From nobody Mon Sep 28 23:55:25 2026 Received: from mail-pg1-f200.google.com (mail-pg1-f200.google.com [209.85.215.200]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 3D04D4756A1 for ; Fri, 14 Aug 2026 14:33:02 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.215.200 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786717985; cv=none; b=AzeKsjdYUpwr+gw4JXQpqoZFWqkg5aF/fRb5/ahadFmlGqgEK6Lfc69n4VKB7AOOhsiXqx5bswtjtfraeXMcm2neV8SUHBJP1zvzycybpg8lOcFMlTJ1mKkzp7453hnPcVgKuPfC43ujvdkKBgx109ZSTdCD9E8vspbYlo/+5vk= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786717985; c=relaxed/simple; bh=WcCEN0IarFa7dqxqnB42Skg89skLls/g2t8xZYfpEjo=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=QF7g8uMJQtyWeQ2NFNgEmmqBUeRIOb980ndupXMRpghqqq9MX+0tAFAlqEMGGFDapLo9tIDYGSgTbeHsqFJcwvNDkBVcxNQJtcsi4zGeFNtEn92fWSF3TndvmFmZ+xSZuphHvYWrz6UQdQ3my807jpEU7ckXk//ZmlZW99uEk5o= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--praan.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=Wy0cKaXA; arc=none smtp.client-ip=209.85.215.200 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--praan.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="Wy0cKaXA" Received: by mail-pg1-f200.google.com with SMTP id 41be03b00d2f7-cb5cc1e13f8so875100a12.3 for ; Fri, 14 Aug 2026 07:33:01 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20251104; t=1786717980; x=1787322780; darn=vger.kernel.org; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:from:to:cc:subject:date:message-id:reply-to :content-type; bh=+KJskUJlBd536n7h/sL/C3PYmacUse+7O8i0d/ARYdU=; b=Wy0cKaXAS64hT59OaHU/kDs0qnyYTLTbcEc+STE+1Wh9rty3twlnVaw9oGsN2hVDpw i9i4A5TJBCnnmdsE42/xxhQzhW3Oltzmf8cYmmYFgG+t/3J0RO2PfYtoaDg/52IlxDgX 0hqMjkm1RubKVwQNXBM1XL97W+y3VjLqduPFOOrnV86YRi0kzfqAESgVKRsPhh6bGvoO b54vqc3B8cgkkfpQxPqcUcfhIffAdf0rcbdF8j9xuerK3uU60bIBahnn8ZFvug1tD4I/ wQvCAQzVkt7lYx7x2sap5QmoX6Z9kQdLITOxxjAAdOf6L7NKu2oIyjibQsiM7uLfKMWg g+nw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1786717980; x=1787322780; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to:content-type; bh=+KJskUJlBd536n7h/sL/C3PYmacUse+7O8i0d/ARYdU=; b=Xt7wEcES0Mu4uzJWTp/GdhZO6Tbc/b6L2JZ4RhRR8TExaiavtaL/uaEmRm2LakZ2op J9I2+SEsUdZjXnZe/ZHHIPXFw1eKTwPkXAxaZkdZs+IjleozYUdKwmcJXSkdj/gVlIM0 h1AtPYqWh5wau4PhCYieFnAWQlq8QukrjQhM9FbTV7NwzPghrfebFDbEuXHoMN00HLKn mkTYvROCRV1l/u+DI6gDlsqXe440k6V9IOvIvvUqNn3YtxzMGemMzHNcIeQ8NbfCshv5 LMofd7P9m7DDFW8dtbAP5b5ciLNZyC1+B4ax3PuAZDxhUBtH6oPkzoreK2zvZTqNl6lj cAGA== X-Forwarded-Encrypted: i=1; AHgh+RqdVGCo+RLvheNS6LN7K/l1th8CG4JpwdYe78oUQAz8m1MkkGdXEFAyJCXCVJsNa/u+wj5KeX4KiVJyR34=@vger.kernel.org X-Gm-Message-State: AOJu0YwtPyoCPwwMpXk0iSKuF4bEtOJZItjI82NzVrBjy8ypa1Nqg6bG f1yFcOImL6L05TFO8FFaqR6nxym7fzkhpOH2OY21TbXhsKRwZ4RJMmTabTtWLKU5clPmnWHvV5C Oww== X-Received: from dyt26.prod.google.com ([2002:a05:693c:809a:b0:314:314:df67]) (user=praan job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6a20:3d0b:b0:3cc:3d08:da3e with SMTP id adf61e73a8af0-3cc71992e2bmr7604586637.2.1786717980237; Fri, 14 Aug 2026 07:33:00 -0700 (PDT) Date: Fri, 14 Aug 2026 14:32:51 +0000 In-Reply-To: <20260814143255.861084-1-praan@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20260814143255.861084-1-praan@google.com> X-Mailer: git-send-email 2.55.0.691.gc56d675ccc-goog Message-ID: <20260814143255.861084-2-praan@google.com> Subject: [PATCH v6 1/5] nfs: make nfs_page pin-aware From: Pranjal Shrivastava To: linux-nfs@vger.kernel.org, Trond Myklebust , Anna Schumaker Cc: Chuck Lever , Jeff Layton , linux-kernel@vger.kernel.org, Christoph Hellwig , Logan Gunthorpe , Jason Gunthorpe , linux-pci@vger.kernel.org, linux-rdma@vger.kernel.org, Shivaji Kant , Pranjal Shrivastava Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Modernizing the NFS Direct I/O path to use iov_iter_extract_pages() introduces page pinning (GUP) instead of standard page referencing. To handle this correctly, nfs_page must track whether it holds a pin or a standard reference. Introduce a new flag, PG_PINNED, to struct nfs_page. Update the creation path (nfs_page_create_from_page and nfs_page_create_from_folio) to accept a pinned bool and set the flag accordingly. If the page is pinned, we skip the existing reference increment (get_page/folio_get) as the pin itself acts as a reference. Update nfs_clear_request() & nfs_direct_release_pages() to use unpin_user_page() or unpin_user_folio() instead of only refcount decrement (put_page) when PG_PINNED flag is set. Finally, ensure subrequests inherit the pinning status from their parent request. Signed-off-by: Pranjal Shrivastava Reviewed-by: Christoph Hellwig --- fs/nfs/direct.c | 22 +++++++++++++------- fs/nfs/pagelist.c | 44 +++++++++++++++++++++++++++++++--------- fs/nfs/read.c | 2 +- fs/nfs/write.c | 2 +- include/linux/nfs_page.h | 3 +++ 5 files changed, 54 insertions(+), 19 deletions(-) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index e626c72495e6..19792a38c924 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -165,11 +165,17 @@ int nfs_swap_rw(struct kiocb *iocb, struct iov_iter *= iter) return 0; } =20 -static void nfs_direct_release_pages(struct page **pages, unsigned int npa= ges) +static void nfs_direct_release_pages(struct page **pages, unsigned int npa= ges, + bool pinned) { unsigned int i; - for (i =3D 0; i < npages; i++) - put_page(pages[i]); + + if (pinned) { + unpin_user_pages(pages, npages); + } else { + for (i =3D 0; i < npages; i++) + put_page(pages[i]); + } } =20 