https://bugs.kde.org/show_bug.cgi?id=372347
--- Comment #22 from Paul Floyd <pa...@free.fr> --- Filtering out the free was enough to get the test to pass on the 3 OSes that I tested on [Linux, Solaris and FreeBSD]. You are right that it is not necessary for the test, and I'll add malloc to the filter. I want to keep the prereq because in theory someone could be building with a very old GCC that does not support -fsized-deallocation. The test is currently amd64 specific. The signature includes size_t, which is a synonym for unsigned int on x86 (_ZdlPvj) and unsigned long on amd64 (_ZdlPvm). That's for the scalar form of delete, there's also the array form with the same difference. I'll move it up to memcheck/tests and add an x86 expected. -- You are receiving this mail because: You are watching all bug changes.
