Rust follows the train release model via the stable, beta and nightly channels,
by default we build rust from the stable channel, however there are certain
features which are only available in the beta or nightly channels.
Make these channels available by setting a RUST_CHANNEL variable which defaults
to stable making this change transparent to the user.
The snapshot version used by rust during its compilation wont necessarily match
the version being built, specially if were building from an unstable channel,
to avoid confusion rename this to SNAPSHOT_VERSION and use RUST_VERSION for the
version to be built, which is automatically defined to PV.
Append -beta or -nightly to rusts PV for signature awareness.
It is important to note that this does not build rust from the beta/nightly
published tarball (which today build rust v1.67.0 and v1.68.0 respectively),
instead this builds rust from the current selected version (1.66.0) and enables
the beta/nightly features for that version.
Setting the variable RUST_CHANNEL=nightly results in the following:
$ rustc -Vv
rustc 1.66.0-nightly
(From OE-Core rev: 807a52686682d0d0a151ea3dadd99880feb67cc0)
Signed-off-by: Alejandro Hernandez Samaniego <alejandro@enedino.org>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As part of the merging builds, installation of target.json was lost
causing `rustc` to fail on startup with:
error: Error loading target specification: Could not find specification for target "aarch64-poky-linux-gnu". Run `rustc --print target-list` for a list of built-in targets
We know the full target specification, so just install it directly
rather than using the glob approach from previously.
Fixes: b9b0cd99cdc7 ("rust: Enable nativesdk and target builds + replace rust-tools-cross-canadian")
(From OE-Core rev: b187185ceecf1c852b1fdc47451d157a41962d69)
Signed-off-by: Alex Kiernan <alex.kiernan@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meta-clang has options when it comes to C++ runtime, default is to use
gnu runtime, other options are llvm runtime and android runtime. This
patch helps when a distro is using llvm runtime for C/C++ runtime. It
informs the rust build system about right C++ runtime to configure for
when such a setting is used.
(From OE-Core rev: 521872ab2cac092c88446730772cb8c4e22b0cdc)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
All architectures may not support same set of tools, therefore use a
variable to specify this. E.g. on riscv32 rustfmt is not buildable
right now.
(From OE-Core rev: 332df9ebbd75a1825ac99abf1311d2e692d398d7)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is a fix for the fix in YOCTO #14878. When the shebang is more than
128 characters the default shell /bin/sh is used instead of SDK shell as
a fallback, which causes problems with LD_LIBRARY_PATH. With this patch
shell usage is avoided as we use a C wrapper and unset LD_LIBRARY_PATH
that way.
[YOCTO #14892]
(From OE-Core rev: 7cd6faf4e0147eef557f83fb266a25935e26efff)
Signed-off-by: Sundeep KOKKONDA <sundeep.kokkonda@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Emits complete details regarding the test results to understand
and debug from logfiles
(From OE-Core rev: 78aecd6d4e443cb467e9a8007f2db29943dfdd1f)
Signed-off-by: pgowda <pgowda.cve@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Adds the cross compilation linker for build, host and target.
Ensure the main rust config has the appropriate sections added
to match the configurations.
(From OE-Core rev: 02133f92c56c90a5365985b515953e4136e112aa)
Signed-off-by: pgowda <pgowda.cve@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
crossbeam-utils tries to use the triplet to look up whether the target
supports various forms of atomics. We use TARGET_VENDOR and not "-unknown"
in the target case which means this fails and breaks platforms like mips
and powerpc 32 bit. Add a patch to handle TARGET_VENDOR in this case.
(From OE-Core rev: 5d8c01dfed4c82fcc5d504c728a51fb98a262300)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Latest rust has started using ENOTSUP define, which is not available in
the older libc that current release of compiler is using therefore
backport the needed patch. Eventually when vendored version of libc
bumps to 1.33+ we should not need this patch.
(From OE-Core rev: 097e38f28990229c3f54f4e6cb74ca8be0b806e6)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Make unique shell variable names for cargo and rust
setup scripts. This change will make it easier to append to the
scripts in a bbappend file by using the variable for the script.
Before this change it was only possible for the last script as
they shared the same variable name.
