[Buildroot] [PATCH 1/1] package/rust: use cargo environment during installation

[Buildroot] [PATCH 1/1] package/rust: use cargo environment during installation

[James Hilliard]

Pass HOST_PKG_CARGO_ENV to x.py dist, as is already done for x.py build.
This ensures Cargo uses Buildroot's configured CARGO_HOME instead of
falling back to the user's default cache, and keeps the host RUSTFLAGS
consistent between the build and installation steps.

Fixes:
 - https://autobuild.buildroot.net/results/02b/02b2fc6ba7329852d7bf6ae4935c120c4de47b3e

Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
---
 package/rust/rust.mk | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/package/rust/rust.mk b/package/rust/rust.mk
index ca68de589c..7546ad39fe 100644
--- a/package/rust/rust.mk
+++ b/package/rust/rust.mk
@@ -79,7 +79,8 @@ endef
 endif
 
 define HOST_RUST_INSTALL_CMDS
-	cd $(@D); $(HOST_MAKE_ENV) $(HOST_DIR)/bin/python$(PYTHON3_VERSION_MAJOR) x.py dist
+	cd $(@D); $(HOST_MAKE_ENV) $(HOST_PKG_CARGO_ENV) \
+		$(HOST_DIR)/bin/python$(PYTHON3_VERSION_MAJOR) x.py dist
 	$(HOST_RUST_INSTALL_RUSTC)
 	$(HOST_RUST_INSTALL_LIBSTD_TARGET)
 endef
-- 
2.53.0

_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] [PATCH 1/1] package/rust: use cargo environment during installation

[Thomas Petazzoni via buildroot]

On Tue, Oct 06, 2026 at 01:59:33PM -0600, James Hilliard wrote:
> Pass HOST_PKG_CARGO_ENV to x.py dist, as is already done for x.py build.
> This ensures Cargo uses Buildroot's configured CARGO_HOME instead of
> falling back to the user's default cache, and keeps the host RUSTFLAGS
> consistent between the build and installation steps.
> 
> Fixes:
>  - https://autobuild.buildroot.net/results/02b/02b2fc6ba7329852d7bf6ae4935c120c4de47b3e
> 
> Signed-off-by: James Hilliard <james.hilliard1@gmail.com>

Thanks for the patch. Since when does this problem exist? Of course it
exists since x.py support was introduced, but did it become visible
because of a new package, or package update?

As it's a fix, we need to have sufficient information in the commit
message about which commit introduced the issue, so that our LTS
branch maintainers can understand whether it needs to be backported or
not.

Thanks a lot!

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] [PATCH 1/1] package/rust: use cargo environment during installation

[James Hilliard]

On Wed, Oct 7, 2026 at 2:09 AM Thomas Petazzoni via buildroot
<buildroot@buildroot.org> wrote:
>
> On Tue, Oct 06, 2026 at 01:59:33PM -0600, James Hilliard wrote:
> > Pass HOST_PKG_CARGO_ENV to x.py dist, as is already done for x.py build.
> > This ensures Cargo uses Buildroot's configured CARGO_HOME instead of
> > falling back to the user's default cache, and keeps the host RUSTFLAGS
> > consistent between the build and installation steps.
> >
> > Fixes:
> >  - https://autobuild.buildroot.net/results/02b/02b2fc6ba7329852d7bf6ae4935c120c4de47b3e
> >
> > Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
>
> Thanks for the patch. Since when does this problem exist? Of course it
> exists since x.py support was introduced, but did it become visible
> because of a new package, or package update?

I think it's caused by this: https://github.com/rust-lang/rust/pull/156196

>
> As it's a fix, we need to have sufficient information in the commit
> message about which commit introduced the issue, so that our LTS
> branch maintainers can understand whether it needs to be backported or

I think that would be this commit:
https://github.com/buildroot/buildroot/commit/e8c8bd9bc56a827db90c3bb2e71e4613a5a2c730

> not.
>
> Thanks a lot!
>
> Thomas
> --
> Thomas Petazzoni, co-owner and CEO, Bootlin
> Embedded Linux and Kernel engineering and training
> https://bootlin.com
> _______________________________________________
> buildroot mailing list
> buildroot@buildroot.org
> https://lists.buildroot.org/mailman/listinfo/buildroot
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] [PATCH 1/1] package/rust: use cargo environment during installation

