} I think the bigger question here is: Why did this happen in the first place?
}
} Cort,
}
} Is there anyone at FSM Labs know why this occurred?
The problem was with the psc module. Well, actually it wasn't the psc
module it was bad user-land programs.
The PSC module grabs a free syscall when it's loaded. With older kernels
we were getting a free syscall that happens to correspond with a syscall
added to later kernels to set/check privilege levels. Some user-programs
would make a direct call to syscall #201 (not through glibc) and assume it
was the correct one. On kernels that didn't support this syscall the
user-program would get an error. When psc grabbed that syscall the
user-program would end up calling a syscall it didn't intend rather than
the privilege syscall.
We've changed the syscall hook code in psc so that it searches top-down
rather than bottom up. That fixes the problem.
In short, if you have doubts just don't load psc. It's not needed unless
you're using the user-level realtime features of RTLinux.
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
--
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/
