Message ID | 20241104172721.180255-1-pbonzini@redhat.com (mailing list archive) |
---|---|
State | New |
Headers | show |
On Mon, 4 Nov 2024 at 17:35, Paolo Bonzini <pbonzini@redhat.com> wrote: > > The following changes since commit 15195de6a93438be99fdf9a90992c4228527130d: > > ci: enable rust in the Fedora system build job (2024-10-30 16:30:56 +0100) > > are available in the Git repository at: > > https://gitlab.com/bonzini/qemu.git tags/for-upstream-rust > > for you to fetch changes up to d20feaa9a5af597bd20630d041e5dc7808612be1: > > ci: enable rust in the Debian and Ubuntu system build job (2024-10-31 18:39:52 +0100) > > ---------------------------------------------------------------- > * rust: cleanups > * rust: integration tests > * rust/pl011: add support for migration > * rust/pl011: add TYPE_PL011_LUMINARY device > * rust: add support for older compilers and bindgen > * rust: enable rust in the Debian, Fedora and Ubuntu system build job > > ---------------------------------------------------------------- This probably isn't something worth not merging this for, but I noticed while testing (via vm-build-openbsd) that Meson complains: Compiler for language rust for the host machine not found. Program bindgen skipped: feature rust disabled ../meson.build:111: WARNING: bindgen not found, disabling Rust compilation. Message: To use Rust you can install a new version with "cargo install bindgen-cli" Rust is still disabled-by-default, so why is meson probing for bindgen? It would be nice if we could avoid printing WARNING messages for the normal case. (I'm continuing with the CI test run.) thanks -- PMM
On Wed, Nov 6, 2024 at 2:10 PM Peter Maydell <peter.maydell@linaro.org> wrote: > > On Mon, 4 Nov 2024 at 17:35, Paolo Bonzini <pbonzini@redhat.com> wrote: > > > > The following changes since commit 15195de6a93438be99fdf9a90992c4228527130d: > > > > ci: enable rust in the Fedora system build job (2024-10-30 16:30:56 +0100) > > > > are available in the Git repository at: > > > > https://gitlab.com/bonzini/qemu.git tags/for-upstream-rust > > > > for you to fetch changes up to d20feaa9a5af597bd20630d041e5dc7808612be1: > > > > ci: enable rust in the Debian and Ubuntu system build job (2024-10-31 18:39:52 +0100) > > > > ---------------------------------------------------------------- > > * rust: cleanups > > * rust: integration tests > > * rust/pl011: add support for migration > > * rust/pl011: add TYPE_PL011_LUMINARY device > > * rust: add support for older compilers and bindgen > > * rust: enable rust in the Debian, Fedora and Ubuntu system build job > > > > ---------------------------------------------------------------- > > This probably isn't something worth not merging this for, but I > noticed while testing (via vm-build-openbsd) that Meson complains: > > Compiler for language rust for the host machine not found. > Program bindgen skipped: feature rust disabled > ../meson.build:111: WARNING: bindgen not found, disabling Rust compilation. > Message: To use Rust you can install a new version with "cargo install > bindgen-cli" > > Rust is still disabled-by-default, so why is meson probing for bindgen? It's not probing it ("Program bindgen skipped"), but I was a bit too happy about printing warnings. This line: if not bindgen.found() or bindgen.version().version_compare('<0.60.0') should simply have had an "if not have_rust", or something like that. If you want I can resend. I know that Linaro people are in Dublin, so whatever is easiest for you. Paolo
On Wed, 6 Nov 2024 at 13:14, Paolo Bonzini <pbonzini@redhat.com> wrote: > > On Wed, Nov 6, 2024 at 2:10 PM Peter Maydell <peter.maydell@linaro.org> wrote: > > > > On Mon, 4 Nov 2024 at 17:35, Paolo Bonzini <pbonzini@redhat.com> wrote: > > > > > > The following changes since commit 15195de6a93438be99fdf9a90992c4228527130d: > > > > > > ci: enable rust in the Fedora system build job (2024-10-30 16:30:56 +0100) > > > > > > are available in the Git repository at: > > > > > > https://gitlab.com/bonzini/qemu.git tags/for-upstream-rust > > > > > > for you to fetch changes up to d20feaa9a5af597bd20630d041e5dc7808612be1: > > > > > > ci: enable rust in the Debian and Ubuntu system build job (2024-10-31 18:39:52 +0100) > > > > > > ---------------------------------------------------------------- > > > * rust: cleanups > > > * rust: integration tests > > > * rust/pl011: add support for migration > > > * rust/pl011: add TYPE_PL011_LUMINARY device > > > * rust: add support for older compilers and bindgen > > > * rust: enable rust in the Debian, Fedora and Ubuntu system build job > > > > > > ---------------------------------------------------------------- > > > > This probably isn't something worth not merging this for, but I > > noticed while testing (via vm-build-openbsd) that Meson complains: > > > > Compiler for language rust for the host machine not found. > > Program bindgen skipped: feature rust disabled > > ../meson.build:111: WARNING: bindgen not found, disabling Rust compilation. > > Message: To use Rust you can install a new version with "cargo install > > bindgen-cli" > > > > Rust is still disabled-by-default, so why is meson probing for bindgen? > > It's not probing it ("Program bindgen skipped"), but I was a bit too > happy about printing warnings. This line: > > if not bindgen.found() or bindgen.version().version_compare('<0.60.0') > > should simply have had an "if not have_rust", or something like that. > > If you want I can resend. I know that Linaro people are in Dublin, so > whatever is easiest for you. (I'm not in Dublin, as it happens.) I don't think this needs to be fixed in this pullreq; it's fine to send a patch to fix this cosmetic issue and we'll apply it sometime during freeze. However, this does seem to be causing the functional-tests to timeout on the CI job that now enables Rust: https://gitlab.com/qemu-project/qemu/-/jobs/8284623145 https://gitlab.com/qemu-project/qemu/-/jobs/8283637798 both fail with 01-tests/avocado/boot_linux_console.py:BootLinuxConsole.test_arm_virt: INTERRUPTED 13-tests/avocado/replay_kernel.py:ReplayKernelNormal.test_arm_virt: INTERRUPTED 15-tests/avocado/replay_kernel.py:ReplayKernelNormal.test_arm_vexpressa9: INTERRUPTED (which I suspect will turn out to be "all the functional tests that use a pl011"). Could you have a look, please? thanks -- PMM
On Wed, Nov 6, 2024 at 4:00 PM Peter Maydell <peter.maydell@linaro.org> wrote: > However, this does seem to be causing the functional-tests to timeout > on the CI job that now enables Rust: > > https://gitlab.com/qemu-project/qemu/-/jobs/8284623145 > https://gitlab.com/qemu-project/qemu/-/jobs/8283637798 > > both fail with > > 01-tests/avocado/boot_linux_console.py:BootLinuxConsole.test_arm_virt: > INTERRUPTED > 13-tests/avocado/replay_kernel.py:ReplayKernelNormal.test_arm_virt: INTERRUPTED > 15-tests/avocado/replay_kernel.py:ReplayKernelNormal.test_arm_vexpressa9: > INTERRUPTED > > (which I suspect will turn out to be "all the functional tests that > use a pl011"). > > Could you have a look, please? Drat, I'll just drop the patch that enables Rust in CI. Not really in the mood of debugging device code. Paolo