On Tue, May 20, 2025 at 12:29:48PM +0200, Fiona Ebner wrote: > Previous discussion [0]. >
> NOTE: > there are pre-existing test failures on current master, e.g. '240' for > all formats, '295 296 inactive-node-nbd luks-detached-header' for luks > and 'mirror-sparse' for raw. For me, the failures do not change after > this series. I have not yet reproduced the mirror-sparse failure locally; I suspect the culprit may be a difference in filesystems and how well (or not) SEEK_HOLE and FALLOC_FL_PUNCH_HOLE work on different filesystems. Can I get more details on the failure as you see it, so I can work on skipping the test when the filesystem does not support the test at hand? (It might be better to reply in context to the pull request that introduced the file [2]) [2] https://lists.gnu.org/archive/html/qemu-devel/2025-05/msg03623.html -- Eric Blake, Principal Software Engineer Red Hat, Inc. Virtualization: qemu.org | libguestfs.org
