Aaron Puchert bd8d245b13 Add two-byte opcodes and fix some issues in x86.new ... Contains: - Eliminated unnecessary parameter in readNumber - Don't choose opcode if we don't have enough parameters - Add ability to read numbered registers - Such as FPU, MMX or SSE registers. - Correctly assemble some spec-field encoded operations - If we have a spec field specifying the operation and a register-only operand, we have to move the operand first. This is common for FPU instructions. - Use size_t instead of int where appropriate - Add tests for XMM registers - Added all two-byte opcodes - Add tests for new operations - Separate flags for assembly and machine language - Sometimes operations have two register-only operands and we can't easily find out which to encode how. That's why we now have flags describing what we expect in the assembly and separate flags describing how to encode an operand. This also makes the assembler code cleaner.		2015-02-15 13:32:50 +01:00
..
lops.sh	Fix 'ad', 'a8' and add more test cases for x86/arm assemblers	2013-02-21 11:31:04 +01:00
test.arm	Pass t/test.arm, new cmd !123 to run historic command	2013-03-14 02:32:53 +01:00
test.cmp	Implement 'cmp' with memory ref on x86.nz assembler	2013-10-20 23:17:38 +02:00
test.new	Add two-byte opcodes and fix some issues in x86.new	2015-02-15 13:32:50 +01:00
test.nz	Fixes shl,sar,.. in x86-64 assembler	2013-10-28 02:43:47 +01:00
test.nz2	Fix #206 - Implement more 'lea' variants in x86.nz assembler	2013-09-03 18:57:41 +02:00