From: Richard Biener <rguenther@suse.de>
Date: Tue, 10 Mar 2026 13:39:24 +0000 (+0100)
Subject: Add comment to vect_estimate_min_profitable_iters
X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=0c7dc43c8de6a42bcf581420fc75bf425faeff6e;p=thirdparty%2Fgcc.git

Add comment to vect_estimate_min_profitable_iters

The following adds a comment how it's awkward to add the scalar loop
stmt cost vectors with scaled count to the vector loop cost vector
to estimate peeling costs.

	* tree-vect-loop.cc (vect_estimate_min_profitable_iters):
	Add comment about costing of prologue/epilogue.
---

diff --git a/gcc/tree-vect-loop.cc b/gcc/tree-vect-loop.cc
index 94f3d00d60e..ee540028ba9 100644
--- a/gcc/tree-vect-loop.cc
+++ b/gcc/tree-vect-loop.cc
@@ -4130,6 +4130,11 @@ vect_estimate_min_profitable_iters (loop_vec_info loop_vinfo,
 	epilogue_need_br_taken_cost = true;
     }
 
+  /* The way we cummulate peeling costs into the vector prologue/epilogue
+     cost is a bit awkward given we cannot reuse scalar_costs which is
+     already computed and also because it cannot take into account any
+     epilogue vectorization we'll carry out in the end.  */
+
   stmt_info_for_cost *si;
   int j;
   /* Add costs associated with peel_iters_prologue.  */