[Git][ghc/ghc][wip/hadrian-windows-bindist] 26 commits: JS: implement openat(AT_FDCWD...) (#23697)

Tue Aug 22 11:25:39 UTC 2023

Matthew Pickering pushed to branch wip/hadrian-windows-bindist at Glasgow Haskell Compiler / GHC

Commits:
e2b38115 by Sylvain Henry at 2023-08-17T07:54:06-04:00
JS: implement openat(AT_FDCWD...) (#23697)

Use `openSync` to implement `openat(AT_FDCWD...)`.

- - - - -
a975c663 by sheaf at 2023-08-17T07:54:47-04:00
Use unsatisfiable for missing methods w/ defaults

When a class instance has an Unsatisfiable constraint in its context
and the user has not explicitly provided an implementation of a method,
we now always provide a RHS of the form `unsatisfiable @msg`, even
if the method has a default definition available. This ensures that,
when deferring type errors, users get the appropriate error message
instead of a possible runtime loop, if class default methods were
defined recursively.

Fixes #23816

- - - - -
45ca51e5 by Ben Gamari at 2023-08-17T15:16:41-04:00
ghc-internal: Initial commit of the skeleton

- - - - -
88bbf8c5 by Ben Gamari at 2023-08-17T15:16:41-04:00
ghc-experimental: Initial commit

- - - - -
664468c0 by Ben Gamari at 2023-08-17T15:17:17-04:00
testsuite/cloneStackLib: Fix incorrect format specifiers

- - - - -
eaa835bb by Ben Gamari at 2023-08-17T15:17:17-04:00
rts/ipe: Fix const-correctness of IpeBufferListNode

Both info tables and the string table should be `const`

- - - - -
78f6f6fd by Ben Gamari at 2023-08-17T15:17:17-04:00
nonmoving: Drop dead debugging utilities

These are largely superceded by support in the ghc-utils GDB extension.

- - - - -
3f6e8f42 by Ben Gamari at 2023-08-17T15:17:17-04:00
nonmoving: Refactor management of mark thread

Here we refactor that treatment of the worker thread used by the
nonmoving GC for concurrent marking, avoiding creating a new thread with
every major GC cycle. As well, the new scheme is considerably easier to
reason about, consolidating all state in one place, accessed via a small
set of accessors with clear semantics.

- - - - -
88c32b7d by Ben Gamari at 2023-08-17T15:17:17-04:00
testsuite: Skip T23221 in nonmoving GC ways

This test is very dependent upon GC behavior.

- - - - -
381cfaed by Ben Gamari at 2023-08-17T15:17:17-04:00
ghc-heap: Don't expose stack dirty and marking fields

These are GC metadata and are not relevant to the end-user. Moreover,
they are unstable which makes ghc-heap harder to test than necessary.

- - - - -
16828ca5 by Luite Stegeman at 2023-08-21T18:42:53-04:00
bump process submodule to include macOS fix and JS support

- - - - -
b4d5f6ed by Matthew Pickering at 2023-08-21T18:43:29-04:00
ci: Add support for triggering test-primops pipelines

This commit adds 4 ways to trigger testing with test-primops.

1. Applying the ~test-primops label to a validate pipeline.
2. A manually triggered job on a validate pipeline
3. A nightly pipeline job
4. A release pipeline job

Fixes #23695

- - - - -
32c50daa by Matthew Pickering at 2023-08-21T18:43:29-04:00
Add test-primops label support

The test-primops CI job requires some additional builds in the
validation pipeline, so we make sure to enable these jobs when
test-primops label is set.

- - - - -
73ca8340 by Matthew Pickering at 2023-08-21T18:43:29-04:00
Revert "Aarch ncg: Optimize immediate use for address calculations"

This reverts commit 8f3b3b78a8cce3bd463ed175ee933c2aabffc631.

See #23793

- - - - -
5546ad9e by Matthew Pickering at 2023-08-21T18:43:29-04:00
Revert "AArch NCG: Pure refactor"

This reverts commit 00fb6e6b06598752414a0b9a92840fb6ca61338d.

See #23793

- - - - -
02dfcdc2 by Matthew Pickering at 2023-08-21T18:43:29-04:00
Revert "Aarch64 NCG: Use encoded immediates for literals."

This reverts commit 40425c5021a9d8eb5e1c1046e2d5fa0a2918f96c.

See #23793

-------------------------
Metric Increase:
    T4801
    T5321FD
    T5321Fun
-------------------------

- - - - -
7be4a272 by Matthew Pickering at 2023-08-22T08:55:20+01:00
ci: Remove manually triggered test-ci job

This doesn't work on slimmed down pipelines as the needed jobs don't
exist.

If you want to run test-primops then apply the label.

- - - - -
f1ab028d by Matthew Pickering at 2023-08-22T12:19:58+01:00
hadrian: Uniformly pass buildOptions to all builders in runBuilder

In Builder.hs, runBuilderWith mostly ignores the buildOptions in BuildInfo.

This leads to hard to diagnose bugs as any build options you pass with
runBuilderWithCmdOptions are ignored for many builders.

Solution: Uniformly pass buildOptions to the invocation of cmd.

Fixes #23845

- - - - -
8e83c5a6 by Matthew Pickering at 2023-08-22T12:19:58+01:00
Abstract windows toolchain setup

This commit splits up the windows toolchain setup logic into two
functions.

