Hello All:
This patch improve determine_suggested_unroll_factor in finish_cost
with reduction factor of loads/stores/non_load_stores.
Return unroll factor calculated as per reduction factor
with number of loads/stores/non_load_stores (general_ops).
Bootstrapped and regtested on powerpc64-linux-gnu.
Expected gains with spec 2017 benchmarks.
Thanks & Regards
Ajit
rs6000: Improve suggested unroll factor in finish_cost
Improve determine_suggested_unroll_factor in finish_cost
with reduction factor of loads/stores/non_load_stores.
Return unroll factor calculated as per reduction factor
with number of loads/stores/non_load_stores (general_ops).
2024-07-22 Ajit Kumar Agarwal <aagar...@linux.ibm.com>
gcc/ChangeLog:
* config/rs6000/rs6000.cc: Improve
determine_suggested_unroll_factor with reduction factor
of load/stores/general_ops.
---
gcc/config/rs6000/rs6000.cc | 60 ++++++++++---------------------------
1 file changed, 16 insertions(+), 44 deletions(-)
diff --git a/gcc/config/rs6000/rs6000.cc b/gcc/config/rs6000/rs6000.cc
index 5ed64b1e686..0d69ec4cfbe 100644
--- a/gcc/config/rs6000/rs6000.cc
+++ b/gcc/config/rs6000/rs6000.cc
@@ -5458,7 +5458,6 @@ rs6000_cost_data::adjust_vect_cost_per_loop
(loop_vec_info loop_vinfo)
- estimated iteration count when iteration count is unknown;
*/
-
unsigned int
rs6000_cost_data::determine_suggested_unroll_factor (loop_vec_info loop_vinfo)
{
@@ -5483,53 +5482,26 @@ rs6000_cost_data::determine_suggested_unroll_factor
(loop_vec_info loop_vinfo)
unsigned int issue_width = rs6000_vect_unroll_issue;
unsigned int uf = CEIL (reduc_factor * issue_width, nstmts_nonldst);
uf = MIN ((unsigned int) rs6000_vect_unroll_limit, uf);
- /* Make sure it is power of 2. */
- uf = 1 << ceil_log2 (uf);
+ unsigned int temp;
- /* If the iteration count is known, the costing would be exact enough,
- don't worry it could be worse. */
- if (LOOP_VINFO_NITERS_KNOWN_P (loop_vinfo))
- return uf;
-
- /* Inspired by SPEC2017 parest_r, we want to aggressively unroll the
- loop if either condition is satisfied:
- - reduction factor exceeds the threshold;
- - emulated gather load adopted. */
- if (reduc_factor > (unsigned int) rs6000_vect_unroll_reduc_threshold
- || m_gather_load)
- return uf;
-
- /* Check if we can conclude it's good to unroll from the estimated
- iteration count. */
- HOST_WIDE_INT est_niter = get_estimated_loop_iterations_int (loop);
- unsigned int vf = vect_vf_for_cost (loop_vinfo);
- unsigned int unrolled_vf = vf * uf;
- if (est_niter == -1 || est_niter < unrolled_vf)
- /* When the estimated iteration of this loop is unknown, it's possible
- that we are able to vectorize this loop with the original VF but fail
- to vectorize it with the unrolled VF any more if the actual iteration
- count is in between. */
- return 1;
- else
+ if (m_nstores > 0)
{
- unsigned int epil_niter_unr = est_niter % unrolled_vf;
- unsigned int epil_niter = est_niter % vf;
- /* Even if we have partial vector support, it can be still inefficent
- to calculate the length when the iteration count is unknown, so
- only expect it's good to unroll when the epilogue iteration count
- is not bigger than VF (only one time length calculation). */
- if (LOOP_VINFO_CAN_USE_PARTIAL_VECTORS_P (loop_vinfo)
- && epil_niter_unr <= vf)
- return uf;
- /* Without partial vector support, conservatively unroll this when
- the epilogue iteration count is less than the original one
- (epilogue execution time wouldn't be longer than before). */
- else if (!LOOP_VINFO_CAN_USE_PARTIAL_VECTORS_P (loop_vinfo)
- && epil_niter_unr <= epil_niter)
- return uf;
+ temp = CEIL (reduc_factor * rs6000_vect_unroll_issue,
+ m_nstores);
+ uf = MIN (uf, temp);
}
- return 1;
+ if (m_nloads > 0)
+ {
+ temp = CEIL (reduc_factor * rs6000_vect_unroll_issue,
+ m_nloads + m_nstores);
+ uf = MIN (uf, temp);
+ }
+
+ /* Make sure it is power of 2. */
+ uf = 1 << ceil_log2 (uf);
+
+ return uf;
}
void
--
2.43.5
