ld: fix and simplify ELF symbol generation

I started looking at this code because the nm in GNU
binutils was ignoring the first symbol in the .symtab
section.  Apparently, the System V ABI reserves the
first entry and requires all fields inside to be set
to zero.

The list of changes is as follows:

  · reserve the first symbol entry (as noted above)
  · fix the section indices for .data and .bss symbols
  · factor out common code for Elf32 and Elf64
  · remove the special case for elfsymo in [568]l/asm.c:/^asmb
  · add the "etext" symbol in 6l
  · add static symbols

R=rsc
CC=golang-dev
https://golang.org/cl/4524075

src/cmd/5l/asm.c

@@ -34,8 +34,6 @@
 #include	"../ld/lib.h"
 #include	"../ld/elf.h"
 
-int32	OFFSET;
-
 static Prog *PP;
 
 char linuxdynld[] = "/lib/ld-linux.so.2";
@@ -295,7 +293,7 @@ asmb(void)
 {
 	int32 t;
 	int a, dynsym;
-	uint32 fo, symo, startva, elfsymo, elfstro, elfsymsize;
+	uint32 fo, symo, startva;
 	ElfEhdr *eh;
 	ElfPhdr *ph, *pph;
 	ElfShdr *sh;
@@ -305,10 +303,6 @@ asmb(void)
 		Bprint(&bso, "%5.2f asmb\n", cputime());
 	Bflush(&bso);
 
-	elfsymsize = 0;
-	elfstro = 0;
-	elfsymo = 0;
-
 	sect = segtext.sect;
 	seek(cout, sect->vaddr - segtext.vaddr + segtext.fileoff, 0);
 	codeblk(sect->vaddr, sect->len);
@@ -361,27 +355,22 @@ asmb(void)
 			debug['s'] = 1;
 			break;
 		case Hplan9x32:
-			OFFSET = HEADR+textsize+segdata.filelen;
+			symo = HEADR+segtext.len+segdata.filelen;
-			seek(cout, OFFSET, 0);
 			break;
 		case Hnetbsd:
-			OFFSET += rnd(segdata.filelen, 4096);
+			symo = rnd(segdata.filelen, 4096);
-			seek(cout, OFFSET, 0);
 			break;
 		ElfSym:
 			symo = rnd(HEADR+segtext.filelen, INITRND)+segdata.filelen;
 			symo = rnd(symo, INITRND);
 			break;
 		}
+		seek(cout, symo, 0);
 		if(iself) {
 			if(debug['v'])
 			       Bprint(&bso, "%5.2f elfsym\n", cputime());
-			elfsymo = symo+8+symsize+lcsize;
+			asmelfsym();
-			seek(cout, elfsymo, 0);
-			asmelfsym32();
 			cflush();
-			elfstro = seek(cout, 0, 1);
-			elfsymsize = elfstro - elfsymo;
 			ewrite(cout, elfstrdat, elfstrsize);
 
 			// if(debug['v'])
@@ -397,8 +386,7 @@ asmb(void)
 	if(debug['v'])
 		Bprint(&bso, "%5.2f header\n", cputime());
 	Bflush(&bso);
-	OFFSET = 0;
+	seek(cout, 0L, 0);
-	seek(cout, OFFSET, 0);
 	switch(HEADTYPE) {
 	case Hnoheader:	/* no header */
 		break;
@@ -599,15 +587,15 @@ asmb(void)
 
 			sh = newElfShdr(elfstr[ElfStrSymtab]);
 			sh->type = SHT_SYMTAB;
-			sh->off = elfsymo;
+			sh->off = symo;
-			sh->size = elfsymsize;
+			sh->size = symsize;
 			sh->addralign = 4;
 			sh->entsize = 16;
 			sh->link = eh->shnum;	// link to strtab
 
 			sh = newElfShdr(elfstr[ElfStrStrtab]);
 			sh->type = SHT_STRTAB;
-			sh->off = elfstro;
+			sh->off = symo+symsize;
 			sh->size = elfstrsize;
 			sh->addralign = 1;
 

src/cmd/6l/asm.c

@@ -698,7 +698,7 @@ asmb(void)
 {
 	int32 magic;
 	int a, dynsym;
-	vlong vl, startva, symo, elfsymo, elfstro, elfsymsize, machlink;
+	vlong vl, startva, symo, machlink;
 	ElfEhdr *eh;
 	ElfPhdr *ph, *pph;
 	ElfShdr *sh;
@@ -709,9 +709,6 @@ asmb(void)
 	Bflush(&bso);
 
 	elftextsh = 0;
-	elfsymsize = 0;
-	elfstro = 0;
-	elfsymo = 0;
 	
