On Wed, Mar 11, 2026 at 1:33 PM Mathieu Dubois-Briand <[email protected]> wrote: > > On Wed Mar 11, 2026 at 5:39 PM CET, Joshua Watt wrote: > > On Wed, Mar 11, 2026 at 7:55 AM Mathieu Dubois-Briand > > <[email protected]> wrote: > >> > >> On Tue Mar 10, 2026 at 7:38 PM CET, Joshua Watt via lists.openembedded.org > >> wrote: > >> > Changes the SPDX 3 output to include a "recipe" package that describe > >> > static information available at parse time (without building). This is > >> > primarily useful for gathering SPDX 3 VEX information about some or all > >> > recipes, enabling SPDX 3 to be used in place of cve_check.bbclass and > >> > vex.bbclass. > >> > > >> > Special thanks to Benjamin Robin <[email protected]> for > >> > helping work through this. > >> > > >> > V2: Fixes a bug where do_populate_sysroot was running when it should not > >> > be. Drops the patch to ignore ASSUME_PROVIDES recipes, since this is > >> > incorrect (this is already handled by bitbake in the taskgraph, and > >> > doesn't need to be manually removed). > >> > > >> > V3: Fixes a bug where meta-world-recipe-sbom was reporting a circular > >> > dependency. meta-world-recipe-sbom also no longer runs in world builds, > >> > as there's no reason to this. Finally, fixes a bug where > >> > NO_GENERIC_LICENSE files would fail to be found in do_create_spdx > >> > (because do_unpack was not run). > >> > > >> > V4: Fixes test cases. Adds SPDX_PACKAGE_INCLUDE_VEX to control if VEX > >> > information is linked to binary packages, or just recipes. Defaults to > >> > "0" to significantly reduce the size of the SPDX output. > >> > > >> > V5: Fixes dummy-sdk-packages to not generate SPDX output, since it > >> > does funny things with its arch which prevents it from rebuilding SPDX > >> > data properly, and no SPDX data is needed for it anyway > >> > > >> > V6: Fixes a bug where SPDX task would not correctly re-run when they > >> > change, which would cause errors about missing SPDX document. Also > >> > updates to the latest version of the SPDX bindings which improves > >> > performance > >> > > >> > >> Hi Joshua, > >> > >> Ok, we are almost there! > >> > >> I suspect it would work fine on master, but we have a fail on two tests > >> that were recently added by Stefano, and were not merged so far. > >> > >> As both series might still evolve or get reviews, I will probably keep > >> both in my branch, but some changes are needed if we want to merge both > >> series. > > > > This is actually semi-intentional. I renamed the "recipe-" SPDX files > > to "build-". The fix is simple, but it either needs to be applied to > > my changes or Stephanos, depending on the order. Do you have a > > preference? > > > > I do not have any particular preference, so please do as is easier for > you.
Stefano has some feedback to address, so please take my patch series first and Stefano can fix it in the next revision. Thanks > > -- > Mathieu Dubois-Briand, Bootlin > Embedded Linux and Kernel engineering > https://bootlin.com >
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#232918): https://lists.openembedded.org/g/openembedded-core/message/232918 Mute This Topic: https://lists.openembedded.org/mt/118246384/21656 Group Owner: [email protected] Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
