llvmPackages_15: init at 15.0.7#194634
Conversation
c11356f to
30373c3
Compare
79aa175 to
8484b4d
Compare
8484b4d to
123aa5e
Compare
|
With ab9d3df, After that:
|
|
Fixed one
Unfortunately there doesn't seem to be a way to communicate this to the I opened a PR for this (#194994) but given that |
123aa5e to
d0c37e4
Compare
|
With 974ea1c this now builds on Testing Update:
|
None of the patches required any touch-up; the only change of note is:
- due to changes in the libc++/libc++abi build
(https://reviews.llvm.org/D120719 and https://reviews.llvm.org/D131037)
we have to add an extra build option to the libc++ header only
build that sidesteps bits of the libc++ build config that assume
libc++-abi is present in the build:
https://github.com/llvm/llvm-project/blob/4f827318e3e8ccab4ff131e06234caa827e91e4e/libcxx/src/CMakeLists.txt#L255-L256
Rather than maintaining a precise set of build options that let us dodge
referencing libc++-abi variables in the libc++ header only build, we set
`LIBCXX_CXX_ABI` to `none`, as suggested by @lovesegfault.
More discussion about this here: #194634 (comment)
Co-authored-by: Bernardo Meurer <[email protected]>
See the comments here for context: #194634 (comment) Co-authored-by: Weijia Wang <[email protected]>
this is a backport of 8afa321 from NixOS#194634 (llvmPackages_15)
…as overridable args backport of d231d18 from NixOS#194634 (llvmPackages_15)
Port of bc4dbee ("llvmPackages_15: updates for LLVM 15"). None of the patches required any touch-up; the only change of note is: - due to changes in the libc++/libc++abi build (https://reviews.llvm.org/D120719 and https://reviews.llvm.org/D131037) we have to add an extra build option to the libc++ header only build that sidesteps bits of the libc++ build config that assume libc++-abi is present in the build: https://github.com/llvm/llvm-project/blob/4f827318e3e8ccab4ff131e06234caa827e91e4e/libcxx/src/CMakeLists.txt#L255-L256 Rather than maintaining a precise set of build options that let us dodge referencing libc++-abi variables in the libc++ header only build, we set `LIBCXX_CXX_ABI` to `none`, as suggested by @lovesegfault. More discussion about this here: #194634 (comment) Co-authored-by: Bernardo Meurer <[email protected]>
Port of 4fabcf4 ("llvmPackages_15: apply #211230 to llvmPackages_15"). See the comments here for context: #194634 (comment) Co-authored-by: Weijia Wang <[email protected]>
Partial port of 00839fe ("llvmPackages_15.compiler-rt: fixes for Darwin") see: #194634 (comment)
Partial port of 00839fe ("llvmPackages_15.compiler-rt: fixes for Darwin") see: NixOS#194634 (comment)
Port of bc4dbee ("llvmPackages_15: updates for LLVM 15"). None of the patches required any touch-up; the only change of note is: - due to changes in the libc++/libc++abi build (https://reviews.llvm.org/D120719 and https://reviews.llvm.org/D131037) we have to add an extra build option to the libc++ header only build that sidesteps bits of the libc++ build config that assume libc++-abi is present in the build: https://github.com/llvm/llvm-project/blob/4f827318e3e8ccab4ff131e06234caa827e91e4e/libcxx/src/CMakeLists.txt#L255-L256 Rather than maintaining a precise set of build options that let us dodge referencing libc++-abi variables in the libc++ header only build, we set `LIBCXX_CXX_ABI` to `none`, as suggested by @lovesegfault. More discussion about this here: NixOS#194634 (comment) Co-authored-by: Bernardo Meurer <[email protected]>
Port of 4fabcf4 ("llvmPackages_15: apply NixOS#211230 to llvmPackages_15"). See the comments here for context: NixOS#194634 (comment) Co-authored-by: Weijia Wang <[email protected]>
this is a backport of 8afa321 from NixOS#194634 (llvmPackages_15)
…as overridable args backport of d231d18 from NixOS#194634 (llvmPackages_15)
Description of changes
The commit messages in this PR have more details but at a high level this PR contains a few different kinds of changes:
llvmPackages_15clangfrom this package set to bootstraplibcxx, sidestepping some of the issues we've seen with older LLVM package setsninjawhere possiblex86_64andaarch64)lldb, manpages) and updates (LLVM license, dropping some flags we don't need anymore)TODOs:
stagingfrom upstreammasterpkgsLLVMon darwin evals; I think that's good enough for this PRllvmPackages_latest(targetingstaging); 9088010 was dropped from this PRllvmPackages_15llvmPackages_{9,10,11,12,13,14}-DCMAKE_CXX_FLAGSfromclang15.0.7#194634 (comment))15.0.7#194634 (comment))15.0.7#194634 (comment))staging, try dropping this patch (postponed)x86_64-darwin(llvmPackages_15: init at15.0.7#194634 (comment))llvmPackages_{13,14}(llvmPackages_{13,14}.lldb: fix build on x86 macOS #195013)15.0.7#194634 (comment))litllvmPackages_{13,14,15}(llvmPackages_15: init at15.0.7#194634 (comment))x86_64-darwin: 3 failing LLVM testsaarch64-darwinlinking against hostCoreFoundation?roundevenfonly on x86_64? are we expected to have this symbol? (this test doesn't fail onaarch64-darwin; perhaps this is a manifestation of the host frameworks, i.e. mayberoundevenfis available in newer SDKs?)enablePollyinllvmPackages_{12,13,14}libclang-rtand test againstclangNoLibc(with lsan/asan/ubsan/fuzzer variants tested too)libc++abi.sodidn't seem to be linked against olderlibc++abion non-darwin anyways (where the default stdenv wasn't LLVM based) and now we unconditionally use the clang from the same package set which haslibcxxset tonulland thus doesn't try to link in anylibc++abienabledSharedinlibcxxabi) that looks at the output ofllvm-readelf --needed-libs lib++abi.{dylib/so}and scans it forlibc++abideps that aren't thelibc++abibeing scannedllvmPackages_15; if this PR is merged first we should update these PRs to changellvmPackages_15) — not an exhaustive list:_14yet)Checks:
llvmPackages_15.*nix path-info --file ./test.nix --derivationnix path-info --file ./test.nix --derivation --arg systems '[ "aarch64-darwin" "x86_64-darwin" ]' --arg testStdenvPackages '[]'pkgsCross.x86_64-freebsd.boost(withllvmPackages = llvmPackages_15)nix build --expr '(import ./. { system = "x86_64-linux"; overlays = [(f: p: { llvmPackages = f.llvmPackages_15; })]; }).pkgsCross.x86_64-freebsd.boost'libcxxnixpkgs-reviewllvmPackages_latestCloses #191132.
Things done
sandbox = trueset 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.shto update generated release notes