Quoting Chris Wilson (2017-10-31 15:41:52)
> Quoting Mika Kuoppala (2017-10-31 15:27:33)
> > +static inline struct execlist_port *
> > +execlists_port_next(struct intel_engine_execlists * const execlists,
> > + const struct execlist_port * const port)
> > +{
> > + const unsigned int n = __port_add(port_index(port, execlists),
> > + 1,
> > + execlists->port_mask);
>
> How does this compare to
>
> if (port++ == execlists->port + execlists->port_mask)
> port = execlists->port;
>
> return port;
> ?
Rough estimate from bloat-o-meter
patch:
intel_lrc_irq_handler 1734 1926 +192
i915_guc_irq_handler 1522 1620 +98
execlists_cancel_port_requests 81 129 +48
intel_engine_dump 2030 2077 +47
intel_engine_init_cmd_parser 1132 1136 +4
capture 5633 5620 -13
delta:
i915_guc_irq_handler 1620 1617 -3
intel_lrc_irq_handler 1926 1790 -136
overall:
i915_guc_irq_handler 1522 1617 +95
intel_lrc_irq_handler 1734 1790 +56
execlists_cancel_port_requests 81 129 +48
intel_engine_dump 2030 2077 +47
intel_engine_init_cmd_parser 1132 1136 +4
capture 5633 5620 -13
But still +56 in irq_handler, that's mostly dequeue I guess. Can we do
better?
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
