[M68k] (Patch 4/8) MC layer and object file support

[myhsu]

Summary

1. Add the M68k-specific MC layer implementation
2. Add ELF support for M68k
3. Add M68k-specifc CC and reloc

[myhsu]

Part of the restructing of this patch series. Now this patch contains the MC layer and ELF support

[myhsu]

Update licenses

[myhsu]

Update to reflect changes regarding getMCInstrBeads (formally getGenInstrBeads)

[craig.topper]

llvm/include/llvm/module.modulemap

85

Alphabetize with other files

llvm/lib/MC/MCExpr.cpp

228 ↗

(On Diff #296012)

Put return on same line for consistency

378 ↗

(On Diff #296012)

tabs?

llvm/lib/Target/M680x0/MCTargetDesc/M680x0AsmBackend.cpp

177 ↗

(On Diff #296012)

Can this be !isInt<16>(Value)?

187 ↗

(On Diff #296012)

isInt<8>?

llvm/lib/Target/M680x0/MCTargetDesc/M680x0BaseInfo.h

251 ↗

(On Diff #296012)

Can we remove commented out code or put in the patch that needs it?

llvm/lib/Target/M680x0/MCTargetDesc/M680x0ELFObjectWriter.cpp

84 ↗

(On Diff #296012)

Commented out code

llvm/lib/Target/M680x0/MCTargetDesc/M680x0InstPrinter.cpp

35 ↗

(On Diff #296012)

Is the StringRef conversion needed?

llvm/lib/Target/M680x0/MCTargetDesc/M680x0MCAsmInfo.cpp

29 ↗

(On Diff #296012)

Is this number correct for 68K? That looks like X86's NOP encoding.

llvm/lib/Target/M680x0/MCTargetDesc/M680x0MCCodeEmitter.cpp

178 ↗

(On Diff #296012)

Can this use isUIntN from MathExtras.h?

182 ↗

(On Diff #296012)

isIntN from MathExtras?

189 ↗

(On Diff #296012)

Commented out code?

344 ↗

(On Diff #296012)

Commented out code

407 ↗

(On Diff #296012)

Commented out code

llvm/lib/Target/M680x0/MCTargetDesc/M680x0MCTargetDesc.cpp

65 ↗

(On Diff #296012)

I'm not sure the result of a concatenation can be a "SingleStringRef"

[myhsu]

Addressed feedbacks

llvm/lib/Target/M680x0/MCTargetDesc/M680x0MCAsmInfo.cpp

29 ↗

(On Diff #296012)

good catch, 68K's NOP should be different

[jrtc27]

llvm/include/llvm/BinaryFormat/ELFRelocs/m680x0.def

5 ↗

(On Diff #297733)

As I said on D88389:

They're all R_68K_FOO in system headers, please just use that name otherwise it gets confusing.

llvm/include/llvm/Object/ELFObjectFile.h

1145

As I said on D88389:

This gets reported in the file format line of llvm-objdump so should match what binutils has, which is elf32-m68k, though even if that weren't the case it should at least be in keeping with the style of all the others here.

[glaubitz]

llvm/include/llvm/Object/ELFObjectFile.h

1145

Yeah, I agree this should definitely match with what GNU is using there.

I would still prefer the backend being called "M680x0" and therefore the patches should be prefixed with "[M680x0]", similar to "SystemZ" and "s390x".

Naming the "M680x0" instead of "M68K" improves the readability in my personal opinion as it's easier to tell when you are talking about the backend and when you're talking about the architecture and GNU triplet.

[jrtc27]

llvm/include/llvm/Object/ELFObjectFile.h

1145

The difference there is "IBM System z9" etc are the actual product names, and it's not a really clumsy name to type like M680x0. NXP's own site categorises its M68K-derived processors as "68K Processors (Legacy)", and the manuals say things like:

The MCF5407 extends the legacy of Motorola’s 68K family

GCC's own manage uses the M680x0 term in the following ways:

These are the -m options defined for M680x0 and ColdFire processors.

Generate code for a specific M680x0 or ColdFire instruction set architecture. Permissible values of arch for M680x0 architectures are: 68000, 68010, 68020, 68030, 68040, 68060 and cpu32. ColdFire architectures are selected according to Freescale's ISA classification and the permissible values are: isaa, isaaplus, isab and isac.

So the name M680x0 would actually be *more* narrow than what M68K means in practice, with the latter being the general term for any M68000-derived processor and the former being only for the 68000 through 68060 processors and *not* including the ColdFire extensions, yet there's no reason why our backend can't support that just like GCC does.

[glaubitz]

llvm/include/llvm/Object/ELFObjectFile.h

1145

Meh, I think this is really a bike-shedding contest. As I said, I like the name "M680x0" because it clearly tells me we're talking about the backend and not the architecture. It makes reading the code easier in my opinion.

GCC's own manage uses the M680x0 term in the following ways:

Which is titled with "3.19.25 M680x0 Options" ;-)

See: https://gcc.gnu.org/onlinedocs/gcc/M680x0-Options.html

[jrtc27]

llvm/include/llvm/Object/ELFObjectFile.h

1145

Which is titled with "3.19.25 M680x0 Options" ;-)

Which I think is a bug (perhaps historical, if that was the name chosen before ColdFire was added) given the language used throughout. I've been considering sending a patch to change that title though.

[glaubitz]

llvm/include/llvm/Object/ELFObjectFile.h

1145

Well, ok. If LLVM upstream insists on the name "M68k" for the backend, I'm not going fight it. In the end, I want the backend to succeed and I don't want something like a naming dispute to block it.

But please call it "M68k" (with a lower "k"), not "M68K" which would be incorrect as "kilo" is spelled all lower case.

[myhsu]

• [NFC] Rename M680x0 to M68k
• Change ELF reloc's prefix from "R_M680x0_" to "R_68K_" to be in consistent with GCC

[rengolin]

A few nits, but this is looking good.

llvm/include/llvm/MC/MCExpr.h

202 ↗

(On Diff #302184)

Does this get exposed in any way? Would it break a previous enum order?

llvm/lib/Target/M68k/MCTargetDesc/M68kAsmBackend.cpp

217

Isn't there a better way to emit this?

llvm/lib/Target/M68k/MCTargetDesc/M68kBaseInfo.h

2

Other header comments mention "m68k", this should too.

217

Is this comment really meaningful? Or is it commented by accident?

Or is "SP" == "A7"? If so, I wouldn't comment like code because it looks like an accident or bad form.

llvm/lib/Target/M68k/MCTargetDesc/M68kELFObjectWriter.cpp

13

This file is quite light on comments.

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

12

This file is also a bit light on comments. Some assumptions may not be obvious to non-m68k developers.

[myhsu]

• Addressed some of the feedbacks
• Change dummy/placeholder file names

[myhsu]

llvm/include/llvm/MC/MCExpr.h

202 ↗

(On Diff #302184)

You're right, I don't think any of the M68k code use it at this time point. Will remove it for now

[myhsu]

• Fix incorrect relaxation decision

[MaskRay]

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.cpp

85

for (int s = 0; s < 8; s += 8) ?

87

https://llvm.org/docs/CodingStandards.html#don-t-use-braces-on-simple-single-statement-bodies-of-if-else-loop-statements

Applies to many braces in this file.

182

The PCREL_OPERAND style print* functions are usually used this way: D77853

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

59

encodeBits

Newer backends should stick with the function naming coding standard

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

70

clang-format recognized spelling is /*TuneCPU=*/CPU

83

There should be some .cfi_* tests with this change.

[myhsu]

• Addressed some of the feedbacks
• [NFC] Removed '#<number>' in the comments
• [NFC] Fixed minor formatting issues

[myhsu]

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.cpp

85

good catch, thanks

182

Well...actually Motorola is using its own assembly language (and where this (N,%pc) syntax comes from). We're trying to conform with that in order to have better coordination with other existing toolchains

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

83

Correct, it's on the TODO list now

[myhsu]

Fixed incorrect formats (redundant commas) when printing move mask

[RKSimon]

a few minors

llvm/include/llvm/IR/CallingConv.h

248

Why is this 1000 instead of 101?

llvm/include/llvm/module.modulemap

73

M68k.def ?

llvm/lib/Target/M68k/MCTargetDesc/M68kFixupKinds.h

25

This looks dodgy - please can you double check it?

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

43

todo what?

[jrtc27]

llvm/include/llvm/BinaryFormat/ELF.h

761

File should be named the same as the directory in Target, ie with a capital M.

llvm/include/llvm/BinaryFormat/ELFRelocs/m68k.def

5 ↗

(On Diff #312017)

No space before (

llvm/include/llvm/Object/ELFObjectFile.h

1145

This specific string still needs fixing to match binutils's name for the file format as all lowercase.

llvm/lib/Target/M68k/MCTargetDesc/M68kBaseInfo.h

196

The existence of this is surprising; one would expect to be able to pass any valid register number to a function called isAddressRegister and just get back false for special registers. Do you need this assert? If so, please change the function name to reflect that it's only valid for the "general-purpose" (if they're called that on m68k given it has split A and D) registers.

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.cpp

15

Would be helpful to state (either here at the top or in the places where the code gets it wrong) the ways in which it doesn't conform. Also is that the Motorola ASM syntax what GAS uses (which is the most important thing to implement) or do they differ?

182

The objdump output is not the same as the asm input/codegen output in this specific case. What does binutils's objdump do for m68k?

llvm/lib/Target/M68k/MCTargetDesc/M68kMCAsmInfo.cpp

23

What does this mean?

30

Which is what? A specific instruction? A special bit pattern that's never a valid instruction?

40

Clang doesn't do and has never done anything for m68k. This line can go.

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

43

I assume "implement the ParseM68kTriple function" so it's not just a stub... but yes.

88–92

If your hardware-provided call instruction saves the return address in memory like x86, then yes. If your hardware-provided call instruction saves the return address in a register like most other architectures, then no, and the prologue code will ensure the return address gets proper DWARF info if spilled, just like any other register. Given m68k is like x86 and jsr/rts/etc save/restore PC to/from the stack, yes, you do need this I believe.

[jrtc27]

llvm/lib/Target/M68k/MCTargetDesc/M68kBaseInfo.h

18

Old name in the header guard. Repeated several times in this revision and later revisions.

[myhsu]

• Addressed some of the feedbacks
  • I need a little more time double checking the (motorola) assembly syntax

[RKSimon]

llvm/lib/Target/M68k/MCTargetDesc/CMakeLists.txt

11

Fix sorting

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.cpp

82

(style) Move comment into assert message

assert((Mask & 0xFFFF) == Mask && "Mask should be 16 bits");

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.h

45

why is this unsigned when all the other methods that have 'opNum' use int ?

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

179

Merge asserts:

assert((Size + Offset <= 64) && isUIntN(Size, Val) && "Value does not fit");

336

Should we test for null pointer as well here?

if (!Beads || !*Beads) {

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

1–12

We should probably have the boilerplate comments in the initial version of the file?

llvm/lib/Target/M68k/TargetInfo/M68kTargetInfo.cpp

1–12

We should probably have the boilerplate comments in the initial version of the file?

[myhsu]

• Addressed feedbacks

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.cpp

182

I think I can answer this question now: GCC, GNU AS and this M68k backend all use Motorola's own assembly syntax.
objdump, on the hand, doesn't. I'm not clean what they use but it looks like reusing the existing ATT syntax or so.

llvm/lib/Target/M68k/MCTargetDesc/M68kMCAsmInfo.cpp

23

seems to be a stale comments, removing it now

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

1–12

sorry I don't quite get what you said: what is the boilerplate comments here? is it the license + "This file provides M68k target specific descriptions." on the right?

