Make profiling work for ELF. gprof now autodetects the format of

the executable file, so it will work for both a.out and ELF format
files.  I have split the object format specific code into separate
source files.  It's cleaner than it was before, but it's still
pretty crufty.

Don't cheat on your make world for this update.  A lot of things
have to be rebuilt for it to work, including the compiler and all
of the profiled libraries.

6 files changed, 345 insertions, 208 deletions

diff --git a/usr.bin/gprof/Makefile b/usr.bin/gprof/Makefile
index 9aba213..29291cc 100644
--- a/usr.bin/gprof/Makefile
+++ b/usr.bin/gprof/Makefile
@@ -1,7 +1,7 @@
 #	@(#)Makefile	8.1 (Berkeley) 6/29/93
 
 PROG=	gprof
-SRCS=	gprof.c arcs.c dfn.c lookup.c ${MACHINE_ARCH}.c hertz.c \
+SRCS=	gprof.c aout.c arcs.c dfn.c elf.c lookup.c ${MACHINE_ARCH}.c hertz.c \
 	printgprof.c printlist.c
 
 beforeinstall:
diff --git a/usr.bin/gprof/aout.c b/usr.bin/gprof/aout.c
new file mode 100644
index 0000000..940becb
--- /dev/null
+++ b/usr.bin/gprof/aout.c
@@ -0,0 +1,201 @@
+#include <a.out.h>
+
+#include "gprof.h"
+
+static void getstrtab(FILE *, const char *);
+static void getsymtab(FILE *, const char *);
+static void gettextspace(FILE *);
+static bool funcsymbol(struct nlist *);
+
+static char	*strtab;		/* string table in core */
+static long	ssiz;			/* size of the string table */
+static struct	exec xbuf;		/* exec header of a.out */
+
+/* Things which get -E excluded by default. */
+static char	*excludes[] = { "mcount", "__mcleanup", NULL };
+
+    /*
+     * Set up string and symbol tables from a.out.
+     *	and optionally the text space.
+     * On return symbol table is sorted by value.
+     *
+     * Returns 0 on success, -1 on failure.
+     */
+int
+aout_getnfile(const char *filename, char ***defaultEs)
+{
+    FILE	*nfile;
+    int		valcmp();
+
+    nfile = fopen( filename ,"r");
+    if (nfile == NULL) {
+	perror( filename );
+	done();
+    }
+    fread(&xbuf, 1, sizeof(xbuf), nfile);
+    if (N_BADMAG(xbuf)) {
+	fclose(nfile);
+	return -1;
+    }
+    getstrtab(nfile, filename);
+    getsymtab(nfile, filename);
+    gettextspace( nfile );
+    fclose(nfile);
+#   ifdef DEBUG
+	if ( debug & AOUTDEBUG ) {
+	    register int j;
+
+	    for (j = 0; j < nname; j++){
+		printf("[getnfile] 0X%08x\t%s\n", nl[j].value, nl[j].name);
+	    }
+	}
+#   endif DEBUG
+    *defaultEs = excludes;
+    return 0;
+}
+
+static void
+getstrtab(FILE *nfile, const char *filename)
+{
+
+    fseek(nfile, (long)(N_SYMOFF(xbuf) + xbuf.a_syms), 0);
+    if (fread(&ssiz, sizeof (ssiz), 1, nfile) == 0) {
+	warnx("%s: no string table (old format?)" , filename );
+	done();
+    }
+    strtab = calloc(ssiz, 1);
+    if (strtab == NULL) {
+	warnx("%s: no room for %d bytes of string table", filename , ssiz);
+	done();
+    }
+    if (fread(strtab+sizeof(ssiz), ssiz-sizeof(ssiz), 1, nfile) != 1) {
+	warnx("%s: error reading string table", filename );
+	done();
+    }
+}
+
+    /*
+     * Read in symbol table
+     */
+static void
+getsymtab(FILE *nfile, const char *filename)
+{
+    register long	i;
+    int			askfor;
+    struct nlist	nbuf;
+
+    /* pass1 - count symbols */
+    fseek(nfile, (long)N_SYMOFF(xbuf), 0);
+    nname = 0;
+    for (i = xbuf.a_syms; i > 0; i -= sizeof(struct nlist)) {
+	fread(&nbuf, sizeof(nbuf), 1, nfile);
+	if ( ! funcsymbol( &nbuf ) ) {
+	    continue;
+	}
+	nname++;
+    }
+    if (nname == 0) {
+	warnx("%s: no symbols", filename );
