Well, as I knew it would, $rbp-40 is coming back to haunt me.
commit2c87565d656e939cf77459d00e3def854f6a2844
authorstrange <kawk256@gmail.com>
Wed, 20 Jan 2010 02:27:23 +0000 (19 19:27 -0700)
committerstrange <kawk256@gmail.com>
Wed, 20 Jan 2010 02:27:23 +0000 (19 19:27 -0700)
tree02de2d45082d049b2bd49ed2e7a0d2b5cc65aac8
parentaa1fce58ef44fb58b1d6bfccaeb2d0e927c42846
Well, as I knew it would, $rbp-40 is coming back to haunt me.

It would appear that the return address of malloc changes depending on where
and how it is called. $rbp-40, for malloc_test, just happened to be the right
address. Grr.

libc is defying all I know of assembly, and I don't think that's a good thing.
I suppose it is because libc is probably the most-heavily optimized piece of
software in the world . . .

It's research time.
src/monitor/ptrace/MallocObserver.cpp
tests/.gitignore
tests/CMakeLists.txt
tests/interactive_malloc_test.c [new file with mode: 0644]