A user has found a seg fault in the Ghostscript that could be used to launch a denial of service attack. The issue will be described in a private attachment.
Created attachment 5845 [details] description.txt
Created attachment 5846 [details] testg.109277045.pdf
Created attachment 5850 [details] 691043-more.txt Added an attachment with some more observations, private again.
Assigning to me.
Fixed in revision 10590, patch here: http://ghostscript.com/pipermail/gs-cvs/2010-January/010333.html As noted in the submission log this is not a totally comprehensive fix which would require a fairly major inspection and overhaul of both the type 1 and type 2 font interpreter code, as well as the code in pdfwrite which performs similar functions.