void nfs_init_cinfo_from_dreq(struct nfs_commit_info *cinfo, @@ -371,7 +377,8 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nf= s_direct_req *dreq, unsigned int req_len =3D min_t(size_t, bytes, PAGE_SIZE - pgbase); /* XXX do we need to do the eof zeroing found in async_filler? */ req =3D nfs_page_create_from_page(dreq->ctx, pagevec[i], - pgbase, pos, req_len); + false, pgbase, pos, + req_len); if (IS_ERR(req)) { result =3D PTR_ERR(req); break; @@ -386,7 +393,7 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nf= s_direct_req *dreq, requested_bytes +=3D req_len; pos +=3D req_len; } - nfs_direct_release_pages(pagevec, npages); + nfs_direct_release_pages(pagevec, npages, false); kvfree(pagevec); if (result < 0) break; @@ -907,7 +914,8 @@ static ssize_t nfs_direct_write_schedule_iovec(struct n= fs_direct_req *dreq, unsigned int req_len =3D min_t(size_t, bytes, PAGE_SIZE - pgbase); =20 req =3D nfs_page_create_from_page(dreq->ctx, pagevec[i], - pgbase, pos, req_len); + false, pgbase, pos, + req_len); if (IS_ERR(req)) { result =3D PTR_ERR(req); break; @@ -950,7 +958,7 @@ static ssize_t nfs_direct_write_schedule_iovec(struct n= fs_direct_req *dreq, desc.pg_error =3D 0; defer =3D true; } - nfs_direct_release_pages(pagevec, npages); + nfs_direct_release_pages(pagevec, npages, false); kvfree(pagevec); if (result < 0) break; diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 7dd478ffc2fa..a562cfe2a126 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -404,20 +404,28 @@ static struct nfs_page *nfs_page_create(struct nfs_lo= ck_context *l_ctx, return req; } =20 -static void nfs_page_assign_folio(struct nfs_page *req, struct folio *foli= o) +static void nfs_page_assign_folio(struct nfs_page *req, struct folio *foli= o, + bool pinned) { if (folio !=3D NULL) { req->wb_folio =3D folio; - folio_get(folio); + if (pinned) + set_bit(PG_PINNED, &req->wb_flags); + else + folio_get(folio); set_bit(PG_FOLIO, &req->wb_flags); } } =20 -static void nfs_page_assign_page(struct nfs_page *req, struct page *page) +static void nfs_page_assign_page(struct nfs_page *req, struct page *page, + bool pinned) { if (page !=3D NULL) { req->wb_page =3D page; - get_page(page); + if (pinned) + set_bit(PG_PINNED, &req->wb_flags); + else + get_page(page); } } =20 @@ -425,6 +433,7 @@ static void nfs_page_assign_page(struct nfs_page *req, = struct page *page) * nfs_page_create_from_page - Create an NFS read/write request. * @ctx: open context to use * @page: page to write + * @pinned: true if page is pinned * @pgbase: starting offset within the page for the write * @offset: file offset for the write * @count: number of bytes to read/write @@ -435,6 +444,7 @@ static void nfs_page_assign_page(struct nfs_page *req, = struct page *page) */ struct nfs_page *nfs_page_create_from_page(struct nfs_open_context *ctx, struct page *page, + bool pinned, unsigned int pgbase, loff_t offset, unsigned int count) { @@ -446,7 +456,7 @@ struct nfs_page *nfs_page_create_from_page(struct nfs_o= pen_context *ctx, ret =3D nfs_page_create(l_ctx, pgbase, offset >> PAGE_SHIFT, offset_in_page(offset), count); if (!IS_ERR(ret)) { - nfs_page_assign_page(ret, page); + nfs_page_assign_page(ret, page, pinned); nfs_page_group_init(ret, NULL); } nfs_put_lock_context(l_ctx); @@ -457,6 +467,7 @@ struct nfs_page *nfs_page_create_from_page(struct nfs_o= pen_context *ctx, * nfs_page_create_from_folio - Create an NFS read/write request. * @ctx: open context to use * @folio: folio to write + * @pinned: true if folio is pinned * @offset: starting offset within the folio for the write * @count: number of bytes to read/write * @@ -466,6 +477,7 @@ struct nfs_page *nfs_page_create_from_page(struct nfs_o= pen_context *ctx, */ struct nfs_page *nfs_page_create_from_folio(struct nfs_open_context *ctx, struct folio *folio, + bool pinned, unsigned int offset, unsigned int count) { @@ -476,7 +488,7 @@ struct nfs_page *nfs_page_create_from_folio(struct nfs_= open_context *ctx, return ERR_CAST(l_ctx); ret =3D nfs_page_create(l_ctx, offset, folio->index, offset, count); if (!IS_ERR(ret)) { - nfs_page_assign_folio(ret, folio); + nfs_page_assign_folio(ret, folio, pinned); nfs_page_group_init(ret, NULL); } nfs_put_lock_context(l_ctx); @@ -498,9 +510,11 @@ nfs_create_subreq(struct nfs_page *req, offset, count); if (!IS_ERR(ret)) { if (folio) - nfs_page_assign_folio(ret, folio); + nfs_page_assign_folio(ret, folio, + test_bit(PG_PINNED, &req->wb_flags)); else - nfs_page_assign_page(ret, page); + nfs_page_assign_page(ret, page, + test_bit(PG_PINNED, &req->wb_flags)); /* find the last request */ for (last =3D req->wb_head; last->wb_this_page !=3D req->wb_head; @@ -552,11 +566,21 @@ static void nfs_clear_request(struct nfs_page *req) struct nfs_open_context *ctx; =20 if (folio !=3D NULL) { - folio_put(folio); + if (test_and_clear_bit(PG_PINNED, &req->wb_flags)) { + if (req =3D=3D req->wb_head) + unpin_user_folio(folio, 1); + } else { + folio_put(folio); + } req->wb_folio =3D NULL; clear_bit(PG_FOLIO, &req->wb_flags); } else if (page !=3D NULL) { - put_page(page); + if (test_and_clear_bit(PG_PINNED, &req->wb_flags)) { + if (req =3D=3D req->wb_head) + unpin_user_page(page); + } else { + put_page(page); + } req->wb_page =3D NULL; } if (l_ctx !