+	done();
+    }
+    askfor = nname + 1;
+    nl = (nltype *) calloc( askfor , sizeof(nltype) );
+    if (nl == 0) {
+	warnx("no room for %d bytes of symbol table", askfor * sizeof(nltype) );
+	done();
+    }
+
+    /* pass2 - read symbols */
+    fseek(nfile, (long)N_SYMOFF(xbuf), 0);
+    npe = nl;
+    nname = 0;
+    for (i = xbuf.a_syms; i > 0; i -= sizeof(struct nlist)) {
+	fread(&nbuf, sizeof(nbuf), 1, nfile);
+	if ( ! funcsymbol( &nbuf ) ) {
+#	    ifdef DEBUG
+		if ( debug & AOUTDEBUG ) {
+		    printf( "[getsymtab] rejecting: 0x%x %s\n" ,
+			    nbuf.n_type , strtab + nbuf.n_un.n_strx );
+		}
+#	    endif DEBUG
+	    continue;
+	}
+	npe->value = nbuf.n_value;
+	npe->name = strtab+nbuf.n_un.n_strx;
+#	ifdef DEBUG
+	    if ( debug & AOUTDEBUG ) {
+		printf( "[getsymtab] %d %s 0x%08x\n" ,
+			nname , npe -> name , npe -> value );
+	    }
+#	endif DEBUG
+	npe++;
+	nname++;
+    }
+    npe->value = -1;
+}
+
+    /*
+     *	read in the text space of an a.out file
+     */
+static void
+gettextspace(FILE *nfile)
+{
+
+    if ( cflag == 0 ) {
+	return;
+    }
+    textspace = (u_char *) malloc( xbuf.a_text );
+    if ( textspace == 0 ) {
+	warnx("ran out room for %d bytes of text space: can't do -c" ,
+		  xbuf.a_text );
+	return;
+    }
+    (void) fseek( nfile , N_TXTOFF( xbuf ) , 0 );
+    if ( fread( textspace , 1 , xbuf.a_text , nfile ) != xbuf.a_text ) {
+	warnx("couldn't read text space: can't do -c");
+	free( textspace );
+	textspace = 0;
+	return;
+    }
+}
+
+static bool
+funcsymbol(struct nlist *nlistp)
+{
+    char	*name, c;
+
+	/*
+	 *	must be a text symbol,
+	 *	and static text symbols don't qualify if aflag set.
+	 */
+    if ( ! (  ( nlistp -> n_type == ( N_TEXT | N_EXT ) )
+	   || ( ( nlistp -> n_type == N_TEXT ) && ( aflag == 0 ) ) ) ) {
+	return FALSE;
+    }
+	/*
+	 *	name must start with an underscore if uflag is set.
+	 *	can't have any `funny' characters in name,
+	 *	where `funny' means `.' (.o file names)
+	 *	need to make an exception for sparc .mul & co.
+	 *	perhaps we should just drop this code entirely...
+	 */
+    name = strtab + nlistp -> n_un.n_strx;
+    if ( uflag && *name != '_' )
+	return FALSE;
+#ifdef sparc
+    if ( *name == '.' ) {
+	char *p = name + 1;
+	if ( *p == 'u' )
+	    p++;
+	if ( strcmp ( p, "mul" ) == 0 || strcmp ( p, "div" ) == 0 ||
+	     strcmp ( p, "rem" ) == 0 )
+		return TRUE;
+    }
+#endif
+    while ( c = *name++ ) {
+	if ( c == '.' ) {
+	    return FALSE;
+	}
+    }
+    return TRUE;
+}
diff --git a/usr.bin/gprof/elf.c b/usr.bin/gprof/elf.c
new file mode 100644
index 0000000..3a59f53
--- /dev/null
+++ b/usr.bin/gprof/elf.c
@@ -0,0 +1,106 @@
+#include <sys/types.h>
+#include <sys/mman.h>
+#include <sys/stat.h>
+
+#include <elf.h>
+#include <err.h>
+#include <fcntl.h>
+#include <string.h>
+#include <unistd.h>
+
+#include "gprof.h"
+
+static bool wantsym(const Elf_Sym *, const char *);
+
+/* Things which get -E excluded by default. */
+static char	*excludes[] = { ".mcount", "_mcleanup", NULL };
+
+int
+elf_getnfile(const char *filename, char ***defaultEs)
+{
+    int fd;
+    Elf_Ehdr h;
+    struct stat s;
+    void *mapbase;
+    const char *base;
+    const Elf_Shdr *shdrs;
+    const Elf_Shdr *sh_symtab;
+    const Elf_Shdr *sh_strtab;
+    const char *strtab;
+    const Elf_Sym *symtab;
+    int symtabct;
+    int i;
+
+    if ((fd = open(filename, O_RDONLY)) == -1)
+	err(1, "%s", filename);
