https://bugs.exim.org/show_bug.cgi?id=2618
--- Comment #19 from Carlo Marcelo Arenas Belón <carenas@???> ---
there is a proposed[1] fix to sljit upstream that needs testing in BigSur.
for testing with PCRE it would be easier to apply it on top of TRUNK and I have
to admit that I am concerned it might not work with notarized applications that
fork(), based on the constrains we saw before with MAP_JIT and my (obviously
probably not accurate) of the public XNU code for Catalina.
if anyone could provide access to an Apple Silicon BigSur instance we would be
more confident, but if not I could provide some additional code (will require
to have a developer certificate for codesigning though) to run artificial
probes to try to make sure the functionality we use behaves as expected.
the (AFAIK undocumented) constrain on MAP_JIT pages to be inherited
(ironically, only if using the hardened runtime) after fork() is specially
concerning and if it is not fixed in Apple Silicon BigSur will likely result in
a lot of code failing to work, which will be important to raise with Apple.
[1]
https://github.com/zherczeg/sljit/pull/90
--
You are receiving this mail because:
You are on the CC list for the bug.