=3D NULL) { diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 2b70bd2b934b..e7497b029d6c 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -324,7 +324,7 @@ int nfs_read_add_folio(struct nfs_pageio_descriptor *pg= io, =20 aligned_len =3D min_t(unsigned int, ALIGN(len, rsize), fsize); =20 - new =3D nfs_page_create_from_folio(ctx, folio, 0, aligned_len); + new =3D nfs_page_create_from_folio(ctx, folio, false, 0, aligned_len); if (IS_ERR(new)) { error =3D PTR_ERR(new); if (nfs_netfs_folio_unlock(folio)) diff --git a/fs/nfs/write.c b/fs/nfs/write.c index d2b03ceaeb4f..e565b811d8b7 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -1088,7 +1088,7 @@ static struct nfs_page *nfs_setup_write_request(struc= t nfs_open_context *ctx, req =3D nfs_try_to_update_request(folio, offset, bytes); if (req !=3D NULL) goto out; - req =3D nfs_page_create_from_folio(ctx, folio, offset, bytes); + req =3D nfs_page_create_from_folio(ctx, folio, false, offset, bytes); if (IS_ERR(req)) goto out; nfs_inode_add_request(req); diff --git a/include/linux/nfs_page.h b/include/linux/nfs_page.h index 4b9a35dbc062..fd7aafe7cb54 100644 --- a/include/linux/nfs_page.h +++ b/include/linux/nfs_page.h @@ -38,6 +38,7 @@ enum { PG_REMOVE, /* page group sync bit in write path */ PG_CONTENDED1, /* Is someone waiting for a lock? */ PG_CONTENDED2, /* Is someone waiting for a lock? */ + PG_PINNED, /* page is pinned by GUP */ }; =20 struct nfs_inode; @@ -125,11 +126,13 @@ struct nfs_pageio_descriptor { =20 extern struct nfs_page *nfs_page_create_from_page(struct nfs_open_context = *ctx, struct page *page, + bool pinned, unsigned int pgbase, loff_t offset, unsigned int count); extern struct nfs_page *nfs_page_create_from_folio(struct nfs_open_context= *ctx, struct folio *folio, + bool pinned, unsigned int offset, unsigned int count); extern void nfs_release_request(struct nfs_page *); --=20 2.55.0.691.gc56d675ccc-goog From nobody Mon Sep 28 23:55:25 2026 Received: from mail-pg1-f198.google.com (mail-pg1-f198.google.com [209.85.215.198]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 5E94C476CD8 for ; Fri, 14 Aug 2026 14:33:05 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.215.198 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786717990; cv=none; b=iy7ymKJrfB4YEqoBEvq/0qZ/tVW0VsZfBv/+WHzwF9FpOTZdi/pYXOm6cM4y8fGW27anF6TiPDH+h1GySUlAroIdVtDrLjn82cNbRZchoc8rbo9LcE1AC6S0Ceb58+raHO0gA4LT9jkPQKRKGm3opZCQ7pb3LWBbFzvAnie9H+4= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786717990; c=relaxed/simple; bh=u0MKNRKgDVwVPGRV+6Ey10B0Yv0u0Jn1oxVRulYF4u0=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=VKhNb4FjZaR/g+SlDDQ8/MIuKR8wmc/fOMapbaVXu06e9sXc0i7NxAJq99ceFvDjbmqrXmEHfemswtrvwI2TIS1y/9aRymzYGJNEBdCsBhphOj5N1T+mI0UvywtGHz7kCAl90k0JUfmqX370Ts3l1OEgDGOqRps4OEOzthHsWtA= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--praan.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=N093Uj+4; arc=none smtp.client-ip=209.85.215.198 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--praan.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="N093Uj+4" Received: by mail-pg1-f198.google.com with SMTP id 41be03b00d2f7-c85798977dcso1524131a12.0 for ; Fri, 14 Aug 2026 07:33:05 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20251104; t=1786717984; x=1787322784; darn=vger.kernel.org; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:from:to:cc:subject:date:message-id:reply-to :content-type; bh=VC79kGoXaLgBrqdL04SZQWwyFvo3s0PHCta3qnTuB/A=; b=N093Uj+4IjrFgJc79huP/de34mPTNZp6wxRZ9DEFJTnqjRJineALuS/FHHiZPLoWd8 zfFczokF79jY7V8jB54JKboRE7DDKpmADkm6/ajUFn3N7HzzZ9urUhIS6Yht7Cki4zkc I1mUf6b+efJFVDnAU1Mpqg04wFeYxBgjepKBkamOImCTUW9uidUVhDOpcJTHpG4uh1u0 kBVITEOkIeNl4J3IwxXYVcm5XNvKZZ4abheXvzPLb2MT4/pFvNmTkIviHIG0e/9AyvNB cYLZzVwedMU0++Y64T7127b+RLpOgd5IJJkdeh49w3yOyhaUwUyqbutp2Ws+5l4KV6TJ QYlw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1786717984; x=1787322784; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to:content-type; bh=VC79kGoXaLgBrqdL04SZQWwyFvo3s0PHCta3qnTuB/A=; b=gC6dxDd04vcSMHVYDWpChP2ny0orCkm0pS9A+BaKS6goOmlL66Kcg/GuVVTK24sGFx VoMde8DejoM9XVUAPpRfZxjVtMS81MR+bmrka8gAaMmh1FAsrVJnVI/iLjCLDysPwxjU uDJXsQHqRTDOXktEYhbb5Au5PSM5qyM/HHFfQ3T/5QEObVvOSiVthD6oEngbqJU7MxvD TsuxKJEADSr+E8KaXlL5EwiMdgyRORQaMdAGYwwwOxoh7LzmxJXR3ILyndy8at/IM3aV OlrSKQZZpmURvFUD3UEPR12/g0GdkiWAIj+JCvymezr6dSGFrnrf370MfN94wDB0ykD8 OYkw== X-Forwarded-Encrypted: i=1; AHgh+Rq7BJU0/WJpVUb/wA8eFcmda1nJ4YS6BukMt2wygi2MCNZLS2V4tw8h+v2+yTtE5EbuCza9whbnVHVrOO8=@vger.kernel.org X-Gm-Message-State: AOJu0YwiBWQn/GCidHJnats4dWgaUCKXRZgG+PIkaA24Y6A5IFIAgvKF wFgyrx9fIpRJs4mVpP5j5P2ez8ioP2DTd1J+lToSdNgReLRQJlq27SQhWKglnVN7DCfDGFhEf1j lJQ== X-Received: from pjbnl15.prod.google.com ([2002:a17:90b:384f:b0:38f:a3cc:f600]) (user=praan job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:1b47:b0:38e:5964:97a8 with SMTP id 98e67ed59e1d1-3933e63167amr6649209a91.16.1786717983830; Fri, 14 Aug 2026 07:33:03 -0700 (PDT) Date: Fri, 14 Aug 2026 14:32:52 +0000 In-Reply-To: <20260814143255.861084-1-praan@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20260814143255.861084-1-praan@google.com> X-Mailer: git-send-email 2.55.0.691.gc56d675ccc-goog Message-ID: <20260814143255.861084-3-praan@google.com> Subject: [PATCH v6 2/5] nfs: track number of pinned pages in nfs_page From: Pranjal Shrivastava To: linux-nfs@vger.kernel.org, Trond Myklebust , Anna Schumaker Cc: Chuck Lever , Jeff Layton , linux-kernel@vger.kernel.org, Christoph Hellwig , Logan Gunthorpe , Jason Gunthorpe , linux-pci@vger.kernel.org, linux-rdma@vger.kernel.org, Shivaji Kant , Pranjal Shrivastava Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Track