From nobody Thu Dec 18 01:35:38 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 4EA04CDB47E for ; Fri, 13 Oct 2023 16:26:41 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231934AbjJMQ0j (ORCPT ); Fri, 13 Oct 2023 12:26:39 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58046 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232005AbjJMQ0N (ORCPT ); Fri, 13 Oct 2023 12:26:13 -0400 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 50ABA2684 for ; Fri, 13 Oct 2023 09:06:42 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1697213201; 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=uT/Mhdx8sQf5PESeJ8Rr2IswwVlQVrc+dpLVfyRgQ2c=; b=XkrV7UdOuN2BAFXbw9eVP/ZxZWuyHqNVAM2LavDrwIMMmBh4yLd75IW83GNNBLrjJrc6NM BcyqyanwlIIT/1Z3+VPeoCnjVIoeGPjs/HSDHkqlsiJEOkF3SGxwPZxJOECJV2R1g1p18U z37fEtc4e6eKEyWMBV5Rykk1WfJTsjo= Received: from mimecast-mx02.redhat.com (mimecast-mx02.redhat.com [66.187.233.88]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id us-mta-640-oOqr7qHpOt-wp8G5blDJHw-1; Fri, 13 Oct 2023 12:06:23 -0400 X-MC-Unique: oOqr7qHpOt-wp8G5blDJHw-1 Received: from smtp.corp.redhat.com (int-mx07.intmail.prod.int.rdu2.redhat.com [10.11.54.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mimecast-mx02.redhat.com (Postfix) with ESMTPS id 259D588B7B8; Fri, 13 Oct 2023 16:06:22 +0000 (UTC) Received: from warthog.procyon.org.uk (unknown [10.42.28.226]) by smtp.corp.redhat.com (Postfix) with ESMTP id 61A841C060DF; Fri, 13 Oct 2023 16:06:19 +0000 (UTC) From: David Howells To: Jeff Layton , Steve French Cc: David Howells , Matthew Wilcox , Marc Dionne , Paulo Alcantara , Shyam Prasad N , Tom Talpey , Dominique Martinet , Ilya Dryomov , Christian Brauner , linux-afs@lists.infradead.org, linux-cifs@vger.kernel.org, linux-nfs@vger.kernel.org, ceph-devel@vger.kernel.org, v9fs@lists.linux.dev, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-cachefs@redhat.com Subject: [RFC PATCH 33/53] netfs: Provide minimum blocksize parameter Date: Fri, 13 Oct 2023 17:04:02 +0100 Message-ID: <20231013160423.2218093-34-dhowells@redhat.com> In-Reply-To: <20231013160423.2218093-1-dhowells@redhat.com> References: <20231013160423.2218093-1-dhowells@redhat.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Scanned-By: MIMEDefang 3.4.1 on 10.11.54.7 Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Add a parameter for minimum blocksize in the netfs_i_context struct. This can be used, for instance, to force I/O alignment for content encryption. It also requires the use of an RMW cycle if a write we want to do doesn't meet the block alignment requirements. Signed-off-by: David Howells cc: Jeff Layton cc: linux-cachefs@redhat.com cc: linux-fsdevel@vger.kernel.org cc: linux-mm@kvack.org --- fs/netfs/buffered_read.c | 26 ++++++++++++++++++++++---- fs/netfs/buffered_write.c | 3 ++- fs/netfs/direct_read.c | 3 ++- include/linux/netfs.h | 2 ++ 4 files changed, 28 insertions(+), 6 deletions(-) diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c index ab9f8e123245..e06461ef0bfa 100644 --- a/fs/netfs/buffered_read.c +++ b/fs/netfs/buffered_read.c @@ -527,14 +527,26 @@ int netfs_prefetch_for_write(struct file *file, struc= t folio *folio, struct address_space *mapping =3D folio_file_mapping(folio); struct