mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-12-30 17:33:24 +00:00
72062f5744
This patch adds support for AArch64 (ARM's 64-bit architecture) to
LLVM in the "experimental" category. Currently, it won't be built
unless requested explicitly.
This initial commit should have support for:
+ Assembly of all scalar (i.e. non-NEON, non-Crypto) instructions
(except the late addition CRC instructions).
+ CodeGen features required for C++03 and C99.
+ Compilation for the "small" memory model: code+static data <
4GB.
+ Absolute and position-independent code.
+ GNU-style (i.e. "__thread") TLS.
+ Debugging information.
The principal omission, currently, is performance tuning.
This patch excludes the NEON support also reviewed due to an outbreak of
batshit insanity in our legal department. That will be committed soon bringing
the changes to precisely what has been approved.
Further reviews would be gratefully received.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@174054 91177308-0d34-0410-b5e6-96231b3b80d8
44 lines
2.7 KiB
Plaintext
44 lines
2.7 KiB
Plaintext
# These spawn another process so they're rather expensive. Not many.
|
|
|
|
# Instructions notionally in the add/sub (extended register) sheet, but with
|
|
# invalid shift amount or "opt" field.
|
|
# RUN: echo "0x00 0x10 0xa0 0x0b" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x00 0x10 0x60 0x0b" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x00 0x14 0x20 0x0b" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# Instructions notionally in the add/sub (immediate) sheet, but with
|
|
# invalid "shift" field.
|
|
# RUN: echo "0xdf 0x3 0x80 0x91" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0xed 0x8e 0xc4 0x31" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x62 0xfc 0xbf 0x11" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x3 0xff 0xff 0x91" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# Instructions notionally in the load/store (unsigned immediate) sheet.
|
|
# Only unallocated (int-register) variants are: opc=0b11, size=0b10, 0b11
|
|
# RUN: echo "0xd7 0xfc 0xff 0xb9" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0xd7 0xfc 0xcf 0xf9" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# Instructions notionally in the floating-point <-> fixed-point conversion
|
|
# Scale field is 64-<imm> and <imm> should be 1-32 for a 32-bit int register.
|
|
# RUN: echo "0x23 0x01 0x18 0x1e" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x23 0x25 0x42 0x1e" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# Instructions notionally in the logical (shifted register) sheet, but with out
|
|
# of range shift: w-registers can only have 0-31.
|
|
# RUN: echo "0x00 0x80 0x00 0x0a" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# Instructions notionally in the move wide (immediate) sheet, but with out
|
|
# of range shift: w-registers can only have 0 or 16.
|
|
# RUN: echo "0x00 0x00 0xc0 0x12" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x12 0x34 0xe0 0x52" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# Data-processing instructions are undefined when S=1 and for the 0b0000111 value in opcode:sf
|
|
# RUN: echo "0x00 0x00 0xc0 0x5f" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x56 0x0c 0xc0 0x5a" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# Data-processing instructions (2 source) are undefined for a value of 0001xx:0:x or 0011xx:0:x for opcode:S:sf
|
|
# RUN: echo "0x00 0x30 0xc1 0x1a" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
# RUN: echo "0x00 0x10 0xc1 0x1a" | llvm-mc -triple=aarch64 -disassemble 2>&1 | FileCheck %s
|
|
|
|
# CHECK: invalid instruction encoding
|