On 27.01.25 17:56, Matthew Wilcox wrote:

On Mon, Jan 27, 2025 at 05:20:25PM +0100, David Hildenbrand wrote:

quoted

quoted

quoted

       14.04%  postgres         [kernel.kallsyms]          [k] try_grab_folio_fast
               |
                --14.04%--try_grab_folio_fast
                          gup_fast_fallback
                          |
                           --13.85%--iov_iter_extract_pages
                                     bio_iov_iter_get_pages
                                     iomap_dio_bio_iter
                                     __iomap_dio_rw
                                     iomap_dio_rw
                                     xfs_file_dio_read
                                     xfs_file_read_iter
                                     __io_read
                                     io_read
                                     io_issue_sqe
                                     io_submit_sqes
                                     __do_sys_io_uring_enter
                                     do_syscall_64

BTW, two things that come to mind:


(1) We always fallback to GUP-slow, I wonder why. GUP-fast would go via
try_grab_folio_fast().

I don't think we do?

iov_iter_extract_pages() calls
iov_iter_extract_user_pages() calls
pin_user_pages_fast() calls
gup_fast_fallback() calls
gup_fast() calls
gup_fast_pgd_range() calls
gup_fast_p4d_range() calls
gup_fast_pud_range() calls
gup_fast_pud_leaf() calls
try_grab_folio_fast() which is where we see the contention.

If that were to fail, we'd see contention in __get_user_pages_locked(),
right?

Sorry, my brain stripped the "_fast" in the callchain above :( -- it 
*is* try_grab_folio_fast() :)

So yes, we are in GUP-fast as one would expect.

-- 
Cheers,

David / dhildenb