Contributed by jose on from the error-checking dept.
"" Slashdot has a blurb about two Stanford researchers, Dawson Engler and Yichen Xie have written a paper (pdf) that looks at redundant code, dead code, etc. which might suggest errors in the kernel (and possibly security holes?) An interesting read, but it raises the question -- how much, if any, redundant code is in the OpenBSD kernel?"Englar and Xie has written some interesting software analysis tools which they periodically turn loose on OpenBSD, Linux, and other large projects. They find real bugs all the time, but like any analysis engine they also get false positives. This paper (PS) is a comparison of operating system errors broken down by subtypes and reveals some interesting large scale facets about Linux and OpenBSD. Give these papers (and in fact all of their papers in this vein) a good read.
(Comments are closed)
By Chad Loder () on
How about sha1 (it's at least 3 -- the BSD version, the Apache version, and the OpenSSH version. There may be others).
What are other good examples of stuff that gets duplicated all over the place?
By fansipans () on http://dub.gmu.edu/~fansipans/
By zil0g () on
it did sound a tad unfair towards OpenBSD tho, to "warn" the Linux hackers at 2.3.99 so they could fix some 100 bugs before comparing it to OBSD 2.8 -- but that's just me of course :-|