[myhsu]

• [NFC] Fixed minor formatting issue

[RKSimon]

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

1–12

Yes - the comment block containing the license etc.

[myhsu]

• [NFC] Addressed feedbacks

[RKSimon]

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

1–12

Please can you update D88389 with these changes?

[myhsu]

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

1–12

right, almost forget, will do

[RKSimon]

A few minors, but overall I think this is good enough to go in.

@rengolin @jrtc27 Any more comments?

llvm/lib/Target/M68k/MCTargetDesc/M68kELFObjectWriter.cpp

52

(style) Drop the default case and move llvm_unreachable after the switch statement

79

(style) Drop the default case and move llvm_unreachable after the switch statement

90

(style) Drop the default case and move llvm_unreachable after the switch statement

102

(style) Drop the default case and move llvm_unreachable after the switch statement

113

(style) Drop the default case and move llvm_unreachable after the switch statement

llvm/lib/Target/M68k/MCTargetDesc/M68kFixupKinds.h

24

(style) Drop the default case and move llvm_unreachable after the switch statement

43

(style) Drop the default case and move llvm_unreachable after the switch statement

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

118

Can you get uninitialized variables further down? The switch statement isn't exhaustive so static analyzers will complain, add a default llvm_unreachable?

213

drop braces

[jrtc27]

llvm/lib/Target/M68k/MCTargetDesc/M68kBaseInfo.h

117

Commented-out code.

llvm/lib/Target/M68k/MCTargetDesc/M68kMCAsmInfo.cpp

30

This doesn't work; it's assumed to be a single byte (MCAsmStreamer passes ValueSize as 1 when using it), but if it did it would presumably need to be dealt with very carefully wrt endianness.

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

206

Capitalise

282

?

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

80

Capitalise

[myhsu]

• Addressed feedbacks

[myhsu]

llvm/lib/Target/M68k/MCTargetDesc/M68kMCAsmInfo.cpp

30

good catch, and you're right non-single byte fill value will never be used. I'll just remove this line to let MC use the default value.

[RKSimon]

LGTM

[jrtc27]

So, uh, correct me if I'm wrong but none of these patches add assembler or disassembler support? Should that not be present before a target is accepted? Otherwise there's no way to test any of the MC layer without including CodeGen.

[myhsu]

In D88390#2528951, @jrtc27 wrote:

So, uh, correct me if I'm wrong but none of these patches add assembler or disassembler support? Should that not be present before a target is accepted? Otherwise there's no way to test any of the MC layer without including CodeGen.

We don't have AsmParser and disassembler right now. We're using MIR as input to test our integrated assembler, so you're right, we can't test MC layer without CodeGen. But why would that will be a hard blocker?

[rengolin]

In D88390#2529089, @myhsu wrote:

We don't have AsmParser and disassembler right now. We're using MIR as input to test our integrated assembler, so you're right, we can't test MC layer without CodeGen. But why would that will be a hard blocker?

The real requirement is being able to produce code at the end. If that relies on (free, accessible) third-party tools, then it should be find for the first import. Some production targets have (free) proprietary tools that are required to continue the code generation (ex. NVPTX). Other targets have assemblers but can be tested (no available hardware). Having specific hard requirements would make it really hard for LLVM to have new backends.

What we need for the first batch is to know that the target can be used to generate correct code. It's less relevant if that can be shown with an integrated assembler or other tools at the end.

It would be important, however, to know if the m86k community intends to upstream the assembler before moving to production or if the target requires use of third-party closed source tools to work. This will help guide this and other reviews after the first merge, into production.

[myhsu]

In D88390#2530647, @rengolin wrote:

In D88390#2529089, @myhsu wrote:

We don't have AsmParser and disassembler right now. We're using MIR as input to test our integrated assembler, so you're right, we can't test MC layer without CodeGen. But why would that will be a hard blocker?

The real requirement is being able to produce code at the end. If that relies on (free, accessible) third-party tools, then it should be find for the first import. Some production targets have (free) proprietary tools that are required to continue the code generation (ex. NVPTX). Other targets have assemblers but can be tested (no available hardware). Having specific hard requirements would make it really hard for LLVM to have new backends.

Agree.

What we need for the first batch is to know that the target can be used to generate correct code. It's less relevant if that can be shown with an integrated assembler or other tools at the end.

That is also what I'm thinking: it's true that the current work need to use external assembler (GNU AS) to deal with cases like inline assembly, but at the same time it can handle common cases with integrated assembler just fine. We also have test cases on testing that part.

It would be important, however, to know if the m86k community intends to upstream the assembler before moving to production or if the target requires use of third-party closed source tools to work. This will help guide this and other reviews after the first merge, into production.

Speaking of that. Few days ago there is a (draft) PR sent to our GitHub repo: https://github.com/M680x0/M680x0-mono-repo/pull/20
Author of that PR is working on the AsmParser and disassembler (that PR is just a placeholder to inform us in order to prevent overlapping works). I'm pretty excited and looking forward to seeing it appearing upstream before becoming an official backend.

[jrtc27]

In D88390#2529089, @myhsu wrote:

In D88390#2528951, @jrtc27 wrote:

So, uh, correct me if I'm wrong but none of these patches add assembler or disassembler support? Should that not be present before a target is accepted? Otherwise there's no way to test any of the MC layer without including CodeGen.

We don't have AsmParser and disassembler right now. We're using MIR as input to test our integrated assembler, so you're right, we can't test MC layer without CodeGen. But why would that will be a hard blocker?

Ok, MIR is fine and makes sense, if rather clunky compared to assembly. Are those tests cleanly separated from actual CodeGen tests and labeled as such (so once there is an AsmParser and associated tests they can be removed)?

[glaubitz]

In D88390#2530878, @myhsu wrote:

In D88390#2530647, @rengolin wrote:

It would be important, however, to know if the m86k community intends to upstream the assembler before moving to production or if the target requires use of third-party closed source tools to work. This will help guide this and other reviews after the first merge, into production.

Speaking of that. Few days ago there is a (draft) PR sent to our GitHub repo: https://github.com/M680x0/M680x0-mono-repo/pull/20
Author of that PR is working on the AsmParser and disassembler (that PR is just a placeholder to inform us in order to prevent overlapping works). I'm pretty excited and looking forward to seeing it appearing upstream before becoming an official backend.

I just learned about that (because I hadn't any Github notifications enabled for the repo yet) and I'm super excited to see that PR.

It's happening what I was hoping and also expecting - external people just coming by and sending such improvements because they love the retro architecture and enjoy working on LLVM.

So, I think that should make @jrtc27 happy regarding this question :-).

[rengolin]

In D88390#2531077, @glaubitz wrote:

It's happening what I was hoping and also expecting - external people just coming by and sending such improvements because they love the retro architecture and enjoy working on LLVM.

Ideally, we should get them contributing to the upstream LLVM backend directly, but for that, we need the target in tree.

I think this is a good example of how active the community is and what we should expect to get before moving the target to production.

[glaubitz]

In D88390#2532131, @rengolin wrote:

In D88390#2531077, @glaubitz wrote:

It's happening what I was hoping and also expecting - external people just coming by and sending such improvements because they love the retro architecture and enjoy working on LLVM.

Ideally, we should get them contributing to the upstream LLVM backend directly, but for that, we need the target in tree.

Yep, and that's why I think it would be great to get the backend merged soonish and then let the community do it's work :).

I think this is a good example of how active the community is and what we should expect to get before moving the target to production.

Great to hear that and I fully agree. And all that despite of the age of the architecture :D.

[jrtc27]

llvm/lib/Target/M68k/MCTargetDesc/M68kELFObjectWriter.cpp

64

This is odd; processors shouldn't care about fixups. But regardless of the actual problem, more info would be helpful as this isn't particularly useful for anyone who doesn't already know the problem.

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.cpp

87

Still many instances of this

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

343–345

[myhsu]

• [NFC] Addressed some of the feedbacks

[jrtc27]

A few minor remaining issues, but once those are fixed I believe this is fine to land.

llvm/lib/Target/M68k/MCTargetDesc/M68kAsmBackend.cpp

217

This is how RISC-V does it (well, with OS.write("\x13\0\0\0", 4); and a similar one for 2-byte compressed instructions), so this is fine?

240

Commented-out code

llvm/lib/Target/M68k/MCTargetDesc/M68kELFObjectWriter.cpp

64

Please clarify this comment or delete if it it's wrong

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

379–382

should work as a nicer way to write that (using llvm/Support/EndianStream.h)

[myhsu]

• Addressed feedbacks

[myhsu]

llvm/lib/Target/M68k/MCTargetDesc/M68kELFObjectWriter.cpp

64

I confirm this comment is outdated

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

379–382

good to know, thanks

llvm/include/llvm/BinaryFormat/ELF.h

 #include "ELFRelocs/AMDGPU.def"
 };
 
 // ELF Relocation types for BPF
 enum {
 #include "ELFRelocs/BPF.def"
 };
 
+// ELF Relocation types for M68k
+enum {
+#include "ELFRelocs/M68k.def"
+};
+
 // MSP430 specific e_flags
 enum : unsigned {
   EF_MSP430_MACH_MSP430x11 = 11,
   EF_MSP430_MACH_MSP430x11x1 = 110,
   EF_MSP430_MACH_MSP430x12 = 12,
   EF_MSP430_MACH_MSP430x13 = 13,
   EF_MSP430_MACH_MSP430x14 = 14,
   EF_MSP430_MACH_MSP430x15 = 15,

llvm/include/llvm/BinaryFormat/ELFRelocs/M68k.def

+#ifndef ELF_RELOC
+#error "ELF_RELOC must be defined"
+#endif
+
+ELF_RELOC(R_68K_NONE,          0)  /* No reloc */
+ELF_RELOC(R_68K_32,            1)  /* Direct 32 bit  */
+ELF_RELOC(R_68K_16,            2)  /* Direct 16 bit  */
+ELF_RELOC(R_68K_8,             3)  /* Direct 8 bit  */
+ELF_RELOC(R_68K_PC32,          4)  /* PC relative 32 bit */
+ELF_RELOC(R_68K_PC16,          5)  /* PC relative 16 bit */
+ELF_RELOC(R_68K_PC8,           6)  /* PC relative 8 bit */
+ELF_RELOC(R_68K_GOTPCREL32,    7)  /* 32 bit PC relative GOT entry */
+ELF_RELOC(R_68K_GOTPCREL16,    8)  /* 16 bit PC relative GOT entry */
+ELF_RELOC(R_68K_GOTPCREL8,     9)  /* 8 bit PC relative GOT entry */
+ELF_RELOC(R_68K_GOTOFF32,      10) /* 32 bit GOT offset */
+ELF_RELOC(R_68K_GOTOFF16,      11) /* 16 bit GOT offset */
+ELF_RELOC(R_68K_GOTOFF8,       12) /* 8 bit GOT offset */
+ELF_RELOC(R_68K_PLT32,         13) /* 32 bit PC relative PLT address */
+ELF_RELOC(R_68K_PLT16,         14) /* 16 bit PC relative PLT address */
+ELF_RELOC(R_68K_PLT8,          15) /* 8 bit PC relative PLT address */
+ELF_RELOC(R_68K_PLTOFF32,      16) /* 32 bit PLT offset */
+ELF_RELOC(R_68K_PLTOFF16,      17) /* 16 bit PLT offset */
+ELF_RELOC(R_68K_PLTOFF8,       18) /* 8 bit PLT offset */
+ELF_RELOC(R_68K_COPY,          19) /* Copy symbol at runtime */
+ELF_RELOC(R_68K_GLOB_DAT,      20) /* Create GOT entry */
+ELF_RELOC(R_68K_JMP_SLOT,      21) /* Create PLT entry */
+ELF_RELOC(R_68K_RELATIVE,      22) /* Adjust by program base */
+/* These are GNU extensions to enable C++ vtable garbage collection.  */
+ELF_RELOC(R_68K_GNU_VTINHERIT, 23)
+ELF_RELOC(R_68K_GNU_VTENTRY,   24)
+/* TLS static relocations.  */
+ELF_RELOC(R_68K_TLS_GD32,      25)
+ELF_RELOC(R_68K_TLS_GD16,      26)
+ELF_RELOC(R_68K_TLS_GD8,       27)
+ELF_RELOC(R_68K_TLS_LDM32,     28)
+ELF_RELOC(R_68K_TLS_LDM16,     29)
+ELF_RELOC(R_68K_TLS_LDM8,      30)
+ELF_RELOC(R_68K_TLS_LDO32,     31)
+ELF_RELOC(R_68K_TLS_LDO16,     32)
+ELF_RELOC(R_68K_TLS_LDO8,      33)
+ELF_RELOC(R_68K_TLS_IE32,      34)
+ELF_RELOC(R_68K_TLS_IE16,      35)
+ELF_RELOC(R_68K_TLS_IE8,       36)
+ELF_RELOC(R_68K_TLS_LE32,      37)
+ELF_RELOC(R_68K_TLS_LE16,      38)
+ELF_RELOC(R_68K_TLS_LE8,       39)
+ELF_RELOC(R_68K_TLS_DTPMOD32,  40)
+ELF_RELOC(R_68K_TLS_DTPREL32,  41)
+ELF_RELOC(R_68K_TLS_TPREL32,   42)