netfs_inode *ctx =3D netfs_inode(mapping->host); unsigned long long start =3D folio_pos(folio); - size_t flen =3D folio_size(folio); + unsigned long long i_size, rstart, end; + size_t rlen; int ret; =20 - _enter("%zx @%llx", flen, start); + DEFINE_READAHEAD(ractl, file, NULL, mapping, folio_index(folio)); + + _enter("%zx @%llx", len, start); =20 ret =3D -ENOMEM; =20 - rreq =3D netfs_alloc_request(mapping, file, start, flen, + i_size =3D i_size_read(mapping->host); + end =3D round_up(start + len, 1U << ctx->min_bshift); + if (end > i_size) { + unsigned long long limit =3D round_up(start + len, PAGE_SIZE); + end =3D max(limit, round_up(i_size, PAGE_SIZE)); + } + rstart =3D round_down(start, 1U << ctx->min_bshift); + rlen =3D end - rstart; + + rreq =3D netfs_alloc_request(mapping, file, rstart, rlen, NETFS_READ_FOR_WRITE); if (IS_ERR(rreq)) { ret =3D PTR_ERR(rreq); @@ -548,7 +560,13 @@ int netfs_prefetch_for_write(struct file *file, struct= folio *folio, goto error_put; =20 netfs_stat(&netfs_n_rh_write_begin); - trace_netfs_read(rreq, start, flen, netfs_read_trace_prefetch_for_write); + trace_netfs_read(rreq, rstart, rlen, netfs_read_trace_prefetch_for_write); + + /* Expand the request to meet caching requirements and download + * preferences. + */ + ractl._nr_pages =3D folio_nr_pages(folio); + netfs_rreq_expand(rreq, &ractl); =20 /* Set up the output buffer */ iov_iter_xarray(&rreq->iter, ITER_DEST, &mapping->i_pages, diff --git a/fs/netfs/buffered_write.c b/fs/netfs/buffered_write.c index d5a5a315fbd3..7163fcc05206 100644 --- a/fs/netfs/buffered_write.c +++ b/fs/netfs/buffered_write.c @@ -80,7 +80,8 @@ static enum netfs_how_to_modify netfs_how_to_modify(struc= t netfs_inode *ctx, if (file->f_mode & FMODE_READ) return NETFS_JUST_PREFETCH; =20 - if (netfs_is_cache_enabled(ctx)) + if (netfs_is_cache_enabled(ctx) || + ctx->min_bshift > 0) return NETFS_JUST_PREFETCH; =20 if (!finfo) diff --git a/fs/netfs/direct_read.c b/fs/netfs/direct_read.c index 1d26468aafd9..52ad8fa66dd5 100644 --- a/fs/netfs/direct_read.c +++ b/fs/netfs/direct_read.c @@ -185,7 +185,8 @@ static ssize_t netfs_unbuffered_read_iter_locked(struct= kiocb *iocb, struct iov_ * will then need to pad the request out to the minimum block size. */ if (test_bit(NETFS_RREQ_USE_BOUNCE_BUFFER, &rreq->flags)) { - start =3D rreq->start; + min_bsize =3D 1ULL << ctx->min_bshift; + start =3D round_down(rreq->start, min_bsize); end =3D min_t(unsigned long long, round_up(rreq->start + rreq->len, min_bsize), ctx->remote_i_size); diff --git a/include/linux/netfs.h b/include/linux/netfs.h index fb4f4f826b93..6244f7a9a44a 100644 --- a/include/linux/netfs.h +++ b/include/linux/netfs.h @@ -141,6 +141,7 @@ struct netfs_inode { unsigned long flags; #define NETFS_ICTX_ODIRECT 0 /* The file has DIO in progress */ #define NETFS_ICTX_UNBUFFERED 1 /* I/O should not use the pagecache */ + unsigned char min_bshift; /* log2 min block size for bounding box or 0 */ }; =20 /* @@ -462,6 +463,7 @@ static inline void netfs_inode_init(struct netfs_inode = *ctx, ctx->remote_i_size =3D i_size_read(&ctx->inode); ctx->zero_point =3D ctx->remote_i_size; ctx->flags =3D 0; + ctx->min_bshift =3D 0; #if IS_ENABLED(CONFIG_FSCACHE) ctx->cache =3D NULL; #endif