https://bugs.kde.org/show_bug.cgi?id=511717
--- Comment #37 from Philippe Waroquiers <[email protected]> --- (In reply to Mark Wielaard from comment #36) > (In reply to Philippe Waroquiers from comment #33) > > With the below patch, glibc falls back to mprotect for the guard pages > > (i.e. this implements the first proposed (ugly?) solution). > > > > Can you check if this patch solves (or rather bypasses) the problem ? > > This also fixes memcheck/tests/descr_belowsp.vgtest which tests that > memcheck can detect the stack guard protected page, and > drd/tests/tls_threads.vgtest on glibc-2.42-5.fc43.x86_64 > > So I propose we start with this "ugly" solution and backport it to the 3.26 > branch. We discussed this proposal on IRC and agreed to push and backport as a (likely ? hopefully ?) temporary solution waiting for a possibly cleaner approach in the adress space manager. -- You are receiving this mail because: You are watching all bug changes.