llvm/include/llvm/IR/CallingConv.h

enum {
     /// Calling convention for emscripten __invoke_* functions. The first
     /// argument is required to be the function ptr being indirectly called.
     /// The remainder matches the regular calling convention.
     WASM_EmscriptenInvoke = 99,
 
     /// Calling convention used for AMD graphics targets.
     AMDGPU_Gfx = 100,
 
+    /// M68k_INTR - Calling convention used for M68k interrupt routines.
+    M68k_INTR = 101,
+
     /// The highest possible calling convention ID. Must be some 2^k - 1.
     MaxID = 1023
   };
 
 } // end namespace CallingConv
 
 } // end namespace llvm
 
 #endif // LLVM_IR_CALLINGCONV_H

llvm/include/llvm/Object/ELFObjectFile.h

 }
 
 template <class ELFT>
 StringRef ELFObjectFile<ELFT>::getFileFormatName() const {
   bool IsLittleEndian = ELFT::TargetEndianness == support::little;
   switch (EF.getHeader().e_ident[ELF::EI_CLASS]) {
   case ELF::ELFCLASS32:
     switch (EF.getHeader().e_machine) {
+    case ELF::EM_68K:
+      return "elf32-m68k";
     case ELF::EM_386:
       return "elf32-i386";
     case ELF::EM_IAMCU:
       return "elf32-iamcu";
     case ELF::EM_X86_64:
       return "elf32-x86-64";
     case ELF::EM_ARM:
       return (IsLittleEndian ? "elf32-littlearm" : "elf32-bigarm");
default:
     // FIXME: Proper error handling.
     report_fatal_error("Invalid ELFCLASS!");
   }
 }
 
 template <class ELFT> Triple::ArchType ELFObjectFile<ELFT>::getArch() const {
   bool IsLittleEndian = ELFT::TargetEndianness == support::little;
   switch (EF.getHeader().e_machine) {
+  case ELF::EM_68K:
+    return Triple::m68k;
   case ELF::EM_386:
   case ELF::EM_IAMCU:
     return Triple::x86;
   case ELF::EM_X86_64:
     return Triple::x86_64;
   case ELF::EM_AARCH64:
     return IsLittleEndian ? Triple::aarch64 : Triple::aarch64_be;
   case ELF::EM_ARM:

llvm/include/llvm/module.modulemap

module LLVM_BinaryFormat {
     textual header "BinaryFormat/ELFRelocs/ARM.def"
     textual header "BinaryFormat/ELFRelocs/ARC.def"
     textual header "BinaryFormat/ELFRelocs/AVR.def"
     textual header "BinaryFormat/ELFRelocs/BPF.def"
     textual header "BinaryFormat/ELFRelocs/CSKY.def"
     textual header "BinaryFormat/ELFRelocs/Hexagon.def"
     textual header "BinaryFormat/ELFRelocs/i386.def"
     textual header "BinaryFormat/ELFRelocs/Lanai.def"
+    textual header "BinaryFormat/ELFRelocs/M68k.def"
     textual header "BinaryFormat/ELFRelocs/Mips.def"
     textual header "BinaryFormat/ELFRelocs/MSP430.def"
     textual header "BinaryFormat/ELFRelocs/PowerPC64.def"
     textual header "BinaryFormat/ELFRelocs/PowerPC.def"
     textual header "BinaryFormat/ELFRelocs/RISCV.def"
     textual header "BinaryFormat/ELFRelocs/Sparc.def"
     textual header "BinaryFormat/ELFRelocs/SystemZ.def"
     textual header "BinaryFormat/ELFRelocs/VE.def"
     textual header "BinaryFormat/ELFRelocs/x86_64.def"
     textual header "BinaryFormat/WasmRelocs.def"
     textual header "BinaryFormat/MsgPack.def"
 }
 
 module LLVM_Config {
   requires cplusplus
   umbrella "Config"
   extern module LLVM_Extern_Config_Def "module.extern.modulemap"
   module * { export * }
 }
 

llvm/lib/Object/ELF.cpp

#define STRINGIFY_ENUM_CASE(ns, name)                                          \
   case ns::name:                                                               \
     return #name;
 
 #define ELF_RELOC(name, value) STRINGIFY_ENUM_CASE(ELF, name)
 
 StringRef llvm::object::getELFRelocationTypeName(uint32_t Machine,
                                                  uint32_t Type) {
   switch (Machine) {
+  case ELF::EM_68K:
+    switch (Type) {
+#include "llvm/BinaryFormat/ELFRelocs/M68k.def"
+    default:
+      break;
+    }
+    break;
   case ELF::EM_X86_64:
     switch (Type) {
 #include "llvm/BinaryFormat/ELFRelocs/x86_64.def"
     default:
       break;
     }
     break;
   case ELF::EM_386:

llvm/lib/ObjectYAML/ELFYAML.cpp

 #include "llvm/BinaryFormat/ELFRelocs/VE.def"
     break;
   case ELF::EM_CSKY:
 #include "llvm/BinaryFormat/ELFRelocs/CSKY.def"
     break;
   case ELF::EM_PPC64:
 #include "llvm/BinaryFormat/ELFRelocs/PowerPC64.def"
     break;
+  case ELF::EM_68K:
+#include "llvm/BinaryFormat/ELFRelocs/M68k.def"
+    break;
   default:
     // Nothing to do.
     break;
   }
 #undef ELF_RELOC
   IO.enumFallback<Hex32>(Value);
 }
 

llvm/lib/Target/M68k/MCTargetDesc/CMakeLists.txt

 add_llvm_component_library(LLVMM68kDesc
+  M68kAsmBackend.cpp
+  M68kELFObjectWriter.cpp
+  M68kInstPrinter.cpp
+  M68kMCAsmInfo.cpp
+  M68kMCCodeEmitter.cpp
   M68kMCTargetDesc.cpp
 
   LINK_COMPONENTS
   MC
   MCDisassembler
   Support
   M68kInfo
 
   ADD_TO_COMPONENT
   M68k
 )

