home *** CD-ROM | disk | FTP | other *** search
- Path: news.iddis.com!jarnot
- From: jarnot@iddis.com (Kevin J. Jarnot)
- Newsgroups: comp.unix.questions,comp.lang.c,comp.unix.programming
- Subject: Re: SIGBUS - What are the possible causes?
- Date: 13 Jan 1996 05:08:00 GMT
- Organization: IDD Information Services
- Message-ID: <JARNOT.96Jan13000800@box10.dstar.iddis.com>
- References: <4ct4es$3k7@news1.wolfe.net>
- Reply-To: jarnot@iddis.com
- NNTP-Posting-Host: box10.dstar.iddis.com
- To: mcguire@wolfe.net (McGuire)
- In-reply-to: mcguire@wolfe.net's message of Tue, 09 Jan 1996 07:11:16 GMT
-
- >>>>> "McGuire" == McGuire <mcguire@wolfe.net> writes:
-
- McGuire> Hello all, I am trying to compile a program under gcc
- McGuire> 2.7.0 that works under the native HP-UX 9.05 cc. In one
- McGuire> routine I get a SIGBUS error. When I ignore the SIGBUS
- McGuire> error, I get a segmentation violation. When I add some
- McGuire> code to test what might be happening, the location at
- McGuire> which the error occurs appears to change (I am using gdb
- McGuire> under xemacs). Does anyone have any ideas as to what the
- McGuire> possible causes of such an error might be.
-
-
- Generally, a SIGBUS is the result of unaligned access to a memory
- location (ie. grabbing a long from the target address+1). I wouldn't
- ignore the signal, just as I wouldn't ignore a SIGSEGV.
-
- I'll bet that memory has become munged, making it look like the
- error is occurring all over the place. You may have to go through
- the code (probably somewhere near where you see the majority
- of errors) by hand.
-
- Kevin
-
-
-
- --
- Kevin J. Jarnot - Senior Software Engineer | mailto:jarnot@iddis.com
- IDD Information Services Waltham, MA | http://www-cs.canisius.edu/~jarnot
-