[XEN] Make labels _start,_end,_stext,_etext,_sinittext,_einittext generic.

Also sync the tools/symbol.c symbol table generator with Linux.
Make section names generic (e.e.g, .init.text, .init.data, ...).
Signed-off-by: Keir Fraser <keir@xensource.com>

xen/arch/ia64/xen/xenmisc.c

@@ -56,26 +56,6 @@ is_platform_hp_ski(void)
 
 struct pt_regs *guest_cpu_user_regs(void) { return vcpu_regs(current); }
 
-///////////////////////////////
-// from arch/ia64/traps.c
-///////////////////////////////
-
-int is_kernel_text(unsigned long addr)
-{
-	extern char _stext[], _etext[];
-	if (addr >= (unsigned long) _stext &&
-	    addr <= (unsigned long) _etext)
-	    return 1;
-
-	return 0;
-}
-
-unsigned long kernel_text_end(void)
-{
-	extern char _etext[];
-	return (unsigned long) _etext;
-}
-
 ///////////////////////////////
 // from common/keyhandler.c
 ///////////////////////////////

xen/arch/powerpc/setup.c

@@ -91,19 +91,6 @@ static struct domain *idle_domain;
 
 volatile struct processor_area * volatile global_cpu_table[NR_CPUS];
 
-int is_kernel_text(unsigned long addr)
-{
-    if (addr >= (unsigned long) &_start &&
-        addr <= (unsigned long) &_etext)
-        return 1;
-    return 0;
-}
-
-unsigned long kernel_text_end(void)
-{
-    return (unsigned long) &_etext;
-}
-
 static void __init do_initcalls(void)
 {
     initcall_t *call;

xen/arch/powerpc/xen.lds.S

@@ -113,10 +113,10 @@ SECTIONS
 
   . = ALIGN(32);
   __setup_start = .;
-  .setup.init : { *(.setup.init) }
+  .init.setup : { *(.init.setup) }
   __setup_end = .;
   __initcall_start = .;
-  .initcall.init : { *(.initcall.init) }
+  .initcall.init : { *(.initcall1.init) }
   __initcall_end = .;
   __inithcall_start = .;
   .inithcall.text : { *(.inithcall.text) }

xen/arch/x86/boot/x86_32.S

@@ -11,8 +11,6 @@
         .text
 
 ENTRY(start)
-ENTRY(stext)
-ENTRY(_stext)
         jmp __start
 
         .align 4

xen/arch/x86/boot/x86_64.S

@@ -14,8 +14,6 @@
 #define SYM_PHYS(sym) (sym - __PAGE_OFFSET)
 
 ENTRY(start)
-ENTRY(stext)
-ENTRY(_stext)
         jmp __start
 
         .org    0x004

xen/arch/x86/traps.c

@@ -115,22 +115,6 @@ integer_param("debug_stack_lines", debug_stack_lines);
 #define ESP_BEFORE_EXCEPTION(regs) ((unsigned long *)regs->rsp)
 #endif
 
-int is_kernel_text(unsigned long addr)
-{
-    extern char _stext, _etext;
-    if (addr >= (unsigned long) &_stext &&
-        addr <= (unsigned long) &_etext)
-        return 1;
-    return 0;
-
-}
-
-unsigned long kernel_text_end(void)
-{
-    extern char _etext;
-    return (unsigned long) &_etext;
-}
-
 static void show_guest_stack(struct cpu_user_regs *regs)
 {
     int i;
@@ -187,7 +171,7 @@ static void show_trace(struct cpu_user_regs *regs)
     while ( ((long)stack & (STACK_SIZE-BYTES_PER_LONG)) != 0 )
     {
         addr = *stack++;
-        if ( is_kernel_text(addr) )
+        if ( is_kernel_text(addr) || is_kernel_inittext(addr) )
         {
             printk("[<%p>]", _p(addr));
             print_symbol(" %s\n   ", addr);
@@ -316,7 +300,7 @@ void show_stack_overflow(unsigned long esp)
     while ( ((long)stack & (STACK_SIZE-BYTES_PER_LONG)) != 0 )
     {
         addr = *stack++;
-        if ( is_kernel_text(addr) )
+        if ( is_kernel_text(addr) || is_kernel_inittext(addr) )
         {
             printk("%p: [<%p>]", stack, _p(addr));
             print_symbol(" %s\n   ", addr);

xen/arch/x86/x86_32/xen.lds.S

@@ -19,7 +19,8 @@ PHDRS
 SECTIONS
 {
   . = 0xFF000000 + 0x100000;
-  _text = .;			/* Text and read-only data */
+  _start = .;
+  _stext = .;			/* Text and read-only data */
   .text : {
 	*(.text)
 	*(.fixup)
@@ -51,14 +52,16 @@ SECTIONS
 