(From OE-Core rev: 74307f57ad0b3c538be0add11028e4e6199a7662)
Signed-off-by: Peter Bergin <peter@berginkonsult.se>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These match the default from the class so drop them. We then always
generate all targets so remove the configuration from the class.
(From OE-Core rev: e4d56256936c55bab2bf2934ccbde9157ef7dc57)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
rust sets LD_LIBRARY_PATH which causes a lot of problems with host
binaries conflicting with SDK libraries. By using sh/bash from the
SDK we avoid those issues.
[YOCTO #14878]
(From OE-Core rev: dabd97e78a599a81c056c0f3291244f5afae44b0)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Update the README to match the current status and remove unclear/obsolete
information and fix formatting.
(From OE-Core rev: fa5d16692dda655a54653df7b4bf98410090ec5c)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To ensure buildpaths aren't encoded in the debug symbol output, ensure
we pass our prefix remapping compiler options in.
(From OE-Core rev: c7a529743560c37973787e4ab2433cb7485afb86)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Musl targets need a sysroot set to find the musl static libraries. Set this
appropriately in musl builds.
(From OE-Core rev: 1c5ab57ddda393631ce2da97ef5a876a2020d51f)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Enable rust target builds as well as nativesdk-rust for the sdk. Merge
the builds of rust-tools components into the rust build, packaged separately
since this is a lot more efficient and saves rebuilding core rust multiple
times. The tools are not target specific so nativesdk-rust-tools suffices
and we can drop the cross canadian piece.
(From OE-Core rev: b9b0cd99cdc77e7a90e5fd5711e706ebe64c7b6b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Now that target config json is provided by rust-target-config.bbclass,
the need for the cross and crosssdk recipes is removed. Drop them and
simplify dependencies accordingly.
(From OE-Core rev: 4b54f5f52b33db4d2fe95c5faef033b6c6b37b7d)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Since a copy of rust can target any target and doesn't need to be target
specific, we can simplify rust-cross-canadian to simply contain the json
configuration and envirionment setup scripts.
(From OE-Core rev: 6ec18de9d6f49bccff72f2665821dbd732b69b48)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The cargo-cross-candian variant made no sense as one version of cargo in
the SDK can work for all targets. Replace it with nativesdk-cargo instead.
Move the SDK env to rust-cross-canadian.
(From OE-Core rev: 6d6d135924eff5993736ee58ba8cc5d00ca635f3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
For cross compilation, build, host and target may be different. Ensure
the main rust config has the appropriate sections added to match the
configurations.
(From OE-Core rev: e2cdbf9df8d3f90b5a7ba35288c5db4f252d134d)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
To allow nativesdk variants of the tools to build, enable a nativesdk
variant of rust-llvm.
(From OE-Core rev: b4b24dc53fdb86624da71b854dfe018923a203fe)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Improve the common class to handle BUILD_SYS as well as host and target,
removing the need to a workaround in the rust recipe.
(From OE-Core rev: 0dc9635c28c269f782ad4ed6ea06b4e4de1beb8e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Move the "unknown" vendor workaround used during bootstrap to a
central location so it is applied consistently to all RUST_BUILD_SYS
values rather than some subset.
(From OE-Core rev: d02c28b2219d736c9598a13fead7a03eaa3256a6)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of generating target configuration files centrally and often getting
it wrong, or having trouble finding the right set, generate them dynamically
from the bbclass into WORKDIR per recipe.
(From OE-Core rev: 9160e4a37561d8ac882057450a818621bec13bed)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This variable is no longer used by the llvm patches.
(From OE-Core rev: 35e1ecf88cf9725e1ba43c948e90359df983002f)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We need llvm-config for the target but the cross compiled target binary is
no good. We can copy the native one into the target location where it will
then return the target values though.
(From OE-Core rev: f355e55c1fcef57f288279ffb47ebeb52a72b689)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of reusing a "target_section" variable, name them host and build
section instead. This patch sets things up for other improvementsm,
separating out the renaming.
