|
From: Julian S. <js...@ac...> - 2005-04-02 01:21:40
|
Greetings, all. I'm currently mashing Memcheck/Addrcheck internals about. I'd like to get rid of the --sloppy-malloc=yes flag if possible. Having to check the flag at every allocation wastes time in allocation-intensive programs, and I don't believe anybody really uses it. So, does anybody use this flag? J |
|
From: Dennis L. <pla...@tz...> - 2005-04-02 13:19:19
|
Am Samstag, den 02.04.2005, 02:21 +0100 schrieb Julian Seward: > Greetings, all. > > I'm currently mashing Memcheck/Addrcheck internals about. I'd > like to get rid of the --sloppy-malloc=yes flag if possible. > Having to check the flag at every allocation wastes time in > allocation-intensive programs, and I don't believe anybody > really uses it. > > So, does anybody use this flag? I used it from time to time to cope with some messed up libraries, but havent thought about it for over a year or so now, so I personally think noone really needs it, since valgrind is an error detection program and not a program to run programs that are full of errors... greets Dennis PS: I dont know much of the internals of valgrind, but for similar cases where I have runtime flags, I use function pointers and load them with different functions so I need to check the flag only once... |