   . = ALIGN(4096);		/* Init code and data */
   __init_begin = .;
-  .text.init : { *(.text.init) } :text
-  .data.init : { *(.data.init) } :text
+  _sinittext = .;
+  .init.text : { *(.init.text) } :text
+  _einittext = .;
+  .init.data : { *(.init.data) } :text
   . = ALIGN(32);
   __setup_start = .;
-  .setup.init : { *(.setup.init) } :text
+  .init.setup : { *(.init.setup) } :text
   __setup_end = .;
   __initcall_start = .;
-  .initcall.init : { *(.initcall.init) } :text
+  .initcall.init : { *(.initcall1.init) } :text
   __initcall_end = .;
   . = ALIGN(PAGE_SIZE);
   __init_end = .;
@@ -80,8 +83,8 @@ SECTIONS
 
   /* Sections to be discarded */
   /DISCARD/ : {
-	*(.text.exit)
-	*(.data.exit)
+	*(.exit.text)
+	*(.exit.data)
 	*(.exitcall.exit)
 	}
 

xen/arch/x86/x86_64/xen.lds.S

@@ -17,7 +17,8 @@ PHDRS
 SECTIONS
 {
   . = 0xFFFF830000100000;
-  _text = .;			/* Text and read-only data */
+  _start = .;
+  _stext = .;			/* Text and read-only data */
   .text : {
 	*(.text)
 	*(.fixup)
@@ -49,14 +50,16 @@ SECTIONS
 
   . = ALIGN(4096);		/* Init code and data */
   __init_begin = .;
-  .text.init : { *(.text.init) } :text
-  .data.init : { *(.data.init) } :text
+  _sinittext = .;
+  .init.text : { *(.init.text) } :text
+  _einittext = .;
+  .init.data : { *(.init.data) } :text
   . = ALIGN(32);
   __setup_start = .;
-  .setup.init : { *(.setup.init) } :text
+  .init.setup : { *(.init.setup) } :text
   __setup_end = .;
   __initcall_start = .;
-  .initcall.init : { *(.initcall.init) } :text
+  .initcall.init : { *(.initcall1.init) } :text
   __initcall_end = .;
   . = ALIGN(PAGE_SIZE);
   __init_end = .;
@@ -78,8 +81,8 @@ SECTIONS
 
   /* Sections to be discarded */
   /DISCARD/ : {
-	*(.text.exit)
-	*(.data.exit)
+	*(.exit.text)
+	*(.exit.data)
 	*(.exitcall.exit)
 	}
 

xen/common/kexec.c

@@ -150,12 +150,10 @@ static int kexec_get_reserve(xen_kexec_range_t *range)
     return 0;
 }
 
-extern unsigned long _text;
-
 static int kexec_get_xen(xen_kexec_range_t *range)
 {
-    range->start = virt_to_maddr(&_text);
-    range->size = (unsigned long)&_end - (unsigned long)&_text;
+    range->start = virt_to_maddr(_start);
+    range->size = (unsigned long)_end - (unsigned long)_start;
     return 0;
 }
 

xen/common/symbols.c

@@ -12,6 +12,7 @@
 
 #include <xen/config.h>
 #include <xen/symbols.h>
+#include <xen/kernel.h>
 #include <xen/init.h>
 #include <xen/lib.h>
 #include <xen/string.h>
@@ -99,7 +100,7 @@ const char *symbols_lookup(unsigned long addr,
     namebuf[KSYM_NAME_LEN] = 0;
     namebuf[0] = 0;
 
-    if (!is_kernel_text(addr))
+    if (!is_kernel_text(addr) && !is_kernel_inittext(addr))
         return NULL;
 
         /* do a binary search on the sorted symbols_addresses array */
@@ -130,7 +131,8 @@ const char *symbols_lookup(unsigned long addr,
 
     /* if we found no next symbol, we use the end of the section */
     if (!symbol_end)
-        symbol_end = kernel_text_end();
+        symbol_end = is_kernel_inittext(addr) ?
+            (unsigned long)_einittext : (unsigned long)_etext;
 
     *symbolsize = symbol_end - symbols_addresses[low];
     *offset = addr - symbols_addresses[low];

xen/include/asm-ia64/init.h

@@ -1,29 +1,4 @@
 #ifndef _XEN_ASM_INIT_H
 #define _XEN_ASM_INIT_H
 
-/*
- * Mark functions and data as being only used at initialization
- * or exit time.
- */
-#define __init       \
-    __attribute__ ((__section__ (".init.text")))
-#define __exit       \
-    __attribute_used__ __attribute__ ((__section__(".text.exit")))
-#define __initdata   \
-    __attribute__ ((__section__ (".init.data")))
-#define __exitdata   \
-    __attribute_used__ __attribute__ ((__section__ (".data.exit")))
-#define __initsetup  \
-    __attribute_used__ __attribute__ ((__section__ (".init.setup")))
-#define __init_call  \
-    __attribute_used__ __attribute__ ((__section__ (".initcall1.init")))
-#define __exit_call  \
-    __attribute_used__ __attribute__ ((__section__ (".exitcall.exit")))
-
-/* For assembly routines
-#define __INIT		.section	".text.init","ax"
-#define __FINIT		.previous
-#define __INITDATA	.section	".data.init","aw"
-*/
-
 #endif /* _XEN_ASM_INIT_H */

