Messages in this thread | | | Subject | Re: [PATCH v4 5/6] iova: Add iova_len argument to init_iova_domain() | From | John Garry <> | Date | Mon, 2 Aug 2021 18:18:45 +0100 |
| |
On 02/08/2021 17:40, John Garry wrote: > On 02/08/2021 17:16, Robin Murphy wrote: >> On 2021-08-02 17:06, John Garry wrote: >>> On 02/08/2021 16:06, Will Deacon wrote: >>>> On Wed, Jul 14, 2021 at 06:36:42PM +0800, John Garry wrote: >>>>> Add max opt argument to init_iova_domain(), and use it to set the >>>>> rcaches >>>>> range. >>>>> >>>>> Also fix up all users to set this value (at 0, meaning use default). >>>> Wrap that in init_iova_domain_defaults() to avoid the mysterious 0? >>> >>> Sure, I can do something like that. I actually did have separate >>> along those lines in v3 before I decided to change it. >> >> Y'know, at this point I'm now starting to seriously wonder whether >> moving the rcaches into iommu_dma_cookie wouldn't make a whole lot of >> things simpler... :/ > > As I see, the rcache stuff isn't really specific to IOVA anyway, so it > seems sane. > >> >> Does that sound like crazy talk to you, or an idea worth entertaining? > > If you're going to start moving things, has anyone considered putting > rcache support in lib as a generic solution to "Magazines and Vmem: .." > paper?
Having said that, I still think that the rcache code has certain scalability issues, as discussed before. So making more generic and then discarding would be less than ideal.
Thanks, john
| |