kbuild: Speed up kallsyms generation

This patch basically just replaces the scripts/kallsyms script by
a scripts/kallsyms.c C program, which does the same thing much faster.

It also removes duplicates and entries which are not between _stext and
_etext, as they would not get used anyway. This saves about 290KB in
vmlinux with my .config, more than 50% of the kallsyms bloat ;)

Makefile

@@ -352,12 +352,12 @@ ifdef CONFIG_KALLSYMS
 kallsyms.o := .tmp_kallsyms2.o
 
 quiet_cmd_kallsyms = KSYM    $@
-cmd_kallsyms = sh $(KALLSYMS) $< $@
+cmd_kallsyms = $(NM) -n $< | scripts/kallsyms > $@
 
-.tmp_kallsyms1.o: .tmp_vmlinux1
-	$(call cmd,kallsyms)
+.tmp_kallsyms1.o .tmp_kallsyms2.o: %.o: %.S scripts FORCE
+	$(call if_changed_dep,as_o_S)
 
-.tmp_kallsyms2.o: .tmp_vmlinux2
+.tmp_kallsyms%.S: .tmp_vmlinux%
 	$(call cmd,kallsyms)
 
 .tmp_vmlinux1: $(vmlinux-objs) arch/$(ARCH)/vmlinux.lds.s FORCE
@@ -796,7 +796,7 @@ help:
 	@echo  ''
 	@echo  'Execute "make" or "make all" to build all targets marked with [*] '
 	@echo  'For further info browse Documentation/kbuild/*'
- 
+
 
 # Documentation targets
 # ---------------------------------------------------------------------------
@@ -842,6 +842,9 @@ a_flags = -Wp,-MD,$(depfile) $(AFLAGS) $(NOSTDINC_FLAGS) \
 quiet_cmd_as_s_S = CPP     $@
 cmd_as_s_S       = $(CPP) $(a_flags)   -o $@ $< 
 
+quiet_cmd_as_o_S = AS      $@
+cmd_as_o_S       = $(CC) $(a_flags) -c -o $@ $<
+
 # read all saved command lines
 
 targets := $(wildcard $(sort $(targets)))

kernel/kallsyms.c

@@ -45,14 +45,11 @@ const char *kallsyms_lookup(unsigned long addr,
 		for (i = 0; i < best; i++)
 			name += strlen(name)+1;
 
-		/* Base symbol size on next symbol, but beware aliases. */
-		symbol_end = (unsigned long)_etext;
-		for (i = best+1; i < kallsyms_num_syms; i++) {
-			if (kallsyms_addresses[i] != kallsyms_addresses[best]){
-				symbol_end = kallsyms_addresses[i];
-				break;
-			}
-		}
+		/* Base symbol size on next symbol. */
+		if (best + 1 < kallsyms_num_syms)
+			symbol_end = kallsyms_addresses[best + 1];
+		else
+			symbol_end = (unsigned long)_etext;
 
 		*symbolsize = symbol_end - kallsyms_addresses[best];
 		*modname = NULL;

scripts/Makefile

@@ -8,12 +8,11 @@
 # docproc: 	 Preprocess .tmpl file in order to generate .sgml documentation
 # conmakehash:	 Create arrays for initializing the kernel console tables
 
-host-progs    := fixdep split-include conmakehash docproc
+host-progs    := fixdep split-include conmakehash docproc kallsyms
 build-targets := $(host-progs)
 
 # Let clean descend into subdirs
 subdir-	:= lxdialog kconfig
 
 # fixdep is needed to compile other host programs
-$(obj)/split-include $(obj)/docproc \
-$(obj)/conmakehash: $(obj)/fixdep
+$(obj)/split-include $(obj)/docproc $(obj)/conmakehash: $(obj)/fixdep