xen/include/asm-powerpc/init.h

@@ -21,25 +21,6 @@
 #ifndef _XEN_ASM_INIT_H
 #define _XEN_ASM_INIT_H
 
-/*
- * Mark functions and data as being only used at initialization
- * or exit time.
- */
-#define __init       \
-    __attribute__ ((__section__ (".init.text")))
-#define __exit       \
-    __attribute_used__ __attribute__ ((__section__(".text.exit")))
-#define __initdata   \
-    __attribute__ ((__section__ (".init.data")))
-#define __exitdata   \
-    __attribute_used__ __attribute__ ((__section__ (".data.exit")))
-#define __initsetup  \
-    __attribute_used__ __attribute__ ((__section__ (".setup.init")))
-#define __init_call  \
-    __attribute_used__ __attribute__ ((__section__ (".initcall.init")))
-#define __exit_call  \
-    __attribute_used__ __attribute__ ((__section__ (".exitcall.exit")))
-
 struct cpu_user_regs;
 typedef void (*hcall_handler_t)(struct cpu_user_regs *regs);
 

xen/include/asm-x86/config.h

@@ -84,10 +84,6 @@
 
 #define CONFIG_DMA_BITSIZE 30
 
-#ifndef __ASSEMBLY__
-extern unsigned long _end; /* standard ELF symbol */
-#endif /* __ASSEMBLY__ */
-
 #if defined(__x86_64__)
 
 #define CONFIG_X86_64 1

xen/include/asm-x86/init.h

@@ -1,29 +1,4 @@
 #ifndef _XEN_ASM_INIT_H
 #define _XEN_ASM_INIT_H
 
-/*
- * Mark functions and data as being only used at initialization
- * or exit time.
- */
-#define __init       \
-    __attribute__ ((__section__ (".init.text")))
-#define __exit       \
-    __attribute_used__ __attribute__ ((__section__(".text.exit")))
-#define __initdata   \
-    __attribute__ ((__section__ (".init.data")))
-#define __exitdata   \
-    __attribute_used__ __attribute__ ((__section__ (".data.exit")))
-#define __initsetup  \
-    __attribute_used__ __attribute__ ((__section__ (".setup.init")))
-#define __init_call  \
-    __attribute_used__ __attribute__ ((__section__ (".initcall.init")))
-#define __exit_call  \
-    __attribute_used__ __attribute__ ((__section__ (".exitcall.exit")))
-
-/* For assembly routines
-#define __INIT		.section	".text.init","ax"
-#define __FINIT		.previous
-#define __INITDATA	.section	".data.init","aw"
-*/
-
 #endif /* _XEN_ASM_INIT_H */

xen/include/xen/init.h

@@ -4,6 +4,25 @@
 #include <xen/config.h>
 #include <asm/init.h>
 
+/*
+ * Mark functions and data as being only used at initialization
+ * or exit time.
+ */
+#define __init       \
+    __attribute__ ((__section__ (".init.text")))
+#define __exit       \
+    __attribute_used__ __attribute__ ((__section__(".exit.text")))
+#define __initdata   \
+    __attribute__ ((__section__ (".init.data")))
+#define __exitdata   \
+    __attribute_used__ __attribute__ ((__section__ (".exit.data")))
+#define __initsetup  \
+    __attribute_used__ __attribute__ ((__section__ (".init.setup")))
+#define __init_call  \
+    __attribute_used__ __attribute__ ((__section__ (".initcall1.init")))
+#define __exit_call  \
+    __attribute_used__ __attribute__ ((__section__ (".exitcall.exit")))
+
 /* These macros are used to mark some functions or 
  * initialized data (doesn't apply to uninitialized data)
  * as `initialization' functions. The kernel can take this

xen/include/xen/kernel.h

@@ -56,6 +56,23 @@
 	1;                                      \
 })
 
+extern char _start[], _end[];
+#define is_kernel(p) ({                         \
+    char *__p = (char *)(unsigned long)(p);     \
+    (__p >= _start) && (__p <= _end);           \
+})
+
+extern char _stext[], _etext[];
+#define is_kernel_text(p) ({                    \
+    char *__p = (char *)(unsigned long)(p);     \
+    (__p >= _stext) && (__p <= _etext);         \
+})
+
+extern char _sinittext[], _einittext[];
+#define is_kernel_inittext(p) ({                \
+    char *__p = (char *)(unsigned long)(p);     \
+    (__p >= _sinittext) && (__p <= _einittext); \
+})
 
