-
-
Notifications
You must be signed in to change notification settings - Fork 15.1k
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
llvmPackages_git.*: Bump to newer commit #154465
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The question whether to update or not to, I'd say depends on how much hassle it is for you to rebase the patches etc.?
Haven't looked at the patches or tried building yet, can take a closer look later.
@@ -10,12 +12,26 @@ | |||
, headersOnly ? false | |||
}: | |||
|
|||
let | |||
basename = "libcxx"; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nit: it's basename
here and baseName
in compiler-rt
.
inherit version; | ||
|
||
inherit src; | ||
sourceRoot = "source/libcxx"; | ||
src = runCommand "${pname}-src-${version}" {} '' |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can we maybe generalize this runCommand
thing and replace it with some kind of function (passed in from the top level) which gets a dir and a list of extra paths?
I don't think we can use filterSource for this sadly though.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah I think that would be good, just didn't do it yet. I did something similar in #82131 i should perhaps pull out.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah, that looks right!
@@ -22,7 +23,7 @@ | |||
|| stdenv.isAarch32 # broken for the armv7l builder | |||
) | |||
, enablePolly ? false | |||
}: | |||
} @args: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
args
is unused.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ah thanks, yeah that was before I renamed src
to monorepoSrc
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Removing it in #268812
Thanks a lot @Ericson2314 and @sternenseemann! :) Unfortunately,
I wouldn't mind merging it. The patches situation isn't ideal (not to say can be a bit frustrating at times :P) but I guess for LLVM it should be fine (and this doesn't make it worse anyway, it just adds a bit more content to Nixpkgs). Hopefully the rest of the big patches will land soon... :) Anyway: @Ericson2314 and @sternenseemann wouldn't you be interested to add yourself officially as LLVM package maintainers? AFAIK you're currently doing by far most of the work anyway and usually end up reviewing other PRs as well. So it seems like that would really make sense, unless you don't want to for some reason. (AFAIK most listed LLVM maintainers aren't that active btw, e.g., I'm only in there since I ended up having to add new versions for Chromium.) |
Yeah I think adding us would make sense. |
Can we merge this now so that we can package LLVM 14.0.0-rc1? (Probably makes sense to pick a revision that doesn't require additional changes for 14.0.0-rc1 (e.g., that exact revision or the commit where LLVM 14 was branched off from master) but maybe that's already the case for the current revision.) I'm a bit late as it looks like it's already required for Chromium M99 which will be released on Mar 1. |
|
Motivation for this change
Not sure whether we want to merge this now, or wait for more patches to land upstream first, but it was good to at least confirm the CMake is still working
Things done
sandbox = true
set innix.conf
? (See Nix manual)nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD"
. Note: all changes have to be committed, also see nixpkgs-review usage./result/bin/
)nixos/doc/manual/md-to-db.sh
to update generated release notes