dvadura@watdragon.waterloo.edu (Dennis Vadura) (02/09/89)
Machine DN3500, SYSTYPE=bsd4.3, COMPILER=/bin/cc, OS=SR10.1
Trying to compile gcc-1.33, while compiling stmt.c, I get the following:
Fatal error in /usr/apollo/lib/cc
Status 023
*** Exit 19
The compiler does make it through the 4000 line source file (I put a syntax
error at the end which it found), so the error is comming from a later pass.
Command line used is:
cc -g -O -U__STDC__ -DSHORT_ENUM_BUG -c stmt.c
If anyone has seen this and has found a fix I would be glad to hear from you.
If this is truly an internal bug, then could someone from Apollo indicate
what causes the above message so that a workaround can be found, I need gcc
yesterday. I did go through the C language reference and did not find any
mention of Fatal errors, perhaps this could be added to the manual, with some
indication of what the cause is so that I and others could conceive a work
around, but not having any hints as to the cause is totally useless.
-thanks
-dennis
P.S> Is anyone keeping a list of bugs in 10.1, I'm building up quite a nice
collection. The best one so far: dbx appears to be useful for
debugging *correct* programs only :-).
--
--------------------------------------------------------------------------------
Dennis Vadura, Computer Science Dept., UUCP,BITNET: dvadura@water
University of Waterloo, Waterloo, Ont. EDU,CDN,CSNET: dvadura@waterloo
================================================================================markg@CAEN.ENGIN.UMICH.EDU (Mark Giuffrida) (02/10/89)
Machine DN3500, SYSTYPE=bsd4.3, COMPILER=/bin/cc, OS=SR10.1 Trying to compile gcc-1.33, while compiling stmt.c, I get the following: Fatal error in /usr/apollo/lib/cc Status 023 *** Exit 19 Yes. It really has nothing to do with your code. I have seen it happen on a 10 line source file. I don't know how to explain it, but I do know a fix. The fix is to open a *new* shell and recompile it. Something in the environment of your old shell is causing it to throw up. I used to notice this a lot in presr10.1. I run sr10.1 and have rarely ever seen it crop up. Mark Giuffrida University of Michigan markg@caen.engin.umich.edu
pha@CAEN.ENGIN.UMICH.EDU (Paul H. Anderson) (02/10/89)
Regarding bugs in compilers or opsystems: If you are able to send mail to the apollo mailing list about bugs, you should consider sending APR's to apollo directly via email. Read the man page on mkapr (sr10.0 or later only). It contains the e-mail address, and enough information on using the program to make it a quite painless process. I have done this several times, and it seems to work reasonably well, although you do have to be patient awaiting a reply (via US Postal mail). Paul Anderson CAEN
dvadura@watdragon.waterloo.edu (Dennis Vadura) (02/13/89)
In article <11392@watdragon.waterloo.edu> dvadura@watdragon.waterloo.edu (Dennis Vadura) writes: > >Machine DN3500, SYSTYPE=bsd4.3, COMPILER=/bin/cc, OS=SR10.1 >Trying to compile gcc-1.33, while compiling stmt.c, I get the following: > > Fatal error in /usr/apollo/lib/cc > Status 023 > >Command line used is: > cc -g -O -U__STDC__ -DSHORT_ENUM_BUG -c stmt.c Well, a number of people have suggested various solutions to this problem, thanks to all those that responded, here is the fix. *DON'T mix -g and -O options on the command line* Other suggestions included opening a new DM window/pad, or freeing disk space. Doing it in a new window produced a consistent error, so that wasn't it, and I have apx 170 Megs of free disk space, and if this thing ran out of space with that much free, it would be on it's way to the land of dead workstations. The reason I'm posting this is that you don't always get the error, *some* files get compiled. Almost all sane compilers that I know of don't allow the mixing of -g and -O, and print a warning, that they're using -g only. Lot's of makefiles that I have seen make this assumption when giving flags for debugging versions of programs, so I figure somebody out there could be interested in this. This might also explain my dbx problem with another program. It too was compiled with -g and -O both set, and the result was that it put dbx into a tail spin. -sigh (I hope mkapr takes < 5 min) -dennis -- -------------------------------------------------------------------------------- Dennis Vadura, Computer Science Dept., UUCP,BITNET: dvadura@water University of Waterloo, Waterloo, Ont. EDU,CDN,CSNET: dvadura@waterloo ================================================================================