From 8e1bd2ba044400e28b3219ba13d05d594825ceea Mon Sep 17 00:00:00 2001
From: Peter Johnson <peter@tortall.net>
Date: Thu, 3 Nov 2005 05:29:42 +0000
Subject: [PATCH] Handle instruction and prefix identifiers properly when used
 in other places in GAS input.  Do this by adding a tokenizer state that turns
 off insn and prefix generation when inside an instruction or directive, AND
 adding a special case for labels.

* gas-parser.h (state): Add INSTDIR.
* gas-token.re: Switch state back to INITIAL on ';' or newline; set state
when entering instruction or directive, add special case for labels.
* gas-bison.y: Remove non-working attempt at translating INSN and PREFIX
into string token; add special case for LABEL identifiers (generated from
special case for labels in tokenizer).

svn path=/trunk/yasm/; revision=1309
---
 modules/parsers/gas/gas-bison.y               |  20 +-
 modules/parsers/gas/gas-parser.h              |   3 +-
 modules/parsers/gas/gas-token.re              | 143 ++++---
 modules/parsers/gas/tests/Makefile.inc        |   3 +
 modules/parsers/gas/tests/gas-instlabel.asm   |   3 +
 .../parsers/gas/tests/gas-instlabel.errwarn   |   0
 modules/parsers/gas/tests/gas-instlabel.hex   | 376 ++++++++++++++++++
 7 files changed, 476 insertions(+), 72 deletions(-)
 create mode 100644 modules/parsers/gas/tests/gas-instlabel.asm
 create mode 100644 modules/parsers/gas/tests/gas-instlabel.errwarn
 create mode 100644 modules/parsers/gas/tests/gas-instlabel.hex

diff --git a/modules/parsers/gas/gas-bison.y b/modules/parsers/gas/gas-bison.y
index 81cec4ee..442983c6 100644
--- a/modules/parsers/gas/gas-bison.y
+++ b/modules/parsers/gas/gas-bison.y
@@ -102,7 +102,7 @@ static void gas_parser_directive
 %token <int_info> RESERVE_SPACE
 %token <arch_data> INSN PREFIX REG REGGROUP SEGREG TARGETMOD
 %token LEFT_OP RIGHT_OP
-%token <str_val> ID DIR_ID
+%token <str_val> ID DIR_ID LABEL
 %token LINE
 %token DIR_2BYTE DIR_4BYTE DIR_ALIGN DIR_ASCII DIR_ASCIZ DIR_BALIGN
 %token DIR_BSS DIR_BYTE DIR_COMM DIR_DATA DIR_DOUBLE DIR_ENDR DIR_EXTERN
@@ -169,6 +169,14 @@ lineexp: instr
 	$$ = $3;
 	define_label(parser_gas, $1, 0);
     }
+    | LABEL		{
+	$$ = (yasm_bytecode *)NULL;
+	define_label(parser_gas, $1, 0);
+    }
+    | LABEL instr	{
+	$$ = $2;
+	define_label(parser_gas, $1, 0);
+    }
     /* Line directive */
     | DIR_LINE INTNUM {