scripts/kallsyms

-#! /bin/sh
-# Written by Rusty Russell <rusty@rustcorp.com.au> 2002.
-
-if [ $# -ne 2 ]; then
-    echo Usage: kallsyms vmlinux objfile >&2
-
-    echo Adds a .kallsyms section containing symbol info.
-    exit 1
-fi
-
-set -e
-
-# Clean up on exit.
-trap "rm -f kallsyms.map kallsyms.c $2" 0
-
-# Takes nm output from $1, produces a .c file on standard output.
-encode_symbols()
-{
-    # First take addresses.
-    echo "unsigned long kallsyms_addresses[] = {"
-    sed 's/^[	]*\([A-Fa-f0-9]*\).*/0x\1UL,/' < $1
-    echo "};"
-
-    # Now output size.
-    echo "unsigned long kallsyms_num_syms = `wc -l < $1`;"
-
-    # Now output names.
-    echo "char kallsyms_names[] = ";
-    sed 's/^[         ]*[A-Fa-f0-9]*[         ]*.[    ]\(.*\)/"\1\\0"/' < $1
-    echo ";"
-}
-
-# FIXME: Use System.map as input, and regenerate each time in Makefile.
-$NM -n $1 | grep -v '\(compiled\)\|\(\.o$$\)\|\( [aUw] \)\|\(\.\.ng$$\)\|\(LASH[RL]DI\)' | sort > kallsyms.map
-
-encode_symbols kallsyms.map > kallsyms.c
-$CC $CFLAGS -c -o $2 kallsyms.c
-
-trap "rm -f kallsyms.map kallsyms.c" 0
-exit 0

scripts/kallsyms.c

+/* Generate assembler source containing symbol information
+ *
+ * Copyright 2002       by Kai Germaschewski
+ *
+ * This software may be used and distributed according to the terms
+ * of the GNU General Public License, incorporated herein by reference.
+ *
+ * Usage: nm -n vmlinux | scripts/kallsyms > symbols.S
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+struct sym_entry {
+	unsigned long long addr;
+	char type;
+	char *sym;
+};
+
+
+static struct sym_entry *table;
+static int size, cnt;
+static unsigned long long _stext, _etext;
+
+static void
+usage(void)
+{
+	fprintf(stderr, "Usage: kallsyms < in.map > out.S\n");
+	exit(1);
+}
+
+static int
+read_symbol(FILE *in, struct sym_entry *s)
+{
+	char str[500];
+	int rc;
+
+	rc = fscanf(in, "%llx %c %499s\n", &s->addr, &s->type, str);
+	if (rc != 3) {
+		if (rc != EOF) {
+			/* skip line */
+			fgets(str, 500, in);
+		}
+		return -1;
+	}
+	s->sym = strdup(str);
+	return 0;
+}
+
+static int
+symbol_valid(struct sym_entry *s)
+{
+	if (s->addr < _stext)
+		return 0;
+
+	if (s->addr > _etext)
+		return 0;
+
+	if (strstr(s->sym, "_compiled."))
+		return 0;
+
+	return 1;
+}
+
+static void
+read_map(FILE *in)
+{
+	int i;
+
+	while (!feof(in)) {
+		if (cnt >= size) {
+			size += 10000;
+			table = realloc(table, sizeof(*table) * size);
+			if (!table) {
+				fprintf(stderr, "out of memory\n");
+				exit (1);
+			}
+		}
+		if (read_symbol(in, &table[cnt]) == 0)
+			cnt++;
+	}
+	for (i = 0; i < cnt; i++) {
+		if (strcmp(table[i].sym, "_stext") == 0)
+			_stext = table[i].addr;
+		if (strcmp(table[i].sym, "_etext") == 0)
+			_etext = table[i].addr;
+	}
+}
+
+static void
+write_src(void)
+{
+	unsigned long long last_addr;
+	int i, valid = 0;
+
+	printf(".data\n");
+
+	printf(".globl kallsyms_addresses\n");
+	printf("\t.align 8\n");
+	printf("kallsyms_addresses:\n");
+	for (i = 0, last_addr = 0; i < cnt; i++) {
+		if (!symbol_valid(&table[i]))
+			continue;
+		
+		if (table[i].addr == last_addr)
+			continue;
+
+		printf("\t.long\t%#llx\n", table[i].addr);
+		valid++;
+		last_addr = table[i].addr;
+	}
+	printf("\n");
+
+	printf(".globl kallsyms_num_syms\n");
+	printf("\t.align 8\n");
+	printf("\t.long\t%d\n", valid);
+	printf("\n");
+
+	printf(".globl kallsyms_names\n");
+	printf(".data\n");
+	printf("\t.align 8\n");
+	printf("kallsyms_names:\n");
+	for (i = 0, last_addr = 0; i < cnt; i++) {
+		if (!symbol_valid(&table[i]))
+			continue;
+		
+		if (table[i].addr == last_addr)
+			continue;
+
+		printf("\t.string\t\"%s\"\n", table[i].sym);
+		last_addr = table[i].addr;
+	}
+	printf("\n");
+}
+
+int
+main(int argc, char **argv)
+{
+	if (argc != 1)
+		usage();
+
+	read_map(stdin);
+	write_src();
+
+	return 0;
+}
+