 	if(debug['v'])
 		Bprint(&bso, "%5.2f codeblk\n", cputime());
@@ -790,36 +787,13 @@ asmb(void)
 			symo = rnd(symo, PEFILEALIGN);
 			break;
 		}
+		seek(cout, symo, 0);
 		switch(HEADTYPE) {
 		default:
 			if(iself) {
-				/*
-				 * the symbol information is stored as
-				 *	32-bit symbol table size
-				 *	32-bit line number table size
-				 *	symbol table
-				 *	line number table
-				 */
-				seek(cout, symo+8, 0);
-				if(debug['v'])
-					Bprint(&bso, "%5.2f sp\n", cputime());
-				Bflush(&bso);
-				if(debug['v'])
-					Bprint(&bso, "%5.2f pc\n", cputime());
-				Bflush(&bso);
-				if(!debug['s'])
-					strnput("", INITRND-(8+symsize+lcsize)%INITRND);
-				cflush();
 				seek(cout, symo, 0);
-				lputl(symsize);
+				asmelfsym();
-				lputl(lcsize);
 				cflush();
-				elfsymo = symo+8+symsize+lcsize;
-				seek(cout, elfsymo, 0);
-				asmelfsym64();
-				cflush();
-				elfstro = seek(cout, 0, 1);
-				elfsymsize = elfstro - elfsymo;
 				ewrite(cout, elfstrdat, elfstrsize);
 
 				if(debug['v'])
@@ -830,7 +804,6 @@ asmb(void)
 			break;
 		case Hdarwin:
 		case Hwindows:
-			seek(cout, symo, 0);
 			if(debug['v'])
 			       Bprint(&bso, "%5.2f dwarf\n", cputime());
 
@@ -1054,15 +1027,15 @@ asmb(void)
 
 			sh = newElfShdr(elfstr[ElfStrSymtab]);
 			sh->type = SHT_SYMTAB;
-			sh->off = elfsymo;
+			sh->off = symo;
-			sh->size = elfsymsize;
+			sh->size = symsize;
 			sh->addralign = 8;
 			sh->entsize = 24;
 			sh->link = eh->shnum;	// link to strtab
 
 			sh = newElfShdr(elfstr[ElfStrStrtab]);
 			sh->type = SHT_STRTAB;
-			sh->off = elfstro;
+			sh->off = symo+symsize;
 			sh->size = elfstrsize;
 			sh->addralign = 1;
 
@@ -1149,6 +1122,10 @@ genasmsym(void (*put)(Sym*, char*, int, vlong, vlong, int, Sym*))
 	Auto *a;
 	Sym *s;
 
+	s = lookup("etext", 0);
+	if(s->type == STEXT)
+		put(s, s->name, 'T', s->value, s->size, s->version, 0);
+
 	for(s=allsym; s!=S; s=s->allsym) {
 		if(s->hide)
 			continue;

src/cmd/8l/asm.c

@@ -663,7 +663,7 @@ asmb(void)
 {
 	int32 v, magic;
 	int a, dynsym;
-	uint32 symo, startva, machlink, elfsymo, elfstro, elfsymsize;
+	uint32 symo, startva, machlink;
 	ElfEhdr *eh;
 	ElfPhdr *ph, *pph;
 	ElfShdr *sh;
@@ -675,10 +675,6 @@ asmb(void)
 		Bprint(&bso, "%5.2f asmb\n", cputime());
 	Bflush(&bso);
 
-	elfsymsize = 0;
-	elfstro = 0;
-	elfsymo = 0;
-
 	sect = segtext.sect;
 	seek(cout, sect->vaddr - segtext.vaddr + segtext.fileoff, 0);
 	codeblk(sect->vaddr, sect->len);
@@ -724,10 +720,10 @@ asmb(void)
 			if(iself)
 				goto Elfsym;
 		case Hgarbunix:
-			seek(cout, rnd(HEADR+segtext.filelen, 8192)+segdata.filelen, 0);
+			symo = rnd(HEADR+segtext.filelen, 8192)+segdata.filelen;
 			break;
 		case Hunixcoff:
-			seek(cout, rnd(HEADR+segtext.filelen, INITRND)+segdata.filelen, 0);
+			symo = rnd(HEADR+segtext.filelen, INITRND)+segdata.filelen;
 			break;
 		case Hplan9x32:
 			symo = HEADR+segtext.filelen+segdata.filelen;
@@ -749,17 +745,14 @@ asmb(void)
 			symo = rnd(symo, PEFILEALIGN);
 			break;
 		}
+		seek(cout, symo, 0);
 		switch(HEADTYPE) {
 		default:
 			if(iself) {
 				if(debug['v'])
 				       Bprint(&bso, "%5.2f elfsym\n", cputime());
-				elfsymo = symo+8+symsize+lcsize;
+				asmelfsym();
-				seek(cout, elfsymo, 0);
-				asmelfsym32();
 				cflush();
-				elfstro = seek(cout, 0, 1);
-				elfsymsize = elfstro - elfsymo;
 				ewrite(cout, elfstrdat, elfstrsize);
 
