On Thu Sep 24, 2026 at 7:21 PM CEST, Danilo Krummrich wrote: > So, I think we should change dma_len() to just return u32 (I'm also fine with > a > new type, but it might be slighly overkill). > > I'd usually pick this patch regardless and leave that for a follow-up, as it > is > a separate issue. But if we change dma_len() first, we only have to update tyr > once, whereas if change it after we end up touching nova and tyr twice.
Makes sense, thanks. I sent v4 with a patch in front that makes dma_len() return u32, so the newtype patch no longer touches tyr or nova-core: https://lore.kernel.org/r/[email protected] Heads-up: patch 1 interacts with Alexandre's nova_num on one line in gsp.rs. Whichever lands second needs a two-line fixup; details in the cover letter.
