From c595699ca75f88245cdecb86c91d53d11bbf69f6 Mon Sep 17 00:00:00 2001
From: Ben Gras <ben@minix3.org>
Date: Wed, 26 Sep 2012 18:29:11 +0200
Subject: [PATCH] unstack, sort: cleanup and improvement

lets unstack
(a) know about in-kernel ipc entry points and
(b) be able handle >2GB symbol offsets.

	. sort: add -x for hex numerical sort
	. unstack: gnm is obsolete
	. unstack: datasizes is obsolete (use nm --size-sort instead)
	. unstack: add ipc entry points read from procfs (hex)
	. unstack: use sort -x to sort symbol order so the procfs ones are
	  sorted independent of position and original ordering
---
 commands/sort/sort.c        | 26 ++++++++++++++++
 commands/unstack/unstack.sh | 61 ++++++++++++++++---------------------
 man/man1/sort.1             |  5 ++-
 3 files changed, 56 insertions(+), 36 deletions(-)

diff --git a/commands/sort/sort.c b/commands/sort/sort.c
index e8497d433..5481d2779 100644
--- a/commands/sort/sort.c
+++ b/commands/sort/sort.c
@@ -96,6 +96,7 @@ typedef struct {
   BOOL fold_case;
   BOOL ascii;
   BOOL numeric;
+  BOOL hexmode;
 } FIELD;
 
 /* Field declarations. A total of FILEDS_LIMIT is allowed */
@@ -150,6 +151,7 @@ char *skip_fields(char *str, int nf);
 int compare(char *el1, char *el2);
 int cmp(unsigned char *el1, unsigned char *el2, FIELD * field);
 int digits(char *str1, char *str2, BOOL check_sign);
+int hexits(char *str1, char *str2);
 void files_merge(int file_cnt);
 void merge(int start_file, int limit_file);
 void put_line(char *line);
@@ -250,6 +252,10 @@ register FIELD *field;
 	field->numeric = TRUE;
 	field->blanks = TRUE;
 	break;
+      case 'x':
+	field->hexmode = TRUE;
+	field->blanks = TRUE;
+	break;
       case 'r':			/* Reverse comparisons */
 	field->reverse = TRUE;
 	break;
@@ -761,6 +767,8 @@ FIELD *field;
   }
   if (field->numeric)		/* Compare numeric */
 	return digits((char *) el1, (char *) el2, TRUE);
+  if (field->hexmode)		/* Compare hex */
+	return hexits((char *) el1, (char *) el2);
 
   for (;;) {
 	while (*el1 == *el2) {
@@ -811,6 +819,24 @@ FIELD *field;
   /* NOTREACHED */
 }
 
+int hexits(char *str1, char *str2)
+{
+	unsigned long v1, v2;
+	int r1, r2;
+	r1 = sscanf(str1, "0x%lx", &v1);
+	r2 = sscanf(str2, "0x%lx", &v2);
+
+	/* ordering based on reasonable hex number */
+	if(r1 == 1 && r2 != 1) return HIGHER;
+	if(r1 != 1 && r2 == 1) return LOWER;
+	if(r1 != 1 && r2 != 1) return SAME;
+
+	if(v1 > v2) return HIGHER;
+	if(v1 < v2) return LOWER;
+
+	return SAME;
+}
+
 /*
  * Digits compares () the two strings that point to a number of digits followed
  * by an optional decimal point.
diff --git a/commands/unstack/unstack.sh b/commands/unstack/unstack.sh
index bb781c624..118913216 100644
--- a/commands/unstack/unstack.sh
+++ b/commands/unstack/unstack.sh
@@ -3,17 +3,23 @@
 # Look at /usr/pkg/bin first in case there is an old nm in /usr/bin
 PATH=/usr/pkg/bin:$PATH:/usr/gnu/bin
 
+# Does procfs give us some extra 'symbols'?
+IPCVECS=/proc/ipcvecs
+if [ -f $IPCVECS ]
+then	EXTRANM="cat $IPCVECS"
+fi
+
 # Check usage
 if [ $# -lt 1 ]
 then	echo "Usage: unstack <executable> [0x... [0x... ] ]"
-	echo "       datasizes <executable>"
 	exit 1
 fi
 
 # Check invocation mode
 case "`basename $0`" in
 	datasizes)
-		mode=data
+		echo "datasizes is obsolete; please use nm --size-sort instead."
+		exit 1
 		;;
 	unstack)
 		mode=stack
@@ -28,46 +34,31 @@ esac
 executable=$1
 shift
 
-# gnu nm can be gnm or nm
-if which gnm >/dev/null 2>&1
-then	GNM=gnm
-else	GNM=nm
+if ! which gawk >/dev/null 2>&1
+then	echo "Please install gawk."
+	exit 1
 fi
 
-# Invoke gnu nm or ack nm?
-if file $executable | grep NSYM >/dev/null 2>&1
-then	NM="$GNM --radix=d"
-elif file $executable | grep ELF >/dev/null 2>&1
-then	NM="$GNM --radix=d"
-else	NM="acknm -d"
+# Invoke binutils nm or ack nm?
+if file $executable | grep ELF >/dev/null 2>&1
+then	NM="nm"
+else	NM="acknm"
 fi
 
-# Invoked as unstack?
-if [ $mode = stack ]
-then
-	while [ $# -gt 0 ]
-	do	dec="`printf %d $1`"
-		$NM -n $executable | grep ' [Tt] [^.]' | awk '
-		  {  if($1 > '$dec') { printf "%s+0x%x\n", name, '$dec'-offset; exit }
-		     name=$3; offset=$1
-		  }'
-		shift
-	done
+SYMLIST=/tmp/unstack.$$
 
-	exit 0
-fi
+# store sorted, filtered nm output once
+( $NM $executable ; $EXTRANM ) | sed 's/^/0x/' | sort -x | grep ' [Tt] [^.]' >$SYMLIST
 
-# Invoked as datasizes?
-if [ $mode = data ]
-then
-	$NM -n $executable |
-		grep ' [bBdD] [^.]' | awk '{ if (lastpos) printf "%10ld kB  %s\n", ($1-lastpos)/1024, lastname; lastpos=$1; lastname=$3 }' | sort -n
+while [ $# -gt 0 ]
+do	 gawk <$SYMLIST --non-decimal-data -v symoffset=$1 '
+	  {  if($1 > symoffset) { printf "%s+0x%x\n", name, symoffset-prevoffset; exit }
+	     name=$3; prevoffset=$1;
+	  }'
+	shift
+done
 
-	exit 0
-fi
-
-# Can't happen.
-echo "Impossible invocation."
+rm -f $SYMLIST
 
 exit 1
 
diff --git a/man/man1/sort.1 b/man/man1/sort.1
index 1bc30edd9..9d276ae9c 100644
--- a/man/man1/sort.1
+++ b/man/man1/sort.1
@@ -35,7 +35,10 @@ sort \- sort a file of ASCII lines
 # Merge presorted files
 .TP 5
 .B \-n
-# Numeric sort order
+# Numeric sort order (decimal)
+.TP 5
+.B \-x
+# Numeric sort order (hex)
 .TP 5
 .B \-o
 # Next argument is output file