On Fri, May 20, 2005 at 04:05:53PM +0200, Bartosz Fenski aka fEnIo wrote: > Since I don't have access to alpha desktop it would be quite hard to debug > the problem.
I doubt this is relevant for the bug, but Valgrind shows several errors in the memory handling. For instance, asc allocates memory with new[] some places and then later free it with asc_free() (graphicset.cpp:199 and sg.cpp:309); another problem is that it deletes "void*"-pointers, which gives undefined behaviour (g++ complains about this). Second, there are several uses of uninitialised values (which could make bugs crop up on one platform but not another); I won't list them all here, but simply running "valgrind asc -w" (on x86, of course) shows quite a lot of them. (Some are due to X, but the others should really be fixed.) Also note that there are several places where g++ warns "array subscript has type char". As the signedness of char is undefined, this may very well be a pile of portability bugs waiting to happen... BTW, setting DEB_BUILD_OPTIONS="noopt" doesn't work, since you only override CFLAGS, not CXXFLAGS. /* Steinar */ -- Homepage: http://www.sesse.net/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