* FP_INSTALL_WINDOWS_TOOLCHAIN - deals with downloading the toolchain if
  it isn't already downloaded
* FP_SETUP_WINDOWS_TOOLCHAIN - sets the environment variables to point
  to the correct place

FP_SETUP_WINDOWS_TOOLCHAIN is abstracted from the location of the mingw
toolchain and also the eventual location where we will install the
toolchain in the installed bindist.

This is the first step towards #23608

- - - - -
ca6f1aea by Matthew Pickering at 2023-08-22T12:19:58+01:00
Generate build.mk for bindists

The config.mk.in script was relying on some variables which were
supposed to be set by build.mk but therefore never were when used to
install a bindist.

Specifically

* BUILD_PROF_LIBS to determine whether we had profiled libraries or not
* DYNAMIC_GHC_PROGRAMS to determine whether we had shared libraries or
  not

Not only were these never set but also not really accurate because you
could have shared libaries but still statically linked ghc executable.

In addition variables like GhcLibWays were just never used, so those
have been deleted from the script.

Now instead we generate a build.mk file which just directly specifies
which RtsWays we have supplied in the bindist and whether we have
DYNAMIC_GHC_PROGRAMS.

- - - - -
0c8ce334 by Matthew Pickering at 2023-08-22T12:19:58+01:00
hadrian: Add reloc-binary-dist-* targets

This adds a command line option to build a "relocatable" bindist.

The bindist is created by first creating a normal bindist and then
installing it using the `RelocatableBuild=YES` option. This creates a
bindist without any wrapper scripts pointing to the libdir.

The motivation for this feature is that we want to ship relocatable
bindists on windows and this method is more uniform than the ad-hoc
method which lead to bugs such as #23608 and #23476

The relocatable bindist can be built with the "reloc-binary-dist" target
and supports the same suffixes as the normal "binary-dist" command to
specify the compression style.

- - - - -
12c76320 by Matthew Pickering at 2023-08-22T12:19:58+01:00
packaging: Fix installation scripts on windows/RelocatableBuild case

This includes quite a lot of small fixes which fix the installation
makefile to work on windows properly. This also required fixing the
RelocatableBuild variable which seemed to have been broken for a long
while.

Sam helped me a lot writing this patch by providing a windows machine to
test the changes. Without him it would have taken ages to tweak
everything.

Co-authored-by: sheaf <sam.derbyshire at gmail.com>

- - - - -
3e6bc784 by Matthew Pickering at 2023-08-22T12:19:58+01:00
ci: Build relocatable bindist on windows

We now build the relocatable bindist target on windows, which means we
test and distribute the new method of creating a relocatable bindist.

- - - - -
559f65f4 by Matthew Pickering at 2023-08-22T12:19:58+01:00
hadrian: Add error when trying to build binary-dist target on windows

The binary dist produced by `binary-dist` target doesn't work on windows
because of the wrapper script the makefile installs. In order to not
surprise any packagers we just give an error if someone tries to build
the old binary-dist target rather than the reloc-binary-dist target.

- - - - -
0a80b057 by Matthew Pickering at 2023-08-22T12:19:58+01:00
hadrian: Remove query' logic to use tooldir

- - - - -
6bce4b89 by Matthew Pickering at 2023-08-22T12:19:58+01:00
configure: Set WindresCmd directly and removed unused variables

For some reason there was an indirection via the Windres variable before
setting WindresCmd. That indirection led to #23855.

I then also noticed that these other variables were just not used
anywhere when trying to work out what the correct condition was for this
big of the configure script.

- - - - -

30 changed files:

- .gitlab-ci.yml
- .gitlab/ci.sh
- .gitlab/generate-ci/gen_ci.hs
- .gitlab/jobs.yaml
- compiler/GHC/CmmToAsm/AArch64/CodeGen.hs
- compiler/GHC/CmmToAsm/AArch64/Instr.hs
- compiler/GHC/CmmToAsm/AArch64/Ppr.hs
- compiler/GHC/CmmToAsm/AArch64/Regs.hs
- compiler/GHC/Tc/Errors.hs
- compiler/GHC/Tc/TyCl/Instance.hs
- configure.ac
- distrib/configure.ac.in
- docs/users_guide/9.10.1-notes.rst
- hadrian/README.md
- hadrian/bindist/Makefile
- hadrian/bindist/config.mk.in
- hadrian/src/Builder.hs
- hadrian/src/Hadrian/Builder/Ar.hs
- hadrian/src/Packages.hs
- hadrian/src/Rules/BinaryDist.hs
- hadrian/src/Rules/Generate.hs
- hadrian/src/Rules/Lint.hs
- hadrian/src/Settings/Default.hs
- libraries/base/base.cabal
- libraries/base/jsbits/base.js
- libraries/base/jsbits/errno.js
- libraries/base/tests/Makefile
- libraries/base/tests/System/all.T
- + libraries/base/tests/T23697.hsc
- + libraries/base/tests/T23697.stdout

The diff was not included because it is too large.

View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b78e53b4ee9312255b2b737ee4142561fd3b721b...6bce4b890dcbd965d7301da562cfb1d4ffdbab5b

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/b78e53b4ee9312255b2b737ee4142561fd3b721b...6bce4b890dcbd965d7301da562cfb1d4ffdbab5b
You're receiving this email because of your account on gitlab.haskell.org.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20230822/4c96573f/attachment-0001.html>