cmd/link: run generator functions along with symbol writing

Like we do for applying relocations, for generator symbols, run
the generator function along with symbol writing. This will
probably have better locality and parallelism.

Linking cmd/compile,

Asmb          29.9ms ± 5%    19.1ms ±12%  -36.18%  (p=0.000 n=10+9)
TotalTime      351ms ± 3%     339ms ± 2%   -3.51%  (p=0.000 n=11+10)

Change-Id: I9cda6718bf70b3bcf1b7a501a845d6136234d2ee
Reviewed-on: https://go-review.googlesource.com/c/go/+/263640
Trust: Cherry Zhang <[email protected]>
Run-TryBot: Cherry Zhang <[email protected]>
TryBot-Result: Go Bot <[email protected]>
Reviewed-by: Jeremy Faller <[email protected]>
Reviewed-by: Than McIntosh <[email protected]>

Author: cherrymui

src/cmd/link/internal/ld/data.go

@@ -951,6 +951,9 @@ func writeBlock(ctxt *Link, out *OutBuf, ldr *loader.Loader, syms []loader.Sym,
 		}
 		P := out.WriteSym(ldr, s)
 		st.relocsym(s, P)
+		if f, ok := ctxt.generatorSyms[s]; ok {
+			f(ctxt, s)
+		}
 		addr += int64(len(P))
 		siz := ldr.SymSize(s)
 		if addr < val+siz {

src/cmd/link/internal/ld/main.go

@@ -36,14 +36,12 @@ import (
 	"cmd/internal/objabi"
 	"cmd/internal/sys"
 	"cmd/link/internal/benchmark"
-	"cmd/link/internal/loader"
 	"flag"
 	"log"
 	"os"
 	"runtime"
 	"runtime/pprof"
 	"strings"
-	"sync"
 )
 
 var (
@@ -331,16 +329,6 @@ func Main(arch *sys.Arch, theArch Arch) {
 	// will be applied directly there.
 	bench.Start("Asmb")
 	asmb(ctxt)
-	// Generate large symbols.
-	var wg sync.WaitGroup
-	for s, f := range ctxt.generatorSyms {
-		wg.Add(1)
-		go func(f generatorFunc, s loader.Sym) {
-			defer wg.Done()
-			f(ctxt, s)
-		}(f, s)
-	}
-	wg.Wait()
 
 	// Generate additional symbols for the native symbol table just prior
 	// to code generation.