From nobody Thu Feb 12 20:28:53 2026 Received: from mout-p-101.mailbox.org (mout-p-101.mailbox.org [80.241.56.151]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 15CA419B3D3; Fri, 7 Jun 2024 14:59:29 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=80.241.56.151 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1717772371; cv=none; b=gpIWsCPO19bwDESepigZ3S1TbjhIe3rNMh2x4rfyrVLMiWJbWCDvvs1GilGIjKoZdsNZfl8Ihwk9FNla3f7B49eVi9g/Q/vGqr1PZvifvlJKQl4J59MbdgGdg/+y7PT9PLF/630WnZxgUe9N3Hui2gxT1l5jaGqOKAXgccHPLOQ= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1717772371; c=relaxed/simple; bh=NEU1YKn8XdRoPS5041H7/m6H9yP7woEBYkHVucW3IbI=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=SZJQhk4P+QJR2e1AGRG1TnvhfTNUNwNajqGDY42q5bg9FwrMV8PFHPnjdmpQwB86PhB7/mRphievfrso4YtT2JhPU7olgopj3y1iPiZ4O6m+4G5IXeW3umL32UyviNK8YupXs4bHOnH1AWnbA0hJU0RGQuhBz3nZKk668yAevKA= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=pankajraghav.com; spf=pass smtp.mailfrom=pankajraghav.com; dkim=pass (2048-bit key) header.d=pankajraghav.com header.i=@pankajraghav.com header.b=HNg2Zlay; arc=none smtp.client-ip=80.241.56.151 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=pankajraghav.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=pankajraghav.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=pankajraghav.com header.i=@pankajraghav.com header.b="HNg2Zlay" Received: from smtp102.mailbox.org (smtp102.mailbox.org [10.196.197.102]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by mout-p-101.mailbox.org (Postfix) with ESMTPS id 4Vwkqy3GMxz9smQ; Fri, 7 Jun 2024 16:59:26 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=pankajraghav.com; s=MBO0001; t=1717772366; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=gcGImSYw6myhhr+aFoG6zYcMMyebatE1gThNoSCMqDI=; b=HNg2Zlay2Z+TJiNAAL/h42UPCFPyyjecycy9Rzn3UPRWBgUjnJgbsCSIjM4WoHM1JTXBMX hCfT1g6FDmQQYumow62hfgcTTyqWLaWigc8MLjlBgOs9U6Udn2Ufg9QXj3KCcR3HpLIQOn EzIjcnhiQRKvFbHJz8ZhAcXMXtn+kf3u7YBF9+W6zny1XIDPiilEMmKiSzCUBL2MDZ8V5O TZYFf27k5Kci3MKtZI63RCU84Fiw6g4YpkFOSs+k6O61zJw84mFhgXfNM3j2tCeWzGVWe5 gTVlByOdsdDkNx/NdTOKUzebiyJ03ma21W5BwlHvtkWZI9H2UZeReiue4sdvlg== From: "Pankaj Raghav (Samsung)" To: david@fromorbit.com, djwong@kernel.org, chandan.babu@oracle.com, brauner@kernel.org, akpm@linux-foundation.org, willy@infradead.org Cc: mcgrof@kernel.org, linux-mm@kvack.org, hare@suse.de, linux-kernel@vger.kernel.org, yang@os.amperecomputing.com, Zi Yan , linux-xfs@vger.kernel.org, p.raghav@samsung.com, linux-fsdevel@vger.kernel.org, kernel@pankajraghav.com, hch@lst.de, gost.dev@samsung.com, cl@os.amperecomputing.com, john.g.garry@oracle.com Subject: [PATCH v7 04/11] readahead: allocate folios with mapping_min_order in readahead Date: Fri, 7 Jun 2024 14:58:55 +0000 Message-ID: <20240607145902.1137853-5-kernel@pankajraghav.com> In-Reply-To: <20240607145902.1137853-1-kernel@pankajraghav.com> References: <20240607145902.1137853-1-kernel@pankajraghav.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" From: Pankaj Raghav page_cache_ra_unbounded() was allocating single pages (0 order folios) if there was no folio found in an index. Allocate mapping_min_order folios as we need to guarantee the minimum order if it is set. When read_pages() is triggered and if a page is already present, check for truncation and move the ractl->_index by mapping_min_nrpages if that folio was truncated. This is done to ensure we keep the alignment requirement while adding a folio to the page cache. page_cache_ra_order() tries to allocate folio to the page cache with a higher order if the index aligns with that order. Modify it so that the order does not go below the mapping_min_order requirement of the page cache. This function will do the right thing