i965/fs: Take the shared mathbox into account in instruction scheduling.

I don't have evidence for this amounting to any improvement,
but it does codify a bit more what we understand so far about
the pipeline.
This commit is contained in:
Eric Anholt 2011-01-18 21:26:53 -08:00
parent 382c2d99da
commit b41d323c90
1 changed files with 15 additions and 0 deletions

View File

@ -438,6 +438,21 @@ instruction_scheduler::schedule_instructions(fs_inst *next_block_header)
instructions.push_tail(child);
}
}
/* Shared resource: the mathbox. There's one per EU (on later
* generations, it's even more limited pre-gen6), so if we send
* something off to it then the next math isn't going to make
* progress until the first is done.
*/
if (chosen->inst->is_math()) {
foreach_iter(exec_list_iterator, iter, instructions) {
schedule_node *n = (schedule_node *)iter.get();
if (n->inst->is_math())
n->unblocked_time = MAX2(n->unblocked_time,
time + chosen->latency);
}
}
}
assert(instructions_to_schedule == 0);