Re: [PATCH v14 05/21] fsverity: improve flushing performance of fsverity_fill_zerohash
Christoph Hellwig <[email protected]> Tue, 4 Aug 2026 19:42:23 +0200
| Newsgroups | gmane.linux.file-systems.f2fs,gmane.linux.file-systems,gmane.comp.file-systems.ext4,gmane.comp.file-systems.btrfs |
|---|---|
| Message-ID | <[email protected]> |
On Mon, Aug 03, 2026 at 10:07:55PM +0200, Andrey Albershteyn wrote: > The current version calls flush_dcache_folio(), in memcpy_to_folio(), to > flush whole folio on every digest (which is 128 for 4k) on the HIGHMEM > systems. Open code folio mapping and flushing to copy all digests at > once. This looks correct, although to me optimizing for this feels like premature optimizations not worth the ugly code unless we have numbers to justify it. If Eric wants it: Reviewed-by: Christoph Hellwig <[email protected]> > + if (folio_test_partial_kmap(folio) && > + off > PAGE_SIZE - offset_in_page(offset)) > + off = PAGE_SIZE - offset_in_page(offset); > + for (; to < (vaddr + off); to += vi->tree_params.digest_size) Style nitpick: no need for braces when comparing with simple integer arithmetics like this. > + for (off = offset; off < (offset + len); Same here.