cmd/compile: inline constant sized memclrNoHeapPointers calls on PPC64

Update the function isInlinableMemclr for ppc64 and ppc64le
to enable inlining for the constant sized cases < 512.

Larger cases can use dcbz which performs better but requires
alignment checking so it is best to continue using memclrNoHeapPointers
for those cases.

Results on p10:

MemclrKnownSize1         2.07ns ± 0%     0.57ns ± 0%   -72.59%
MemclrKnownSize2         2.56ns ± 5%     0.57ns ± 0%   -77.82%
MemclrKnownSize4         5.15ns ± 0%     0.57ns ± 0%   -89.00%
MemclrKnownSize8         2.23ns ± 0%     0.57ns ± 0%   -74.57%
MemclrKnownSize16        2.23ns ± 0%     0.50ns ± 0%   -77.74%
MemclrKnownSize32        2.28ns ± 0%     0.56ns ± 0%   -75.28%
MemclrKnownSize64        2.49ns ± 0%     0.72ns ± 0%   -70.95%
MemclrKnownSize112       2.97ns ± 2%     1.14ns ± 0%   -61.72%
MemclrKnownSize128       4.64ns ± 6%     2.45ns ± 1%   -47.17%
MemclrKnownSize192       5.45ns ± 5%     2.79ns ± 0%   -48.87%
MemclrKnownSize248       4.51ns ± 0%     2.83ns ± 0%   -37.12%
MemclrKnownSize256       6.34ns ± 1%     3.58ns ± 0%   -43.53%
MemclrKnownSize512       3.64ns ± 0%     3.64ns ± 0%    -0.03%
MemclrKnownSize1024      4.73ns ± 0%     4.73ns ± 0%    +0.01%
MemclrKnownSize4096      17.1ns ± 0%     17.1ns ± 0%    +0.07%
MemclrKnownSize512KiB    2.12µs ± 0%     2.12µs ± 0%      ~     (all equal)

Change-Id: If1abf5749f4802c64523a41fe0058bd144d0ea46
Reviewed-on: https://go-review.googlesource.com/c/go/+/464340
Run-TryBot: Lynn Boger <[email protected]>
Reviewed-by: Jakub Ciolek <[email protected]>
Reviewed-by: Archana Ravindar <[email protected]>
Reviewed-by: Ian Lance Taylor <[email protected]>
Reviewed-by: Carlos Eduardo Seo <[email protected]>
TryBot-Result: Gopher Robot <[email protected]>
Reviewed-by: Than McIntosh <[email protected]>

Author: laboger

src/cmd/compile/internal/ssa/_gen/generic.rules

@@ -2070,7 +2070,7 @@
 // Turn known-size calls to memclrNoHeapPointers into a Zero.
 // Note that we are using types.Types[types.TUINT8] instead of sptr.Type.Elem() - see issue 55122 and CL 431496 for more details.
 (SelectN [0] call:(StaticCall {sym} sptr (Const(64|32) [c]) mem))
-  && isInlinableMemclr(config)
+  && isInlinableMemclr(config, int64(c))
   && isSameCall(sym, "runtime.memclrNoHeapPointers")
   && call.Uses == 1
   && clobber(call)

src/cmd/compile/internal/ssa/rewrite.go

@@ -1365,10 +1365,16 @@ func zeroUpper56Bits(x *Value, depth int) bool {
 	return false
 }
 
-func isInlinableMemclr(c *Config) bool {
+func isInlinableMemclr(c *Config, sz int64) bool {
 	// TODO: expand this check to allow other architectures
 	// see CL 454255 and issue 56997
-	return c.arch == "amd64" || c.arch == "arm64"
+	switch c.arch {
+	case "amd64", "arm64":
+		return true
+	case "ppc64le", "ppc64":
+		return sz < 512
+	}
+	return false
 }
 
 // isInlinableMemmove reports whether the given arch performs a Move of the given size

src/cmd/compile/internal/ssa/rewritegeneric.go

@@ -47,7 +47,7 @@ func SliceExtensionConst(s []int) []int {
 	// amd64:-`.*runtime\.makeslice`
 	// amd64:-`.*runtime\.panicmakeslicelen`
 	// amd64:"MOVUPS\tX15"
-	// ppc64x:`.*runtime\.memclrNoHeapPointers`
+	// ppc64x:-`.*runtime\.memclrNoHeapPointers`
 	// ppc64x:-`.*runtime\.makeslice`
 	// ppc64x:-`.*runtime\.panicmakeslicelen`
 	return append(s, make([]int, 1<<2)...)
@@ -58,7 +58,7 @@ func SliceExtensionConstInt64(s []int) []int {
 	// amd64:-`.*runtime\.makeslice`
 	// amd64:-`.*runtime\.panicmakeslicelen`
 	// amd64:"MOVUPS\tX15"
-	// ppc64x:`.*runtime\.memclrNoHeapPointers`
+	// ppc64x:-`.*runtime\.memclrNoHeapPointers`
 	// ppc64x:-`.*runtime\.makeslice`
 	// ppc64x:-`.*runtime\.panicmakeslicelen`
 	return append(s, make([]int, int64(1<<2))...)
@@ -69,7 +69,7 @@ func SliceExtensionConstUint64(s []int) []int {
 	// amd64:-`.*runtime\.makeslice`
 	// amd64:-`.*runtime\.panicmakeslicelen`
 	// amd64:"MOVUPS\tX15"
-	// ppc64x:`.*runtime\.memclrNoHeapPointers`
+	// ppc64x:-`.*runtime\.memclrNoHeapPointers`
 	// ppc64x:-`.*runtime\.makeslice`
 	// ppc64x:-`.*runtime\.panicmakeslicelen`
 	return append(s, make([]int, uint64(1<<2))...)
@@ -80,12 +80,20 @@ func SliceExtensionConstUint(s []int) []int {
 	// amd64:-`.*runtime\.makeslice`
 	// amd64:-`.*runtime\.panicmakeslicelen`
 	// amd64:"MOVUPS\tX15"
-	// ppc64x:`.*runtime\.memclrNoHeapPointers`
+	// ppc64x:-`.*runtime\.memclrNoHeapPointers`
 	// ppc64x:-`.*runtime\.makeslice`
 	// ppc64x:-`.*runtime\.panicmakeslicelen`
 	return append(s, make([]int, uint(1<<2))...)
 }
 
+// On ppc64x continue to use memclrNoHeapPointers
+// for sizes >= 512.
+func SliceExtensionConst512(s []int) []int {
+	// amd64:-`.*runtime\.memclrNoHeapPointers`
+	// ppc64x:`.*runtime\.memclrNoHeapPointers`
+	return append(s, make([]int, 1<<9)...)
+}
+
 func SliceExtensionPointer(s []*int, l int) []*int {
 	// amd64:`.*runtime\.memclrHasPointers`
 	// amd64:-`.*runtime\.makeslice`