+    if (read(fd, &h, sizeof h) != sizeof h || !IS_ELF(h)) {
+	close(fd);
+	return -1;
+    }
+    if (fstat(fd, &s) == -1)
+	err(1, "Cannot fstat %s", filename);
+    if ((mapbase = mmap(0, s.st_size, PROT_READ, MAP_SHARED, fd, 0)) ==
+      MAP_FAILED)
+	err(1, "Cannot mmap %s", filename);
+    close(fd);
+
+    base = (const char *)mapbase;
+    shdrs = (const Elf_Shdr *)(base + h.e_shoff);
+
+    /* Find the symbol table and associated string table section. */
+    for (i = 1;  i < h.e_shnum;  i++)
+	if (shdrs[i].sh_type == SHT_SYMTAB)
+	    break;
+    if (i == h.e_shnum)
+	errx(1, "%s has no symbol table", filename);
+    sh_symtab = &shdrs[i];
+    sh_strtab = &shdrs[sh_symtab->sh_link];
+
+    symtab = (const Elf_Sym *)(base + sh_symtab->sh_offset);
+    symtabct = sh_symtab->sh_size / sh_symtab->sh_entsize;
+    strtab = (const char *)(base + sh_strtab->sh_offset);
+
+    /* Count the symbols that we're interested in. */
+    nname = 0;
+    for (i = 1;  i < symtabct;  i++)
+	if (wantsym(&symtab[i], strtab))
+	    nname++;
+
+    /* Allocate memory for them, plus a terminating entry. */
+    if ((nl = (nltype *)calloc(nname + 1, sizeof(nltype))) == NULL)
+	errx(1, "Insufficient memory for symbol table");
+
+    /* Read them in. */
+    npe = nl;
+    for (i = 1;  i < symtabct;  i++) {
+	const Elf_Sym *sym = &symtab[i];
+
+	if (wantsym(sym, strtab)) {
+	    npe->value = sym->st_value;
+	    npe->name = strtab + sym->st_name;
+	    npe++;
+	}
+    }
+    npe->value = -1;
+
+    *defaultEs = excludes;
+    return 0;
+}
+
+static bool
+wantsym(const Elf_Sym *sym, const char *strtab)
+{
+    int type;
+    int bind;
+
+    type = ELF_ST_TYPE(sym->st_info);
+    bind = ELF_ST_BIND(sym->st_info);
+
+    if (type != STT_FUNC ||
+      bind == STB_WEAK ||
+      (aflag && bind == STB_LOCAL) ||
+      (uflag && strchr(strtab + sym->st_name, '.') != NULL))
+	return 0;
+
+    return 1;
+}
diff --git a/usr.bin/gprof/gprof.1 b/usr.bin/gprof/gprof.1
index 76b5543..dc6f369 100644
--- a/usr.bin/gprof/gprof.1
+++ b/usr.bin/gprof/gprof.1
@@ -113,6 +113,7 @@ The static call graph of the program is discovered by a heuristic
 that examines the text space of the object file.
 Static-only parents or children are shown
 with call counts of 0.
+This option is not supported on some architectures.
 .It Fl C Ar count
 Find a minimal set of arcs that can be broken to eliminate all cycles with
 .Ar count
@@ -210,8 +211,14 @@ to accumulate profile data across several runs of an
 .Pa a.out
 file.
 .It Fl u
-Suppresses the printing of functions whose name does not begin with
-an underscore.
+Suppresses the printing of functions whose names are not visible to
+C programs.  For the ELF object format, this means names that
+contain the
+.Ql \&.
+character.  For the a.out object format, it means names that do not
+begin with a
+.Ql _
+character.
 All relevant information about such functions belongs to the
 (non-suppressed) function with the next lowest address.
 This is useful for eliminating "functions" that are just labels
diff --git a/usr.bin/gprof/gprof.c b/usr.bin/gprof/gprof.c
index 185e2c6..0d4d63f 100644
--- a/usr.bin/gprof/gprof.c
+++ b/usr.bin/gprof/gprof.c
@@ -42,19 +42,16 @@ static const char copyright[] =
 static char sccsid[] = "@(#)gprof.c	8.1 (Berkeley) 6/6/93";
 #endif
 static const char rcsid[] =
-	"$Id: gprof.c,v 1.6 1997/07/15 08:04:40 charnier Exp $";
+	"$Id: gprof.c,v 1.7 1998/08/08 17:48:26 jdp Exp $";
 #endif /* not lint */
 