 #endif /* _LINUX_KERNEL_H */
 

xen/include/xen/symbols.h

@@ -6,9 +6,6 @@
 
 #define KSYM_NAME_LEN 127
 
-extern int is_kernel_text(unsigned long addr);
-extern unsigned long kernel_text_end(void);
-
 /* Lookup an address. */
 const char *symbols_lookup(unsigned long addr,
                            unsigned long *symbolsize,
@@ -16,7 +13,7 @@ const char *symbols_lookup(unsigned long addr,
                            char *namebuf);
 
 /* Replace "%s" in format with address, if found */
-extern void __print_symbol(const char *fmt, unsigned long address);
+void __print_symbol(const char *fmt, unsigned long address);
 
 /* This macro allows us to keep printk typechecking */
 static void __check_printsym_format(const char *fmt, ...)

xen/tools/symbols.c

@@ -5,7 +5,7 @@
  * This software may be used and distributed according to the terms
  * of the GNU General Public License, incorporated herein by reference.
  *
- * Usage: nm -n <object-file> | scripts/symbols [--all-symbols] > symbols.S
+ * Usage: nm -n vmlinux | scripts/symbols [--all-symbols] > symbols.S
  *
  * ChangeLog:
  *
@@ -24,75 +24,37 @@
  *
  */
 
+#define _GNU_SOURCE
+
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #include <ctype.h>
 
-/* maximum token length used. It doesn't pay to increase it a lot, because
- * very long substrings probably don't repeat themselves too often. */
-#define MAX_TOK_SIZE		11
 #define KSYM_NAME_LEN		127
 
-/* we use only a subset of the complete symbol table to gather the token count,
- * to speed up compression, at the expense of a little compression ratio */
-#define WORKING_SET		1024
-
-/* first find the best token only on the list of tokens that would profit more
- * than GOOD_BAD_THRESHOLD. Only if this list is empty go to the "bad" list.
- * Increasing this value will put less tokens on the "good" list, so the search
- * is faster. However, if the good list runs out of tokens, we must painfully
- * search the bad list. */
-#define GOOD_BAD_THRESHOLD	10
-
-/* token hash parameters */
-#define HASH_BITS		18
-#define HASH_TABLE_SIZE		(1 << HASH_BITS)
-#define HASH_MASK		(HASH_TABLE_SIZE - 1)
-#define HASH_BASE_OFFSET	2166136261U
-#define HASH_FOLD(a)		((a)&(HASH_MASK))
-
-/* flags to mark symbols */
-#define SYM_FLAG_VALID		1
-#define SYM_FLAG_SAMPLED	2
 
 struct sym_entry {
 	unsigned long long addr;
-	char type;
-	unsigned char flags;
-	unsigned char len;
+	unsigned int len;
 	unsigned char *sym;
 };
 
 
 static struct sym_entry *table;
-static int size, cnt;
-static unsigned long long _stext, _etext;
+static unsigned int table_size, table_cnt;
+static unsigned long long _stext, _etext, _sinittext, _einittext, _sextratext, _eextratext;
 static int all_symbols = 0;
 static char symbol_prefix_char = '\0';
 
-struct token {
-	unsigned char data[MAX_TOK_SIZE];
-	unsigned char len;
-	/* profit: the number of bytes that could be saved by inserting this
-	 * token into the table */
-	int profit;
-	struct token *next;	/* next token on the hash list */
-	struct token *right;	/* next token on the good/bad list */
-	struct token *left;    /* previous token on the good/bad list */
-	struct token *smaller; /* token that is less one letter than this one */
-	};
-
-struct token bad_head, good_head;
-struct token *hash_table[HASH_TABLE_SIZE];
+int token_profit[0x10000];
 
 /* the table that holds the result of the compression */
-unsigned char best_table[256][MAX_TOK_SIZE+1];
+unsigned char best_table[256][2];
 unsigned char best_table_len[256];
 
 
-static void
-usage(void)
+static void usage(void)
 {
 	fprintf(stderr, "Usage: symbols [--all-symbols] [--symbol-prefix=<prefix char>] < in.map > out.S\n");
 	exit(1);
@@ -102,21 +64,19 @@ usage(void)
  * This ignores the intensely annoying "mapping symbols" found
  * in ARM ELF files: $a, $t and $d.
  */
-static inline int
-is_arm_mapping_symbol(const char *str)
+static inline int is_arm_mapping_symbol(const char *str)
 {
 	return str[0] == '$' && strchr("atd", str[1])
 	       && (str[2] == '\0' || str[2] == '.');
 }
 
-static int
-read_symbol(FILE *in, struct sym_entry *s)
+static int read_symbol(FILE *in, struct sym_entry *s)
 {
 	char str[500];
-	char *sym;
+	char *sym, stype;
 	int rc;
 
