On Tue, Jul 11, 2017 at 18:59:37 +0100, Alex Bennée wrote:
> Previously DISAS_JUMP did ensure this but with the optimisation of
> 8a6b28c7 (optimize indirect branches) we might not leave the loop.
> This means if any pending interrupts are cleared by changing IRQ flags
> we might never get around to servicing them. You usually notice this
> by seeing the lookup_tb_ptr() helper gainfully chaining TBs together
> while cpu->interrupt_request remains high and the exit_request has not
> been set.
> 
> This breaks amongst other things the OPTEE test suite which executes
> an eret from the secure world after a non-secure world IRQ has gone
> pending which then never gets serviced.
> 
> Instead of using the previously implied semantics of DISAS_JUMP we use
> DISAS_EXIT which will always exit the run-loop.
> 
> CC: Etienne Carriere <[email protected]>
> CC: Joakim Bech <[email protected]>
> CC: Peter Maydell <[email protected]>
> CC: Emilio G. Cota <[email protected]>
> Signed-off-by: Alex Bennée <[email protected]>
> Reviewed-by: Richard Henderson <[email protected]>

Reviewed-by: Emilio G. Cota <[email protected]>

                E.

Reply via email to