Fwd: Re: New mremap bug

Archaic archaic at indy.rr.com
Wed Feb 18 17:36:15 PST 2004

> http://isec.pl/vulnerabilities/isec-0014-mremap-unmap.txt
> For those of you who'd want to test for this vulnerability,
> I've written some simple exploit code:
> http://linuxfromscratch.org/~devine/mremap_poc_2.c

Here is the beauty of patching for certain types of problems before
known vulnerabilities are found. I built the code and get this:

mmap: Cannot allocate memory
created ~65860 VMAs
now mremapping 0x4050D000 at 0x40509000
kernel may not be vulnerable

I am using the openwall patch here. Granted, though, I will build 2.4.25
when openwall rewrites the patch (since the 2.4.24 one doesn't apply),
but I am in no harm waiting to do it.


CVS is not *intended* to inflict great pain on any who use it, it is
just *possible* that it may do so.

- Bill Maltby (in a post on the LFS mailing lists)

More information about the hlfs-dev mailing list