py/asmarm: Fix locals address loading code generation with large imm.

This commit fixes code generation for loading a local's address if its
index is greater than 63.

The old code blindly encoded the offset into an `ADD Rd, Rn, #imm` opcode,
but only the lowest 8 bits would be put into the opcode itself.  This
commit instead generates a two-opcodes sequence, a constant load into R8,
and then an `ADD Rd, Rn, R8` opcode.

This fixes `tests/float/math_domain.py` for the qemu/SABRELITE board.

Signed-off-by: Alessandro Gatti <a.gatti@frob.it>

1 files changed, 9 insertions, 2 deletions

diff --git a/py/asmarm.c b/py/asmarm.c
index b20bdebde..a1c15d0fa 100644
--- a/py/asmarm.c
+++ b/py/asmarm.c
@@ -292,8 +292,15 @@ void asm_arm_orr_reg_reg_reg(asm_arm_t *as, uint rd, uint rn, uint rm) {
 }
 
 void asm_arm_mov_reg_local_addr(asm_arm_t *as, uint rd, int local_num) {
-    // add rd, sp, #local_num*4
-    emit_al(as, asm_arm_op_add_imm(rd, ASM_ARM_REG_SP, local_num << 2));
+    if (local_num >= 0x40) {
+        // mov r8, #local_num*4
+        // add rd, sp, r8
+        asm_arm_mov_reg_i32_optimised(as, ASM_ARM_REG_R8, local_num << 2);
+        emit_al(as, asm_arm_op_add_reg(rd, ASM_ARM_REG_SP, ASM_ARM_REG_R8));
+    } else {
+        // add rd, sp, #local_num*4
+        emit_al(as, asm_arm_op_add_imm(rd, ASM_ARM_REG_SP, local_num << 2));
+    }
 }
 
 void asm_arm_mov_reg_pcrel(asm_arm_t *as, uint reg_dest, uint label) {