 #include <err.h>
 #include "gprof.h"
 
-    /*
-     *	things which get -E excluded by default.
-     */
-char	*defaultEs[] = { "mcount" , "__mcleanup" , 0 };
+static int valcmp(const void *, const void *);
+
 
 static struct gmonhdr	gmonhdr;
-static	bool	uflag;
 static int lflag;
 static int Lflag;
 
@@ -64,6 +61,7 @@ main(argc, argv)
 {
     char	**sp;
     nltype	**timesortnlp;
+    char	**defaultEs;
 
     --argc;
     argv++;
@@ -160,19 +158,25 @@ main(argc, argv)
 	gmonname = GMONNAME;
     }
 	/*
+	 *	get information from the executable file.
+	 */
+    if (elf_getnfile(a_outname, &defaultEs) == -1 &&
+      aout_getnfile(a_outname, &defaultEs) == -1)
+	errx(1, "%s: bad format", a_outname);
+	/*
+	 *	sort symbol table.
+	 */
+    qsort(nl, nname, sizeof(nltype), valcmp);
+	/*
 	 *	turn off default functions
 	 */
-    for ( sp = &defaultEs[0] ; *sp ; sp++ ) {
+    for ( sp = defaultEs ; *sp ; sp++ ) {
 	Eflag = TRUE;
 	addlist( Elist , *sp );
 	eflag = TRUE;
 	addlist( elist , *sp );
     }
 	/*
-	 *	get information about a.out file.
-	 */
-    getnfile();
-	/*
 	 *	get information about mon.out file(s).
 	 */
     do	{
@@ -223,146 +227,6 @@ main(argc, argv)
 }
 