llvm/lib/Target/M68k/MCTargetDesc/M68kAsmBackend.cpp

+//===-- M68kAsmBackend.cpp - M68k Assembler Backend ---------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains definitions for M68k assembler backend.
+///
+//===----------------------------------------------------------------------===//
+
+#include "MCTargetDesc/M68kBaseInfo.h"
+#include "MCTargetDesc/M68kFixupKinds.h"
+
+#include "llvm/ADT/StringSwitch.h"
+#include "llvm/BinaryFormat/ELF.h"
+#include "llvm/BinaryFormat/MachO.h"
+#include "llvm/MC/MCAsmBackend.h"
+#include "llvm/MC/MCELFObjectWriter.h"
+#include "llvm/MC/MCExpr.h"
+#include "llvm/MC/MCFixupKindInfo.h"
+#include "llvm/MC/MCInst.h"
+#include "llvm/MC/MCMachObjectWriter.h"
+#include "llvm/MC/MCObjectWriter.h"
+#include "llvm/MC/MCRegisterInfo.h"
+#include "llvm/MC/MCSectionCOFF.h"
+#include "llvm/MC/MCSectionELF.h"
+#include "llvm/MC/MCSectionMachO.h"
+#include "llvm/MC/MCSubtargetInfo.h"
+#include "llvm/Support/ErrorHandling.h"
+#include "llvm/Support/MathExtras.h"
+#include "llvm/Support/TargetRegistry.h"
+#include "llvm/Support/raw_ostream.h"
+
+using namespace llvm;
+
+namespace {
+
+class M68kAsmBackend : public MCAsmBackend {
+
+public:
+  M68kAsmBackend(const Target &T) : MCAsmBackend(support::big) {}
+
+  unsigned getNumFixupKinds() const override { return 0; }
+
+  void applyFixup(const MCAssembler &Asm, const MCFixup &Fixup,
+                  const MCValue &Target, MutableArrayRef<char> Data,
+                  uint64_t Value, bool IsResolved,
+                  const MCSubtargetInfo *STI) const override {
+    unsigned Size = 1 << getFixupKindLog2Size(Fixup.getKind());
+
+    assert(Fixup.getOffset() + Size <= Data.size() && "Invalid fixup offset!");
+
+    // Check that uppper bits are either all zeros or all ones.
+    // Specifically ignore overflow/underflow as long as the leakage is
+    // limited to the lower bits. This is to remain compatible with
+    // other assemblers.
+    assert(isIntN(Size * 8 + 1, Value) &&
+           "Value does not fit in the Fixup field");
+
+    // Write in Big Endian
+    for (unsigned i = 0; i != Size; ++i)
+      Data[Fixup.getOffset() + i] = uint8_t(Value >> ((Size - i - 1) * 8));
+  }
+
+  bool mayNeedRelaxation(const MCInst &Inst,
+                         const MCSubtargetInfo &STI) const override;
+
+  bool fixupNeedsRelaxation(const MCFixup &Fixup, uint64_t Value,
+                            const MCRelaxableFragment *DF,
+                            const MCAsmLayout &Layout) const override;
+
+  void relaxInstruction(MCInst &Inst,
+                        const MCSubtargetInfo &STI) const override;
+
+  /// Returns the minimum size of a nop in bytes on this target. The assembler
+  /// will use this to emit excess padding in situations where the padding
+  /// required for simple alignment would be less than the minimum nop size.
+  unsigned getMinimumNopSize() const override { return 2; }
+
+  /// Write a sequence of optimal nops to the output, covering \p Count bytes.
+  /// \return - true on success, false on failure
+  bool writeNopData(raw_ostream &OS, uint64_t Count) const override;
+};
+} // end anonymous namespace
+
+/// cc—Carry clear      GE—Greater than or equal
+/// LS—Lower or same    PL—Plus
+/// CS—Carry set        GT—Greater than
+/// LT—Less than
+/// EQ—Equal            HI—Higher
+/// MI—Minus            VC—Overflow clear
+///                     LE—Less than or equal
+/// NE—Not equal        VS—Overflow set
+static unsigned getRelaxedOpcodeBranch(const MCInst &Inst) {
+  unsigned Op = Inst.getOpcode();
+  switch (Op) {
+  default:
+    return Op;
+  case M68k::BRA8:
+    return M68k::BRA16;
+  case M68k::Bcc8:
+    return M68k::Bcc16;
+  case M68k::Bls8:
+    return M68k::Bls16;
+  case M68k::Blt8:
+    return M68k::Blt16;
+  case M68k::Beq8:
+    return M68k::Beq16;
+  case M68k::Bmi8:
+    return M68k::Bmi16;
+  case M68k::Bne8:
+    return M68k::Bne16;
+  case M68k::Bge8:
+    return M68k::Bge16;
+  case M68k::Bcs8:
+    return M68k::Bcs16;
+  case M68k::Bpl8:
+    return M68k::Bpl16;
+  case M68k::Bgt8:
+    return M68k::Bgt16;
+  case M68k::Bhi8:
+    return M68k::Bhi16;
+  case M68k::Bvc8:
+    return M68k::Bvc16;
+  case M68k::Ble8:
+    return M68k::Ble16;
+  case M68k::Bvs8:
+    return M68k::Bvs16;
+  }
+}
+
+static unsigned getRelaxedOpcodeArith(const MCInst &Inst) {
+  unsigned Op = Inst.getOpcode();
+  switch (Op) {
+  default:
+    return Op;
+    // NOTE there will be some relaxations for PCD and ARD mem for x20
+  }
+}
+
+static unsigned getRelaxedOpcode(const MCInst &Inst) {
+  unsigned R = getRelaxedOpcodeArith(Inst);
+  if (R != Inst.getOpcode())
+    return R;
+  return getRelaxedOpcodeBranch(Inst);
+}
+
+bool M68kAsmBackend::mayNeedRelaxation(const MCInst &Inst,
+                                       const MCSubtargetInfo &STI) const {
+  // Branches can always be relaxed in either mode.
+  if (getRelaxedOpcodeBranch(Inst) != Inst.getOpcode())
+    return true;
+
+  // Check if this instruction is ever relaxable.
+  if (getRelaxedOpcodeArith(Inst) == Inst.getOpcode())
+    return false;
+
+  // Check if the relaxable operand has an expression. For the current set of
+  // relaxable instructions, the relaxable operand is always the last operand.
+  // NOTE will change for x20 mem
+  unsigned RelaxableOp = Inst.getNumOperands() - 1;
+  if (Inst.getOperand(RelaxableOp).isExpr())
+    return true;
+
+  return false;
+}
+
+bool M68kAsmBackend::fixupNeedsRelaxation(const MCFixup &Fixup, uint64_t Value,
+                                          const MCRelaxableFragment *DF,
+                                          const MCAsmLayout &Layout) const {
+  // TODO Newer CPU can use 32 bit offsets, so check for this when ready
+  if (!isInt<16>(Value)) {
+    llvm_unreachable("Cannot relax the instruction, value does not fit");
+  }
+  // Relax if the value is too big for a (signed) i8. This means that byte-wide
+  // instructions have to matched by default
+  //
+  // NOTE
+  // A branch to the immediately following instruction automatically
+  // uses the 16-bit displacement format because the 8-bit
+  // displacement field contains $00 (zero offset).
+  return Value == 0 || !isInt<8>(Value);
+}
+
+// NOTE Can tblgen help at all here to verify there aren't other instructions
+// we can relax?
+void M68kAsmBackend::relaxInstruction(MCInst &Inst,
+                                      const MCSubtargetInfo &STI) const {
+  // The only relaxations M68k does is from a 1byte pcrel to a 2byte PCRel.
+  unsigned RelaxedOp = getRelaxedOpcode(Inst);
+
+  if (RelaxedOp == Inst.getOpcode()) {
+    SmallString<256> Tmp;
+    raw_svector_ostream OS(Tmp);
+    Inst.dump_pretty(OS);
+    OS << "\n";
+    report_fatal_error("unexpected instruction to relax: " + OS.str());
+  }
+
+  Inst.setOpcode(RelaxedOp);
+}
+
+bool M68kAsmBackend::writeNopData(raw_ostream &OS, uint64_t Count) const {
+  // Cannot emit NOP with size being not multiple of 16 bits.
+  if (Count % 2 != 0)
+    return false;
+
+  uint64_t NumNops = Count / 2;
+  for (uint64_t i = 0; i != NumNops; ++i) {
+    OS << "\x4E\x71";
+  }
+
+  return true;
+}
+
+namespace {
+
+class M68kELFAsmBackend : public M68kAsmBackend {
+public:
+  uint8_t OSABI;
+  M68kELFAsmBackend(const Target &T, uint8_t OSABI)
+      : M68kAsmBackend(T), OSABI(OSABI) {}
+
+  std::unique_ptr<MCObjectTargetWriter>
+  createObjectTargetWriter() const override {
+    return createM68kELFObjectWriter(OSABI);
+  }
+};
+
+} // end anonymous namespace
+
+MCAsmBackend *llvm::createM68kAsmBackend(const Target &T,
+                                         const MCSubtargetInfo &STI,
+                                         const MCRegisterInfo &MRI,
+                                         const MCTargetOptions &Options) {
+  const Triple &TheTriple = STI.getTargetTriple();
+  uint8_t OSABI = MCELFObjectTargetWriter::getOSABI(TheTriple.getOS());
+  return new M68kELFAsmBackend(T, OSABI);
+}

llvm/lib/Target/M68k/MCTargetDesc/M68kBaseInfo.h

+//===-- M68kBaseInfo.h - Top level definitions for M68k MC --*- C++ -*-----===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains small standalone helper functions and enum definitions
+/// for the M68k target useful for the compiler back-end and the MC
+/// libraries.  As such, it deliberately does not include references to LLVM
+/// core code gen types, passes, etc..
+///
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_TARGET_M68K_MCTARGETDESC_M68KBASEINFO_H
+#define LLVM_LIB_TARGET_M68K_MCTARGETDESC_M68KBASEINFO_H
+
+#include "M68kMCTargetDesc.h"
+
+#include "llvm/MC/MCExpr.h"
+#include "llvm/Support/DataTypes.h"
+#include "llvm/Support/ErrorHandling.h"
+
+#define GET_INSTRINFO_MI_OPS_INFO
+#define GET_INSTRINFO_OPERAND_TYPES_ENUM
+#define GET_INSTRINFO_LOGICAL_OPERAND_SIZE_MAP
+#include "M68kGenInstrInfo.inc"
+
+namespace llvm {
+
+namespace M68k {
+
+/// Enums for memory operand decoding. Supports these forms:
+/// (d,An)
+/// (d,An,Xn)
+/// ([bd,An],Xn,od)
+/// ([bd,An,Xn],od)
+/// TODO Implement scaling other than 1
+enum { MemDisp = 0, MemBase = 1, MemIndex = 2, MemOuter = 3 };
+
+/// Enums for pc-relative memory operand decoding. Supports these forms:
+/// (d,PC)
+/// (d,PC,Xn)
+/// ([bd,PC],Xn,od)
+/// ([bd,PC,Xn],od)
+enum { PCRelDisp = 0, PCRelIndex = 1, PCRelOuter = 2 };
+} // namespace M68k
+
+namespace M68kBeads {
+enum {
+  Ctrl = 0x0,
+  Bits1 = 0x1,
+  Bits2 = 0x2,
+  Bits3 = 0x3,
+  Bits4 = 0x4,
+  DAReg = 0x5,
+  DA = 0x6,
+  Reg = 0x7,
+  Disp8 = 0x8,
+  Imm8 = 0x9,
+  Imm16 = 0xA,
+  Imm32 = 0xB,
+  Imm3 = 0xC,
+};
+
+// Ctrl payload
+enum {
+  Term = 0x0,
+  Ignore = 0x1,
+};
+} // namespace M68kBeads
+
+/// This namespace holds all of the target specific flags that instruction info
+/// tracks.
+namespace M68kII {
+/// Target Operand Flag enum.
+enum TOF {
+
+  MO_NO_FLAG,
+
+  /// On a symbol operand this indicates that the immediate is the absolute
+  /// address of the symbol.
+  MO_ABSOLUTE_ADDRESS,
+
+  /// On a symbol operand this indicates that the immediate is the pc-relative
+  /// address of the symbol.
+  MO_PC_RELATIVE_ADDRESS,
+
+  /// On a symbol operand this indicates that the immediate is the offset to
+  /// the GOT entry for the symbol name from the base of the GOT.
+  ///
+  ///    name@GOT
+  MO_GOT,
+
+  /// On a symbol operand this indicates that the immediate is the offset to
+  /// the location of the symbol name from the base of the GOT.
+  ///
+  ///    name@GOTOFF
+  MO_GOTOFF,
+
+  /// On a symbol operand this indicates that the immediate is offset to the
+  /// GOT entry for the symbol name from the current code location.
+  ///
+  ///    name@GOTPCREL
+  MO_GOTPCREL,
+
+  /// On a symbol operand this indicates that the immediate is offset to the
+  /// PLT entry of symbol name from the current code location.
+  ///
+  ///    name@PLT
+  MO_PLT,
+}; // enum TOF
+
+/// Return true if the specified TargetFlag operand is a reference to a stub
+/// for a global, not the global itself.
+inline static bool isGlobalStubReference(unsigned char TargetFlag) {
+  switch (TargetFlag) {
+  default:
+    return false;
+  case M68kII::MO_GOTPCREL: // pc-relative GOT reference.
+  case M68kII::MO_GOT:      // normal GOT reference.
+    return true;
+  }
+}
+
+/// Return True if the specified GlobalValue is a direct reference for a
+/// symbol.
+inline static bool isDirectGlobalReference(unsigned char Flag) {
+  switch (Flag) {
+  default:
+    return false;
+  case M68kII::MO_NO_FLAG:
+  case M68kII::MO_ABSOLUTE_ADDRESS:
+  case M68kII::MO_PC_RELATIVE_ADDRESS:
+    return true;
+  }
+}
+
+/// Return true if the specified global value reference is relative to a 32-bit
+/// PIC base (M68kISD::GLOBAL_BASE_REG). If this is true, the addressing mode
+/// has the PIC base register added in.
+inline static bool isGlobalRelativeToPICBase(unsigned char TargetFlag) {
+  switch (TargetFlag) {
+  default:
+    return false;
+  case M68kII::MO_GOTOFF: // isPICStyleGOT: local global.
+  case M68kII::MO_GOT:    // isPICStyleGOT: other global.
+    return true;
+  }
+}
+
+/// Return True if the specified GlobalValue requires PC addressing mode.
+inline static bool isPCRelGlobalReference(unsigned char Flag) {
+  switch (Flag) {
+  default:
+    return false;
+  case M68kII::MO_GOTPCREL:
+  case M68kII::MO_PC_RELATIVE_ADDRESS:
+    return true;
+  }
+}
+
+/// Return True if the Block is referenced using PC
+inline static bool isPCRelBlockReference(unsigned char Flag) {
+  switch (Flag) {
+  default:
+    return false;
+  case M68kII::MO_PC_RELATIVE_ADDRESS:
+    return true;
+  }
+}
+
+static inline bool isAddressRegister(unsigned RegNo) {
+  switch (RegNo) {
+  case M68k::WA0:
+  case M68k::WA1:
+  case M68k::WA2:
+  case M68k::WA3:
+  case M68k::WA4:
+  case M68k::WA5:
+  case M68k::WA6:
+  case M68k::WA7:
+  case M68k::A0:
+  case M68k::A1:
+  case M68k::A2:
+  case M68k::A3:
+  case M68k::A4:
+  case M68k::A5:
+  case M68k::A6:
+  case M68k::A7:
+  case M68k::SP:
+    return true;
+  default:
+    return false;
+  }
+}
+
+static inline bool hasMultiMIOperands(unsigned Op, unsigned LogicalOpIdx) {
+  return M68k::getLogicalOperandSize(Op, LogicalOpIdx) > 1;
+}
+
+static inline unsigned getMaskedSpillRegister(unsigned order) {
+  switch (order) {
+  default:
+    return 0;
+  case 0:
+    return M68k::D0;
+  case 1:
+    return M68k::D1;
+  case 2:
+    return M68k::D2;
+  case 3:
+    return M68k::D3;
+  case 4:
+    return M68k::D4;
+  case 5:
+    return M68k::D5;
+  case 6:
+    return M68k::D6;
+  case 7:
+    return M68k::D7;
+  case 8:
+    return M68k::A0;
+  case 9:
+    return M68k::A1;
+  case 10:
+    return M68k::A2;
+  case 11:
+    return M68k::A3;
+  case 12:
+    return M68k::A4;
+  case 13:
+    return M68k::A5;
+  case 14:
+    return M68k::A6;
+  case 15:
+    return M68k::A7;
+  }
+}
+
+} // namespace M68kII
+
+} // namespace llvm
+
+#endif

