blbates@AERO4.LARC.NASA.GOV ("Brent L. Bates AAD/TAB MS361 x42854") (11/09/90)
I am trying to port some software that was writen for a 3130 to a 4D/VGX. Everything compiles ok, however when I try to execute it, I immediatly get a segmentation violation. Using edge to debug it, I found it dies upon the very first executable line. The line was originally ginit(); I tried changing this to winopen, but that didn't change things. Anyone have any ideas why it would die on the very first executable line? There is nothing to debug, because it hasn't done anything. -- Brent
ghelms@sgi.com (Gretchen Helms) (11/10/90)
("Brent L. Bates AAD/TAB MS361 x42854") writes: > I am trying to port some software that was writen for a 3130 to a 4D/VGX. >Everything compiles ok, however when I try to execute it, I immediatly get >a segmentation violation. Using edge to debug it, I found it dies upon the >very first executable line. The line was originally ginit(); I tried >changing this to winopen, but that didn't change things. One thing that changed from 3.2 to 3.3 was that if you had "winopen();" in your program, it would die under 3.3. Seems 3.3 prefers that you use "winopen("");" instead. One thing you definitely want to check is to make sure that your winopen has either a name like winopen("Strange problem"); or else is called like winopen(""); -- G. "Murdock" Helms The further on the edge, Silicon Graphics The hotter the intensity. Product Support Engineer ghelms@csd.sgi.com
steve@olympus.wrcr.unr.edu (Stephen Wheatcraft) (11/12/90)
In article <9011081612.AA02433@aero4.larc.nasa.gov> blbates@AERO4.LARC.NASA.GOV ("Brent L. Bates AAD/TAB MS361 x42854") writes: > > I am trying to port some software that was writen for a 3130 to a 4D/VGX. >Everything compiles ok, however when I try to execute it, I immediatly get >a segmentation violation. Using edge to debug it, I found it dies upon the >very first executable line. The line was originally ginit(); I tried >changing this to winopen, but that didn't change things. > Anyone have any ideas why it would die on the very first executable line? >There is nothing to debug, because it hasn't done anything. >-- It may not really be dying on the first line, despite what edge tells you. You can't always depend on edge (or dbxtool on Suns, which is very similar) to "tell the truth". I have had several debugging problems with both in which I was ultimately able to determine that the debugger was leading me astray. Cost me several days of wild goose chases a few times. I could go into details, but I don't think they would help in your case and would just waste bandwidth. Just remember not to put all your trust in edge - try some other debugging methods just to be sure. Steve Wheatcraft Internet: steve@olympus.wrc.unr.edu Desert Research Institute Bitnet : steve@unsvax.bitnet University of Nevada System P.O. Box 60220 AT&T : (702) 673-7393 Reno, NV 89506 FAX : (702) 673-7397
merritt@iris613.gsfc.nasa.gov (John H Merritt) (11/12/90)
In article <1990Nov9.230351.28413@odin.corp.sgi.com> ghelms@sgi.com (Gretchen Helms) writes: > >One thing that changed from 3.2 to 3.3 was that >if you had "winopen();" in your program, it would >die under 3.3. Seems 3.3 prefers that you use >"winopen("");" instead. Wrong, winopen() failed to work in earlier versions, I whould get core dumps; you were just treading on thin ice. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ John H. Merritt # Yesterday I knew nothing, Applied Research Corporation # Today I know that. merritt@iris613.gsfc.nasa.gov #