-	rc = fscanf(in, "%llx %c %499s\n", &s->addr, &s->type, str);
+	rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, str);
 	if (rc != 3) {
 		if (rc != EOF) {
 			/* skip line */
@@ -135,7 +95,15 @@ read_symbol(FILE *in, struct sym_entry *s)
 		_stext = s->addr;
 	else if (strcmp(sym, "_etext") == 0)
 		_etext = s->addr;
-	else if (toupper(s->type) == 'A')
+	else if (strcmp(sym, "_sinittext") == 0)
+		_sinittext = s->addr;
+	else if (strcmp(sym, "_einittext") == 0)
+		_einittext = s->addr;
+	else if (strcmp(sym, "_sextratext") == 0)
+		_sextratext = s->addr;
+	else if (strcmp(sym, "_eextratext") == 0)
+		_eextratext = s->addr;
+	else if (toupper(stype) == 'A')
 	{
 		/* Keep these useful absolute symbols */
 		if (strcmp(sym, "__kernel_syscall_via_break") &&
@@ -145,22 +113,24 @@ read_symbol(FILE *in, struct sym_entry *s)
 			return -1;
 
 	}
-	else if (toupper(s->type) == 'U' ||
+	else if (toupper(stype) == 'U' ||
 		 is_arm_mapping_symbol(sym))
 		return -1;
+	/* exclude also MIPS ELF local symbols ($L123 instead of .L123) */
+	else if (str[0] == '$')
+		return -1;
 
 	/* include the type field in the symbol name, so that it gets
 	 * compressed together */
 	s->len = strlen(str) + 1;
-	s->sym = (unsigned char *) malloc(s->len + 1);
+	s->sym = malloc(s->len + 1);
 	strcpy((char *)s->sym + 1, str);
-	s->sym[0] = s->type;
+	s->sym[0] = stype;
 
 	return 0;
 }
 
-static int
-symbol_valid(struct sym_entry *s)
+static int symbol_valid(struct sym_entry *s)
 {
 	/* Symbols which vary between passes.  Passes 1 and 2 must have
 	 * identical symbol lists.  The symbols_* symbols below are only added
@@ -189,7 +159,9 @@ symbol_valid(struct sym_entry *s)
 	/* if --all-symbols is not specified, then symbols outside the text
 	 * and inittext sections are discarded */
 	if (!all_symbols) {
-		if (s->addr < _stext || s->addr > _etext)
+		if ((s->addr < _stext || s->addr > _etext)
+		    && (s->addr < _sinittext || s->addr > _einittext)
+		    && (s->addr < _sextratext || s->addr > _eextratext))
 			return 0;
 		/* Corner case.  Discard any symbols with the same value as
 		 * _etext _einittext or _eextratext; they can move between pass
@@ -197,7 +169,9 @@ symbol_valid(struct sym_entry *s)
 		 * move then they may get dropped in pass 2, which breaks the
 		 * symbols rules.
 		 */
-		if (s->addr == _etext && strcmp((char *)s->sym + offset, "_etext"))
+		if ((s->addr == _etext && strcmp((char*)s->sym + offset, "_etext")) ||
+		    (s->addr == _einittext && strcmp((char*)s->sym + offset, "_einittext")) ||
+		    (s->addr == _eextratext && strcmp((char*)s->sym + offset, "_eextratext")))
 			return 0;
 	}
 
@@ -212,20 +186,19 @@ symbol_valid(struct sym_entry *s)
 	return 1;
 }
 
-static void
-read_map(FILE *in)
+static void read_map(FILE *in)
 {
 	while (!feof(in)) {
-		if (cnt >= size) {
-			size += 10000;
-			table = realloc(table, sizeof(*table) * size);
+		if (table_cnt >= table_size) {
+			table_size += 10000;
+			table = realloc(table, sizeof(*table) * table_size);
 			if (!table) {
 				fprintf(stderr, "out of memory\n");
 				exit (1);
 			}
 		}
-		if (read_symbol(in, &table[cnt]) == 0)
-			cnt++;
+		if (read_symbol(in, &table[table_cnt]) == 0)
+			table_cnt++;
 	}
 }
 
@@ -269,10 +242,9 @@ static int expand_symbol(unsigned char *data, int len, char *result)
 	return total;
 }
 
-static void
-write_src(void)
+static void write_src(void)
 {
-	int i, k, off, valid;
+	unsigned int i, k, off;
 	unsigned int best_idx[256];
 	unsigned int *markers;
 	char buf[KSYM_NAME_LEN+1];
@@ -289,33 +261,24 @@ write_src(void)
 	printf(".data\n");
 
 	output_label("symbols_addresses");
-	valid = 0;
-	for (i = 0; i < cnt; i++) {
-		if (table[i].flags & SYM_FLAG_VALID) {
-			printf("\tPTR\t%#llx\n", table[i].addr);
-			valid++;
-		}
+	for (i = 0; i < table_cnt; i++) {
+		printf("\tPTR\t%#llx\n", table[i].addr);
 	}
 	printf("\n");
 
 	output_label("symbols_num_syms");
-	printf("\tPTR\t%d\n", valid);
+	printf("\tPTR\t%d\n", table_cnt);
 	printf("\n");
 