the number of pinned pages in nfs_page to handle unpinning correctly, ensuring that only primary requests perform the final unpinning operation, preventing subrequests from incorrectly performing unpinning on behalf of their parent requests. Add wb_nr_pinned to struct nfs_page to store the count of pinned pages owned by the request. Update request creation and cleanup helpers to initialize and use wb_nr_pinned for primary requests. Use the nfs_page_array_len() helper to calculate the number of pages spanned by a request's offset and length. Signed-off-by: Pranjal Shrivastava Reviewed-by: Christoph Hellwig --- fs/nfs/pagelist.c | 13 +++++++++---- include/linux/nfs_page.h | 1 + 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index a562cfe2a126..b9ccf2a87e3c 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -457,6 +457,8 @@ struct nfs_page *nfs_page_create_from_page(struct nfs_o= pen_context *ctx, offset_in_page(offset), count); if (!IS_ERR(ret)) { nfs_page_assign_page(ret, page, pinned); + if (pinned) + ret->wb_nr_pinned =3D 1; nfs_page_group_init(ret, NULL); } nfs_put_lock_context(l_ctx); @@ -489,6 +491,9 @@ struct nfs_page *nfs_page_create_from_folio(struct nfs_= open_context *ctx, ret =3D nfs_page_create(l_ctx, offset, folio->index, offset, count); if (!IS_ERR(ret)) { nfs_page_assign_folio(ret, folio, pinned); + if (pinned) + ret->wb_nr_pinned =3D nfs_page_array_len(offset_in_page(offset), + count); nfs_page_group_init(ret, NULL); } nfs_put_lock_context(l_ctx); @@ -567,8 +572,8 @@ static void nfs_clear_request(struct nfs_page *req) =20 if (folio !=3D NULL) { if (test_and_clear_bit(PG_PINNED, &req->wb_flags)) { - if (req =3D=3D req->wb_head) - unpin_user_folio(folio, 1); + if (req->wb_nr_pinned > 0) + unpin_user_folio(folio, req->wb_nr_pinned); } else { folio_put(folio); } @@ -576,8 +581,8 @@ static void nfs_clear_request(struct nfs_page *req) clear_bit(PG_FOLIO, &req->wb_flags); } else if (page !=3D NULL) { if (test_and_clear_bit(PG_PINNED, &req->wb_flags)) { - if (req =3D=3D req->wb_head) - unpin_user_page(page); + if (req->wb_nr_pinned > 0) + unpin_user_pages(&page, req->wb_nr_pinned); } else { put_page(page); } diff --git a/include/linux/nfs_page.h b/include/linux/nfs_page.h index fd7aafe7cb54..080fa3e23580 100644 --- a/include/linux/nfs_page.h +++ b/include/linux/nfs_page.h @@ -59,6 +59,7 @@ struct nfs_page { struct nfs_page *wb_this_page; /* list of reqs for this page */ struct nfs_page *wb_head; /* head pointer for req list */ unsigned short wb_nio; /* Number of I/O attempts */ + unsigned int wb_nr_pinned; /* Number of pinned pages */ }; =20 struct nfs_pgio_mirror; --=20 2.55.0.691.gc56d675ccc-goog From nobody Mon Sep 28 23:55:25 2026 Received: from mail-pj1-f69.google.com (mail-pj1-f69.google.com [209.85.216.69]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C3202476040 for ; Fri, 14 Aug 2026 14:33:10 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.69 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786717994; cv=none; b=f1gsN+FPuQ1KnwtkspLD3C357T/20lM6rRpz2bQhPKu74YEAUFnvKCnrHlI9EAx463+NutjpkIbUBjymUaBrgHCBrQwxmuKroyh3RzqKk09DV3+62aZHygj5e2xYEsEufBMjea7CYBvagD1GCabvjQ1Dnvj6uQgXbkbHUTeaZ0U= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786717994; c=relaxed/simple; bh=iznWeDCf/xVMjXe2dlj/pF4NEFalcZlXzwz9e4tzuwU=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=bepMBVyEQsy/y5azv1we3wB6w3hSVqJ9SVZW7TXG5T0D1O0nFiSYOGBkPmc5HfoRDzvuIg6klu7RlSx5tArr8tzUnqFU5CzkLNOwM1CRVGDsVrqSViOIXpTNsgAwIlf9ncuqTqzO2cs/QISY5tAPOESvUZZuUd0MVFNZhXpSmxE= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--praan.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=J+JFzrRJ; arc=none smtp.client-ip=209.85.216.69 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--praan.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="J+JFzrRJ" Received: by mail-pj1-f69.google.com with SMTP id 98e67ed59e1d1-385d2703b64so2175756a91.1 for ; Fri, 14 Aug 2026 07:33:10 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20251104; t=1786717989; x=1787322789; darn=vger.kernel.org; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:from:to:cc:subject:date:message-id:reply-to :content-type; bh=Dgzo1tF03PdbXxJfwkTkV/zmue9zVVPZETKq5+mYuVg=; b=J+JFzrRJTgvYfIImJMx4kIo38OqTdn+S4tEc5CUekyyvyTezCdT254LY9rdmhUxt71 vBPIVS//Fg0PSG2P5ChqymvRU+bRg/N53YjtdTEWB6MnFvCJn2i09cn/GPM8ExGrbkrb XcL9p/rDKmAkxd2psyJqiSHWruz2xTFOL2A/kLMTp4deotLl42HnnMMfdQQW2s853hje EEI+BIUKXwoK1Td8cNCnRgIDnL9BTd+ekbjg82EOaFryAicg0VUS4ebYxu6Xa9rGB3G4 p/gLTZnhsNpPk1GIQ1NWO52wZlqNNtxkNju6KMS2Me7pxg/01pcBCO2TkuuTqNTkg3sY Mzpw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1786717989; x=1787322789; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to:content-type; bh=Dgzo1tF03PdbXxJfwkTkV/zmue9zVVPZETKq5+mYuVg=; b=r4CseD6VFXbboxKiaGolF+AoMuRTOkMxkZQhYg3xtUn9tuEhdW1nOVtuCUCtAFVj4d bbC+nbC6Bj1ynuHulvADs3dmaOqTiwrXG72moNsAaM+qJ//J7rlMmxc3BS56NrzyfKe+ U2P1QHQV+3mKi2UHgBqvnNEyTKQGJAURCI7dfCEsdcMeXK863r9VnobKnKRi5VR286YF aBd79bYGOht/I4v1RRoNnk/U+vbLilA+Wx8Pd6vLelr9ILO9DDb4Q4EEsxxtQVPQG4+z ZhXPHKuNiiNvawLOz6UKPLDkvSlcSfH2ZqCr0KgVwbEkJz1bZQg9FQZFAXuU44jkIHkw F2+Q== X-Forwarded-Encrypted: i=1; AHgh+RpxCapol3r0C0lAviDk3rxBKBygjXcYCDtAe/4pbZkonnSaZ5WEVYc6BWYGq9e8z9CcYYn3skETjr5jgFg=@vger.kernel.org X-Gm-Message-State: AOJu0YwroLAJ+R9u1u3PL+tozM+khIcbuweSekgxs8jH2CJl8XFg8pOW 05c1cEmRxa1LfQc9pJ93BEvsrd8Dd3aoDPBnCvVj+sUXXSHlsS6JWlWgkzHv2uP+copQtxD2Jg9 zTQ== X-Received: from pjkx5.prod.google.com ([2002:a17:90a:7885:b0:38e:a556:bb27]) (user=praan job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90a:d407:b0:38e:4f42:d0d6 with SMTP id 98e67ed59e1d1-3933b09f4f3mr4912072a91.12.1786717988724; Fri, 14 Aug 2026 07:33:08 -0700 (PDT) Date: Fri, 14 Aug 2026 14:32:53 +0000 In-Reply-To: <20260814143255.861084-1-praan@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20260814143255.861084-1-praan@google.com> X-Mailer: git-send-email 2.55.0.691.gc56d675ccc-goog Message-ID: <20260814143255.861084-4-praan@google.com> Subject: [PATCH v6 3/5] nfs: introduce nfs_release_request_list helper From: Pranjal Shrivastava To: linux-nfs@vger.kernel.org, Trond Myklebust , Anna Schumaker Cc: Chuck Lever , Jeff Layton , linux-kernel@vger.kernel.org, Christoph Hellwig , Logan Gunthorpe , Jason Gunthorpe , linux-pci@vger.kernel.org, linux-rdma@vger.kernel.org, Shivaji Kant , Pranjal Shrivastava Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Introduce a centralized helper, nfs_release_request_list, to handle the bulk release of nfs_page requests from a list. This serves as a preparatory step for two upcoming improvements: 1. Pin-Aware Cleanup: As we migrate to iov_iter_extract_* API, requests will hold pins (GUP) instead of standard references. The helper ensures that the correct unpinning logic gets applied consistently across all requests in a list. 2. Folio Support: In subsequent patches where nfs_page structures will cover multi-page folios, this helper provides a clean infrastructure to unlock these larger units of I/O in bulk during completion, similar to the pattern in bio_release_pages. Additionally, refactor nfs_read_sync_pgio_error() to utilize this new helper. Signed-off-by: Pranjal Shrivastava Reviewed-by: Shivaji Kant Reviewed-by: Christoph Hellwig --- fs/nfs/direct.c | 8 +------- fs/nfs/pagelist.c | 17 +++++++++++++++++ include/linux/nfs_page.h | 4 ++-- 3 files changed, 20 insertions(+), 9 deletions(-) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 19792a38c924..96995736fac2 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -314,13 +314,7 @@ static void nfs_direct_read_completion(struct nfs_pgio= _header *hdr) =20 static void nfs_read_sync_pgio_error(struct list_head *head, int error) { - struct nfs_page *req; - - while (!list_empty(head)) { - req =3D nfs_list_entry(head->next); - nfs_list_remove_request(req); - nfs_release_request(req); - } + nfs_release_request_list(head); } =20 static void nfs_direct_pgio_init(struct nfs_pgio_header *hdr) diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index b9ccf2a87e3c..71f0ce2bc4ea 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -628,6 +628,23 @@ void nfs_release_request(struct nfs_page *req) } EXPORT_SYMBOL_GPL(nfs_release_request); =20 +/* + * nfs_release_request_list - Release a list of NFS read/write requests + * @head: list of requests to release + * + * Removes each request from the list and drops it's refcount. + */ +void nfs_release_request_list(struct list_head *head) +{ + while (!list_empty(head)) { + struct nfs_page *req =3D nfs_list_entry(head->next); + + nfs_list_remove_request(req); + nfs_release_request(req); + } +} +EXPORT_SYMBOL_GPL(nfs_release_request_list); + /* * nfs_generic_pg_test - determine if requests can be coalesced * @desc: pointer to descriptor diff --git a/include/linux/nfs_page.h b/include/linux/nfs_page.h index 080fa3e23580..c38e4b380be5 100644 --- a/include/linux/nfs_page.h +++ b/include/linux/nfs_page.h @@ -136,8 +136,8 @@ extern struct nfs_page *nfs_page_create_from_folio(stru= ct nfs_open_context *ctx, bool pinned, unsigned int offset, unsigned int count); -extern void nfs_release_request(struct nfs_page *); - +void nfs_release_request(struct nfs_page *req); +void nfs_release_request_list(struct list_head *head); =20 extern void nfs_pageio_init(struct nfs_pageio_descriptor *desc, struct inode *inode, --=20 2.55.0.691.gc56d675ccc-goog From nobody Mon Sep 28 23:55:25 2026 Received: from mail-pf1-f198.google.com (mail-pf1-f198.google.com [209.85.210.198]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 3907D476689 for ; Fri, 14 Aug 2026 14:33:14 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.210.198 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786718000; cv=none; b=A8NKlJZhIP0ECoNRvUwx1ld2DyI6Epkwn2DEpGR7kBSH5yT9vXyKCRSS8Szu6cjCsgQoclmOKNLSS2IWHph1hjHGRtYBqhfS+bi5DyA8itXZRhGsBvJB5ek7alM4Jd6eY+bWOxzE8WAF0IiKuz8KAsc8TGBm9Dwou1VaxUC3BqA= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786718000; c=relaxed/simple; bh=ex6pRePhYMA39HcNB1bKP2BgdIOPCiqjQAlJZqxWR5c=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=LwiHOSGmzB5+RxDhoASEbDttOTHhDWtG1w8qhRRwCzOQ2Qg5zVFyqqSBMSBnf9rbEq9yHgiNcvuzPQ13TV34EYkI6zSQ5gvSgh+mzkIgsIzItauDiL/yGTjwSWCUKAF9bxfmnazLUdcmK2+Rz1mLywqNbdWBYdv8hcgvAWz1pkg= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--praan.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=n210P0bC; arc=none smtp.client-ip=209.85.210.198 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--praan.