On Fri, Jun 07, 2024 at 04:45:14PM +0100, Steven Price wrote: > On 06/06/2024 19:38, Catalin Marinas wrote: > > Anyway, we could do some hacking around gen_pool as a temporary solution > > (maybe as a set of patches on top of this series to be easier to revert) > > and start investigating a proper decrypted page allocator in parallel. > > We just need to find a victim that has the page allocator fresh in mind > > (Ryan or Alexandru ;)). > > Thanks for the suggestions Catalin. I had a go at implementing something > with gen_pool - the below (very lightly tested) hack seems to work. This > is on top of the current series. > > I *think* it should also be safe to drop the whole alignment part with > this custom allocator, which could actually save memory. But I haven't > quite got my head around that yet. Thanks Steven. It doesn't look too complex and it solves the memory wasting. We don't actually free the pages from gen_pool but I don't think it matters much, the memory would get reused if devices are removed and re-added. -- Catalin
