| Message ID | 20210430145549.2662354-1-willy@infradead.org (mailing list archive) |
|---|---|
| Headers | show |
| Series | Folio Prequel patches | expand |
On Fri, 30 Apr 2021 15:55:41 +0100 "Matthew Wilcox (Oracle)" <willy@infradead.org> wrote: > These patches have all been posted before and not picked up yet. I've > built the folio patches on top of them; while they are not necessarily > prerequisites in the conceptual sense, I'm not convinced that the > folio patches will build without them. The nth_page patch is purely > an efficiency question, while patch 5 ("Make compound_head > const-preserving") is required for the current implementation of > page_folio(). Patch 8 ("Fix struct page layout on 32-bit systems") > is required for the struct folio layout to match struct page layout > on said 32-bit systems (arm, mips, ppc). > > They are on top of next-20210430 > I'm running them since a couple days on an arm64 machine which uses the page_pool API. No issues so far. Regards,
On Mon, May 10, 2021 at 06:12:45PM +0200, Matteo Croce wrote: > On Fri, 30 Apr 2021 15:55:41 +0100 > "Matthew Wilcox (Oracle)" <willy@infradead.org> wrote: > > > These patches have all been posted before and not picked up yet. I've > > built the folio patches on top of them; while they are not necessarily > > prerequisites in the conceptual sense, I'm not convinced that the > > folio patches will build without them. The nth_page patch is purely > > an efficiency question, while patch 5 ("Make compound_head > > const-preserving") is required for the current implementation of > > page_folio(). Patch 8 ("Fix struct page layout on 32-bit systems") > > is required for the struct folio layout to match struct page layout > > on said 32-bit systems (arm, mips, ppc). > > > > They are on top of next-20210430 > > > > I'm running them since a couple days on an arm64 machine which uses the > page_pool API. No issues so far. Thanks! Andrew picked up the first seven of these patches last night, and I resent the eighth one today: https://lore.kernel.org/linux-mm/20210510153211.1504886-1-willy@infradead.org/
These patches have all been posted before and not picked up yet. I've built the folio patches on top of them; while they are not necessarily prerequisites in the conceptual sense, I'm not convinced that the folio patches will build without them. The nth_page patch is purely an efficiency question, while patch 5 ("Make compound_head const-preserving") is required for the current implementation of page_folio(). Patch 8 ("Fix struct page layout on 32-bit systems") is required for the struct folio layout to match struct page layout on said 32-bit systems (arm, mips, ppc). They are on top of next-20210430 Matthew Wilcox (Oracle) (8): mm: Optimise nth_page for contiguous memmap mm: Make __dump_page static mm/debug: Factor PagePoisoned out of __dump_page mm/page_owner: Constify dump_page_owner mm: Make compound_head const-preserving mm: Constify get_pfnblock_flags_mask and get_pfnblock_migratetype mm: Constify page_count and page_ref_count mm: Fix struct page layout on 32-bit systems include/linux/mm.h | 4 ++++ include/linux/mm_types.h | 4 ++-- include/linux/mmdebug.h | 3 +-- include/linux/page-flags.h | 10 +++++----- include/linux/page_owner.h | 6 +++--- include/linux/page_ref.h | 4 ++-- include/linux/pageblock-flags.h | 2 +- include/net/page_pool.h | 12 +++++++++++- mm/debug.c | 25 +++++++------------------ mm/page_alloc.c | 16 ++++++++-------- mm/page_owner.c | 2 +- net/core/page_pool.c | 12 +++++++----- 12 files changed, 52 insertions(+), 48 deletions(-)