     /*
-     * Set up string and symbol tables from a.out.
-     *	and optionally the text space.
-     * On return symbol table is sorted by value.
-     */
-getnfile()
-{
-    FILE	*nfile;
-    int		valcmp();
-
-    nfile = fopen( a_outname ,"r");
-    if (nfile == NULL) {
-	perror( a_outname );
-	done();
-    }
-    fread(&xbuf, 1, sizeof(xbuf), nfile);
-    if (N_BADMAG(xbuf)) {
-	warnx("%s: bad format", a_outname );
-	done();
-    }
-    getstrtab(nfile);
-    getsymtab(nfile);
-    gettextspace( nfile );
-    qsort(nl, nname, sizeof(nltype), valcmp);
-    fclose(nfile);
-#   ifdef DEBUG
-	if ( debug & AOUTDEBUG ) {
-	    register int j;
-
-	    for (j = 0; j < nname; j++){
-		printf("[getnfile] 0X%08x\t%s\n", nl[j].value, nl[j].name);
-	    }
-	}
-#   endif DEBUG
-}
-
-getstrtab(nfile)
-    FILE	*nfile;
-{
-
-    fseek(nfile, (long)(N_SYMOFF(xbuf) + xbuf.a_syms), 0);
-    if (fread(&ssiz, sizeof (ssiz), 1, nfile) == 0) {
-	warnx("%s: no string table (old format?)" , a_outname );
-	done();
-    }
-    strtab = calloc(ssiz, 1);
-    if (strtab == NULL) {
-	warnx("%s: no room for %d bytes of string table", a_outname , ssiz);
-	done();
-    }
-    if (fread(strtab+sizeof(ssiz), ssiz-sizeof(ssiz), 1, nfile) != 1) {
-	warnx("%s: error reading string table", a_outname );
-	done();
-    }
-}
-
-    /*
-     * Read in symbol table
-     */
-getsymtab(nfile)
-    FILE	*nfile;
-{
-    register long	i;
-    int			askfor;
-    struct nlist	nbuf;
-
-    /* pass1 - count symbols */
-    fseek(nfile, (long)N_SYMOFF(xbuf), 0);
-    nname = 0;
-    for (i = xbuf.a_syms; i > 0; i -= sizeof(struct nlist)) {
-	fread(&nbuf, sizeof(nbuf), 1, nfile);
-	if ( ! funcsymbol( &nbuf ) ) {
-	    continue;
-	}
-	nname++;
-    }
-    if (nname == 0) {
-	warnx("%s: no symbols", a_outname );
-	done();
-    }
-    askfor = nname + 1;
-    nl = (nltype *) calloc( askfor , sizeof(nltype) );
-    if (nl == 0) {
-	warnx("no room for %d bytes of symbol table", askfor * sizeof(nltype) );
-	done();
-    }
-
-    /* pass2 - read symbols */
-    fseek(nfile, (long)N_SYMOFF(xbuf), 0);
-    npe = nl;
-    nname = 0;
-    for (i = xbuf.a_syms; i > 0; i -= sizeof(struct nlist)) {
-	fread(&nbuf, sizeof(nbuf), 1, nfile);
-	if ( ! funcsymbol( &nbuf ) ) {
-#	    ifdef DEBUG
-		if ( debug & AOUTDEBUG ) {
-		    printf( "[getsymtab] rejecting: 0x%x %s\n" ,
-			    nbuf.n_type , strtab + nbuf.n_un.n_strx );
-		}
-#	    endif DEBUG
-	    continue;
-	}
-	npe->value = nbuf.n_value;
-	npe->name = strtab+nbuf.n_un.n_strx;
-#	ifdef DEBUG
-	    if ( debug & AOUTDEBUG ) {
-		printf( "[getsymtab] %d %s 0x%08x\n" ,
-			nname , npe -> name , npe -> value );
-	    }
-#	endif DEBUG
-	npe++;
-	nname++;
-    }
-    npe->value = -1;
-}
-
-    /*
-     *	read in the text space of an a.out file
-     */
-gettextspace( nfile )
-    FILE	*nfile;
-{
-
-    if ( cflag == 0 ) {
-	return;
-    }
-    textspace = (u_char *) malloc( xbuf.a_text );
-    if ( textspace == 0 ) {
-	warnx("ran out room for %d bytes of text space: can't do -c" ,
-		  xbuf.a_text );
-	return;
-    }
-    (void) fseek( nfile , N_TXTOFF( xbuf ) , 0 );
-    if ( fread( textspace , 1 , xbuf.a_text , nfile ) != xbuf.a_text ) {
-	warnx("couldn't read text space: can't do -c");
-	free( textspace );
-	textspace = 0;
-	return;
-    }
-}
-    /*
      *	information from a gmon.out file is in two parts:
      *	an array of sampling hits within pc ranges,
      *	and the arcs.
@@ -531,9 +395,14 @@ dumpsum( sumfile )
     fclose( sfile );
 }
 
