Messages in this thread |  | | Date | Sun, 29 Oct 2000 15:42:37 -0500 (EST) | From | Alexander Viro <> | Subject | Re: page->mapping == 0 |
| |
On Sun, 29 Oct 2000, Alan Cox wrote:
> > I would expect problems with truncate, mmap, rename, POSIX locks, fasync, > > ptrace and mount go unnoticed for _long_. Ditto for parts of procfs > > Well the ptrace one still has mysteriously breaks usermode linux against it > on my list here. Was that ever explained. It looked like the stack got corrupted > which is weird.
Alan, is it me or usermode port really tends to catch the stack overflows? <thinks> How about we allocate the pages by 4, not by 2 as we do now, and explicitly unmap the pages below the task_struct? That would still leave the chance of stack overflow going unnoticed, but the window would be limited.
Another possibility for ptrace-related screwups: past-the-EOF check in filemap_nopage(). I'm still not convinced that it's right. It may be, but...
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
|  |