Possible race between PTRACE_SETVFPREGS and PTRACE_CONT on ARM?

[Russell King - ARM Linux]

On Tue, May 31, 2016 at 02:52:52PM +0100, Will Deacon wrote:
> The only thing I'm uncertain of is whether or not PTRACE_SEIZE/PTRACE_LISTEN
> allow switching to the child (but even then, how is the parent doing
> to issue such a request?).

I can't see how that would be possible - the parent needs to finish
the vfp_set() call before it can issue any others - and there can
only be one tracer.  IOW, if we're in vfp_set(), the thread must
have attached to the child, at which point PTRACE_SEIZE will be
rejected.

PTRACE_LISTEN also requires the child to be in STOP state as well, and
again, can only be issued from the current tracer.

-- 
RMK's Patch system: http://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.