 	$$ = (yasm_bytecode *)NULL;
@@ -716,16 +724,6 @@ expr_id: label_id
     | DIR_DATA	{ $$ = yasm__xstrdup(".data"); }
     | DIR_TEXT	{ $$ = yasm__xstrdup(".text"); }
     | DIR_BSS	{ $$ = yasm__xstrdup(".bss"); }
-    | INSN	{
-	/* Used as an identifier, grab token to make it one! */
-	$$ = yasm__xstrndup(parser_gas->s.tok,
-			    (size_t)(parser_gas->s.cur - parser_gas->s.tok));
-    }
-    | PREFIX	{
-	/* Used as an identifier, grab token to make it one! */
-	$$ = yasm__xstrndup(parser_gas->s.tok,
-			    (size_t)(parser_gas->s.cur - parser_gas->s.tok));
-    }
 ;
 
 label_id: ID | DIR_ID;
diff --git a/modules/parsers/gas/gas-parser.h b/modules/parsers/gas/gas-parser.h
index d17ee34a..df9421f2 100644
--- a/modules/parsers/gas/gas-parser.h
+++ b/modules/parsers/gas/gas-parser.h
@@ -88,7 +88,8 @@ typedef struct yasm_parser_gas {
     Scanner s;
     enum {
 	INITIAL,
-	SECTION_DIRECTIVE
+	SECTION_DIRECTIVE,
+	INSTDIR
     } state;
 
     int code_section;
diff --git a/modules/parsers/gas/gas-token.re b/modules/parsers/gas/gas-token.re
index e284232f..67bcb6bf 100644
--- a/modules/parsers/gas/gas-token.re
+++ b/modules/parsers/gas/gas-token.re
@@ -334,60 +334,64 @@ scan:
 	">>"			{ RETURN(RIGHT_OP); }
 	"<"			{ RETURN(LEFT_OP); }
 	">"			{ RETURN(RIGHT_OP); }
-	[-+|^!*&/~$():;@=,]	{ RETURN(s->tok[0]); }
+	[-+|^!*&/~$():@=,]	{ RETURN(s->tok[0]); }
+	";"	{
+	    parser_gas->state = INITIAL;
+	    RETURN(s->tok[0]);
+	}
 
 	/* arch-independent directives */
-	'.2byte'	{ RETURN(DIR_2BYTE); }
-	'.4byte'	{ RETURN(DIR_4BYTE); }
-	'.8byte'	{ RETURN(DIR_QUAD); }
-	'.align'	{ RETURN(DIR_ALIGN); }
-	'.ascii'	{ RETURN(DIR_ASCII); }
-	'.asciz'	{ RETURN(DIR_ASCIZ); }
-	'.balign'	{ RETURN(DIR_BALIGN); }
-	'.bss'		{ RETURN(DIR_BSS); }
-	'.byte'		{ RETURN(DIR_BYTE); }
-	'.comm'		{ RETURN(DIR_COMM); }
-	'.data'		{ RETURN(DIR_DATA); }
-	'.double'	{ RETURN(DIR_DOUBLE); }
-	'.endr'		{ RETURN(DIR_ENDR); }
-	'.equ'		{ RETURN(DIR_EQU); }
-	'.extern'	{ RETURN(DIR_EXTERN); }
-	'.file'		{ RETURN(DIR_FILE); }
-	'.float'	{ RETURN(DIR_FLOAT); }
-	'.global'	{ RETURN(DIR_GLOBAL); }
-	'.globl'	{ RETURN(DIR_GLOBAL); }
-	'.hword'	{ RETURN(DIR_SHORT); }
-	'.ident'	{ RETURN(DIR_IDENT); }
-	'.int'		{ RETURN(DIR_INT); }
-	'.lcomm'	{ RETURN(DIR_LCOMM); }
-	'.line'		{ RETURN(DIR_LINE); }
-	'.loc'		{ RETURN(DIR_LOC); }
-	'.long'		{ RETURN(DIR_INT); }
