On Mon, Feb 20, 2017 at 2:02 PM, Jan Hubicka <hubi...@ucw.cz> wrote:
>> > 2017-02-16 Bin Cheng <bin.ch...@arm.com>
>> >
>> > PR tree-optimization/77536
>> > * tree-ssa-loop-manip.c (niter_for_unrolled_loop): New function.
>> > (tree_transform_and_unroll_loop): Use above function to compute the
>> > estimated niter of unrolled loop.
>> > * tree-ssa-loop-manip.h niter_for_unrolled_loop(): New declaration.
>> > * tree-vect-loop.c (scale_profile_for_vect_loop): New function.
>> > (vect_transform_loop): Call above function.
Thanks very much for your suggestions. I don't know profiling logic
very well and have some questions embedded below before I start
revising patch.
>
> +/* Return estimated niter for LOOP after unrolling by FACTOR times. */
> +
> +unsigned
> +niter_for_unrolled_loop (struct loop *loop, unsigned factor)
> +{
> + unsigned est_niter = expected_loop_iterations (loop);
>
> What happens when you have profile and loop iterates very many times?
> Perhaps we want to do all calculation in gcov_type and use
> expected_loop_iterations_unbounded>?
>
> expected_loop_iterations is capping by 10000 that is easy to overflow.
>
> + gcc_assert (factor != 0);
> + unsigned new_est_niter = est_niter / factor;
> +
> + /* Without profile feedback, loops for which we do not know a better
> estimate
> + are assumed to roll 10 times. When we unroll such loop, it appears to
> + roll too little, and it may even seem to be cold. To avoid this, we
> + ensure that the created loop appears to roll at least 5 times (but at
> + most as many times as before unrolling). */
> + if (new_est_niter < 5)
> + {
> + if (est_niter < 5)
> + new_est_niter = est_niter;
> + else
> + new_est_niter = 5;
> + }
> +
> + return new_est_niter;
> +}
>
> I see this code is pre-existing, but please extend it to test if
> loop->header->count is non-zero. Even if we do not have idea about loop
> iteration count estimate we may end up predicting more than 10 iterations when
> predictors combine that way.
If I use expected_loop_iterations_unbounded, then do I need to handle
loop->header->count explicitly here? I suppose not because it has
below code already:
/* If we have no profile at all, use AVG_LOOP_NITER. */
if (profile_status_for_fn (cfun) == PROFILE_ABSENT)
expected = PARAM_VALUE (PARAM_AVG_LOOP_NITER);
else if (loop->latch && (loop->latch->count || loop->header->count))
{
gcov_type count_in, count_latch;
//...
The second question is: looks like it only takes latch->count into
consideration when PROFILE_ABSENT. But according to your comments, we
could have nonzero count sometime?
>
> Perhaps testing estimated-loop_iterations would also make sense, but that
> could be dealt with incrementally.
>
> +static void
> +scale_profile_for_vect_loop (struct loop *loop, unsigned vf)
> +{
> + unsigned freq_h = loop->header->frequency;
> + unsigned freq_e = EDGE_FREQUENCY (loop_preheader_edge (loop));
> + /* Reduce loop iterations by the vectorization factor. */
> + unsigned new_est_niter = niter_for_unrolled_loop (loop, vf);
> +
> + if (freq_h != 0)
> + scale_loop_frequencies (loop, freq_e * (new_est_niter + 1), freq_h);
> +
> I am always trying to avoid propagating small mistakes (i.e. frong freq_h or
> freq_h) into bigger mistakes (i.e. wrong profile of the whole loop) to avoid
> spreading mistakes across cfg.
>
> But I guess here it is sort of safe because vectorized loops are simple.
> You can't just scale down the existing counts/frequencies by vf, because the
> entry edge frequency was adjusted.
I am not 100% follow here, it looks the code avoids changing frequency
counter for preheader/exit edge, otherwise we would need to change all
counters dominated by them?
>
> Also niter_for_unrolled_loop depends on sanity of the profile, so perhaps you
> need to compute it before you start chanigng the CFG by peeling proplogue?
Peeling for prologue doesn't change profiling information of
vect_loop, it is the skip edge from before loop to preferred epilogue
loop that will change profile counters. I guess here exists a dilemma
that niter_for_unrolled_loop is for loop after peeling for prologue?
Thanks,
bin
>
> Finally if freq_e is 0, all frequencies and counts will be probably dropped to
> 0. What about determining fraction by counts if they are available?
>
> Otherwise the patch looks good and thanks a lot for working on this!
>
> Honza
>
>> >
>> > gcc/testsuite/ChangeLog
>> > 2017-02-16 Bin Cheng <bin.ch...@arm.com>
>> >
>> > PR tree-optimization/77536
>> > * gcc.dg/vect/pr79347.c: Revise testing string.
