The NTLM fakeauth_auth helper has a memory leak that may cause it to run out of memory under high load, or if it runs for a very long time. Additionally, a malformed NTLM type 3 message could cause a segmentation violation. http://www.squid-cache.org/Versions/v2/2.5/bugs/#squid-2.5.STABLE7-fakeauth_auth
cyfred, pls include the patch in the patchset ____ http://securitytracker.com/alerts/2005/Jan/1012818.html http://secunia.com/advisories/13789/
Just went into cvs now.
cyfred, pls bump the ebuild, so people can catch the updated patchset We should probably use ~arch keywords, so that we can go through arch testing and don't run into problems like it happened before with a different ebuild.
Revision bump done.
jaervosz and /me voted against GLSA publication on this one, if anyone objects pls reopen
GLSA 200501-25