-	'.octa'		{ RETURN(DIR_OCTA); }
-	'.org'		{ RETURN(DIR_ORG); }
-	'.p2align'	{ RETURN(DIR_P2ALIGN); }
-	'.rept'		{ RETURN(DIR_REPT); }
+	'.2byte'	{ parser_gas->state = INSTDIR; RETURN(DIR_2BYTE); }
+	'.4byte'	{ parser_gas->state = INSTDIR; RETURN(DIR_4BYTE); }
+	'.8byte'	{ parser_gas->state = INSTDIR; RETURN(DIR_QUAD); }
+	'.align'	{ parser_gas->state = INSTDIR; RETURN(DIR_ALIGN); }
+	'.ascii'	{ parser_gas->state = INSTDIR; RETURN(DIR_ASCII); }
+	'.asciz'	{ parser_gas->state = INSTDIR; RETURN(DIR_ASCIZ); }
+	'.balign'	{ parser_gas->state = INSTDIR; RETURN(DIR_BALIGN); }
+	'.bss'		{ parser_gas->state = INSTDIR; RETURN(DIR_BSS); }
+	'.byte'		{ parser_gas->state = INSTDIR; RETURN(DIR_BYTE); }
+	'.comm'		{ parser_gas->state = INSTDIR; RETURN(DIR_COMM); }
+	'.data'		{ parser_gas->state = INSTDIR; RETURN(DIR_DATA); }
+	'.double'	{ parser_gas->state = INSTDIR; RETURN(DIR_DOUBLE); }
+	'.endr'		{ parser_gas->state = INSTDIR; RETURN(DIR_ENDR); }
+	'.equ'		{ parser_gas->state = INSTDIR; RETURN(DIR_EQU); }
+	'.extern'	{ parser_gas->state = INSTDIR; RETURN(DIR_EXTERN); }
+	'.file'		{ parser_gas->state = INSTDIR; RETURN(DIR_FILE); }
+	'.float'	{ parser_gas->state = INSTDIR; RETURN(DIR_FLOAT); }
+	'.global'	{ parser_gas->state = INSTDIR; RETURN(DIR_GLOBAL); }
+	'.globl'	{ parser_gas->state = INSTDIR; RETURN(DIR_GLOBAL); }
+	'.hword'	{ parser_gas->state = INSTDIR; RETURN(DIR_SHORT); }
+	'.ident'	{ parser_gas->state = INSTDIR; RETURN(DIR_IDENT); }
+	'.int'		{ parser_gas->state = INSTDIR; RETURN(DIR_INT); }
+	'.lcomm'	{ parser_gas->state = INSTDIR; RETURN(DIR_LCOMM); }
+	'.line'		{ parser_gas->state = INSTDIR; RETURN(DIR_LINE); }
+	'.loc'		{ parser_gas->state = INSTDIR; RETURN(DIR_LOC); }
+	'.long'		{ parser_gas->state = INSTDIR; RETURN(DIR_INT); }
+	'.octa'		{ parser_gas->state = INSTDIR; RETURN(DIR_OCTA); }
+	'.org'		{ parser_gas->state = INSTDIR; RETURN(DIR_ORG); }
+	'.p2align'	{ parser_gas->state = INSTDIR; RETURN(DIR_P2ALIGN); }
+	'.rept'		{ parser_gas->state = INSTDIR; RETURN(DIR_REPT); }
 	'.section'	{
 	    parser_gas->state = SECTION_DIRECTIVE;
 	    RETURN(DIR_SECTION);
 	}
-	'.set'		{ RETURN(DIR_EQU); }
-	'.short'	{ RETURN(DIR_SHORT); }
-	'.single'	{ RETURN(DIR_FLOAT); }
-	'.size'		{ RETURN(DIR_SIZE); }
-	'.skip'		{ RETURN(DIR_SKIP); }
-	'.sleb128'	{ RETURN(DIR_SLEB128); }
-	'.space'	{ RETURN(DIR_SKIP); }
-	'.string'	{ RETURN(DIR_ASCIZ); }
