| Commit message (Collapse) | Author | Age |
|
|
|
| |
addendum to https://github.com/NixOS/nixpkgs/pull/85945
|
|
|
|
|
|
|
| |
I hate the thing too even though I made it, and rather just get rid of
it. But we can't do that yet. In the meantime, this brings us more
inline with autoconf and will make it slightly easier for me to write a
pkg-config wrapper, which we need.
|
|
|
|
| |
No need for rebuilds on other platforms
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
This is a speculative fix since the llvm_10 build runs OOM currently but it
should also be affected.
See parent commit for more info.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The armv7l builder has has the same issue some of the aarch64 ones have (#56245)
See:
https://hydra.nixos.org/eval/1583749?filter=llvm_&compare=1583176&full=#tabs-still-fail
https://nix-cache.s3.amazonaws.com/log/3ppqyssyqj5d6sh257zfb5rb6km4m6sa-llvm-7.1.0.drv
https://nix-cache.s3.amazonaws.com/log/rmacd5xwkw3lyb6zknc3sgdcb3cvk6k1-llvm-8.0.1.drv
https://nix-cache.s3.amazonaws.com/log/6s8pasqwkrzzv5lgw1l22zy7mrpzx2p3-llvm-9.0.1.drv
|
| |
| |
| |
| |
| | |
LLVM builds frequently time out on hydra at least on AArch64. Marking
big-parallel gives them more resources to finish quicker.
|
| |
| |
| |
| |
| | |
LLVM builds frequently time out on hydra at least on AArch64. Marking
big-parallel gives them more resources to finish quicker.
|
| |
| |
| |
| |
| | |
LLVM builds frequently time out on hydra at least on AArch64. Marking
big-parallel gives them more resources to finish quicker.
|
| |
| |
| |
| |
| | |
LLVM builds frequently time out on hydra at least on AArch64. Marking
big-parallel gives them more resources to finish quicker.
|
| |
| |
| |
| |
| | |
LLVM builds frequently time out on hydra at least on AArch64. Marking
big-parallel gives them more resources to finish quicker.
|
|/
|
|
| |
Fixes #85622
|
|\ |
|
| | |
|
| |
| |
| |
| |
| | |
In 0c73297c075cf4fd87922f621fc2358c65b659f3 I made some changes up to 9
for 20.03 and master, now I do the same for 10 which is just on master.
|
|\| |
|
| | |
|
|/
|
|
| |
also: https://www.cmake.org/Wiki/CMake_RPATH_handling has moved
|
|
|
|
|
| |
updated hashes and url
updated comment regarding version/release_version to make it clearer
|
|\
| |
| | |
llvmPackages_10: rc3 -> rc4 -> rc5, polly enabled by default
|
| |
| |
| |
| | |
updated versino and hashes for new rc
|
| |
| |
| |
| |
| |
| |
| | |
There is no good reason to have separate builds for polly and no-polly
versions. wwwwwThe reason for the two versions was (as far as I can
tell) to defer rebuilds (see ed60483257b62c32b5b58c6e91c9c8cd586f77d0).
Polly is now enabled by default.
|
| |
| |
| |
| |
| | |
Only needed to update hashes and the version.
Updated comment for extension handling patch
|
|\ \
| |/
|/| |
Misc fixes for armv6 bare metal cross
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- Cross to bare metal with GCC works
- Flags are deduplicated
- Darwin bootstrapping for 8 and 0 closer.
- Flags are same across versions.
|
| |
| |
| |
| |
| | |
Lumping it in with the target platform libraries was incorrect, and
caused eval failures when gcc couldn't be built for the target platform.
|
| |
| |
| |
| | |
We only want to refer to GCC under these conditions.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
http://lists.llvm.org/pipermail/llvm-dev/2020-March/139729.html
Additionally cherry-picked 3 commits from `llvm-project/master`:
- llvm/llvm-project@d21664c
- llvm/llvm-project@3a0f6e6
- llvm/llvm-project@87dac7d
such that clang can automatically pick up the polly plugin from the
`llvm-polly` build.
|
|\ \ |
|
| |/ |
|
|/
|
|
|
|
|
|
| |
* starting with rc2
* make `lldb` compilable again on Darwin
* separate out manpage creation for `lldb` into a new derivation
* minor tweaks to the patching of sources,
some of which are backportable to earlier versions
|
|\
| |
| | |
Staging next
|
| |
| |
| |
| |
| | |
This reverts commit f14a4d60b74625547ba7fc0572e42e9b61c0e54f, and thus
applying 558af6fbc7c15f472658339668113f4d5c6440e0.
|
| | |
|
| | |
|
|/ |
|
|
|
|
|
|
|
|
|
|
| |
Naive concatenation of $LD_LIBRARY_PATH can result in an empty
colon-delimited segment; this tells glibc to load libraries from the
current directory, which is definitely wrong, and may be a security
vulnerability if the current directory is untrusted. (See #67234, for
example.) Fix this throughout the tree.
Signed-off-by: Anders Kaseorg <andersk@mit.edu>
|
|\
| |
| | |
darwin-stdenv: bootstrap with python3
|
| | |
|
|\ \
| | |
| | | |
clang_9: Correct ClangTargets-release.cmake
|
| | |
| | |
| | |
| | |
| | | |
Since commit 4271020 libclang-cpp is also moved to $lib. This adds the missing
substitution in ClangTargets-release.cmake.
|
|\ \ \
| |_|/
|/| | |
|
| |\ \
| | | |
| | | | |
llvm: use correct version of tablegen when cross compiling _8 and _9
|
| | |/ |
|
| | | |
|
|/ /
| |
| |
| |
| | |
Fetch from github release page,
as that's where https://releases.llvm.org/9.0.1 redirects.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
As noted in the incorporated patch, name collisions between lambda
arguments and lambda captures cause compilation failures in GCC9[1].
The issue has been fixed in a later version of CLang, unfortunately the
relevant patches weren't backported to the versions 3.9 and 4.0[2].
[1] http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#2211
[2] https://github.com/llvm-mirror/clang/commit/bb177a9307360a211ca0111b49b160698395074c#diff-d63bede6ed85300ad2aaf4eaee571456L6278
|
| | |
|
|\| |
|