Lex the register number with the register name.

This avoids the need for a start condition to prevent for example g1.8<0,1,0>UW
being lexed as GENREG NUMBER LANGLE etc. rather than
GENREG INTEGER DOT INTEGER LANGLE etc.

2 files changed, 38 insertions, 35 deletions

diff --git a/assembler/gram.y b/assembler/gram.y
index 0609d90a..7ed2f41b 100644
--- a/assembler/gram.y
+++ b/assembler/gram.y
@@ -63,7 +63,8 @@
 
 %token <integer> ALIGN1 ALIGN16 MASK_DISABLE EOT
 
-%token GENREG MSGREG ACCREG ADDRESSREG FLAGREG CONTROLREG IPREG
+%token <integer> GENREG MSGREG ACCREG ADDRESSREG
+%token FLAGREG CONTROLREG IPREG
 
 %token MOV
 %token MUL MAC MACH LINE SAD2 SADA2 DP4 DPH DP3 DP2
@@ -497,46 +498,51 @@ subregnum:	DOT INTEGER
 ;
 
 /* 1.4.5: Register files and register numbers */
-directgenreg:	GENREG INTEGER subregnum
+directgenreg:	GENREG subregnum
 		{
 		  /* Returns an instruction with just the destination register
 		   * fields filled in.
 		   */
 		  $$.reg_file = BRW_GENERAL_REGISTER_FILE;
-		  $$.reg_nr = $2;
-		  $$.subreg_nr = $3;
+		  $$.reg_nr = $1;
+		  $$.subreg_nr = $2;
 		}
 
-directmsgreg:	MSGREG INTEGER subregnum
+directmsgreg:	MSGREG subregnum
 		{
 		  /* Returns an instruction with just the destination register
 		   * fields filled in.
 		   */
 		  $$.reg_file = BRW_MESSAGE_REGISTER_FILE;
-		  $$.reg_nr = $2;
-		  $$.subreg_nr = $3;
+		  $$.reg_nr = $1;
+		  $$.subreg_nr = $2;
 		}
 ;
 
-accreg:		ACCREG INTEGER subregnum
+accreg:		ACCREG subregnum
 		{
 		  /* Returns an instruction with just the destination register
 		   * fields filled in.
 		   */
+		  if ($1 > 1) {
+		    fprintf(stderr,
+			    "accumulator register number %d out of range", $1);
+		    YYERROR;
+		  }
 		  $$.reg_file = BRW_ARCHITECTURE_REGISTER_FILE;
-		  $$.reg_nr = BRW_ARF_ACCUMULATOR | $2;
-		  $$.subreg_nr = $3;
+		  $$.reg_nr = BRW_ARF_ACCUMULATOR | $1;
+		  $$.subreg_nr = $2;
 		}
 ;
 
-addrreg:	ADDRESSREG INTEGER subregnum
+addrreg:	ADDRESSREG subregnum
 		{
 		  /* Returns an instruction with just the destination register
 		   * fields filled in.
 		   */
 		  $$.reg_file = BRW_ARCHITECTURE_REGISTER_FILE;
-		  $$.reg_nr = BRW_ARF_ADDRESS | $2;
-		  $$.subreg_nr = $3;
+		  $$.reg_nr = BRW_ARF_ADDRESS | $1;
+		  $$.subreg_nr = $2;
 		}
 ;
 
diff --git a/assembler/lex.l b/assembler/lex.l
index 8982df55..e60b228b 100644
--- a/assembler/lex.l
+++ b/assembler/lex.l
@@ -7,7 +7,6 @@
 int saved_state = INITIAL;
 
 %}
-%s IN_REG
 %x BLOCK_COMMENT
 
 %%
@@ -74,33 +73,31 @@ int saved_state = INITIAL;
 "-" { return MINUS; }
 "(abs)" { return ABS; }
 
- /* XXX: this lexing of register files is shady */
-"acc" {
-	BEGIN(IN_REG);
+"acc"[0-9]+ {
+	yylval.integer = atoi(yytext + 1);
 	return ACCREG;
 }
-"a" {
-	BEGIN(IN_REG);
+"a"[0-9]+ {
+	yylval.integer = atoi(yytext + 1);
 	return ADDRESSREG;
 }
-"m" {
-	BEGIN(IN_REG);
+"m"[0-9]+ {
+	yylval.integer = atoi(yytext + 1);
 	return MSGREG;
 }
-"f" {
-	BEGIN(IN_REG);
+"f"[0-9]+ {
+	yylval.integer = atoi(yytext + 1);
 	return FLAGREG;
 }
-[gr] {
-	BEGIN(IN_REG);
+[gr][0-9]+ {
+	yylval.integer = atoi(yytext + 1);
 	return GENREG;
 }
-"cr" {
-	BEGIN(IN_REG);
+"cr"[0-9]+ {
+	yylval.integer = atoi(yytext + 1);
 	return CONTROLREG;
 }
 "ip" {
-	BEGIN(IN_REG);
 	return IPREG;
 }
 
@@ -108,13 +105,13 @@ int saved_state = INITIAL;
   * Lexing of register types should probably require the ":" symbol specified
   * in the BNF of the assembly, but our existing source didn't use that syntax.
   */
-"UD" { BEGIN(INITIAL); return TYPE_UD; }
-"D" { BEGIN(INITIAL); return TYPE_D; }
-"UW" { BEGIN(INITIAL); return TYPE_UW; }
-"W" { BEGIN(INITIAL); return TYPE_W; }
-"UB" { BEGIN(INITIAL); return TYPE_UB; }
-"B" { BEGIN(INITIAL); return TYPE_B; }
-"F" { BEGIN(INITIAL); return TYPE_F; }
+"UD" { return TYPE_UD; }
+"D" { return TYPE_D; }
+"UW" { return TYPE_UW; }
+"W" { return TYPE_W; }
+"UB" { return TYPE_UB; }
+"B" { return TYPE_B; }
+"F" { return TYPE_F; }
 
 "sat" { return SATURATE; }
 "align1" { return ALIGN1; }