gdb: incomplete backtrace, strange step(?) in C++ (egcs 1.1.2)

From: Achim Bohnet <ach_at_mpe.mpg.de>
Date: Tue, 25 Jul 2000 13:22:18 +0200

Hello,

we use(d) gdb 4.17,4,18 and 5.0 together with egcs 1.1.2 on
V4.0E + duv40eas0004 jumbo patch. Debugging C++ code for, e.g, SEGV
or FPE shows incomplete backtraces. Most of the
time 'bt' lists a Constructor and the methods called from there.
But all frames that called the constructor are lost.

A brute force interative N*next+crash then (N-1)next+step etc
does not work because because 'step' instead of 'next' also crashes with
and empty backtrace and gdb tells me that the 'current line' is
the opening brace of a metthod.

Therefore it's almost impossible to debug :(

Egcs and gdb are build straight forward: configure, make .. etc. Only
egcs uses -fsquangle as default (in gcc/cp/decl2.c from GCC FAQ).

Has anybody a reliably combination of gdb/g++ on Digital Unix?

Will summarize.

Thx,
Achim

-- 
    To me vi is Zen.  To use vi is to practice zen. Every command is
    a koan. Profound to the user, unintelligible to the uninitiated.
    You discover truth everytime you use it.
					-- reddy_at_lion.austin.ibm.com

Received on Tue Jul 25 2000 - 11:23:33 NZST