problem in which
	under certain circumstances, ps and top would report that
	processes had waaaaaaaay more virtual memory than they
	should have.  subsequant looks at swapinfo and the like
	seemed to indicate they they weren't really using the amount
	of swap space that ps and top said they were.

	i received mail a while back saying it was fixed.  however,
	i managed to trip it again tonight on -current as of jan

	basically, any time i use "limit stacksize " on csh or tcsh, any process
	created as a subprocess of that shell [or subrpocess of a subprocess
	etc...] reports that it has the same amount of vm as the 
	limit i set for the stacksize.


	[csh] % limit stacksize 4m
	[csh] % jove
	[csh] % vi 
	[csh] % vmstat 1

	all three processes will report having 4m virtual sizes, with
	normal res sizes.  setting the limit to 8m -or anything greater-
	than 8m causes them to report 8m for some rason.  setting
	it at 2m causes them all to have 2m virt size under ps and top

	[side note---i can -never- set the stacksize greater than

	as soon as i exit the shell, and go back to a parent shell on
	the tty, problem goes away.

	no idea where the bug is or what causes it tho' :-)

							off to bed [i'm sick]
							marc 'em.

Marc Wandschneider					    Seattle, WA
Barney the Dinosaur sings! You faint... Barney sings!  Barney sings! --More--
You Die... --More--