March 16, 2026
2:12 p.m.
On Wed Mar 11 15:59:37 2026 +0000, Joerg Rueppel wrote:
Thanks for the review. I evaluated your suggestion and stumbled upon how to track potential double allocations when the same data object is reached through multiple refs (via Resolve). pobj-\>pdata would then be non NULL already, so we could skip the second alloc, but how do we know when to free it without adding ref counting to that copy, too? This seems more complicated than the lazy copy with the additional field. I don't think it would necessarily require recounting, but yeah, that sounds like it's getting more complicated than necessary.
-- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_132358