even if the new_order passed is less than the mapping_min_order. When adding new folios to the page cache we must also ensure the index used is aligned to the mapping_min_order as the page cache requires the index to be aligned to the order of the folio. readahead_expand() is called from readahead aops to extend the range of the readahead so this function can assume ractl->_index to be aligned with min_order. Reviewed-by: Hannes Reinecke Signed-off-by: Pankaj Raghav --- mm/readahead.c | 85 +++++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 71 insertions(+), 14 deletions(-) diff --git a/mm/readahead.c b/mm/readahead.c index da34b28da02c..389cd802da63 100644 --- a/mm/readahead.c +++ b/mm/readahead.c @@ -206,9 +206,10 @@ void page_cache_ra_unbounded(struct readahead_control = *ractl, unsigned long nr_to_read, unsigned long lookahead_size) { struct address_space *mapping =3D ractl->mapping; - unsigned long index =3D readahead_index(ractl); + unsigned long ra_folio_index, index =3D readahead_index(ractl); gfp_t gfp_mask =3D readahead_gfp_mask(mapping); - unsigned long i =3D 0; + unsigned long mark, i =3D 0; + unsigned int min_nrpages =3D mapping_min_folio_nrpages(mapping); =20 /* * Partway through the readahead operation, we will have added @@ -223,6 +224,22 @@ void page_cache_ra_unbounded(struct readahead_control = *ractl, unsigned int nofs =3D memalloc_nofs_save(); =20 filemap_invalidate_lock_shared(mapping); + index =3D mapping_align_start_index(mapping, index); + + /* + * As iterator `i` is aligned to min_nrpages, round_up the + * difference between nr_to_read and lookahead_size to mark the + * index that only has lookahead or "async_region" to set the + * readahead flag. + */ + ra_folio_index =3D round_up(readahead_index(ractl) + nr_to_read - lookahe= ad_size, + min_nrpages); + mark =3D ra_folio_index - index; + if (index !=3D readahead_index(ractl)) { + nr_to_read +=3D readahead_index(ractl) - index; + ractl->_index =3D index; + } + /* * Preallocate as many pages as we will need. */ @@ -230,7 +247,9 @@ void page_cache_ra_unbounded(struct readahead_control *= ractl, struct folio *folio =3D xa_load(&mapping->i_pages, index + i); int ret; =20 + if (folio && !xa_is_value(folio)) { + long nr_pages =3D folio_nr_pages(folio); /* * Page already present? Kick off the current batch * of contiguous pages before continuing with the @@ -240,12 +259,24 @@ void page_cache_ra_unbounded(struct readahead_control= *ractl, * not worth getting one just for that. */ read_pages(ractl); - ractl->_index +=3D folio_nr_pages(folio); + + /* + * Move the ractl->_index by at least min_pages + * if the folio got truncated to respect the + * alignment constraint in the page cache. + * + */ + if (mapping !=3D folio->mapping) + nr_pages =3D min_nrpages; + + VM_BUG_ON_FOLIO(nr_pages < min_nrpages, folio); + ractl->_index +=3D nr_pages; i =3D ractl->_index + ractl->_nr_pages - index; continue; } =20 - folio =3D filemap_alloc_folio(gfp_mask, 0); + folio =3D filemap_alloc_folio(gfp_mask, + mapping_min_folio_order(mapping)); if (!folio) break; =20 @@ -255,11 +286,11 @@ void page_cache_ra_unbounded(struct readahead_control= *ractl, if (ret =3D=3D -ENOMEM) break; read_pages(ractl); - ractl->_index++; + ractl->_index +=3D min_nrpages; i =3D ractl->_index + ractl->_nr_pages - index; continue; } - if (i =3D=3D nr_to_read - lookahead_size) + if (i =3D=3D mark) folio_set_readahead(folio); ractl->_workingset |=3D folio_test_workingset(folio); ractl->_nr_pages +=3D folio_nr_pages(folio); @@ -493,13 +524,19 @@ void page_cache_ra_order(struct readahead_control *ra= ctl, { struct address_space *mapping =3D ractl->mapping; pgoff_t index =3D readahead_index(ractl); + unsigned int min_order =3D