-valcmp(p1, p2)
-    nltype *p1, *p2;
+static int
+valcmp(v1, v2)
+    const void *v1;
+    const void *v2;
 {
+    const nltype *p1 = (const nltype *)v1;
+    const nltype *p2 = (const nltype *)v2;
+
     if ( p1 -> value < p2 -> value ) {
 	return LESSTHAN;
     }
@@ -711,48 +580,6 @@ alignentries()
     }
 }
 
-bool
-funcsymbol( nlistp )
-    struct nlist	*nlistp;
-{
-    char	*name, c;
-
-	/*
-	 *	must be a text symbol,
-	 *	and static text symbols don't qualify if aflag set.
-	 */
-    if ( ! (  ( nlistp -> n_type == ( N_TEXT | N_EXT ) )
-	   || ( ( nlistp -> n_type == N_TEXT ) && ( aflag == 0 ) ) ) ) {
-	return FALSE;
-    }
-	/*
-	 *	name must start with an underscore if uflag is set.
-	 *	can't have any `funny' characters in name,
-	 *	where `funny' means `.' (.o file names)
-	 *	need to make an exception for sparc .mul & co.
-	 *	perhaps we should just drop this code entirely...
-	 */
-    name = strtab + nlistp -> n_un.n_strx;
-    if ( uflag && *name != '_' )
-	return FALSE;
-#ifdef sparc
-    if ( *name == '.' ) {
-	char *p = name + 1;
-	if ( *p == 'u' )
-	    p++;
-	if ( strcmp ( p, "mul" ) == 0 || strcmp ( p, "div" ) == 0 ||
-	     strcmp ( p, "rem" ) == 0 )
-		return TRUE;
-    }
-#endif
-    while ( c = *name++ ) {
-	if ( c == '.' ) {
-	    return FALSE;
-	}
-    }
-    return TRUE;
-}
-
 done()
 {
 
diff --git a/usr.bin/gprof/gprof.h b/usr.bin/gprof/gprof.h
index 750fd2e..ce164d5 100644
--- a/usr.bin/gprof/gprof.h
+++ b/usr.bin/gprof/gprof.h
@@ -37,7 +37,6 @@
 #include <sys/stat.h>
 #include <sys/gmon.h>
 
-#include <a.out.h>
 #include <stdio.h>
 #include <stdlib.h>
 
@@ -121,7 +120,7 @@ typedef struct arcstruct	arctype;
      * its address, the number of calls and compute its share of cpu time.
      */
 struct nl {
-    char		*name;		/* the name */
+    const char		*name;		/* the name */
     unsigned long	value;		/* the pc entry point */
     unsigned long	svalue;		/* entry point aligned to histograms */
     double		time;		/* ticks in this routine */
@@ -217,9 +216,6 @@ double	totime;			/* total time for all routines */
 double	printtime;		/* total of time being printed */
 double	scale;			/* scale factor converting samples to pc
 				   values: each sample covers scale bytes */
-char	*strtab;		/* string table in core */
-long	ssiz;			/* size of the string table */
-struct	exec xbuf;		/* exec header of a.out */
 unsigned char	*textspace;	/* text space of a.out in core */
 int	cyclethreshold;		/* with -C, minimum cycle size to ignore */
 
@@ -237,6 +233,7 @@ bool	fflag;				/* specific functions requested */
 bool	Fflag;				/* functions requested with time */
 bool	kflag;				/* arcs to be deleted */
 bool	sflag;				/* sum multiple gmon.out files */
+bool	uflag;				/* suppress symbols hidden from C */
 bool	zflag;				/* zero time/called functions, too */
 
     /*
@@ -259,6 +256,7 @@ struct stringlist	*ktolist;
 /*
 		addarc();
 */
+int		aout_getnfile(const char *, char ***);
 int		arccmp();
 arctype		*arclookup();
 /*
@@ -280,17 +278,15 @@ bool		dfn_numbered();
 nltype		**doarcs();
 /*
 		done();
+*/
+int		elf_getnfile(const char *, char ***);
+/*
 		findcalls();
 		flatprofheader();
 		flatprofline();
 */
-bool		funcsymbol();
 /*
-		getnfile();
 		getpfile();
-		getstrtab();
-		getsymtab();
-		gettextspace();
 		gprofheader();
 		gprofline();
 		main();