-2

After installing gnat for darwin (arm) macos, following this guide (see "AdaCore compiler on mac" section), I could not find the gnatmake in the /bin folder that all the other binaries were in. Quite confused. (I wanted to install gnat to build ghdl.)

Simon Wright
  • 25,108
  • 2
  • 35
  • 62
dcxst
  • 173
  • 3
  • 8
  • 1
    The Adacore link provides access to x86_64 GNAT binaries that can be run using Rosetta. I've used Simon Wright's earlier gcc-12.1.0-aarch64 download to build GHDL 3.0.0-dev (2.0.0.r444.ga33efb91d) a year ago. Both have gnatmake and Apple Silicon binaries can be more that twice as fast (the mcode version providing JIT code generation currently only supports x86_64). See the [#Issue 1115 comment](https://github.com/ghdl/ghdl/issues/1115#issuecomment-1176963769), a brief synopsis of the effort. An arm64 version isn't released, ghdl lacking access to a github build platform. – user16145658 Jul 11 '23 at 23:42
  • I forgot to specify you have to navigate to the correct binary from the link. There isn't an explicit link because it's in a dropdown menu. – dcxst Jul 12 '23 at 18:16

1 Answers1

1

I presume you’re running on a Mac with Apple silicon (e.g. M1, M2).

You can run Intel binaries on Apple silicon under Rosetta emulation (it’s very fast, don’t worry). The machine will ask you if that’s OK the first time, thereafter it’s transparent.

For Intel (x86_64) binaries, see here.

For Apple (aarch64, aka arm64) binaries, see here

For the IDE, GNAT Studio, see here. There’s an issue: not all the menu choices respond to mouse clicks, you may have to use the cursor keys and [RET].

Simon Wright
  • 25,108
  • 2
  • 35
  • 62
  • Thanks. I found another workaround, so I haven't tested it. I'm curious, if you downvoted this post (someone did), what was the reason? (would appreciate feedback for future posts) – dcxst Jul 12 '23 at 18:27
  • 2
    SO was originally supposed to be about programming problems; people were supposed to post actual code showing what they’d tried, what the expected to happen, and what actually happened. Your Q clearly doesn’t match _that_ model. On the other hand, there are few enough Ada responders that _I_ try to answer questions like yours. Not that I wouldn’t downvote someone who obviously hadn’t made the least effort! – Simon Wright Jul 13 '23 at 19:22
  • Noted, ty for the feedback – dcxst Jul 18 '23 at 04:27