llvm/lib/Target/M68k/MCTargetDesc/M68kELFObjectWriter.cpp

+//===---------- M68kELFObjectWriter.cpp - M68k ELF Writer ---*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains definitions for M68k ELF Writers
+///
+//===----------------------------------------------------------------------===//
+
+#include "MCTargetDesc/M68kFixupKinds.h"
+#include "MCTargetDesc/M68kMCTargetDesc.h"
+
+#include "llvm/BinaryFormat/ELF.h"
+#include "llvm/MC/MCAsmInfo.h"
+#include "llvm/MC/MCContext.h"
+#include "llvm/MC/MCELFObjectWriter.h"
+#include "llvm/MC/MCExpr.h"
+#include "llvm/MC/MCValue.h"
+#include "llvm/Support/ErrorHandling.h"
+
+using namespace llvm;
+
+namespace {
+class M68kELFObjectWriter : public MCELFObjectTargetWriter {
+public:
+  M68kELFObjectWriter(uint8_t OSABI);
+
+  ~M68kELFObjectWriter() override;
+
+protected:
+  unsigned getRelocType(MCContext &Ctx, const MCValue &Target,
+                        const MCFixup &Fixup, bool IsPCRel) const override;
+};
+} // namespace
+
+M68kELFObjectWriter::M68kELFObjectWriter(uint8_t OSABI)
+    : MCELFObjectTargetWriter(false, OSABI, ELF::EM_68K, /* RELA */ true) {}
+
+M68kELFObjectWriter::~M68kELFObjectWriter() {}
+
+enum M68kRelType { RT_32, RT_16, RT_8 };
+
+static M68kRelType
+getType(unsigned Kind, MCSymbolRefExpr::VariantKind &Modifier, bool &IsPCRel) {
+  switch (Kind) {
+  case FK_Data_4:
+  case FK_PCRel_4:
+    return RT_32;
+  case FK_PCRel_2:
+  case FK_Data_2:
+    return RT_16;
+  case FK_PCRel_1:
+  case FK_Data_1:
+    return RT_8;
+  }
+  llvm_unreachable("Unimplemented");
+}
+
+unsigned M68kELFObjectWriter::getRelocType(MCContext &Ctx,
+                                           const MCValue &Target,
+                                           const MCFixup &Fixup,
+                                           bool IsPCRel) const {
+  MCSymbolRefExpr::VariantKind Modifier = Target.getAccessVariant();
+  unsigned Kind = Fixup.getKind();
+  M68kRelType Type = getType(Kind, Modifier, IsPCRel);
+  switch (Modifier) {
+  default:
+    llvm_unreachable("Unimplemented");
+  case MCSymbolRefExpr::VK_None:
+    switch (Type) {
+    case RT_32:
+      return IsPCRel ? ELF::R_68K_PC32 : ELF::R_68K_32;
+    case RT_16:
+      return IsPCRel ? ELF::R_68K_PC16 : ELF::R_68K_16;
+    case RT_8:
+      return IsPCRel ? ELF::R_68K_PC8 : ELF::R_68K_8;
+    }
+    llvm_unreachable("Unrecognized size");
+  case MCSymbolRefExpr::VK_GOTPCREL:
+    switch (Type) {
+    case RT_32:
+      return ELF::R_68K_GOTPCREL32;
+    case RT_16:
+      return ELF::R_68K_GOTPCREL16;
+    case RT_8:
+      return ELF::R_68K_GOTPCREL8;
+    }
+    llvm_unreachable("Unrecognized size");
+  case MCSymbolRefExpr::VK_GOTOFF:
+    assert(!IsPCRel);
+    switch (Type) {
+    case RT_32:
+      return ELF::R_68K_GOTOFF32;
+    case RT_16:
+      return ELF::R_68K_GOTOFF16;
+    case RT_8:
+      return ELF::R_68K_GOTOFF8;
+    }
+    llvm_unreachable("Unrecognized size");
+  case MCSymbolRefExpr::VK_PLT:
+    switch (Type) {
+    case RT_32:
+      return ELF::R_68K_PLT32;
+    case RT_16:
+      return ELF::R_68K_PLT16;
+    case RT_8:
+      return ELF::R_68K_PLT8;
+    }
+    llvm_unreachable("Unrecognized size");
+  }
+}
+
+std::unique_ptr<MCObjectTargetWriter>
+llvm::createM68kELFObjectWriter(uint8_t OSABI) {
+  return std::make_unique<M68kELFObjectWriter>(OSABI);
+}

llvm/lib/Target/M68k/MCTargetDesc/M68kFixupKinds.h

+//===-- M68kFixupKinds.h - M68k Specific Fixup Entries ------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains M68k specific fixup entries.
+///
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_TARGET_M68k_MCTARGETDESC_M68kFIXUPKINDS_H
+#define LLVM_LIB_TARGET_M68k_MCTARGETDESC_M68kFIXUPKINDS_H
+
+#include "llvm/MC/MCFixup.h"
+
+namespace llvm {
+static inline unsigned getFixupKindLog2Size(unsigned Kind) {
+  switch (Kind) {
+  case FK_PCRel_1:
+  case FK_SecRel_1:
+  case FK_Data_1:
+    return 0;
+  case FK_PCRel_2:
+  case FK_SecRel_2:
+  case FK_Data_2:
+    return 1;
+  case FK_PCRel_4:
+  case FK_SecRel_4:
+  case FK_Data_4:
+    return 2;
+  }
+  llvm_unreachable("invalid fixup kind!");
+}
+
+static inline MCFixupKind getFixupForSize(unsigned Size, bool isPCRel) {
+  switch (Size) {
+  case 8:
+    return isPCRel ? FK_PCRel_1 : FK_Data_1;
+  case 16:
+    return isPCRel ? FK_PCRel_2 : FK_Data_2;
+  case 32:
+    return isPCRel ? FK_PCRel_4 : FK_Data_4;
+  case 64:
+    return isPCRel ? FK_PCRel_8 : FK_Data_8;
+  }
+  llvm_unreachable("Invalid generic fixup size!");
+}
+
+} // namespace llvm
+
+#endif

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.h

+//===-- M68kInstPrinter.h - Convert M68k MCInst to asm ------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains declarations for an M68k MCInst printer.
+///
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_TARGET_M68K_INSTPRINTER_M68KINSTPRINTER_H
+#define LLVM_LIB_TARGET_M68K_INSTPRINTER_M68KINSTPRINTER_H
+
+#include "llvm/MC/MCInstPrinter.h"
+
+namespace llvm {
+
+class TargetMachine;
+
+class M68kInstPrinter : public MCInstPrinter {
+public:
+  M68kInstPrinter(const MCAsmInfo &MAI, const MCInstrInfo &MII,
+                  const MCRegisterInfo &MRI)
+      : MCInstPrinter(MAI, MII, MRI) {}
+
+  // Autogenerated by tblgen.
+  void printInstruction(const MCInst *MI, uint64_t Address, raw_ostream &O);
+  static const char *getRegisterName(unsigned RegNo);
+
+  void printRegName(raw_ostream &OS, unsigned RegNo) const override;
+  void printInst(const MCInst *MI, uint64_t Address, StringRef Annot,
+                 const MCSubtargetInfo &STI, raw_ostream &O) override;
+
+  bool printAliasInstr(const MCInst *MI, uint64_t Address, raw_ostream &OS);
+  void printCustomAliasOperand(const MCInst *MI, unsigned OpIdx,
+                               unsigned PrintMethodIdx, raw_ostream &O);
+
+  std::pair<const char *, uint64_t> getMnemonic(const MCInst *MI) override;
+
+private:
+  void printOperand(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printImmediate(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  /// Print register mask for MOVEM instruction in order D0-D7,A0-A7
+  void printMoveMask(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  /// Print register mask for MOVEM instruction in order A7-A0,D7-D0
+  void printMoveMaskR(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printDisp(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printARIMem(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printARIPIMem(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printARIPDMem(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printARIDMem(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printARIIMem(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printAbsMem(const MCInst *MI, unsigned opNum, raw_ostream &O);
+  void printPCDMem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                   raw_ostream &O);
+  void printPCIMem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                   raw_ostream &O);
+
+  //===----------------------------------------------------------------------===//
+  // Specializations
+  //===----------------------------------------------------------------------===//
+  //
+  void printPCRelImm(const MCInst *MI, uint64_t Address, unsigned opNum,
+                     raw_ostream &O) {
+    printOperand(MI, opNum, O);
+  }
+
+  void printARI8Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIMem(MI, opNum, O);
+  }
+  void printARI16Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIMem(MI, opNum, O);
+  }
+  void printARI32Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIMem(MI, opNum, O);
+  }
+
+  void printARIPI8Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIPIMem(MI, opNum, O);
+  }
+  void printARIPI16Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIPIMem(MI, opNum, O);
+  }
+  void printARIPI32Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIPIMem(MI, opNum, O);
+  }
+
+  void printARIPD8Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIPDMem(MI, opNum, O);
+  }
+  void printARIPD16Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIPDMem(MI, opNum, O);
+  }
+  void printARIPD32Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIPDMem(MI, opNum, O);
+  }
+
+  void printARID8Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIDMem(MI, opNum, O);
+  }
+  void printARID16Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIDMem(MI, opNum, O);
+  }
+  void printARID32Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIDMem(MI, opNum, O);
+  }
+
+  void printARII8Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIIMem(MI, opNum, O);
+  }
+  void printARII16Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIIMem(MI, opNum, O);
+  }
+  void printARII32Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printARIIMem(MI, opNum, O);
+  }
+
+  void printAS8Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printAbsMem(MI, opNum, O);
+  }
+  void printAS16Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printAbsMem(MI, opNum, O);
+  }
+  void printAS32Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printAbsMem(MI, opNum, O);
+  }
+
+  void printAL8Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printAbsMem(MI, opNum, O);
+  }
+  void printAL16Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printAbsMem(MI, opNum, O);
+  }
+  void printAL32Mem(const MCInst *MI, unsigned opNum, raw_ostream &O) {
+    printAbsMem(MI, opNum, O);
+  }
+
+  void printPCD8Mem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                    raw_ostream &O) {
+    printPCDMem(MI, Address, opNum, O);
+  }
+  void printPCD16Mem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                     raw_ostream &O) {
+    printPCDMem(MI, Address, opNum, O);
+  }
+  void printPCD32Mem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                     raw_ostream &O) {
+    printPCDMem(MI, Address, opNum, O);
+  }
+
+  void printPCI8Mem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                    raw_ostream &O) {
+    printPCIMem(MI, Address, opNum, O);
+  }
+  void printPCI16Mem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                     raw_ostream &O) {
+    printPCIMem(MI, Address, opNum, O);
+  }
+  void printPCI32Mem(const MCInst *MI, uint64_t Address, unsigned opNum,
+                     raw_ostream &O) {
+    printPCIMem(MI, Address, opNum, O);
+  }
+};
+} // end namespace llvm
+
+#endif

