New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Boot hangs on M1 macs #4257
Comments
Yep, same problem here. |
M1 macs will require their own release binaries, right now we're only building for x86-64. If any M1 pilots are game, I'd be curious to know if our current build works locally without changes. /cc @brendanhay |
Curious if anyone has tried running it under arch -x86_64
…On Mon, Jan 11, 2021 at 09:49, Joe Bryan ***@***.***> wrote:
M1 macs will require their own release binaries, right now we're only
building for x86-64.
If any M1 pilots are game, I'd be curious to know if our current build
works locally without changes.
/cc @brendanhay <https://github.com/brendanhay>
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#4257 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAOFPBVSBQTTQ2TQY6H7GH3SZM23TANCNFSM4V33MHSQ>
.
|
I use an M1 mac and |
No easy way to build it yet, the aarch64-darwin version of nix is still extremely experimental. |
Looks like it's coming along: NixOS/nixpkgs#95903 (comment) |
Looks like things have matured enough, trying to build it shouldn't be a herculean effort. If somebody wants to follow the steps in this comment and the comment above it, you should be able to get an aarch64-darwin urbit. I'd recommend just trying to build the normal nix, dynamic c binaries. Don't try static versions or urbit-king unless you want extra challenge, and trying to build urbit-king probably just won't work (but, if it did, also might cause a not insignificant amount of ssd wear with the abnormally high swap usage of the m1) There might be problems with the lmdb mapsize with macos on aarch64, just like I discovered with linux on aarch64. Have to see if it runs without modification. |
If nix support on M1 is still too flaky, one could reuse the |
Commit 23634f6 in #4675 ought to do the trick. This build process uses an updated
Incidentally, I found out why boot was hanging: on aarch64-apple-darwin, the location loom was |
This sounds very promising |
Can confirm that changing the loombase as per 23634f6, compiling a static build on a intel mac will result in a working binary on apple silicon. Will be running under rosetta, but it's a start. |
This works for me on my M1, just make sure to use the 23634f6. I just checked out this commit from the |
GitHub Actions say they plan to have M1 runners up by EOY. It's probably best to put off setting up official M1 builds till then, as setting up an M1 build server and hooking it up as an external runner is a bit hairy, and on the other hand x64 Urbit binaries run fine with Rosetta. |
building and running on m1 seems to work now |
It does indeed, closing @joemfb |
when release binaries for m1? |
https://github.com/urbit/vere/releases check out the |
Describe the bug
Startup hangs when urbit is run on an M1-based Apple machine.
To Reproduce
Steps to reproduce the behaviour:
urbit -w sampel-palnet -k /path/to/keyfile
on an M1 macExpected behaviour
Boot should proceed normally.
Screenshots
If applicable, add screenshots to help explain your problem.
System (please supply the following information, if relevant):
%base
hash (use+trouble
to check): n/a, unable to complete boot processAdditional context
Reporting this on behalf of a couple of users on Discord who have encountered this, so my ability to test things is going to be limited.
The text was updated successfully, but these errors were encountered: