runtime: implement procyield properly for ppc64x

The procyield() function should yield the processor as in other
architectures. On ppc64x, this is achieved by setting the Program
Priority Register to 'low priority' prior to the spin loop, and
setting it back to 'medium-low priority' afterwards.

benchmark                          old ns/op     new ns/op     delta
BenchmarkMakeChan/Byte-8           87.7          86.6          -1.25%
BenchmarkMakeChan/Int-8            107           106           -0.93%
BenchmarkMakeChan/Ptr-8            201           204           +1.49%
BenchmarkMakeChan/Struct/0-8       78.2          79.7          +1.92%
BenchmarkMakeChan/Struct/32-8      196           200           +2.04%
BenchmarkMakeChan/Struct/40-8      236           230           -2.54%
BenchmarkChanNonblocking-8         8.64          8.85          +2.43%
BenchmarkChanUncontended-8         5577          5598          +0.38%
BenchmarkChanContended-8           66106         51529         -22.05%
BenchmarkChanSync-8                451           441           -2.22%
BenchmarkChanSyncWork-8            9155          9170          +0.16%
BenchmarkChanProdCons0-8           1585          1083          -31.67%
BenchmarkChanProdCons10-8          1094          838           -23.40%
BenchmarkChanProdCons100-8         831           657           -20.94%
BenchmarkChanProdConsWork0-8       1471          941           -36.03%
BenchmarkChanProdConsWork10-8      1033          721           -30.20%
BenchmarkChanProdConsWork100-8     730           511           -30.00%
BenchmarkChanCreation-8            135           128           -5.19%
BenchmarkChanSem-8                 602           463           -23.09%
BenchmarkChanPopular-8             3017466       2188441       -27.47%

Fixes #25625

Change-Id: Iacb1c888d3c066902152b8367500348fb631c5f9
Reviewed-on: https://go-review.googlesource.com/115376
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

src/cmd/vet/all/whitelist/linux_ppc64x.txt

@@ -2,4 +2,3 @@
 
 runtime/sys_linux_ppc64x.s: [GOARCH] _sigtramp: function _sigtramp missing Go declaration
 runtime/sys_linux_ppc64x.s: [GOARCH] _cgoSigtramp: function _cgoSigtramp missing Go declaration
-runtime/asm_ppc64x.s: [GOARCH] procyield: use of 24(R1) points beyond argument frame

src/runtime/asm_ppc64x.s

@@ -457,7 +457,19 @@ CALLFN(·call268435456, 268435456)
 CALLFN(·call536870912, 536870912)
 CALLFN(·call1073741824, 1073741824)
 
-TEXT runtime·procyield(SB),NOSPLIT,$0-0
+TEXT runtime·procyield(SB),NOSPLIT|NOFRAME,$0-4
+	MOVW	cycles+0(FP), R7
+	// POWER does not have a pause/yield instruction equivalent.
+	// Instead, we can lower the program priority by setting the
+	// Program Priority Register prior to the wait loop and set it
+	// back to default afterwards. On Linux, the default priority is
+	// medium-low. For details, see page 837 of the ISA 3.0.
+	OR	R1, R1, R1	// Set PPR priority to low
+again:
+	SUB	$1, R7
+	CMP	$0, R7
+	BNE	again
+	OR	R6, R6, R6	// Set PPR priority back to medium-low
 	RET
 
 // void jmpdefer(fv, sp);