zu2
zu2
I had same problem on M1 Macbook Monterey 12.4.
repeated_op better. --- ../Fuzix-Compiler-Kit/be-codegen-6800.c 2024-10-07 15:42:20 +++ be-codegen-6800.c 2024-10-11 12:07:15 ``` @@ -747,8 +801,14 @@ } else { /* FIXME: merge this into _off versions of helpers that do the...
``` --- ../Fuzix-Compiler-Kit/be-code-6800.c 2024-10-07 15:42:20 +++ be-code-6800.c 2024-10-09 15:38:05 @@ -341,13 +341,17 @@ } /* TODO: if we save_d we need to keep and valid */ /* Inline */ -...
I tried BYTETABLE. When processing with cmp_direct, the terms on the right side can be omitted, but the terms on the left side have already been processed at the time...
I tried backend-byte.c. It looks good. I need to consider the scope of BYTEOP further. test program: ``` int main(int argc, char **argv) { signed char i=0; signed char k=0;...
one more. ``` --- ../Fuzix-Compiler-Kit/be-codegen-6800.c 2024-11-13 00:32:04 +++ be-codegen-6800.c 2024-11-16 13:57:20 @@ -503,7 +552,7 @@ return 1; } } - if (s == 2 && r->op == T_LOCAL) { +...
The reason the array reference is wrong is because T_NAME was being added in bytes. I fixed that, but other things may need to be fixed. T_EQ/EQEQ seems to be...
In the signed case, this results in better code. use gen_shortcut. ``` @@ -1393,6 +1428,38 @@ return do_xeqop(n, "xoreq"); case T_HATEQ: return do_xeqop(n, "xhateq"); + case T_PLUS: + if (s==2...
Enable byte_label_tree in the current version, cmp_direct generates incorrect code, and the sample program goes into an infinite loop. Even though __cclt is called, only one byte is pushed from...
make_tos_ptr doesn't work, need tsx for now