llvm/lib/Target/M68k/MCTargetDesc/M68kInstPrinter.cpp

+//===-- M68kInstPrinter.cpp - Convert M68k MCInst to asm ----*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains definitions for an M68k MCInst printer.
+///
+//===----------------------------------------------------------------------===//
+
+// TODO Conform with all supported Motorola ASM syntax
+// Motorola's assembly has several syntax variants, especially on
+// addressing modes.
+// For example, you can write pc indirect w/ displacement as
+// `x(%pc)`, where `x` is the displacement imm, or `(x,%pc)`.
+// Currently we're picking the variant that is different from
+// GCC, albeit being recognizable by GNU AS.
+// Not sure what is the impact now (e.g. some syntax might
+// not be recognized by some old consoles' toolchains, in which
+// case we can not use our integrated assembler), but either way,
+// it will be great to support all of the variants in the future.
+
+#include "M68kInstPrinter.h"
+#include "M68kBaseInfo.h"
+
+#include "llvm/ADT/StringExtras.h"
+#include "llvm/MC/MCExpr.h"
+#include "llvm/MC/MCInst.h"
+#include "llvm/MC/MCInstrInfo.h"
+#include "llvm/MC/MCSymbol.h"
+#include "llvm/Support/ErrorHandling.h"
+#include "llvm/Support/raw_ostream.h"
+
+using namespace llvm;
+
+#define DEBUG_TYPE "asm-printer"
+
+#define PRINT_ALIAS_INSTR
+#include "M68kGenAsmWriter.inc"
+
+void M68kInstPrinter::printRegName(raw_ostream &OS, unsigned RegNo) const {
+  OS << "%" << getRegisterName(RegNo);
+}
+
+void M68kInstPrinter::printInst(const MCInst *MI, uint64_t Address,
+                                StringRef Annot, const MCSubtargetInfo &STI,
+                                raw_ostream &O) {
+  if (!printAliasInstr(MI, Address, O))
+    printInstruction(MI, Address, O);
+
+  printAnnotation(O, Annot);
+}
+
+void M68kInstPrinter::printOperand(const MCInst *MI, unsigned OpNo,
+                                   raw_ostream &O) {
+  const MCOperand &MO = MI->getOperand(OpNo);
+  if (MO.isReg()) {
+    printRegName(O, MO.getReg());
+    return;
+  }
+
+  if (MO.isImm()) {
+    printImmediate(MI, OpNo, O);
+    return;
+  }
+
+  assert(MO.isExpr() && "Unknown operand kind in printOperand");
+  MO.getExpr()->print(O, &MAI);
+}
+
+void M68kInstPrinter::printImmediate(const MCInst *MI, unsigned opNum,
+                                     raw_ostream &O) {
+  const MCOperand &MO = MI->getOperand(opNum);
+  if (MO.isImm())
+    O << '#' << MO.getImm();
+  else if (MO.isExpr()) {
+    O << '#';
+    MO.getExpr()->print(O, &MAI);
+  } else
+    llvm_unreachable("Unknown immediate kind");
+}
+
+void M68kInstPrinter::printMoveMask(const MCInst *MI, unsigned opNum,
+                                    raw_ostream &O) {
+  unsigned Mask = MI->getOperand(opNum).getImm();
+  assert((Mask & 0xFFFF) == Mask && "Mask is always 16 bits");
+
+  // A move mask is splitted into two parts:
+  // bits 0 ~ 7  correspond to D0 ~ D7 regs
+  // bits 8 ~ 15 correspond to A0 ~ A7 regs
+  //
+  // In the assembly syntax, we want to use a dash to replace
+  // a continuous range of registers. For example, if the bit
+  // mask is 0b101110, we want to print "D1-D3,D5" instead of
+  // "D1,D2,D3,D4,D5".
+  //
+  // However, we don't want a dash to cross between data registers
+  // and address registers (i.e. there shouldn't be a dash crossing
+  // bit 7 and 8) since that is not really intuitive. So we simply
+  // print the data register part (bit 0~7) and address register part
+  // separately.
+  uint8_t HalfMask;
+  unsigned Reg;
+  for (int s = 0; s < 16; s += 8) {
+    HalfMask = (Mask >> s) & 0xFF;
+    // Print separation comma only if
+    // both data & register parts have bit(s) set
+    if (s != 0 && (Mask & 0xFF) && HalfMask)
+      O << ',';
+
+    for (int i = 0; HalfMask; ++i) {
+      if ((HalfMask >> i) & 0b1) {
+        HalfMask ^= 0b1 << i;
+        Reg = M68kII::getMaskedSpillRegister(i + s);
+        printRegName(O, Reg);
+
+        int j = i;
+        while ((HalfMask >> (j + 1)) & 0b1)
+          HalfMask ^= 0b1 << ++j;
+
+        if (j != i) {
+          O << '-';
+          Reg = M68kII::getMaskedSpillRegister(j + s);
+          printRegName(O, Reg);
+        }
+
+        i = j;
+
+        if (HalfMask)
+          O << ',';
+      }
+    }
+  }
+}
+
+void M68kInstPrinter::printDisp(const MCInst *MI, unsigned opNum,
+                                raw_ostream &O) {
+  const MCOperand &Op = MI->getOperand(opNum);
+  if (Op.isImm()) {
+    O << Op.getImm();
+    return;
+  }
+  assert(Op.isExpr() && "Unknown operand kind in printOperand");
+  Op.getExpr()->print(O, &MAI);
+}
+
+void M68kInstPrinter::printARIMem(const MCInst *MI, unsigned opNum,
+                                  raw_ostream &O) {
+  O << '(';
+  printOperand(MI, opNum, O);
+  O << ')';
+}
+
+void M68kInstPrinter::printARIPIMem(const MCInst *MI, unsigned opNum,
+                                    raw_ostream &O) {
+  O << "(";
+  printOperand(MI, opNum, O);
+  O << ")+";
+}
+
+void M68kInstPrinter::printARIPDMem(const MCInst *MI, unsigned opNum,
+                                    raw_ostream &O) {
+  O << "-(";
+  printOperand(MI, opNum, O);
+  O << ")";
+}
+
+void M68kInstPrinter::printARIDMem(const MCInst *MI, unsigned opNum,
+                                   raw_ostream &O) {
+  O << '(';
+  printDisp(MI, opNum + M68k::MemDisp, O);
+  O << ',';
+  printOperand(MI, opNum + M68k::MemBase, O);
+  O << ')';
+}
+
+void M68kInstPrinter::printARIIMem(const MCInst *MI, unsigned opNum,
+                                   raw_ostream &O) {
+  O << '(';
+  printDisp(MI, opNum + M68k::MemDisp, O);
+  O << ',';
+  printOperand(MI, opNum + M68k::MemBase, O);
+  O << ',';
+  printOperand(MI, opNum + M68k::MemIndex, O);
+  O << ')';
+}
+
+// NOTE forcing (W,L) size available since M68020 only
+void M68kInstPrinter::printAbsMem(const MCInst *MI, unsigned opNum,
+                                  raw_ostream &O) {
+  const MCOperand &MO = MI->getOperand(opNum);
+  if (MO.isImm()) {
+    // ??? Print it in hex?
+    O << (unsigned int)MO.getImm();
+  } else {
+    printOperand(MI, opNum, O);
+  }
+}
+
+void M68kInstPrinter::printPCDMem(const MCInst *MI, uint64_t Address,
+                                  unsigned opNum, raw_ostream &O) {
+  O << '(';
+  printDisp(MI, opNum + M68k::PCRelDisp, O);
+  O << ",%pc)";
+}
+
+void M68kInstPrinter::printPCIMem(const MCInst *MI, uint64_t Address,
+                                  unsigned opNum, raw_ostream &O) {
+  O << '(';
+  printDisp(MI, opNum + M68k::PCRelDisp, O);
+  O << ",%pc,";
+  printOperand(MI, opNum + M68k::PCRelIndex, O);
+  O << ')';
+}

llvm/lib/Target/M68k/MCTargetDesc/M68kMCAsmInfo.h

+//===-- M68kMCAsmInfo.h - M68k Asm Info --------------------*- C++ -*--===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains the declarations of the M68k MCAsmInfo properties.
+///
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_TARGET_M68K_MCTARGETDESC_M68KMCASMINFO_H
+#define LLVM_LIB_TARGET_M68K_MCTARGETDESC_M68KMCASMINFO_H
+
+#include "llvm/MC/MCAsmInfoELF.h"
+
+namespace llvm {
+class Triple;
+
+class M68kELFMCAsmInfo : public MCAsmInfoELF {
+  void anchor() override;
+
+public:
+  explicit M68kELFMCAsmInfo(const Triple &Triple);
+};
+
+} // namespace llvm
+
+#endif

llvm/lib/Target/M68k/MCTargetDesc/M68kMCAsmInfo.cpp

+//===-- M68kMCAsmInfo.cpp - M68k Asm Properties -------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains the definitions of the M68k MCAsmInfo properties.
+///
+//===----------------------------------------------------------------------===//
+
+#include "M68kMCAsmInfo.h"
+
+#include "llvm/ADT/Triple.h"
+
+using namespace llvm;
+
+void M68kELFMCAsmInfo::anchor() {}
+
+M68kELFMCAsmInfo::M68kELFMCAsmInfo(const Triple &T) {
+  CodePointerSize = 4;
+  CalleeSaveStackSlotSize = 4;
+
+  IsLittleEndian = false;
+
+  // Debug Information
+  SupportsDebugInformation = true;
+
+  // Exceptions handling
+  ExceptionsType = ExceptionHandling::DwarfCFI;
+
+  CommentString = ";";
+}

