> > > Here is a patch for all the buffer overflow and potential buffer overflows
> > > in apache that I noticed in my run through the source. First, a few Q&A
> > > that I asked myself.
> >
> > Few of these changes seem to apply to anything but the error code.
> >
> > Based on that, I don't have a big problem with adding them. A few
> > comments though.
> >
> > * Ben added the vbprintf() code awhile back. Seems like a fair chunk
> > of that code could be used to supply an snprintf() that could be
> > used more effectively to keep these sorts of problems from creeping
> > back in.
> >
> > * IF we decide to include these changes, seems that it would be nice
> > to get them in a fair bit _before_ the proposed changes that Ben
> > will be making to the API just to make it a bit easier to pin down
> > problem causers if there are any.
> >
> Xinetd also has strx_nprint() and strx_nprintv() functions to do this. I
> sent mail to Mark about these, if they'd be of help.
>
> It's a tough job, and I'm glad Mark's doing it.
>
> chuck
If I didn't make that point in my earlier mail, I agree. Thanks for
grabbing this Marc. Since Marc seems to be taking on some of these
security issues, I would really appreciate his comments regarding the
suexec stuff as well.
|