On Wed, Feb 01, 2017 at 10:14:28AM +0200, Joonas Lahtinen wrote: > On to, 2017-01-19 at 11:41 +0000, Chris Wilson wrote: > > Just create several batches of requests and expect it to not fall over! > > > > Signed-off-by: Chris Wilson <[email protected]> > > <SNIP> > > > +static int live_nop_request(void *arg) > > +{ > > <SNIP> > > > + for_each_prime_number_from(prime, 1, 8192) { > > #define MAGIC_PRIME_2 8192
I'm sorry, but that is worse than having a very clear arbitrary number. > > + > > + i915_add_request(request); > > + } > > + i915_wait_request(request, > > + I915_WAIT_LOCKED, > > + MAX_SCHEDULE_TIMEOUT); > > + > > + times[1] = ktime_sub(ktime_get_raw(), times[1]); > > + if (prime == 1) > > + times[0] = times[1]; > > Having this as an array is just hiding names and gaining nothing, how > about calling them just time_first, time_last? time_0, time_N? The array is because I did/am considering that the graph may be interesting. There should be a plateau at the point the ring is full - not world shattering, just mildly interesting. Still, probably want the times for cold submit_1 vs warm submit_1. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
