Are you planning to soon introduce an asm parser and disassembler so that you can test the encodings? I recommend doing this early do that you can test those along the way.

Yes, we plan to support asm parser in disassembler, but I would not say these were nearby plans. Currently we can just generate assembler and then use GCC as with Nios2 support to generate object-files. But I'm OK with your recommendation, I think we can focus on that before continue committing new instructions.

Hi Craig,

Fixed a typo when assigning default CPU for Nios2.
Added a test checking for acceptance of Nios2 processors.

Updates according to Craig's comments.

Thanks Eric for your help.

I've prepared another two patches, that include DAG selector, assembly printing and all NiosII instructions. But it makes sense to upload them for review after I commit the current with Subtarget. Could you please take a look?

I made necessary fixes after Craig's comments.

Thanks, Craig!

Changes according to Craig's comments.

Is there enough functional here that there should be tests for? i.e. make sure march/mcpu switches are recognized, that the target is recognized, etc.

Initially I thought about adding such check for the triple is supported. But it was not obvious how to do that and llvm-experienced guys around said it was not a big sense to add tests at this stage.
Even now I use "not not llc" to make my lit test work. Anyway I tried to create such test case, please look. The test checks for -mtriple=nios2 and -march=nios2 are accepted. Further I will expand this tests with -mcpu after adding CPUs.

I've updated patch according to comments from David and Craig.

Thanks Craig for the good advice. You are right I can check llc with MCU triple using option -use-ctors. Please see updated patch.

In D32543#738595, @ahatanak wrote:

Do you need a front-end test? Since the changes are in the backend, I think it's better to add the test to the backend (using llc).