cmd/ld: align function entry on arch-specific boundary

16 seems pretty standard on x86 for function entry.
I don't know if ARM would benefit, so I used just 4
(single instruction alignment).

This has a minor absolute effect on the current timings.
The main hope is that it will make them more consistent from
run to run.

benchmark                 old ns/op    new ns/op    delta
BenchmarkBinaryTree17    4222117400   4140739800   -1.93%
BenchmarkFannkuch11      3462631800   3259914400   -5.85%
BenchmarkGobDecode         20887622     20620222   -1.28%
BenchmarkGobEncode          9548772      9384886   -1.72%
BenchmarkGzip                151687       150333   -0.89%
BenchmarkGunzip                8742         8741   -0.01%
BenchmarkJSONEncode        62730560     65210990   +3.95%
BenchmarkJSONDecode       252569180    249394860   -1.26%
BenchmarkMandelbrot200      5267599      5273394   +0.11%
BenchmarkRevcomp25M       980813500    996013800   +1.55%
BenchmarkTemplate         361259100    360620840   -0.18%

R=ken2
CC=golang-dev
https://golang.org/cl/6244066

src/cmd/5l/l.h

@@ -36,7 +36,8 @@
 enum
 {
 	thechar = '5',
-	PtrSize = 4
+	PtrSize = 4,
+	FuncAlign = 4  // single-instruction alignment
 };
 
 #ifndef	EXTERN

src/cmd/6l/l.h

@@ -40,7 +40,8 @@
 enum
 {
 	thechar = '6',
-	PtrSize = 8
+	PtrSize = 8,
+	FuncAlign = 16
 };
 
 #define	P		((Prog*)0)

src/cmd/8l/l.h

@@ -40,7 +40,8 @@
 enum
 {
 	thechar = '8',
-	PtrSize = 4
+	PtrSize = 4,
+	FuncAlign = 16
 };
 
 #define	P		((Prog*)0)

src/cmd/ld/data.c

@@ -1012,6 +1012,8 @@ textaddress(void)
 			continue;
 		if(sym->align != 0)
 			va = rnd(va, sym->align);
+		else if(sym->text != P)
+			va = rnd(va, FuncAlign);
 		sym->value = 0;
 		for(sub = sym; sub != S; sub = sub->sub) {
 			sub->value += va;