home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!olivea!sgigate!psinntp!psinntp!isc-newsserver!rbm2273
- From: rbm2273@ultb.isc.rit.edu (R.B. Mead )
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: SKsh 2.0 and ixemul.library/GCC 2.2.2
- Message-ID: <1992Aug23.181440.4922@ultb.isc.rit.edu>
- Date: 23 Aug 92 18:14:40 GMT
- References: <1992Aug23.143234.2461@ultb.isc.rit.edu>
- Sender: news@ultb.isc.rit.edu (USENET News System)
- Organization: Rochester Institute of Technology
- Lines: 38
- Originator: rbm2273@ultb
- Nntp-Posting-Host: ultb-gw.isc.rit.edu
-
- In article <1992Aug23.143234.2461@ultb.isc.rit.edu> kxs5829@ultb.isc.rit.edu (K.X. Saunders ) writes:
- > There appears to be a curious interaction between SKsh 2.0 and
- >programs compiled with Markus Wild's GCC 2.2.2 port, which all use
- >ixemul.library. It seems that an ixemul.library using program will lose
- >72 bytes of fast mem when started from SKsh. This does not occur when the
- >program is started from the standard 2.0 shell. This also does not occur
- >if a non-ixemul.library using program is started from SKsh 2.0.
- >
- > Maybe there is something to the fact that SKsh uses System() and
- >the standard 2.0 shell uses RunCommand().
- >
-
- Strange number .. this 72! I just noticed, about 2 hours ago, that every
- time I remake GNU Chess V4 (just the link) I lose *** 72 bytes ***.
-
- It happens only when invoked via:
-
- Make 1.5 Dec 25 1991 12:24:46
- Copyright ) 1991 by Ben Eng, All Rights Reserved
-
- I'm using the standard makefile included in the distribution, with
- minor changes in -D directives. I'm also using GCC V2.2.2 and, of
- course, the IXEMUL library (thanks Markus).
-
- If I just run the link via gcc (exactly as bmake reported) there is
- no memory loss.
-
- If I "bmake gnuchessr", I lose 72 bytes, each time a link is required.
-
- This comes right on the heels of a really silly mistake I made recently
- regarding a memory leak... which turned out to be "operator error"!
-
- Humm... what could be the trouble? I was assuming bmake might be the
- trouble and was about to try another make program.
-
- Regards. [ rbm2273@ultb.isc.rit.edu ]
-
- ps - I'm using "avail flush" when checking for lost memory.
-