 	/* table of offset markers, that give the offset in the compressed stream
 	 * every 256 symbols */
-	markers = (unsigned int *) malloc(sizeof(unsigned int)*((valid + 255) / 256));
+	markers = (unsigned int *) malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256));
 
 	output_label("symbols_names");
-	valid = 0;
 	off = 0;
-	for (i = 0; i < cnt; i++) {
-
-		if (!(table[i].flags & SYM_FLAG_VALID))
-			continue;
-
-		if ((valid & 0xFF) == 0)
-			markers[valid >> 8] = off;
+	for (i = 0; i < table_cnt; i++) {
+		if ((i & 0xFF) == 0)
+			markers[i >> 8] = off;
 
 		printf("\t.byte 0x%02x", table[i].len);
 		for (k = 0; k < table[i].len; k++)
@@ -323,12 +286,11 @@ write_src(void)
 		printf("\n");
 
 		off += table[i].len + 1;
-		valid++;
 	}
 	printf("\n");
 
 	output_label("symbols_markers");
-	for (i = 0; i < ((valid + 255) >> 8); i++)
+	for (i = 0; i < ((table_cnt + 255) >> 8); i++)
 		printf("\tPTR\t%d\n", markers[i]);
 	printf("\n");
 
@@ -338,7 +300,7 @@ write_src(void)
 	off = 0;
 	for (i = 0; i < 256; i++) {
 		best_idx[i] = off;
-		expand_symbol(best_table[i],best_table_len[i],buf);
+		expand_symbol(best_table[i], best_table_len[i], buf);
 		printf("\t.asciz\t\"%s\"\n", buf);
 		off += strlen(buf) + 1;
 	}
@@ -353,153 +315,13 @@ write_src(void)
 
 /* table lookup compression functions */
 
-static inline unsigned int rehash_token(unsigned int hash, unsigned char data)
-{
-	return ((hash * 16777619) ^ data);
-}
-
-static unsigned int hash_token(unsigned char *data, int len)
-{
-	unsigned int hash=HASH_BASE_OFFSET;
-	int i;
-
-	for (i = 0; i < len; i++)
-		hash = rehash_token(hash, data[i]);
-
-	return HASH_FOLD(hash);
-}
-
-/* find a token given its data and hash value */
-static struct token *find_token_hash(unsigned char *data, int len, unsigned int hash)
-{
-	struct token *ptr;
-
-	ptr = hash_table[hash];
-
-	while (ptr) {
-		if ((ptr->len == len) && (memcmp(ptr->data, data, len) == 0))
-			return ptr;
-		ptr=ptr->next;
-	}
-
-	return NULL;
-}
-
-static inline void insert_token_in_group(struct token *head, struct token *ptr)
-{
-	ptr->right = head->right;
-	ptr->right->left = ptr;
-	head->right = ptr;
-	ptr->left = head;
-}
-
-static inline void remove_token_from_group(struct token *ptr)
-{
-	ptr->left->right = ptr->right;
-	ptr->right->left = ptr->left;
-}
-
-
-/* build the counts for all the tokens that start with "data", and have lenghts
- * from 2 to "len" */
-static void learn_token(unsigned char *data, int len)
-{
-	struct token *ptr,*last_ptr;
-	int i, newprofit;
-	unsigned int hash = HASH_BASE_OFFSET;
-	unsigned int hashes[MAX_TOK_SIZE + 1];
-
-	if (len > MAX_TOK_SIZE)
-		len = MAX_TOK_SIZE;
-
-	/* calculate and store the hash values for all the sub-tokens */
-	hash = rehash_token(hash, data[0]);
-	for (i = 2; i <= len; i++) {
-		hash = rehash_token(hash, data[i-1]);
-		hashes[i] = HASH_FOLD(hash);
-	}
-
-	last_ptr = NULL;
-	ptr = NULL;
-
-	for (i = len; i >= 2; i--) {
-		hash = hashes[i];
-
-		if (!ptr) ptr = find_token_hash(data, i, hash);
-
-		if (!ptr) {
-			/* create a new token entry */
-			ptr = (struct token *) malloc(sizeof(*ptr));
-
-			memcpy(ptr->data, data, i);
-			ptr->len = i;
-
-			/* when we create an entry, it's profit is 0 because
-			 * we also take into account the size of the token on
-			 * the compressed table. We then subtract GOOD_BAD_THRESHOLD
-			 * so that the test to see if this token belongs to
-			 * the good or bad list, is a comparison to zero */
-			ptr->profit = -GOOD_BAD_THRESHOLD;
-
-			ptr->next = hash_table[hash];
-			hash_table[hash] = ptr;
-
-			insert_token_in_group(&bad_head, ptr);
-
-			ptr->smaller = NULL;
-		} else {
-			newprofit = ptr->profit + (ptr->len - 1);
-			/* check to see if this token needs to be moved to a
-			 * different list */
-			if((ptr->profit < 0) && (newprofit >= 0)) {
-				remove_token_from_group(ptr);
-				insert_token_in_group(&good_head,ptr);
-			}
-			ptr->profit = newprofit;
-		}
-
-		if (last_ptr) last_ptr->smaller = ptr;
-		last_ptr = ptr;
-
-		ptr = ptr->smaller;
-	}
-}
-
-/* decrease the counts for all the tokens that start with "data", and have lenghts
- * from 2 to "len". This function is much simpler than learn_token because we have
- * more guarantees (tho tokens exist, the ->smaller pointer is set, etc.)
- * The two separate functions exist only because of compression performance */
-static void forget_token(unsigned char *data, int len)
-{
-	struct token *ptr;
-	int i, newprofit;
-	unsigned int hash=0;
-
-	if (len > MAX_TOK_SIZE) len = MAX_TOK_SIZE;
-
-	hash = hash_token(data, len);
-	ptr = find_token_hash(data, len, hash);
-
-	for (i = len; i >= 2; i--) {
-
-		newprofit = ptr->profit - (ptr->len - 1);
-		if ((ptr->profit >= 0) && (newprofit < 0)) {
-			remove_token_from_group(ptr);
-			insert_token_in_group(&bad_head, ptr);
-		}
-		ptr->profit=newprofit;
-
-		ptr=ptr->smaller;
-	}
-}
-
 /* count all the possible tokens in a symbol */
 static void learn_symbol(unsigned char *symbol, int len)
 {
 	int i;
 
