infodump about a gcc bug i encountered on the weekend
>get powerful arm machine to test stuff on>not a mac btw>build segfaults>wtf? it ran fine yesterday>int global = 0;>disassembly showsif (global)
was reading from an address set to 1, and different from global's actual address>stallmanbrokemycompiler.gimp>"culprit" (bug trigger) isglobal = 1;
in a function that runs when an argument flag is set, shouldn't cause problems>change it toglobal = i;
and make function accept i>compiler still pulls the phantom 1 out of its ass>whatthefuck.appimage>remove that line completely>runs fine now>get pissed, defineglobal
as a pointer dereference from avolatile *volatile
and write a 20 line comment and git commit message about how you should use clang (another buggy piece of shit but that's for another day)>worx nowcniles would defend this because trannies or something