[Bug 281710] RegEXP bug in bracket expression [^...] - sed(1), grep(1), re_format(7)

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=281710

--- Comment #14 from Eric <[email protected]> ---
(in reply to Kyle Evans comment #10)

Hi Kyle, I have a question and related request as it seems that NetBSD has
pulled in (Feb 23, 2021) FreeBSD's code of
lib/libc/regex/regcomp.c
wrt the error in function singleton():
https://github.com/NetBSD/src/commit/1ee269c3a208a14da224b6e9917e2e9798961fff

You, or you may know who else within the team of FreeBSD core/developers,
likely know how to best liase with NetBSD and inform them of the problem. I
don't have anything setup OS wise or account wise regarding NetBSD, so I
haven't done any actual testing.


=====
-- NetBSD code situation view
For NetBSD's main "trunk", these parts seem relevant:
https://github.com/NetBSD/src/blob/trunk/lib/libc/regex/regcomp.c#L1139-L1143
https://github.com/NetBSD/src/blob/trunk/lib/libc/regex/regcomp.c#L1733-L1754

and the commit "sync with FreeBSD" as of Feb 23, 2021:
https://github.com/NetBSD/src/commit/1ee269c3a208a14da224b6e9917e2e9798961fff

For NetBSD 10.0, released as of NetBSD 10.0 (Mar 28, 2024), these parts seem
relevant:
https://github.com/NetBSD/src/blob/netbsd-10/lib/libc/regex/regcomp.c#L1141-L1145
https://github.com/NetBSD/src/blob/netbsd-10/lib/libc/regex/regcomp.c#L1735-L1756
and seems to contain the buggy singleton function.

By the looks of it seems that the singleton function issue isn't in NetBSD 9
code:
https://github.com/NetBSD/src/blob/netbsd-9/lib/libc/regex/regcomp.c#L831-L835

However, I'm unfamiliar with their github branch structures of their minor
releases as the latest release of  NetBSD 9 seems to be 9.4 (April 20, 2024):
https://www.netbsd.org/releases/formal-9/NetBSD-9.4.html


-- OpenBSD code situation view
For OpenBSD, based on
https://github.com/openbsd/src/blob/master/lib/libc/regex/regcomp.c#L674-L678
it looks like they haven't implemented any optimization like the singleton()
function.

-- 
You are receiving this mail because:
You are on the CC list for the bug.