Home
| Calendar
| Mail Lists
| List Archives
| Desktop SIG
| Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings Linux Cafe | Meeting Notes | Linux Links | Bling | About BLU |
Another way to detect memory leaks is to build your own malloc/free. I might be a bit presumptuous, but if you build your own to preempt the libc malloc, you might add some statistics to track all allocations and frees. As long as you use a simple algorithm, it should not kill performance too much. You can use mmap(2) or sbrk(2) to allocate your pages. Then you could add a statistics interface as well as a possible freeall function. You could use a simple doubly linked list as your free pool and implement a first-fit or best fit algorithm. Then when freeing up a block, you try to coalesce. Then at certain points in your code, call the statistics package to see where you stand. By doing this, you have complete control over the allocation/deallocation process, and you might find places where you are holding on to memory in a manner that does not fit the definition of memory leak, and may not be detectable by valgrind. -- Jerry Feldman <gaf at blu.org> Boston Linux and Unix user group http://www.blu.org PGP key id:C5061EA9 PGP Key fingerprint:053C 73EC 3AC1 5C44 3E14 9245 FB00 3ED5 C506 1EA9 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: <http://lists.blu.org/pipermail/discuss/attachments/20040716/c68a8300/attachment.sig>
BLU is a member of BostonUserGroups | |
We also thank MIT for the use of their facilities. |