Seems the commercial oh-so-exclusive "stable" @grsecurity kernel was also affected by the aforementioned DoS (it also got updated).
-
@grsecurity The existing code was *fine* but triggered a *false positive* using S_O. The patched code was *broken* *and* triggered S_O.
@grsecurity So not only did you not fix the bug (i.e. patch was not tested), but made things worse by introducing brokenness if !S_O.
