ptw@vaxine.UUCP (P. T. Withington) (05/03/84)
We have recently purchased a number of Venix's and noticed that their csh
seems to grow ad nauseum, eventually blowing out its (small/medium model) data
segment. Thought this was just a bad port on Venturcom's part, but seem to
see the same effect on our Vaxen (of course the data segment there has yet
to blow out...). Has anyone else noticed this bug/feature? No we don't have
our history set to infinity.
Additional detail: it almost looks like csh is just never freeing anything.
One quick way to blow out our Venix is to source a .cshrc full of aliases
a couple of times in a row. I do recall that csh does some funny mumbo-jumbo
in place of malloc/free...
o.o -ptw
~mats@dual.UUCP (Mats Wichmann) (05/07/84)
Our copy of CSH has never seemed to show this behavior - perhaps somebody over at UniSoft might enlighten us as to the nature of the problem, since they seem to have fixed it. On the other hand, maybe not - they may have a competitive advantage here. How `bout it, guys?
esa@kvvax4.UUCP (Esa K Viitala) (05/10/84)
>>We have recently purchased a number of Venix's and noticed that their csh >>seems to grow ad nauseum, eventually blowing out its (small/medium model) data >>segment. Thought this was just a bad port on Venturcom's part...... >The same thing seems to happen under xenix's csh, I have no reason why, >but when it runs out of space, it is not handled very gracefully.... >(i.e., core dump time) Here too, I think. We have 4.1bsd/vax11/780, and occasionally csh says 'assertion botched <something or other>' and then nothing. This comes from a memory allocation routine, I forgot the name, if either the chain of blocks has corrupted or the the chain is empty. I've concluded that it ain't corrupted. Fixes anybody? -- Esa K Viitala {decvax,philabs}!mcvax!kvport!kvvax4!esa A/S Kongsberg Vaapenfabrikk,CTG4,P.O.Box 25,N-3601 Kongsberg,Norway tel. 473-739644 tlx. 71491 vaapn n
kds@intelca.UUCP (Ken Shoemaker) (05/13/84)
>We have recently purchased a number of Venix's and noticed that their csh >seems to grow ad nauseum, eventually blowing out its (small/medium model) data >segment. Thought this was just a bad port on Venturcom's part...... The same thing seems to happen under xenix's csh, I have no reason why, but when it runs out of space, it is not handled very gracefully.... (i.e., core dump time) -- Ken Shoemaker, Intel, Santa Clara, Ca. {pur-ee,hplabs,ucbvax!amd70,ogcvax!omsvax}!intelca!kds