-	'.text'		{ RETURN(DIR_TEXT); }
-	'.tfloat'	{ RETURN(DIR_TFLOAT); }
-	'.type'		{ RETURN(DIR_TYPE); }
-	'.quad'		{ RETURN(DIR_QUAD); }
-	'.uleb128'	{ RETURN(DIR_ULEB128); }
-	'.value'	{ RETURN(DIR_VALUE); }
-	'.weak'		{ RETURN(DIR_WEAK); }
-	'.word'		{ RETURN(DIR_WORD); }
-	'.zero'		{ RETURN(DIR_ZERO); }
+	'.set'		{ parser_gas->state = INSTDIR; RETURN(DIR_EQU); }
+	'.short'	{ parser_gas->state = INSTDIR; RETURN(DIR_SHORT); }
+	'.single'	{ parser_gas->state = INSTDIR; RETURN(DIR_FLOAT); }
+	'.size'		{ parser_gas->state = INSTDIR; RETURN(DIR_SIZE); }
+	'.skip'		{ parser_gas->state = INSTDIR; RETURN(DIR_SKIP); }
+	'.sleb128'	{ parser_gas->state = INSTDIR; RETURN(DIR_SLEB128); }
+	'.space'	{ parser_gas->state = INSTDIR; RETURN(DIR_SKIP); }
+	'.string'	{ parser_gas->state = INSTDIR; RETURN(DIR_ASCIZ); }
+	'.text'		{ parser_gas->state = INSTDIR; RETURN(DIR_TEXT); }
+	'.tfloat'	{ parser_gas->state = INSTDIR; RETURN(DIR_TFLOAT); }
+	'.type'		{ parser_gas->state = INSTDIR; RETURN(DIR_TYPE); }
+	'.quad'		{ parser_gas->state = INSTDIR; RETURN(DIR_QUAD); }
+	'.uleb128'	{ parser_gas->state = INSTDIR; RETURN(DIR_ULEB128); }
+	'.value'	{ parser_gas->state = INSTDIR; RETURN(DIR_VALUE); }
+	'.weak'		{ parser_gas->state = INSTDIR; RETURN(DIR_WEAK); }
+	'.word'		{ parser_gas->state = INSTDIR; RETURN(DIR_WORD); }
+	'.zero'		{ parser_gas->state = INSTDIR; RETURN(DIR_ZERO); }
 
 	/* label or maybe directive */
 	[.][a-zA-Z0-9_$.]* {
@@ -432,22 +436,41 @@ scan:
 	    RETURN(REG);
 	}
 
+	/* label */
+	[a-zA-Z][a-zA-Z0-9_$.]* ws* ':' {
+	    /* strip off colon and any whitespace */
+	    count = TOKLEN-1;
+	    while (s->tok[count] == ' ' || s->tok[count] == '\t'
+		   || s->tok[count] == '\r')
+		count--;
+	    /* Just an identifier, return as such. */
+	    lvalp->str_val = yasm__xstrndup(s->tok, count);
+	    RETURN(LABEL);
+	}
+
 	/* identifier that may be an instruction, etc. */
 	[a-zA-Z][a-zA-Z0-9_$.]* {
-	    savech = s->tok[TOKLEN];
-	    s->tok[TOKLEN] = '\0';
-	    if (yasm_arch_parse_check_insn(parser_gas->arch, lvalp->arch_data,
-					   s->tok, cur_line)) {
-		s->tok[TOKLEN] = savech;
-		RETURN(INSN);
-	    }
-	    if (yasm_arch_parse_check_prefix(parser_gas->arch,
-					     lvalp->arch_data, s->tok,
-					     cur_line)) {
+	    /* Can only be an instruction/prefix when not inside an
+	     * instruction or directive.