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="n210P0bC" Received: by mail-pf1-f198.google.com with SMTP id d2e1a72fcca58-84e3d575d6eso1525322b3a.3 for ; Fri, 14 Aug 2026 07:33:14 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20251104; t=1786717993; x=1787322793; darn=vger.kernel.org; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:from:to:cc:subject:date:message-id:reply-to :content-type; bh=HTq2Yo3MiZtOH+LV5j9JTfmNCmFQW03NdlsC4mRN3vI=; b=n210P0bC1sT+9vl5QbuDFZxoQQkS5rXH6iPKyBBPPE+a8ZZ6rlYtvd9vew8JK8QZBd N87ngX8C+ebspYbYOLKrUYRQl8qhAXAlIE2jPMxI/h1wbsFMGiTd9QrG2T47DEQmurOy wjphxTK+H9D5+Ae/9P/onCp2pIuLuAMFFnR8dLHnAc4Vn2uLi6lHILn1s511m+WKciaQ AWYNFRD34zjOui7Paf5SSCxsy/xDa1GCmuiWsM9Oa70YZzdMhMrXff3OR54YJrd29akW Fe6H4eilx4caj6aIK0mmcAjgE7zqMuq0/Ma7izDv21gdAzYR0SEXIfcNIaO0/XGyKGZU K1eg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1786717993; x=1787322793; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to:content-type; bh=HTq2Yo3MiZtOH+LV5j9JTfmNCmFQW03NdlsC4mRN3vI=; b=hzIMWYPKRHHfrjaqE3aIc7iFDxsZahc2wsanDkxeR7JKhX7NubOiEOXRkcYOrKLith wDOwuFs4v3quhV5yagJfhBezHZqLTGFluA1RWM/RybSa5P7mFolgSoE3nQpOnRii8c+/ nTLoRCKWblK/EoGLRDgIuzE+eTVLq4AJ+7KKpI9IflaL0Vvg+atadU/zo41eQvjwkR9R E5fS3D5DWaNJCbEPuCkEzBrKfBwVfQ8EwKFdpgefLuKTzMb6o3d8Ehr0yZlf4RHVUtuC YoJOStHM98LBOt6DwC4Gd6KaAa/plj9MeUOAQ6flSrp+4gjL/SwCj3U6RsSZ7nbyl2PI OJog== X-Forwarded-Encrypted: i=1; AHgh+RpgFsv5Ghhm7mCl356i4jSd2f+3aFm+iw/FvzXzn0PEOM1Rkr5ZNNL1QCb/S8r0YLdojJTNPHBP46fvMqw=@vger.kernel.org X-Gm-Message-State: AOJu0YxYAK8JtARS1qb3aLXP0TFFOYChDbOo+B2wzOkxg2ZUBChNOLEK 3uocpRBEg5PVFIFc+V2uE/IDufeZ9Wocg6rtmoyfeC9X9MGf2sPNtCzz9wXqVcbT6gTxRVJwgoj J+w== X-Received: from pffy16.prod.google.com ([2002:aa7:93d0:0:b0:847:b0e2:92a3]) (user=praan job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6a00:7487:b0:84f:dfce:966b with SMTP id d2e1a72fcca58-84fdfce99a7mr4924223b3a.8.1786717992732; Fri, 14 Aug 2026 07:33:12 -0700 (PDT) Date: Fri, 14 Aug 2026 14:32:54 +0000 In-Reply-To: <20260814143255.861084-1-praan@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20260814143255.861084-1-praan@google.com> X-Mailer: git-send-email 2.55.0.691.gc56d675ccc-goog Message-ID: <20260814143255.861084-5-praan@google.com> Subject: [PATCH v6 4/5] nfs: migrate direct I/O to iov_iter_extract_pages From: Pranjal Shrivastava To: linux-nfs@vger.kernel.org, Trond Myklebust , Anna Schumaker Cc: Chuck Lever , Jeff Layton , linux-kernel@vger.kernel.org, Christoph Hellwig , Logan Gunthorpe , Jason Gunthorpe , linux-pci@vger.kernel.org, linux-rdma@vger.kernel.org, Shivaji Kant , Pranjal Shrivastava Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Migrate the NFS Direct I/O path away from the legacy iov_iter_get_pages_alloc2() API to the modern iov_iter_extract_pages API. The transition aligns NFS with the modern VFS extraction model and serves as a preparatory step for supporting requirements such as page pinning via GUP for DMA. The migration fixes a bug in the Direct I/O loop where pages were being unpinned immediately after request creation. With the new extraction model, pins are held until the I/O is complete. Manual release in the loop is correspondingly updated to only clean up failed pages. Signed-off-by: Pranjal Shrivastava Reviewed-by: Shivaji Kant Reviewed-by: Christoph Hellwig --- fs/nfs/direct.c | 37 +++++++++++++++++++------------------ 1 file changed, 19 insertions(+), 18 deletions(-) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 96995736fac2..a1479fbf4f9e 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -168,14 +168,8 @@ int nfs_swap_rw(struct kiocb *iocb, struct iov_iter *i= ter) static void nfs_direct_release_pages(struct page **pages, unsigned int npa= ges, bool pinned) { - unsigned int i; - - if (pinned) { + if (pinned) unpin_user_pages(pages, npages); - } else { - for (i =3D 0; i < npages; i++) - put_page(pages[i]); - } } =20 void nfs_init_cinfo_from_dreq(struct nfs_commit_info *cinfo, @@ -354,16 +348,17 @@ static ssize_t nfs_direct_read_schedule_iovec(struct = nfs_direct_req *dreq, inode_dio_begin(inode); =20 while (iov_iter_count(iter)) { - struct page **pagevec; + struct page **pagevec =3D NULL; size_t bytes; size_t pgbase; unsigned npages, i; + bool pinned =3D iov_iter_extract_will_pin(iter); =20 - result =3D iov_iter_get_pages_alloc2(iter, &pagevec, - rsize, &pgbase); + result =3D iov_iter_extract_pages(iter, &pagevec, + rsize, ~0U, 0, &pgbase); if (result < 0) break; -=09 + bytes =3D result; npages =3D (result + pgbase + PAGE_SIZE - 1) / PAGE_SIZE; for (i =3D 0; i < npages; i++) { @@ -371,7 +366,7 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nf= s_direct_req *dreq, unsigned int req_len =3D min_t(size_t, bytes, PAGE_SIZE - pgbase); /* XXX do we need to do the eof zeroing found in async_filler? */ req =3D nfs_page_create_from_page(dreq->ctx, pagevec[i], - false, pgbase, pos, + pinned, pgbase, pos, req_len); if (IS_ERR(req)) { result =3D PTR_ERR(req); @@ -380,6 +375,7 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nf= s_direct_req *dreq, if (!nfs_pageio_add_request(&desc, req)) { result =3D desc.pg_error; nfs_release_request(req); + i++; break; } pgbase =3D 0; @@ -387,7 +383,8 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nf= s_direct_req *dreq, requested_bytes +=3D req_len; pos +=3D req_len; } - nfs_direct_release_pages(pagevec, npages, false); + if (i < npages) + nfs_direct_release_pages(pagevec + i, npages - i, pinned); kvfree(pagevec); if (result < 0) break; @@ -891,13 +888,14 @@ static ssize_t nfs_direct_write_schedule_iovec(struct= nfs_direct_req *dreq, =20 NFS_I(inode)->write_io +=3D iov_iter_count(iter); while (iov_iter_count(iter)) { - struct page **pagevec; + struct page **pagevec =3D NULL; size_t bytes; size_t pgbase; unsigned npages, i; + bool pinned =3D iov_iter_extract_will_pin(iter); =20 - result =3D iov_iter_get_pages_alloc2(iter, &pagevec, - wsize, &pgbase); + result =3D iov_iter_extract_pages(iter, &pagevec, + wsize, ~0U, 0, &pgbase); if (result < 0) break; =20 @@ -908,7 +906,7 @@ static ssize_t nfs_direct_write_schedule_iovec(struct n= fs_direct_req *dreq, unsigned int req_len =3D min_t(size_t, bytes, PAGE_SIZE - pgbase); =20 req =3D nfs_page_create_from_page(dreq->ctx, pagevec[i], - false, pgbase, pos, + pinned, pgbase, pos, req_len); if (IS_ERR(req)) { result =3D PTR_ERR(req); @@ -918,6 +916,7 @@ static ssize_t nfs_direct_write_schedule_iovec(struct n= fs_direct_req *dreq, if (desc.pg_error < 0) { nfs_free_request(req); result =3D desc.pg_error; + i++; break; } =20 @@ -939,6 +938,7 @@ static ssize_t nfs_direct_write_schedule_iovec(struct n= fs_direct_req *dreq, if (desc.pg_error < 0 && desc.pg_error !