 	for (i = 0; i < len - 1; i++)
-		learn_token(symbol + i, len - i);
+		token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++;
 }
 
 /* decrease the count for all the possible tokens in a symbol */
@@ -508,117 +330,90 @@ static void forget_symbol(unsigned char *symbol, int len)
 	int i;
 
 	for (i = 0; i < len - 1; i++)
-		forget_token(symbol + i, len - i);
+		token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--;
 }
 
-/* set all the symbol flags and do the initial token count */
+/* remove all the invalid symbols from the table and do the initial token count */
 static void build_initial_tok_table(void)
 {
-	int i, use_it, valid;
+	unsigned int i, pos;
 
-	valid = 0;
-	for (i = 0; i < cnt; i++) {
-		table[i].flags = 0;
+	pos = 0;
+	for (i = 0; i < table_cnt; i++) {
 		if ( symbol_valid(&table[i]) ) {
-			table[i].flags |= SYM_FLAG_VALID;
-			valid++;
+			if (pos != i)
+				table[pos] = table[i];
+			learn_symbol(table[pos].sym, table[pos].len);
+			pos++;
 		}
 	}
-
-	use_it = 0;
-	for (i = 0; i < cnt; i++) {
-
-		/* subsample the available symbols. This method is almost like
-		 * a Bresenham's algorithm to get uniformly distributed samples
-		 * across the symbol table */
-		if (table[i].flags & SYM_FLAG_VALID) {
-
-			use_it += WORKING_SET;
-
-			if (use_it >= valid) {
-				table[i].flags |= SYM_FLAG_SAMPLED;
-				use_it -= valid;
-			}
-		}
-		if (table[i].flags & SYM_FLAG_SAMPLED)
-			learn_symbol(table[i].sym, table[i].len);
-	}
+	table_cnt = pos;
 }
 
 /* replace a given token in all the valid symbols. Use the sampled symbols
  * to update the counts */
-static void compress_symbols(unsigned char *str, int tlen, int idx)
+static void compress_symbols(unsigned char *str, int idx)
 {
-	int i, len, learn, size;
-	unsigned char *p;
+	unsigned int i, len, size;
+	unsigned char *p1, *p2;
 
-	for (i = 0; i < cnt; i++) {
-
-		if (!(table[i].flags & SYM_FLAG_VALID)) continue;
+	for (i = 0; i < table_cnt; i++) {
 
 		len = table[i].len;
-		learn = 0;
-		p = table[i].sym;
+		p1 = table[i].sym;
+
+		/* find the token on the symbol */
+		p2 = memmem(p1, len, str, 2);
+		if (!p2) continue;
+
+		/* decrease the counts for this symbol's tokens */
+		forget_symbol(table[i].sym, len);
+
+		size = len;
 
 		do {
+			*p2 = idx;
+			p2++;
+			size -= (p2 - p1);
+			memmove(p2, p2 + 1, size);
+			p1 = p2;
+			len--;
+
+			if (size < 2) break;
+
 			/* find the token on the symbol */
-			p = (unsigned char *) strstr((char *) p, (char *) str);
-			if (!p) break;
+			p2 = memmem(p1, size, str, 2);
 
