public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* [Committed] RISC-V: Fix VSETVL PASS fusion bug
@ 2023-09-18 11:08 Juzhe-Zhong
  0 siblings, 0 replies; only message in thread
From: Juzhe-Zhong @ 2023-09-18 11:08 UTC (permalink / raw)
  To: gcc-patches; +Cc: kito.cheng, kito.cheng, jeffreyalaw, rdapp.gcc, Juzhe-Zhong

There is an obvious fusion bug that is exposed by more VLS patterns support.
After more VLS modes support, it cause following FAILs:
FAIL: gcc.target/riscv/rvv/autovec/reduc/reduc_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/reduc/reduc_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/reduc/reduc_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/reduc/reduc_run-2.c execution test

Demand 1: SEW = 64, LMUL = 1, RATIO = 64, demand SEW, demand GE_SEW.
Demand 2: SEW = 64, LMUL = 2, RATIO = 32, demand SEW, demand GE_SEW, demand RATIO.

Before this patch:
merge demand: SEW = 64, LMUL = 1, RATIO = 32, demand SEW, demand LMUL, demand GE_SEW.
It's obvious incorrect of merge LMUL which should be new LMUL = (demand 2 RATIO * greatest SEW) = M2

gcc/ChangeLog:

	* config/riscv/riscv-vsetvl.cc (vlmul_for_greatest_sew_second_ratio): New function.
	* config/riscv/riscv-vsetvl.def (DEF_SEW_LMUL_FUSE_RULE): Fix bug.

---
 gcc/config/riscv/riscv-vsetvl.cc  | 8 ++++++++
 gcc/config/riscv/riscv-vsetvl.def | 4 ++--
 2 files changed, 10 insertions(+), 2 deletions(-)

diff --git a/gcc/config/riscv/riscv-vsetvl.cc b/gcc/config/riscv/riscv-vsetvl.cc
index 5f031c18df5..df980b6770e 100644
--- a/gcc/config/riscv/riscv-vsetvl.cc
+++ b/gcc/config/riscv/riscv-vsetvl.cc
@@ -1308,6 +1308,14 @@ vlmul_for_first_sew_second_ratio (const vector_insn_info &info1,
   return calculate_vlmul (info1.get_sew (), info2.get_ratio ());
 }
 
+static vlmul_type
+vlmul_for_greatest_sew_second_ratio (const vector_insn_info &info1,
+				     const vector_insn_info &info2)
+{
+  return calculate_vlmul (MAX (info1.get_sew (), info2.get_sew ()),
+			  info2.get_ratio ());
+}
+
 static unsigned
 ratio_for_second_sew_first_vlmul (const vector_insn_info &info1,
 				  const vector_insn_info &info2)
diff --git a/gcc/config/riscv/riscv-vsetvl.def b/gcc/config/riscv/riscv-vsetvl.def
index 7289c01efcf..709cc4ee0df 100644
--- a/gcc/config/riscv/riscv-vsetvl.def
+++ b/gcc/config/riscv/riscv-vsetvl.def
@@ -329,8 +329,8 @@ DEF_SEW_LMUL_FUSE_RULE (/*SEW*/ DEMAND_TRUE, /*LMUL*/ DEMAND_FALSE,
 			/*NEW_DEMAND_SEW*/ true,
 			/*NEW_DEMAND_LMUL*/ false,
 			/*NEW_DEMAND_RATIO*/ true,
-			/*NEW_DEMAND_GE_SEW*/ true, greatest_sew, first_vlmul,
-			second_ratio)
+			/*NEW_DEMAND_GE_SEW*/ true, greatest_sew,
+			vlmul_for_greatest_sew_second_ratio, second_ratio)
 DEF_SEW_LMUL_FUSE_RULE (/*SEW*/ DEMAND_TRUE, /*LMUL*/ DEMAND_FALSE,
 			/*RATIO*/ DEMAND_ANY, /*GE_SEW*/ DEMAND_TRUE,
 			/*SEW*/ DEMAND_FALSE, /*LMUL*/ DEMAND_TRUE,
-- 
2.36.3


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2023-09-18 11:08 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-09-18 11:08 [Committed] RISC-V: Fix VSETVL PASS fusion bug Juzhe-Zhong

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).