=3D -EAGAIN) { result =3D desc.pg_error; nfs_unlock_and_release_request(req); + i++; break; } =20 @@ -952,7 +952,8 @@ static ssize_t nfs_direct_write_schedule_iovec(struct n= fs_direct_req *dreq, desc.pg_error =3D 0; defer =3D true; } - nfs_direct_release_pages(pagevec, npages, false); + if (i < npages) + nfs_direct_release_pages(pagevec + i, npages - i, pinned); kvfree(pagevec); if (result < 0) break; --=20 2.55.0.691.gc56d675ccc-goog From nobody Mon Sep 28 23:55:25 2026 Received: from mail-pj1-f72.google.com (mail-pj1-f72.google.com [209.85.216.72]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 8256147A0B6 for ; Fri, 14 Aug 2026 14:33:18 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.72 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786718002; cv=none; b=bPyjQcut4+VxlhndiBZtaMJyCW7zmy81lJ3tUUxuQzqAacYf4h9bEPeMOQvnDVVbasdciE0H4MN46fBFj1jJzkjefx7MhXuEJu/Kv10dBQWAzx1+xusjW2n15OuncEfhVEowQ1gcsINA8KzIhQqPIJpGCqfulCD3o5blaVOotyU= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786718002; c=relaxed/simple; bh=TmXCqyPZhqD8tZHxAFnz34BnUbGwD3NFtE9dcxrrwvs=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=AH2m8ETYT69OCOfD5n2yyQbJxvrtR9pypywBmeciQa6h4hTfLTU/LB2uGWcvNPkG3e0caZ3b+egV3dpNWG3zjeTjHGWwdWM9Lvx2//as5q7uwTnsNqSMgcr1uDVIvHHemW637IxVn6D81OIfiVsIWai1OCKVvudH+xtKkNho1Sw= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--praan.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=oKq5Flzn; arc=none smtp.client-ip=209.85.216.72 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--praan.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="oKq5Flzn" Received: by mail-pj1-f72.google.com with SMTP id 98e67ed59e1d1-38e7b87ce77so2221597a91.0 for ; Fri, 14 Aug 2026 07:33:18 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20251104; t=1786717996; x=1787322796; darn=vger.kernel.org; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:from:to:cc:subject:date:message-id:reply-to :content-type; bh=0SgqAzIQu/0Bu39tyhAL9Ik5goBfz/9VI07g1fTVYzY=; b=oKq5FlzntHvK0Symg2Q7ZFAtfC/q3L57oSLdkLQgJc74r/meGKp/vDvu+wCuD1nwot FeAOWKlRmQoO/9c/RjWxQPvK/M9EfVDHHS7o+59U3V5t+ALLdm9ty/M/nrAACVFIXGWU x8slsx8YGGJhr6ZrwWYNSEcpoUinK2kPytxZ9+fkiUHcBvxZhQ+8s559H4GlSP7q2jOT iSk9XY5ulmLGfvfomNX5c7DihcZxY1P/T13YnhHjBeKRjrUKNlEu/1RS/Y/pPSkAhsLL h5jOLEEjmWENr/L6BW5JF/vcbmRZfFagpRsWpsSn1NPExXXfJb3xP5jyplvenwBKCKYE v19A== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1786717996; x=1787322796; h=content-type:cc:to:from:subject:message-id:references:mime-version :in-reply-to:date:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to:content-type; bh=0SgqAzIQu/0Bu39tyhAL9Ik5goBfz/9VI07g1fTVYzY=; b=qAhS3OXaX2M4HbS15sWn48EIk29tAe81TJDz9NfmQELzrQYhR81FXncrkW4GGC4ix3 yxDlxe/nMX/wfZfAb6DRlmzIG4pLAjDJWF7/v6GOZIO37gNLROGnhbp3hcgIvSZaIK9t sd/jcT5p43Vf2Ri0WYymNE7eKcfrT2M/qv2PpZoLgZn6WWqM2pT8I34w4tcOL8QxhA3S qIfgHTnI4MIR6jdP57QwGvHoUBFMMnConBAkvXzjp6VgGIzL3WSJRGoJ7uDmQMX2StFZ mEyrx/T26WMMoscRhvTlWGhik8+UV8h80NoluuERKV3Ku1GIvBN9oQegen1c4rvrI3fB jpeg== X-Forwarded-Encrypted: i=1; AHgh+Rod1S1syXF7Y2bXpaMDsjDWsPySrTFIrdmaSOV8p3mopOjDx6+fWM1XvwKF+gscogT2EIOfUNa56tk1r6U=@vger.kernel.org X-Gm-Message-State: AOJu0YxarXVrnZQiWxn0vzYkHtDCN8aQVxbSqyOCt6elG4akszn3qaLG wMivxAFkO9hPdNh+CMKvEnZhEqKbXIWgpWNxU2aksieVE1w2gu/Y7v/UrSUZNKoGzoMUkdiCCPX 1Rg== X-Received: from pjab15.prod.google.com ([2002:a17:90a:100f:b0:381:171e:961c]) (user=praan job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:2747:b0:38e:2f21:3bb5 with SMTP id 98e67ed59e1d1-3933baa0da9mr7909118a91.21.1786717995686; Fri, 14 Aug 2026 07:33:15 -0700 (PDT) Date: Fri, 14 Aug 2026 14:32:55 +0000 In-Reply-To: <20260814143255.861084-1-praan@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20260814143255.861084-1-praan@google.com> X-Mailer: git-send-email 2.55.0.691.gc56d675ccc-goog Message-ID: <20260814143255.861084-6-praan@google.com> Subject: [PATCH v6 5/5] nfs: introduce nfs_direct_extract_pages helper From: Pranjal Shrivastava To: linux-nfs@vger.kernel.org, Trond Myklebust , Anna Schumaker Cc: Chuck Lever , Jeff Layton , linux-kernel@vger.kernel.org, Christoph Hellwig , Logan Gunthorpe , Jason Gunthorpe , linux-pci@vger.kernel.org, linux-rdma@vger.kernel.org, Shivaji Kant , Pranjal Shrivastava Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Introduce nfs_direct_extract_pages() in direct.c to centralize page extraction and request creation for the Direct I/O path. The helper manages extraction from the iters and builds a list of nfs_page requests Refactor nfs_direct_read_schedule_iovec() and nfs_direct_write_schedule_iovec() to utilize the new helper, unifying the extraction logic on both paths. Signed-off-by: Pranjal Shrivastava Reviewed-by: Shivaji Kant