[PATCH/RFC future] nptl: reinstate SIGCANCEL/SIGSETXID safeguards

Kevin Cernekee cernekee at gmail.com
Wed May 11 04:16:45 UTC 2011


On Tue, May 10, 2011 at 2:06 PM, Peter Mazinger <ps.m at gmx.net> wrote:
> I am for only one implementation (that should go only into libc), else you have unpredictable/different results depending on which linking order you choose libc/libpthread

Austin's patch in the -future branch does accomplish this, but as it
stands, it completely eliminates the special handling of the
SIGCANCEL/SIGSETXID bits in sigfillset() and sigprocmask().

For sigfillset() this causes a mismatch between glibc behavior (N-2
bits set) and uClibc behavior (N bits set).

My patch reinstates the special handling for both functions, although
this causes the sigprocmask() behavior to no longer match glibc.  If
that is a problem, I can change it and resubmit.


More information about the uClibc mailing list