From: Christophe Leroy > Sent: 01 April 2023 19:45 > > Le 01/04/2023 à 20:33, David Laight a écrit : > > From: Vlastimil Babka > >> Sent: 32 March 2023 10:47 > >> > >> As the SLOB removal is on track and the SLAB removal is planned, I have > >> realized - why should we stop there and not remove also SLUB? What's a > >> slab allocator good for in 2023? The RAM sizes are getting larger and > >> the modules cheaper [1]. The object constructor trick was perhaps > >> interesting in 1994, but not with contemporary CPUs. So all the slab > >> allocator does today is just adding an unnecessary layer of complexity > >> over the page allocator. > > > > Why stop there? > > Remove kmalloc() completely. > > With cheap memory isn't unreasonable to go back to compile-time > > settable fixed size arrays for all items. > > Should make 'use after free' much easier to track down. > > > > While at it, why not also increase the page size to hugepage size, that > would allow us to remove all the complex THP logic, and would remove one > level of page tables. I would also reduce TLB pressure. Good idea, 4k pages were used on systems where the total system memory was a few MB. Memory system are now easily 1000x times larger - so why not 4MB pages .... David - Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