+	     */
+	    if (parser_gas->state != INSTDIR) {
+		savech = s->tok[TOKLEN];
+		s->tok[TOKLEN] = '\0';
+		if (yasm_arch_parse_check_insn(parser_gas->arch,
+					       lvalp->arch_data, s->tok,
+					       cur_line)) {
+		    s->tok[TOKLEN] = savech;
+		    parser_gas->state = INSTDIR;
+		    RETURN(INSN);
+		}
+		if (yasm_arch_parse_check_prefix(parser_gas->arch,
+						 lvalp->arch_data, s->tok,
+						 cur_line)) {
+		    s->tok[TOKLEN] = savech;
+		    RETURN(PREFIX);
+		}
 		s->tok[TOKLEN] = savech;
-		RETURN(PREFIX);
 	    }
-	    s->tok[TOKLEN] = savech;
 	    /* Just an identifier, return as such. */
 	    lvalp->str_val = yasm__xstrndup(s->tok, TOKLEN);
 	    RETURN(ID);
diff --git a/modules/parsers/gas/tests/Makefile.inc b/modules/parsers/gas/tests/Makefile.inc
index 7134b34a..e357c98b 100644
--- a/modules/parsers/gas/tests/Makefile.inc
+++ b/modules/parsers/gas/tests/Makefile.inc
@@ -12,6 +12,9 @@ EXTRA_DIST += modules/parsers/gas/tests/datavis.hex
 EXTRA_DIST += modules/parsers/gas/tests/datavis2.asm
 EXTRA_DIST += modules/parsers/gas/tests/datavis2.errwarn
 EXTRA_DIST += modules/parsers/gas/tests/datavis2.hex
+EXTRA_DIST += modules/parsers/gas/tests/gas-instlabel.asm
+EXTRA_DIST += modules/parsers/gas/tests/gas-instlabel.errwarn
+EXTRA_DIST += modules/parsers/gas/tests/gas-instlabel.hex
 EXTRA_DIST += modules/parsers/gas/tests/gas-semi.asm
 EXTRA_DIST += modules/parsers/gas/tests/gas-semi.errwarn
 EXTRA_DIST += modules/parsers/gas/tests/gas-semi.hex
diff --git a/modules/parsers/gas/tests/gas-instlabel.asm b/modules/parsers/gas/tests/gas-instlabel.asm
new file mode 100644
index 00000000..44c47482
--- /dev/null
+++ b/modules/parsers/gas/tests/gas-instlabel.asm
@@ -0,0 +1,3 @@
+	.globl	SUB
+	.type	SUB, @function
+SUB:
diff --git a/modules/parsers/gas/tests/gas-instlabel.errwarn b/modules/parsers/gas/tests/gas-instlabel.errwarn
new file mode 100644
index 00000000..e69de29b
diff --git a/modules/parsers/gas/tests/gas-instlabel.hex b/modules/parsers/gas/tests/gas-instlabel.hex
new file mode 100644
index 00000000..cba25918
--- /dev/null
+++ b/modules/parsers/gas/tests/gas-instlabel.hex
@@ -0,0 +1,376 @@
+7f 
+45 
+4c 
+46 
+01 
+01 
+01 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+01 
+00 
+03 
+00 
+01 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+b0 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+34 
+00 
+00 
+00 
+00 
+00 
+28 
+00 
+05 
+00 
+01 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+2e 
+74 
+65 
+78 
+74 
+00 
+2e 
+73 
+74 
+72 
+74 
+61 
+62 
+00 
+2e 
+73 
+79 
+6d 
+74 
+61 
+62 
+00 
+2e 
+73 
+68 
+73 
+74 
+72 
+74 
+61 
+62 
+00 
+00 
+00 
+00 
+00 
+2d 
+00 
+53 
+55 
+42 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+01 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+04 
+00 
+f1 
+ff 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+03 
+00 
+04 
+00 
+03 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+12 
+00 
+04 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+17 
+00 
+00 
+00 
+03 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+40 
+00 
+00 
+00 
+21 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+07 
+00 
+00 
+00 
+03 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+64 
+00 
+00 
+00 
+07 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+0f 
+00 
+00 
+00 
+02 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+6c 
+00 
+00 
+00 
+40 
+00 
+00 
+00 
+02 
+00 
+00 
+00 
+03 
+00 
+00 
+00 
+04 
+00 
+00 
+00 
+10 
+00 
+00 
+00 
+01 
+00 
+00 
+00 
+01 
+00 
+00 
+00 
+06 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+40 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
+10 
+00 
+00 
+00 
+00 
+00 
+00 
+00 
-- 
2.40.0