(From OE-Core rev: 279b809e8306b0eb6ae6caeb6b0a3c3c62d2f9d2)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The code was using a mixture of XXX_SYS and RUST_XXX_SYS. Use
RUST_XXX_SYS consistently and add the variables to the global exclsion
on signatures as they're reflected in the directory triplets and trying
to filter them out the hashes separately is too painful.
(From OE-Core rev: ee0c0fdf9c1eba9eece6ed1293fda25bf18964b3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a patch which removes the sections of code which encode buildpaths. Whilst
not ideal, the patches at least show where the problematic data is coming from
and should allow more focused work to resovle it by someone who has a better
understanding of rust and what this code is doing. It does look unlikely we
actually need this code in our usecases anyway.
(From OE-Core rev: c08c522fc29445aef0c64f0dd8df8a3531c04afa)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Remove further code duplication by allowing the main function to be
configurable.
(From OE-Core rev: 2d4b35a0db1220e7a7f1b6776223aa4d40a3f246)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently most of the rust recipes use this code but it is all piecemeal.
Turn the code into a class where things can start to be rationalised. Ultimately
some of the data and python code should be moved to a python library but one step
at a time.
No functionality changes.
(From OE-Core rev: 3795285cbf362e13b8151bfdbe1bce999ac28641)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some of the subtleties in the different codepaths for target rust json generation
were not easy to spot. Start to simplfy the code to make this clearer.
This patch should not have any functionality change although ABIEXTENSION
has to be excluded from the function signature, the triplet would normally
cover anything set there.
(From OE-Core rev: 0f83d959465e0d99f98ade6803281585931d1b02)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Match the code in rust-cross-canadian so that further simplifications
can be considered in future.
(From OE-Core rev: 5cb62d0a33777cb9afb1eea5f736a2580ce50dc7)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
rust-cross had special handling for armv7 targets but we also need this
for cross-canadian. Merge the code into the main function so everything is
consistent.
Also then fix the arm definition to be arm-eabi since ABI is correctly
being looked up.
(From OE-Core rev: ff3c3dbbd2bf1bb7bb70b55cca203e9eedcf14a8)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Based upon a patch from Otavio Salvador <otavio@ossystems.com.br>,
ensure the target json files are written in the correct order with
the most specific last incase it overwrites earlier files if the prefixes
match.
(From OE-Core rev: 1912c4e9e0ecf9655f3b3a41588b54d7956f5899)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When a 'BUILD' target is requested we shouldn't be looking at TARGET_SYS but
at BUILD_SYS. Due to the way rust mangles triplets, we need the HOST_SYS triplet
to work with existing code - fixing that issue is a separate patch.
Also drop the arch_abi argument, it doens't make any sense to a getVar() call
and was a copy and paste error.
Based on a patch from Otavio Salvador <otavio@ossystems.com.br> but separated out
and tweaked.
Fixes: bd36593ba3 ("rust-common: Drop LLVM_TARGET and simplify")
(From OE-Core rev: d554161a045d12411f288394e253c54aa4c1257c)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop the two libstdc patches as they've finally appeared upstream.
Disable the use of libstdc++.a from the host distributions, as it results
in cross-distro contamination in rust-native.
(From OE-Core rev: 94760bc118952160865352c10ca7693680b5ce7e)
Signed-off-by: Alexander Kanavin <alex@linutronix.de>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Commit bd36593ba3db758b3eacc974e48468a665967961 did introduce a
regression when building package rust-cross-canadian-aarch64
on a x86_64 host. This commit will fix that configuration.
Suggested-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(From OE-Core rev: ef566af964e9f9d2c440a3b5771ed801216f30f9)
Signed-off-by: Peter Bergin <peter@berginkonsult.se>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The previous changes cause sstatetests.SStateTests.test_sstate_32_64_same_hash
to fail since RUST_XXX_SYS changes depending on the native architecture. This
is correct but these are accounted for in the layout of paths in TMPDIR so
they should be excluded from the task signatures for the correct behaviour.
(From OE-Core rev: 2b9bb4c07d0ab53f58cbdd8a96896780a90be9a8)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This all seems over complicated for something which is basically always
one of two values. This might even help cross-canadian work on something
which isn't x86-64.
(From OE-Core rev: bd36593ba3db758b3eacc974e48468a665967961)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
