https://bugs.kde.org/show_bug.cgi?id=452274
--- Comment #11 from Paul Floyd <pjfl...@wanadoo.fr> ---
That means that it is trying to start a new syscall whilst another is in
flight.
In the logs I see poll() then sigprocmask(). But that doesn't mean that the new
syscall is sigprocmask, it could be some other syscall caused by an async
signal.
How easy would it be for me to build maiko? I mainly use VirtualBox and just a
shell on platforms other than 13.0-RELEASE amd64.
Otherwise, could you modify syswrap-main.c again
vg_assert(sci->status.what == SsIdle);
getSyscallArgsFromGuestState( &sci->orig_args, &tst->arch.vex, trc );
/* Copy .orig_args to .args. The pre-handler may modify .args, but
we want to keep the originals too, just in case. */
sci->args = sci->orig_args;
/* Save the syscall number in the thread state in case the syscall
is interrupted by a signal. */
sysno = sci->orig_args.sysno;
-> put a VG_(printf) of sysno here, and move the assert after the printf
--
You are receiving this mail because:
You are watching all bug changes.