 				if(debug['v'])
@@ -768,10 +761,9 @@ asmb(void)
 			}
 			break;
 		case Hplan9x32:
-			seek(cout, symo, 0);
 			asmplan9sym();
 			cflush();
-			
+
 			sym = lookup("pclntab", 0);
 			if(sym != nil) {
 				lcsize = sym->np;
@@ -783,7 +775,6 @@ asmb(void)
 			break;
 		case Hdarwin:
 		case Hwindows:
-			seek(cout, symo, 0);
 			if(debug['v'])
 				Bprint(&bso, "%5.2f dwarf\n", cputime());
 			dwarfemitdebugsections();
@@ -1110,15 +1101,15 @@ asmb(void)
 
 			sh = newElfShdr(elfstr[ElfStrSymtab]);
 			sh->type = SHT_SYMTAB;
-			sh->off = elfsymo;
+			sh->off = symo;
-			sh->size = elfsymsize;
+			sh->size = symsize;
 			sh->addralign = 4;
 			sh->entsize = 16;
 			sh->link = eh->shnum;	// link to strtab
 
 			sh = newElfShdr(elfstr[ElfStrStrtab]);
 			sh->type = SHT_STRTAB;
-			sh->off = elfstro;
+			sh->off = symo+symsize;
 			sh->size = elfstrsize;
 			sh->addralign = 1;
 

src/cmd/ld/lib.h

@@ -186,8 +186,7 @@ vlong	addsize(Sym*, Sym*);
 vlong	adduint8(Sym*, uint8);
 vlong	adduint16(Sym*, uint16);
 void	asmsym(void);
-void	asmelfsym32(void);
+void	asmelfsym(void);
-void	asmelfsym64(void);
 void	asmplan9sym(void);
 void	strnput(char*, int);
 void	dodata(void);

src/cmd/ld/symtab.c

@@ -61,11 +61,35 @@ putelfstr(char *s)
 }
 
 void
-putelfsym64(Sym *x, char *s, int t, vlong addr, vlong size, int ver, Sym *go)
+putelfsyment(int off, vlong addr, vlong size, int info, int shndx)
 {
-	int bind, type, shndx, stroff;
+	switch(thechar) {
-	
+	case '6':
-	bind = STB_GLOBAL;
+		LPUT(off);
+		cput(info);
+		cput(0);
+		WPUT(shndx);
+		VPUT(addr);
+		VPUT(size);
+		symsize += ELF64SYMSIZE;
+		break;
+	default:
+		LPUT(off);
+		LPUT(addr);
+		LPUT(size);
+		cput(info);
+		cput(0);
+		WPUT(shndx);
+		symsize += ELF32SYMSIZE;
+		break;
+	}
+}
+
+void
+putelfsym(Sym *x, char *s, int t, vlong addr, vlong size, int ver, Sym *go)
+{
+	int bind, type, shndx, off;
+
 	switch(t) {
 	default:
 		return;
@@ -75,65 +99,27 @@ putelfsym64(Sym *x, char *s, int t, vlong addr, vlong size, int ver, Sym *go)
 		break;
 	case 'D':
 		type = STT_OBJECT;
-		shndx = elftextsh + 1;
+		if((x->type&~SSUB) == SRODATA)
+			shndx = elftextsh + 1;
+		else
+			shndx = elftextsh + 2;
 		break;
 	case 'B':
 		type = STT_OBJECT;
-		shndx = elftextsh + 2;
+		shndx = elftextsh + 3;
 		break;
 	}
-	
+	bind = ver ? STB_LOCAL : STB_GLOBAL;
-	stroff = putelfstr(s);
+	off = putelfstr(s);
-	LPUT(stroff);	// string
+	putelfsyment(off, addr, size, (bind<<4)|(type&0xf), shndx);
-	cput((bind<<4)|(type&0xF));
-	cput(0);
-	WPUT(shndx);
-	VPUT(addr);
-	VPUT(size);
 }
 
 void
-asmelfsym64(void)
+asmelfsym(void)
 {
-	genasmsym(putelfsym64);
+	// the first symbol entry is reserved
-}
+	putelfsyment(0, 0, 0, (STB_LOCAL<<4)|STT_NOTYPE, 0);
-
+	genasmsym(putelfsym);
-void
-putelfsym32(Sym *x, char *s, int t, vlong addr, vlong size, int ver, Sym *go)
-{
-	int bind, type, shndx, stroff;
-	
-	bind = STB_GLOBAL;
-	switch(t) {
-	default:
-		return;
-	case 'T':
-		type = STT_FUNC;
-		shndx = elftextsh + 0;
-		break;
-	case 'D':
-		type = STT_OBJECT;
-		shndx = elftextsh + 1;
-		break;
-	case 'B':
-		type = STT_OBJECT;
-		shndx = elftextsh + 2;
-		break;
-	}
-	
-	stroff = putelfstr(s);
-	LPUT(stroff);	// string
-	LPUT(addr);
-	LPUT(size);
-	cput((bind<<4)|(type&0xF));
-	cput(0);
-	WPUT(shndx);
-}
-
-void
-asmelfsym32(void)
-{
-	genasmsym(putelfsym32);
 }
 
 void