llvm/lib/Target/M68k/MCTargetDesc/M68kMCCodeEmitter.cpp

+//===-- M68kMCCodeEmitter.cpp - Convert M68k code emitter ---*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file contains defintions for M68k code emitter.
+///
+//===----------------------------------------------------------------------===//
+
+#include "MCTargetDesc/M68kBaseInfo.h"
+#include "MCTargetDesc/M68kFixupKinds.h"
+#include "MCTargetDesc/M68kMCTargetDesc.h"
+
+#include "llvm/MC/MCCodeEmitter.h"
+#include "llvm/MC/MCContext.h"
+#include "llvm/MC/MCExpr.h"
+#include "llvm/MC/MCInst.h"
+#include "llvm/MC/MCInstrInfo.h"
+#include "llvm/MC/MCRegisterInfo.h"
+#include "llvm/MC/MCSubtargetInfo.h"
+#include "llvm/MC/MCSymbol.h"
+#include "llvm/Support/Debug.h"
+#include "llvm/Support/EndianStream.h"
+#include "llvm/Support/raw_ostream.h"
+
+namespace llvm {
+namespace M68k {
+// Forward declarations
+const uint8_t *getMCInstrBeads(unsigned);
+} // end namespace M68k
+} // end namespace llvm
+
+using namespace llvm;
+
+#define DEBUG_TYPE "m68k-mccodeemitter"
+
+namespace {
+class M68kMCCodeEmitter : public MCCodeEmitter {
+  M68kMCCodeEmitter(const M68kMCCodeEmitter &) = delete;
+  void operator=(const M68kMCCodeEmitter &) = delete;
+  const MCInstrInfo &MCII;
+  MCContext &Ctx;
+
+public:
+  M68kMCCodeEmitter(const MCInstrInfo &mcii, MCContext &ctx)
+      : MCII(mcii), Ctx(ctx) {}
+
+  ~M68kMCCodeEmitter() override {}
+
+  // TableGen'erated function
+  const uint8_t *getGenInstrBeads(const MCInst &MI) const {
+    return M68k::getMCInstrBeads(MI.getOpcode());
+  }
+
+  unsigned encodeBits(unsigned ThisByte, uint8_t Bead, const MCInst &MI,
+                      const MCInstrDesc &Desc, uint64_t &Buffer,
+                      unsigned Offset, SmallVectorImpl<MCFixup> &Fixups,
+                      const MCSubtargetInfo &STI) const;
+
+  unsigned encodeReg(unsigned ThisByte, uint8_t Bead, const MCInst &MI,
+                     const MCInstrDesc &Desc, uint64_t &Buffer, unsigned Offset,
+                     SmallVectorImpl<MCFixup> &Fixups,
+                     const MCSubtargetInfo &STI) const;
+
+  unsigned encodeImm(unsigned ThisByte, uint8_t Bead, const MCInst &MI,
+                     const MCInstrDesc &Desc, uint64_t &Buffer, unsigned Offset,
+                     SmallVectorImpl<MCFixup> &Fixups,
+                     const MCSubtargetInfo &STI) const;
+
+  void encodeInstruction(const MCInst &MI, raw_ostream &OS,
+                         SmallVectorImpl<MCFixup> &Fixups,
+                         const MCSubtargetInfo &STI) const override;
+};
+
+} // end anonymous namespace
+
+unsigned M68kMCCodeEmitter::encodeBits(unsigned ThisByte, uint8_t Bead,
+                                       const MCInst &MI,
+                                       const MCInstrDesc &Desc,
+                                       uint64_t &Buffer, unsigned Offset,
+                                       SmallVectorImpl<MCFixup> &Fixups,
+                                       const MCSubtargetInfo &STI) const {
+  unsigned Num = 0;
+  switch (Bead & 0xF) {
+  case M68kBeads::Bits1:
+    Num = 1;
+    break;
+  case M68kBeads::Bits2:
+    Num = 2;
+    break;
+  case M68kBeads::Bits3:
+    Num = 3;
+    break;
+  case M68kBeads::Bits4:
+    Num = 4;
+    break;
+  }
+  unsigned char Val = (Bead & 0xF0) >> 4;
+
+  LLVM_DEBUG(dbgs() << "\tEncodeBits"
+                    << " Num: " << Num << " Val: 0x");
+  LLVM_DEBUG(dbgs().write_hex(Val) << "\n");
+
+  Buffer |= (Val << Offset);
+
+  return Num;
+}
+
+unsigned M68kMCCodeEmitter::encodeReg(unsigned ThisByte, uint8_t Bead,
+                                      const MCInst &MI, const MCInstrDesc &Desc,
+                                      uint64_t &Buffer, unsigned Offset,
+                                      SmallVectorImpl<MCFixup> &Fixups,
+                                      const MCSubtargetInfo &STI) const {
+  bool DA, Reg;
+  switch (Bead & 0xF) {
+  default:
+    llvm_unreachable("Unrecognized Bead code for register type");
+  case M68kBeads::DAReg:
+    Reg = true;
+    DA = true;
+    break;
+  case M68kBeads::DA:
+    Reg = false;
+    DA = true;
+    break;
+  case M68kBeads::Reg:
+    Reg = true;
+    DA = false;
+    break;
+  }
+
+  unsigned Op = (Bead & 0x70) >> 4;
+  bool Alt = (Bead & 0x80);
+  LLVM_DEBUG(dbgs() << "\tEncodeReg"
+                    << " Op: " << Op << ", DA: " << DA << ", Reg: " << Reg
+                    << ", Alt: " << Alt << "\n");
+
+  auto MIOpIdx = M68k::getLogicalOperandIdx(MI.getOpcode(), Op);
+  bool IsPCRel = Desc.OpInfo[MIOpIdx].OperandType == MCOI::OPERAND_PCREL;
+
+  MCOperand MCO;
+  if (M68kII::hasMultiMIOperands(MI.getOpcode(), Op)) {
+    if (IsPCRel) {
+      assert(Alt &&
+             "PCRel addresses use Alt bead register encoding by default");
+      MCO = MI.getOperand(MIOpIdx + M68k::PCRelIndex);
+    } else {
+      MCO = MI.getOperand(MIOpIdx + (Alt ? M68k::MemIndex : M68k::MemBase));
+    }
+  } else {
+    assert(!Alt && "You cannot use Alt register with a simple operand");
+    MCO = MI.getOperand(MIOpIdx);
+  }
+
+  unsigned RegNum = MCO.getReg();
+  auto RI = Ctx.getRegisterInfo();
+
+  unsigned Written = 0;
+  if (Reg) {
+    uint32_t Val = RI->getEncodingValue(RegNum);
+    Buffer |= Val << Offset;
+    Offset += 3;
+    Written += 3;
+  }
+
+  if (DA) {
+    Buffer |= (char)M68kII::isAddressRegister(RegNum) << Offset;
+    Written++;
+  }
+
+  return Written;
+}
+
+static unsigned EmitConstant(uint64_t Val, unsigned Size, unsigned Pad,
+                             uint64_t &Buffer, unsigned Offset) {
+  assert(Size + Offset <= 64 && isUIntN(Size, Val) && "Value does not fit");
+
+  // Pad the instruction with zeros if any
+  // FIXME Emit zeros in the padding, since there might be trash in the buffer.
+  Size += Pad;
+
+  // Writing Value in host's endianness
+  Buffer |= Val << Offset;
+  return Size;
+}
+
+unsigned M68kMCCodeEmitter::encodeImm(unsigned ThisByte, uint8_t Bead,
+                                      const MCInst &MI, const MCInstrDesc &Desc,
+                                      uint64_t &Buffer, unsigned Offset,
+                                      SmallVectorImpl<MCFixup> &Fixups,
+                                      const MCSubtargetInfo &STI) const {
+  unsigned ThisWord = ThisByte / 2;
+  unsigned Size = 0;
+  unsigned Pad = 0;
+  unsigned FixOffset = 0;
+  int64_t Addendum = 0;
+  bool NoExpr = false;
+
+  unsigned Type = Bead & 0xF;
+  unsigned Op = (Bead & 0x70) >> 4;
+  bool Alt = (Bead & 0x80);
+
+  auto MIOpIdx = M68k::getLogicalOperandIdx(MI.getOpcode(), Op);
+  bool IsPCRel = Desc.OpInfo[MIOpIdx].OperandType == MCOI::OPERAND_PCREL;
+
+  // The PC value upon instruction reading of a short jump will point to the
+  // next instruction, thus we need to compensate 2 bytes, which is the diff
+  // between the patch point and the PC.
+  if (IsPCRel && ThisWord == 0)
+    Addendum -= 2;
+
+  switch (Type) {
+  // ??? what happens if it is not byte aligned
+  // ??? is it even possible
+  case M68kBeads::Disp8:
+    Size = 8;
+    Pad = 0;
+    FixOffset = ThisByte + 1;
+    Addendum += 1;
+    break;
+  case M68kBeads::Imm8:
+    Size = 8;
+    Pad = 8;
+    FixOffset = ThisByte;
+    break;
+  case M68kBeads::Imm16:
+    Size = 16;
+    Pad = 0;
+    FixOffset = ThisByte;
+    break;
+  case M68kBeads::Imm32:
+    Size = 32;
+    Pad = 0;
+    FixOffset = ThisByte;
+    break;
+  case M68kBeads::Imm3:
+    Size = 3;
+    Pad = 0;
+    NoExpr = true;
+    break;
+  }
+
+  LLVM_DEBUG(dbgs() << "\tEncodeImm"
+                    << " Op: " << Op << ", Size: " << Size << ", Alt: " << Alt
+                    << "\n");
+
+  MCOperand MCO;
+  if (M68kII::hasMultiMIOperands(MI.getOpcode(), Op)) {
+
+    if (IsPCRel) {
+      assert(!Alt && "You cannot use ALT operand with PCRel");
+      MCO = MI.getOperand(MIOpIdx + M68k::PCRelDisp);
+    } else {
+      MCO = MI.getOperand(MIOpIdx + (Alt ? M68k::MemOuter : M68k::MemDisp));
+    }
+
+    if (MCO.isExpr()) {
+      assert(!NoExpr && "Cannot use expression here");
+      const MCExpr *Expr = MCO.getExpr();
+
+      // This only makes sense for PCRel instructions since PC points to the
+      // extension word and Disp8 for example is right justified and requires
+      // correction. E.g. R_68K_PC32 is calculated as S + A - P, P for Disp8
+      // will be EXTENSION_WORD + 1 thus we need to have A equal to 1 to
+      // compensate.
+      // TODO count extension words
+      if (IsPCRel && Addendum != 0) {
+        Expr = MCBinaryExpr::createAdd(
+            Expr, MCConstantExpr::create(Addendum, Ctx), Ctx);
+      }
+
+      Fixups.push_back(MCFixup::create(
+          FixOffset, Expr, getFixupForSize(Size, IsPCRel), MI.getLoc()));
+      // Write zeros
+      return EmitConstant(0, Size, Pad, Buffer, Offset);
+    }
+
+  } else {
+    MCO = MI.getOperand(MIOpIdx);
+    if (MCO.isExpr()) {
+      assert(!NoExpr && "Cannot use expression here");
+      const MCExpr *Expr = MCO.getExpr();
+
+      if (Addendum != 0) {
+        Expr = MCBinaryExpr::createAdd(
+            Expr, MCConstantExpr::create(Addendum, Ctx), Ctx);
+      }
+
+      Fixups.push_back(MCFixup::create(
+          FixOffset, Expr, getFixupForSize(Size, IsPCRel), MI.getLoc()));
+      // Write zeros
+      return EmitConstant(0, Size, Pad, Buffer, Offset);
+    }
+  }
+
+  int64_t I = MCO.getImm();
+
+  // Store 8 as 0, thus making range 1-8
+  if (Type == M68kBeads::Imm3 && Alt) {
+    assert(I && "Cannot encode Alt Imm3 zero value");
+    I %= 8;
+  } else {
+    assert(isIntN(Size, I));
+  }
+
+  uint64_t Imm = I;
+
+  // 32 bit Imm requires HI16 first then LO16
+  if (Size == 32) {
+    Offset += EmitConstant((Imm >> 16) & 0xFFFF, 16, Pad, Buffer, Offset);
+    EmitConstant(Imm & 0xFFFF, 16, Pad, Buffer, Offset);
+    return Size;
+  }
+
+  return EmitConstant(Imm & (UINT64_MAX >> (64 - Size)), Size, Pad, Buffer,
+                      Offset);
+}
+
+#include "M68kGenMCCodeBeads.inc"
+
+void M68kMCCodeEmitter::encodeInstruction(const MCInst &MI, raw_ostream &OS,
+                                          SmallVectorImpl<MCFixup> &Fixups,
+                                          const MCSubtargetInfo &STI) const {
+  unsigned Opcode = MI.getOpcode();
+  const MCInstrDesc &Desc = MCII.get(Opcode);
+
+  LLVM_DEBUG(dbgs() << "EncodeInstruction: " << MCII.getName(Opcode) << "("
+                    << Opcode << ")\n");
+
+  const uint8_t *Beads = getGenInstrBeads(MI);
+  if (!Beads || !*Beads) {
+    llvm_unreachable("*** Instruction does not have Beads defined");
+  }
+
+  uint64_t Buffer = 0;
+  unsigned Offset = 0;
+  unsigned ThisByte = 0;
+
+  for (uint8_t Bead = *Beads; Bead; Bead = *++Beads) {
+    // Check for control beads
+    if (!(Bead & 0xF)) {
+      switch (Bead >> 4) {
+      case M68kBeads::Ignore:
+        continue;
+      }
+    }
+
+    switch (Bead & 0xF) {
+    default:
+      llvm_unreachable("Unknown Bead code");
+      break;
+    case M68kBeads::Bits1:
+    case M68kBeads::Bits2:
+    case M68kBeads::Bits3:
+    case M68kBeads::Bits4:
+      Offset +=
+          encodeBits(ThisByte, Bead, MI, Desc, Buffer, Offset, Fixups, STI);
+      break;
+    case M68kBeads::DAReg:
+    case M68kBeads::DA:
+    case M68kBeads::Reg:
+      Offset +=
+          encodeReg(ThisByte, Bead, MI, Desc, Buffer, Offset, Fixups, STI);
+      break;
+    case M68kBeads::Disp8:
+    case M68kBeads::Imm8:
+    case M68kBeads::Imm16:
+    case M68kBeads::Imm32:
+    case M68kBeads::Imm3:
+      Offset +=
+          encodeImm(ThisByte, Bead, MI, Desc, Buffer, Offset, Fixups, STI);
+      break;
+    }
+
+    // Since M68k is Big Endian we need to rotate each instruction word
+    while (Offset / 16) {
+      support::endian::write<uint16_t>(OS, Buffer, support::big);
+      Buffer >>= 16;
+      Offset -= 16;
+      ThisByte += 2;
+    }
+  }
+
+  assert(Offset == 0 && "M68k Instructions are % 2 bytes");
+  assert((ThisByte && !(ThisByte % 2)) && "M68k Instructions are % 2 bytes");
+}
+
+MCCodeEmitter *llvm::createM68kMCCodeEmitter(const MCInstrInfo &MCII,
+                                             const MCRegisterInfo &MRI,
+                                             MCContext &Ctx) {
+  return new M68kMCCodeEmitter(MCII, Ctx);
+}