-			if (!learn) {
-				/* if this symbol was used to count, decrease it */
-				if (table[i].flags & SYM_FLAG_SAMPLED)
-					forget_symbol(table[i].sym, len);
-				learn = 1;
-			}
+		} while (p2);
 
-			*p = idx;
-			size = (len - (p - table[i].sym)) - tlen + 1;
-			memmove(p + 1, p + tlen, size);
-			p++;
-			len -= tlen - 1;
+		table[i].len = len;
 
-		} while (size >= tlen);
-
-		if(learn) {
-			table[i].len = len;
-			/* if this symbol was used to count, learn it again */
-			if(table[i].flags & SYM_FLAG_SAMPLED)
-				learn_symbol(table[i].sym, len);
-		}
+		/* increase the counts for this symbol's new tokens */
+		learn_symbol(table[i].sym, len);
 	}
 }
 
 /* search the token with the maximum profit */
-static struct token *find_best_token(void)
+static int find_best_token(void)
 {
-	struct token *ptr,*best,*head;
-	int bestprofit;
+	int i, best, bestprofit;
 
 	bestprofit=-10000;
+	best = 0;
 
-	/* failsafe: if the "good" list is empty search from the "bad" list */
-	if(good_head.right == &good_head) head = &bad_head;
-	else head = &good_head;
-
-	ptr = head->right;
-	best = NULL;
-	while (ptr != head) {
-		if (ptr->profit > bestprofit) {
-			bestprofit = ptr->profit;
-			best = ptr;
+	for (i = 0; i < 0x10000; i++) {
+		if (token_profit[i] > bestprofit) {
+			best = i;
+			bestprofit = token_profit[i];
 		}
-		ptr = ptr->right;
 	}
-
 	return best;
 }
 
 /* this is the core of the algorithm: calculate the "best" table */
 static void optimize_result(void)
 {
-	struct token *best;
-	int i;
+	int i, best;
 
 	/* using the '\0' symbol last allows compress_symbols to use standard
 	 * fast string functions */
@@ -632,14 +427,12 @@ static void optimize_result(void)
 			best = find_best_token();
 
 			/* place it in the "best" table */
-			best_table_len[i] = best->len;
-			memcpy(best_table[i], best->data, best_table_len[i]);
-			/* zero terminate the token so that we can use strstr
-			   in compress_symbols */
-			best_table[i][best_table_len[i]]='\0';
+			best_table_len[i] = 2;
+			best_table[i][0] = best & 0xFF;
+			best_table[i][1] = (best >> 8) & 0xFF;
 
 			/* replace this token in all the valid symbols */
-			compress_symbols(best_table[i], best_table_len[i], i);
+			compress_symbols(best_table[i], i);
 		}
 	}
 }
@@ -647,39 +440,28 @@ static void optimize_result(void)
 /* start by placing the symbols that are actually used on the table */
 static void insert_real_symbols_in_table(void)
 {
-	int i, j, c;
+	unsigned int i, j, c;
 
 	memset(best_table, 0, sizeof(best_table));
 	memset(best_table_len, 0, sizeof(best_table_len));
 
-	for (i = 0; i < cnt; i++) {
-		if (table[i].flags & SYM_FLAG_VALID) {
-			for (j = 0; j < table[i].len; j++) {
-				c = table[i].sym[j];
-				best_table[c][0]=c;
-				best_table_len[c]=1;
-			}
+	for (i = 0; i < table_cnt; i++) {
+		for (j = 0; j < table[i].len; j++) {
+			c = table[i].sym[j];
+			best_table[c][0]=c;
+			best_table_len[c]=1;
 		}
 	}
 }
 
 static void optimize_token_table(void)
 {
-	memset(hash_table, 0, sizeof(hash_table));
-
-	good_head.left = &good_head;
-	good_head.right = &good_head;
-
-	bad_head.left = &bad_head;
-	bad_head.right = &bad_head;
-
 	build_initial_tok_table();
 
 	insert_real_symbols_in_table();
 
 	/* When valid symbol is not registered, exit to error */
-	if (good_head.left == good_head.right &&
-	    bad_head.left == bad_head.right) {
+	if (!table_cnt) {
 		fprintf(stderr, "No valid symbol.\n");
 		exit(1);
 	}
@@ -688,8 +470,7 @@ static void optimize_token_table(void)
 }
 
 
-int
-main(int argc, char **argv)
+int main(int argc, char **argv)
 {
 	if (argc >= 2) {
 		int i;