mapping_min_folio_order(mapping); pgoff_t limit =3D (i_size_read(mapping->host) - 1) >> PAGE_SHIFT; pgoff_t mark =3D index + ra->size - ra->async_size; unsigned int nofs; int err =3D 0; gfp_t gfp =3D readahead_gfp_mask(mapping); + unsigned int min_ra_size =3D max(4, mapping_min_folio_nrpages(mapping)); =20 - if (!mapping_large_folio_support(mapping) || ra->size < 4) + /* + * Fallback when size < min_nrpages as each folio should be + * at least min_nrpages anyway. + */ + if (!mapping_large_folio_support(mapping) || ra->size < min_ra_size) goto fallback; =20 limit =3D min(limit, index + ra->size - 1); @@ -508,11 +545,20 @@ void page_cache_ra_order(struct readahead_control *ra= ctl, new_order +=3D 2; new_order =3D min(mapping_max_folio_order(mapping), new_order); new_order =3D min_t(unsigned int, new_order, ilog2(ra->size)); + new_order =3D max(new_order, min_order); } =20 /* See comment in page_cache_ra_unbounded() */ nofs =3D memalloc_nofs_save(); filemap_invalidate_lock_shared(mapping); + /* + * If the new_order is greater than min_order and index is + * already aligned to new_order, then this will be noop as index + * aligned to new_order should also be aligned to min_order. + */ + ractl->_index =3D mapping_align_start_index(mapping, index); + index =3D readahead_index(ractl); + while (index <=3D limit) { unsigned int order =3D new_order; =20 @@ -520,7 +566,7 @@ void page_cache_ra_order(struct readahead_control *ract= l, if (index & ((1UL << order) - 1)) order =3D __ffs(index); /* Don't allocate pages past EOF */ - while (index + (1UL << order) - 1 > limit) + while (order > min_order && index + (1UL << order) - 1 > limit) order--; err =3D ra_alloc_folio(ractl, index, mark, order, gfp); if (err) @@ -784,8 +830,15 @@ void readahead_expand(struct readahead_control *ractl, struct file_ra_state *ra =3D ractl->ra; pgoff_t new_index, new_nr_pages; gfp_t gfp_mask =3D readahead_gfp_mask(mapping); + unsigned long min_nrpages =3D mapping_min_folio_nrpages(mapping); + unsigned int min_order =3D mapping_min_folio_order(mapping); =20 new_index =3D new_start / PAGE_SIZE; + /* + * Readahead code should have aligned the ractl->_index to + * min_nrpages before calling readahead aops. + */ + VM_BUG_ON(!IS_ALIGNED(ractl->_index, min_nrpages)); =20 /* Expand the leading edge downwards */ while (ractl->_index > new_index) { @@ -795,9 +848,11 @@ void readahead_expand(struct readahead_control *ractl, if (folio && !xa_is_value(folio)) return; /* Folio apparently present */ =20 - folio =3D filemap_alloc_folio(gfp_mask, 0); + folio =3D filemap_alloc_folio(gfp_mask, min_order); if (!folio) return; + + index =3D mapping_align_start_index(mapping, index); if (filemap_add_folio(mapping, folio, index, gfp_mask) < 0) { folio_put(folio); return; @@ -807,7 +862,7 @@ void readahead_expand(struct readahead_control *ractl, ractl->_workingset =3D true; psi_memstall_enter(&ractl->_pflags); } - ractl->_nr_pages++; + ractl->_nr_pages +=3D min_nrpages; ractl->_index =3D folio->index; } =20 @@ -822,9 +877,11 @@ void readahead_expand(struct readahead_control *ractl, if (folio && !xa_is_value(folio)) return; /* Folio apparently present */ =20 - folio =3D filemap_alloc_folio(gfp_mask, 0); + folio =3D filemap_alloc_folio(gfp_mask, min_order); if (!folio) return; + + index =3D mapping_align_start_index(mapping, index); if (filemap_add_folio(mapping, folio, index, gfp_mask) < 0) { folio_put(folio); return; @@ -834,10 +891,10 @@ void readahead_expand(struct readahead_control *ractl, ractl->_workingset =3D true; psi_memstall_enter(&ractl->_pflags); } - ractl->_nr_pages++; + ractl->_nr_pages +=3D min_nrpages; if (ra) { - ra->size++; - ra->async_size++; + ra->size +=3D min_nrpages; + ra->async_size +=3D min_nrpages; } } } --=20 2.44.1