[Thomas Petazzoni via buildroot]

Hello,

On Wed, Oct 07, 2026 at 03:45:07AM -0600, James Hilliard wrote:

> > Thanks for the patch. Since when does this problem exist? Of course it
> > exists since x.py support was introduced, but did it become visible
> > because of a new package, or package update?
> 
> I think it's caused by this: https://github.com/rust-lang/rust/pull/156196
> 
> >
> > As it's a fix, we need to have sufficient information in the commit
> > message about which commit introduced the issue, so that our LTS
> > branch maintainers can understand whether it needs to be backported or
> 
> I think that would be this commit:
> https://github.com/buildroot/buildroot/commit/e8c8bd9bc56a827db90c3bb2e71e4613a5a2c730

Do you have a reproducer for this issue?

Your original patch is at
https://patchwork.buildroot.org/project/buildroot/patch/20261006195933.2225477-1-james.hilliard1@gmail.com/.

The failed issue is at:
https://autobuild.buildroot.net/results/02b/02b2fc6ba7329852d7bf6ae4935c120c4de47b3e/build-end.log,
which ended with a timeout. I see:

Dist rustc-1.98.1-src
	finished in 17585.096 seconds

which seems extremely long. I did a build on my (fast) build server,
and it took:

Dist rustc-1.98.1-src-gpl
        finished in 119.230 seconds

but the build completed successfully.

Do you have more details about what's happening, why passing more
flags don't cause a timeout, etc.?

Thanks!

Thomas
-- 
Thomas Petazzoni, co-owner and CEO, Bootlin
Embedded Linux and Kernel engineering and training
https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

Re: [Buildroot] [PATCH 1/1] package/rust: use cargo environment during installation

[James Hilliard]

On Thu, Oct 22, 2026 at 10:13 AM Thomas Petazzoni
<thomas.petazzoni@bootlin.com> wrote:
>
> Hello,
>
> On Wed, Oct 07, 2026 at 03:45:07AM -0600, James Hilliard wrote:
>
> > > Thanks for the patch. Since when does this problem exist? Of course it
> > > exists since x.py support was introduced, but did it become visible
> > > because of a new package, or package update?
> >
> > I think it's caused by this: https://github.com/rust-lang/rust/pull/156196
> >
> > >
> > > As it's a fix, we need to have sufficient information in the commit
> > > message about which commit introduced the issue, so that our LTS
> > > branch maintainers can understand whether it needs to be backported or
> >
> > I think that would be this commit:
> > https://github.com/buildroot/buildroot/commit/e8c8bd9bc56a827db90c3bb2e71e4613a5a2c730
>
> Do you have a reproducer for this issue?
>
> Your original patch is at
> https://patchwork.buildroot.org/project/buildroot/patch/20261006195933.2225477-1-james.hilliard1@gmail.com/.
>
> The failed issue is at:
> https://autobuild.buildroot.net/results/02b/02b2fc6ba7329852d7bf6ae4935c120c4de47b3e/build-end.log,
> which ended with a timeout. I see:
>
> Dist rustc-1.98.1-src
>         finished in 17585.096 seconds
>
> which seems extremely long. I did a build on my (fast) build server,
> and it took:
>
> Dist rustc-1.98.1-src-gpl
>         finished in 119.230 seconds
>
> but the build completed successfully.
>
> Do you have more details about what's happening, why passing more
> flags don't cause a timeout, etc.?

I think it's due to multiple autobuilders holding a lock at the same
time due to using the same CARGO_HOME. This change ensures
each build gets a separate CARGO_HOME. The bug is somewhat
pre-existing in the sense that CARGO_HOME should have been
isolated to begin with. It just wasn't really triggered until the update
which seems to have started locking some resources in CARGO_HOME.

In the autobuilder log this is what's causing it to take so long AFAIU:
Blocking waiting for file lock on package cache

>
> Thanks!
>
> Thomas
> --
> Thomas Petazzoni, co-owner and CEO, Bootlin
> Embedded Linux and Kernel engineering and training
> https://bootlin.com
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot