home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.lang.perl
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!uwm.edu!linac!newsaintmail
- From: crawdad@fncent (Matt Crawford)
- Subject: SunOS 4.1 cc -O loses on perl (4.033) util.c
- Message-ID: <0MQ+CY9O2@linac.fnal.gov>
- Sender: daemon@linac.fnal.gov (The Background Man)
- Nntp-Posting-Host: fncent.fnal.gov
- Organization: FERMILAB, Batavia, IL
- Date: Wed, 23 Dec 1992 16:02:59 GMT
- Lines: 24
-
- I had installed perl 4.0p33 at work under SunOS 4.1.2 and I attempted
- to do the same at home under 4.1. (CPU = 4/75 and 4/60, resp.)
- "Make test" bombed instantly with a core dump in fdopen(). It turned
- out that the optimizer, when working on the fragment
-
- if (p[THIS] != (*mode == 'r')) {
- dup2(p[THIS], *mode == 'r');
- close(p[THIS]);
- }
-
- in mypopen(), clobbered the value of mode with a "ldsb [%i1], %i1"
- instruction. On the last line of mypopen
-
- return fdopen(p[this], mode);
-
- fdopen was passed 'r' instead of "r".
-
- When I compile util.c without "-O", all is well.
-
- Two questions: Does patch 34 or 35 fix this? Where can I find those
- patches? I already looked on devvax.jpl.nasa.gov, ftp.uu.net, and
- convex.com.
- _________________________________________________________
- Matt Crawford crawdad@fncent.fnal.gov Fermilab
-