Re: [RFC PATCH v4 00/16] Reimplement huge pages without hugepd on powerpc (8xx, e500, book3s/64)

[Oscar Salvador <osalvador@suse.com>]

On Mon, May 27, 2024 at 03:29:58PM +0200, Christophe Leroy wrote:
> This is the continuation of the RFC v1 series "Reimplement huge pages
> without hugepd on powerpc 8xx". It now get rid of hugepd completely
> after handling also e500 and book3s/64
> 
> Also see https://github.com/linuxppc/issues/issues/483
> 
> Unlike most architectures, powerpc 8xx HW requires a two-level
> pagetable topology for all page sizes. So a leaf PMD-contig approach
> is not feasible as such.
> 
> Possible sizes on 8xx are 4k, 16k, 512k and 8M.
> 
> First level (PGD/PMD) covers 4M per entry. For 8M pages, two PMD entries
> must point to a single entry level-2 page table. Until now that was
> done using hugepd. This series changes it to use standard page tables
> where the entry is replicated 1024 times on each of the two pagetables
> refered by the two associated PMD entries for that 8M page.
> 
> For e500 and book3s/64 there are less constraints because it is not
> tied to the HW assisted tablewalk like on 8xx, so it is easier to use
> leaf PMDs (and PUDs).
> 
> On e500 the supported page sizes are 4M, 16M, 64M, 256M and 1G. All at
> PMD level on e500/32 (mpc85xx) and mix of PMD and PUD for e500/64. We
> encode page size with 4 available bits in PTE entries. On e300/32 PGD
> entries size is increases to 64 bits in order to allow leaf-PMD entries
> because PTE are 64 bits on e500.
> 
> On book3s/64 only the hash-4k mode is concerned. It supports 16M pages
> as cont-PMD and 16G pages as cont-PUD. In other modes (radix-4k, radix-6k
> and hash-64k) the sizes match with PMD and PUD sizes so that's just leaf
> entries. The hash processing make things a bit more complex. To ease
> things, __hash_page_huge() is modified to bail out when DIRTY or ACCESSED
> bits are missing, leaving it to mm core to fix it.

Ok, I managed to go through the series and provide some feedback.
Sorry you had to bear some dumb questions but I am used to x86 realm where
things are farily easier wrt. hugepage sizes.

I will over v5 when you send it, but I think this would benefit from another
pair of eyes (with more powerpc knowledge than me) having a look.

Anyway, I think this is a great step in the right direction, and definitely
a big help for the upcoming tasks.

I plan to start working on the walk_page API to get rid of hugetlb
specific hooks basing it on this patchset.

Thanks a lot for this work Christophe


-- 
Oscar Salvador
SUSE Labs