On Fri, Oct 09, 2026 at 11:25:32AM +0200, Antoine Lassagne wrote: > Hello Adrian and thank you for the feedback,
Hi Antoine, > I'm sorry. I did realize it was a little weird to close this bug while not > entirely solving it. You mentioned testing arm64 in the bug, and since I > have an arm64 device, I took the "safe" approach. Matteo also suggested me > this approach. as I wrote I only tested that it built on arm64 to confirm that OpenVDB support does not seem to be amd64-only. No runtime testing was done by me. IMHO the safe approach is to default to minimixing architecture differences where reasonably possible. This bug was strictly about that, since such architecture differences are sometimes affecting users and often result in build issues only affecting some architectures. > If you guys all agree that it should be enabled for all architectures, I'm > happy to upload these changes. In any case the bug is still open, since OpenVDB support is currently still disabled even on arm64 - as is common for this kind of architecture delta the architectures are hardcoded in both rules and control. > Antoine cu Adrian > On Fri, 9 Oct 2026 at 10:46, Adrian Bunk <[email protected]> wrote: > > > Control: reopen -1 > > > > Hi Antoine, > > > > I am wondering why you changed only arm64, the main point of my bug > > report was to reduce the delta between architectures. > > > > Case in point, your change did not actually enable OpenVDB support on > > arm64 due to this delta between architectures being in two places: > > > > > > https://buildd.debian.org/status/fetch.php?pkg=blender&arch=arm64&ver=5.0.1%2Bdfsg-7&stamp=1791477118&raw=0 > > ... > > -- Could NOT find OpenVDB (missing: OPENVDB_LIBRARY OPENVDB_INCLUDE_DIR) > > -- OpenVDB not found, disabling WITH_OPENVDB > > ... > > > > cu > > Adrian > >