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.h

+//===-- M68kMCTargetDesc.h - M68k Target Descriptions -------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// This file provides M68k specific target descriptions.
+///
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_TARGET_M68K_MCTARGETDESC_M68KMCTARGETDESC_H
+#define LLVM_LIB_TARGET_M68K_MCTARGETDESC_M68KMCTARGETDESC_H
+
+#include "llvm/MC/MCInstrDesc.h"
+#include "llvm/MC/MCObjectWriter.h"
+#include "llvm/Support/DataTypes.h"
+
+namespace llvm {
+class MCAsmBackend;
+class MCCodeEmitter;
+class MCContext;
+class MCInstrInfo;
+class MCRegisterInfo;
+class MCSubtargetInfo;
+class MCRelocationInfo;
+class MCTargetOptions;
+class Target;
+class Triple;
+class StringRef;
+class raw_ostream;
+class raw_pwrite_stream;
+
+extern Target TheM68kTarget;
+
+MCAsmBackend *createM68kAsmBackend(const Target &T, const MCSubtargetInfo &STI,
+                                   const MCRegisterInfo &MRI,
+                                   const MCTargetOptions &Options);
+
+MCCodeEmitter *createM68kMCCodeEmitter(const MCInstrInfo &MCII,
+                                       const MCRegisterInfo &MRI,
+                                       MCContext &Ctx);
+
+/// Construct an M68k ELF object writer.
+std::unique_ptr<MCObjectTargetWriter> createM68kELFObjectWriter(uint8_t OSABI);
+
+} // namespace llvm
+
+// Defines symbolic names for M68k registers. This defines a mapping from
+// register name to register number.
+#define GET_REGINFO_ENUM
+#include "M68kGenRegisterInfo.inc"
+
+// Defines symbolic names for the M68k instructions.
+#define GET_INSTRINFO_ENUM
+#include "M68kGenInstrInfo.inc"
+
+#define GET_SUBTARGETINFO_ENUM
+#include "M68kGenSubtargetInfo.inc"
+
+#endif

llvm/lib/Target/M68k/MCTargetDesc/M68kMCTargetDesc.cpp

 //===-- M68kMCTargetDesc.cpp - M68k Target Descriptions -----*- C++ -*-===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 ///
 /// \file
 /// This file provides M68k target specific descriptions.
 ///
 //===----------------------------------------------------------------------===//
 
-/// This is just a placeholder to make current
-/// commit buildable. Body of this function will
-/// be filled in later commits
-extern "C" void LLVMInitializeM68kTargetMC() {}
+#include "M68kMCTargetDesc.h"
+
+#include "M68kMCAsmInfo.h"
+
+#include "M68kInstPrinter.h"
+
+#include "llvm/MC/MCELFStreamer.h"
+#include "llvm/MC/MCInstPrinter.h"
+#include "llvm/MC/MCInstrInfo.h"
+#include "llvm/MC/MCRegisterInfo.h"
+#include "llvm/MC/MCSubtargetInfo.h"
+#include "llvm/MC/MCSymbol.h"
+#include "llvm/MC/MachineLocation.h"
+#include "llvm/Support/CommandLine.h"
+#include "llvm/Support/ErrorHandling.h"
+#include "llvm/Support/FormattedStream.h"
+#include "llvm/Support/TargetRegistry.h"
+
+using namespace llvm;
+
+#define GET_INSTRINFO_MC_DESC
+#include "M68kGenInstrInfo.inc"
+
+#define GET_SUBTARGETINFO_MC_DESC
+#include "M68kGenSubtargetInfo.inc"
+
+#define GET_REGINFO_MC_DESC
+#include "M68kGenRegisterInfo.inc"
+
+// TODO Implement feature set parsing logics
+static std::string ParseM68kTriple(const Triple &TT, StringRef CPU) {
+  return "";
+}
+
+static MCInstrInfo *createM68kMCInstrInfo() {
+  MCInstrInfo *X = new MCInstrInfo();
+  InitM68kMCInstrInfo(X); // defined in M68kGenInstrInfo.inc
+  return X;
+}
+
+static MCRegisterInfo *createM68kMCRegisterInfo(const Triple &TT) {
+  MCRegisterInfo *X = new MCRegisterInfo();
+  InitM68kMCRegisterInfo(X, llvm::M68k::A0, 0, 0, llvm::M68k::PC);
+  return X;
+}
+
+static MCSubtargetInfo *createM68kMCSubtargetInfo(const Triple &TT,
+                                                  StringRef CPU, StringRef FS) {
+  std::string ArchFS = ParseM68kTriple(TT, CPU);
+  if (!FS.empty()) {
+    if (!ArchFS.empty()) {
+      ArchFS = (ArchFS + "," + FS).str();
+    } else {
+      ArchFS = FS.str();
+    }
+  }
+  return createM68kMCSubtargetInfoImpl(TT, CPU, /*TuneCPU=*/CPU, ArchFS);
+}
+
+static MCAsmInfo *createM68kMCAsmInfo(const MCRegisterInfo &MRI,
+                                      const Triple &TT,
+                                      const MCTargetOptions &TO) {
+  MCAsmInfo *MAI = new M68kELFMCAsmInfo(TT);
+
+  // Initialize initial frame state.
+  // Calculate amount of bytes used for return address storing
+  int StackGrowth = -4;
+
+  // Initial state of the frame pointer is SP+StackGrowth.
+  // TODO: Add tests for `cfi_*` directives
+  MCCFIInstruction Inst = MCCFIInstruction::cfiDefCfa(
+      nullptr, MRI.getDwarfRegNum(llvm::M68k::SP, true), -StackGrowth);
+  MAI->addInitialFrameState(Inst);
+
+  // Add return address to move list
+  Inst = MCCFIInstruction::createOffset(
+      nullptr, MRI.getDwarfRegNum(M68k::PC, true), StackGrowth);
+  MAI->addInitialFrameState(Inst);
+
+  return MAI;
+}
+
+static MCRelocationInfo *createM68kMCRelocationInfo(const Triple &TheTriple,
+                                                    MCContext &Ctx) {
+  // Default to the stock relocation info.
+  return llvm::createMCRelocationInfo(TheTriple, Ctx);
+}
+
+static MCInstPrinter *createM68kMCInstPrinter(const Triple &T,
+                                              unsigned SyntaxVariant,
+                                              const MCAsmInfo &MAI,
+                                              const MCInstrInfo &MII,
+                                              const MCRegisterInfo &MRI) {
+  return new M68kInstPrinter(MAI, MII, MRI);
+}
+
+extern "C" void LLVMInitializeM68kTargetMC() {
+  Target &T = TheM68kTarget;
+
+  // Register the MC asm info.
+  RegisterMCAsmInfoFn X(T, createM68kMCAsmInfo);
+
+  // Register the MC instruction info.
+  TargetRegistry::RegisterMCInstrInfo(T, createM68kMCInstrInfo);
+
+  // Register the MC register info.
+  TargetRegistry::RegisterMCRegInfo(T, createM68kMCRegisterInfo);
+
+  // Register the MC subtarget info.
+  TargetRegistry::RegisterMCSubtargetInfo(T, createM68kMCSubtargetInfo);
+
+  // Register the code emitter.
+  TargetRegistry::RegisterMCCodeEmitter(T, createM68kMCCodeEmitter);
+
+  // Register the MCInstPrinter.
+  TargetRegistry::RegisterMCInstPrinter(T, createM68kMCInstPrinter);
+
+  // Register the MC relocation info.
+  TargetRegistry::RegisterMCRelocationInfo(T, createM68kMCRelocationInfo);
+
+  // Register the asm backend.
+  TargetRegistry::RegisterMCAsmBackend(T, createM68kAsmBackend);
+}

llvm/lib/Target/M68k/TargetInfo/M68kTargetInfo.cpp

 //===-- M68kTargetInfo.cpp - M68k Target Implementation -----*- C++ -*-===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 ///
 /// \file
 /// This file contains M68k target initializer.
 ///
 //===----------------------------------------------------------------------===//
+#include "MCTargetDesc/M68kMCTargetDesc.h"
+#include "llvm/Support/TargetRegistry.h"
 
-/// This is just a placeholder to make current
-/// commit buildable. Body of this function will
-/// be filled in later commits
-extern "C" void LLVMInitializeM68kTargetInfo() {}
+using namespace llvm;
+
+Target llvm::TheM68kTarget;
+
+extern "C" void LLVMInitializeM68kTargetInfo() {
+  RegisterTarget<Triple::m68k, /*HasJIT=*/true> X(
+      TheM68kTarget, "m68k", "Motorola 68000 family", "M68k");
+}