From 66b9b71ff8fe3d7ff046ac7a2b94fc52c47f5442 Mon Sep 17 00:00:00 2001
From: Michael Tiemann <tiemann@holodeck.cygnus.com>
Date: Wed, 25 Aug 1999 05:28:26 +0000
Subject: [PATCH] loop.c (insert_bct): Replace use of sdiv_optab with asr_optab
and delete comment that code should be...
* loop.c (insert_bct): Replace use of sdiv_optab with asr_optab
and delete comment that code should be rewritten.
From-SVN: r28839
---
gcc/ChangeLog | 5 +++++
gcc/loop.c | 17 ++++-------------
2 files changed, 9 insertions(+), 13 deletions(-)
diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 6597b92dc2f5..7dcf4b92ad06 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,8 @@
+Tue Aug 24 23:26:44 1999 Michael Tiemann <tiemann@holodeck.cygnus.com>
+
+ * loop.c (insert_bct): Replace use of sdiv_optab with asr_optab
+ and delete comment that code should be rewritten.
+
Tue Aug 24 22:56:35 1999 Jeffrey A Law (law@cygnus.com)
* haifa-sched.c (find_rgns): Mark a block found during the DFS search
diff --git a/gcc/loop.c b/gcc/loop.c
index db4664ce1925..4d18f5ee65e7 100644
--- a/gcc/loop.c
+++ b/gcc/loop.c
@@ -9235,19 +9235,10 @@ insert_bct (loop_start, loop_end, loop_info)
NULL_RTX, 0, OPTAB_LIB_WIDEN);
if (increment_value_abs != 1)
- {
- /* ??? This will generate an expensive divide instruction for
- most targets. The original authors apparently expected this
- to be a shift, since they test for power-of-2 divisors above,
- but just naively generating a divide instruction will not give
- a shift. It happens to work for the PowerPC target because
- the rs6000.md file has a divide pattern that emits shifts.
- It will probably not work for any other target. */
- iterations_num_reg = expand_binop (loop_var_mode, sdiv_optab,
- temp_reg,
- GEN_INT (increment_value_abs),
- NULL_RTX, 0, OPTAB_LIB_WIDEN);
- }
+ iterations_num_reg = expand_binop (loop_var_mode, asr_optab,
+ temp_reg,
+ GEN_INT (exact_log2 (increment_value_abs)),
+ NULL_RTX, 0, OPTAB_LIB_WIDEN);
else
iterations_num_reg = temp_reg;
}
--
GitLab