1 files changed, 58 insertions, 0 deletions
diff --git a/whichasm-0.01/classifier_arm.c b/whichasm-0.01/classifier_arm.c
new file mode 100644
index 0000000..b2c127c
--- /dev/null
+++ b/whichasm-0.01/classifier_arm.c
@@ -0,0 +1,58 @@
+/*
+ * Classifier for ARM assembly language
+ *
+ * Copyright (C) 2012 Red Hat, Inc.
+ *
+ */
+
+/* ARM opcodes usually have a format like this:
+ *    mnemonic <Rd>, <Ra>, <Rb>...
+ * Register names include the following:
+ *
+ *    32-bit GPRS: R0-R15 (fp, ip, sp, lr, pc)
+ *    32-bit shadow FIQ mode GPRS: R8_FIQ-R14_FIQ
+ *    32-bit shadow SVC mode GPRS: R13_SVC, R14_SVC
+ *    32-bit shadow ABT mode GPRS: R13_ABT, R14_ABT
+ *    32-bit shadow IRQ mode GPRS: R13_IRQ, R14_IRQ
+ *    32-bit program status registers: APSR, SPSR, CPSR (including shadow regs)
+ *    (also APCS defined a1-a4, v1-v8, s0-s32, d0-d16, etc.)
+ *    (not doing floating point registers at this stage)
+ */
+
+#include "arm.h"
+#include <stdio.h>
+#include <sys/types.h>
+#include <regex.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "classifier.h"
+
+int classifier_arm(char *token_name)
+{
+	const struct arm_opcode *opcode;
+	const struct arm_reg *reg;
+
+	for (opcode=arm_opcodes;opcode->assembler;opcode++) {
+		// TODO - catch variants of assembly language mnemonics
+		// previously limited to opcode->assembler length
+		// need to catch assembly that has appended letters
+		if (strlen(token_name) == strlen(opcode->assembler))
+			if (0 == strncasecmp(token_name, opcode->assembler,
+					     strlen(opcode->assembler))) {
+				//printf("opcode: %s\n", opcode->assembler);
+				return MNEMONIC;
+			}
+	}
+
+	for (reg=arm_regs;reg->assembler;reg++) {
+
+		if (strlen(token_name) == strlen(reg->assembler))
+			if (0 == strncasecmp(reg->assembler,token_name,
+					     strlen(token_name)))
+				//printf("register: %s\n", reg->assembler);
+				return REGISTER;
+	}
+
+	return UNKNOWN;
+}