Reviewed-by: Christoph Hellwig --- fs/nfs/direct.c | 130 ++++++++++++++++++++++++------------------------ 1 file changed, 64 insertions(+), 66 deletions(-) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index a1479fbf4f9e..76c954bbc9af 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -172,6 +172,50 @@ static void nfs_direct_release_pages(struct page **pag= es, unsigned int npages, unpin_user_pages(pages, npages); } =20 +static ssize_t nfs_direct_extract_pages(struct nfs_direct_req *dreq, + struct iov_iter *iter, + size_t size, loff_t *pos, + struct list_head *list) +{ + bool pinned =3D iov_iter_extract_will_pin(iter); + struct page **pagevec =3D NULL; + ssize_t result, bytes =3D 0; + int err =3D 0; + unsigned int npages, i; + size_t pgbase; + + result =3D iov_iter_extract_pages(iter, &pagevec, size, ~0U, 0, &pgbase); + if (result <=3D 0) + return result; + + npages =3D (result + pgbase + PAGE_SIZE - 1) >> PAGE_SHIFT; + for (i =3D 0; i < npages; i++) { + struct nfs_page *req; + unsigned int req_len =3D min_t(size_t, result - bytes, PAGE_SIZE - pgbas= e); + + req =3D nfs_page_create_from_page(dreq->ctx, pagevec[i], + pinned, pgbase, *pos, + req_len); + if (IS_ERR(req)) { + err =3D PTR_ERR(req); + break; + } + + list_add_tail(&req->wb_list, list); + pgbase =3D 0; + bytes +=3D req_len; + *pos +=3D req_len; + } + + if (i < npages) { + iov_iter_revert(iter, result - bytes); + nfs_direct_release_pages(pagevec + i, npages - i, pinned); + } + + kvfree(pagevec); + return bytes ? bytes : err; +} + void nfs_init_cinfo_from_dreq(struct nfs_commit_info *cinfo, struct nfs_direct_req *dreq) { @@ -340,6 +384,7 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nf= s_direct_req *dreq, ssize_t result =3D -EINVAL; size_t requested_bytes =3D 0; size_t rsize =3D max_t(size_t, NFS_SERVER(inode)->rsize, PAGE_SIZE); + LIST_HEAD(nfs_page_list); =20 nfs_pageio_init_read(&desc, dreq->inode, false, &nfs_direct_read_completion_ops); @@ -348,44 +393,23 @@ static ssize_t nfs_direct_read_schedule_iovec(struct = nfs_direct_req *dreq, inode_dio_begin(inode); =20 while (iov_iter_count(iter)) { - struct page **pagevec =3D NULL; - size_t bytes; - size_t pgbase; - unsigned npages, i; - bool pinned =3D iov_iter_extract_will_pin(iter); - - result =3D iov_iter_extract_pages(iter, &pagevec, - rsize, ~0U, 0, &pgbase); + result =3D nfs_direct_extract_pages(dreq, iter, rsize, &pos, &nfs_page_l= ist); if (result < 0) break; =20 - bytes =3D result; - npages =3D (result + pgbase + PAGE_SIZE - 1) / PAGE_SIZE; - for (i =3D 0; i < npages; i++) { - struct nfs_page *req; - unsigned int req_len =3D min_t(size_t, bytes, PAGE_SIZE - pgbase); - /* XXX do we need to do the eof zeroing found in async_filler? */ - req =3D nfs_page_create_from_page(dreq->ctx, pagevec[i], - pinned, pgbase, pos, - req_len); - if (IS_ERR(req)) { - result =3D PTR_ERR(req); - break; - } + while (!list_empty(&nfs_page_list)) { + struct nfs_page *req =3D nfs_list_entry(nfs_page_list.next); + size_t req_len =3D req->wb_bytes; + + nfs_list_remove_request(req); if (!nfs_pageio_add_request(&desc, req)) { result =3D desc.pg_error; nfs_release_request(req); - i++; + nfs_release_request_list(&nfs_page_list); break; } - pgbase =3D 0; - bytes -=3D req_len; requested_bytes +=3D req_len; - pos +=3D req_len; } - if (i < npages) - nfs_direct_release_pages(pagevec + i, npages - i, pinned); - kvfree(pagevec); if (result < 0) break; } @@ -876,6 +900,7 @@ static ssize_t nfs_direct_write_schedule_iovec(struct n= fs_direct_req *dreq, ssize_t result =3D 0; size_t requested_bytes =3D 0; size_t wsize =3D max_t(size_t, NFS_SERVER(inode)->wsize, PAGE_SIZE); + LIST_HEAD(nfs_page_list); bool defer =3D false; =20 trace_nfs_direct_write_schedule_iovec(dreq); @@ -888,57 +913,32 @@ static ssize_t nfs_direct_write_schedule_iovec(struct= nfs_direct_req *dreq, =20 NFS_I(inode)->write_io +=3D iov_iter_count(iter); while (iov_iter_count(iter)) { - struct page **pagevec =3D NULL; - size_t bytes; - size_t pgbase; - unsigned npages, i; - bool pinned =3D iov_iter_extract_will_pin(iter); - - result =3D iov_iter_extract_pages(iter, &pagevec, - wsize, ~0U, 0, &pgbase); + result =3D nfs_direct_extract_pages(dreq, iter, wsize, &pos, &nfs_page_l= ist); if (result < 0) break; =20 - bytes =3D result; - npages =3D (result + pgbase + PAGE_SIZE - 1) / PAGE_SIZE; - for (i =3D 0; i < npages; i++) { - struct nfs_page *req; - unsigned int req_len =3D min_t(size_t, bytes, PAGE_SIZE - pgbase); - - req =3D nfs_page_create_from_page(dreq->ctx, pagevec[i], - pinned, pgbase, pos, - req_len); - if (IS_ERR(req)) { - result =3D PTR_ERR(req); - break; - } - - if (desc.pg_error < 0) { - nfs_free_request(req); - result =3D desc.pg_error; - i++; - break; - } - - pgbase =3D 0; - bytes -=3D req_len; - requested_bytes +=3D req_len; - pos +=3D req_len; + while (!list_empty(&nfs_page_list)) { + struct nfs_page *req =3D nfs_list_entry(nfs_page_list.next); + size_t req_len =3D req->wb_bytes; =20 + nfs_list_remove_request(req); if (defer) { nfs_mark_request_commit(req, NULL, &cinfo, 0); + requested_bytes +=3D req_len; continue; } =20 nfs_lock_request(req); - if (nfs_pageio_add_request(&desc, req)) + if (nfs_pageio_add_request(&desc, req)) { + requested_bytes +=3D req_len; continue; + } =20 /* Exit on hard errors */ if (desc.pg_error < 0 && desc.pg_error !=3D -EAGAIN) { result =3D desc.pg_error; nfs_unlock_and_release_request(req); - i++; + nfs_release_request_list(&nfs_page_list); break; } =20 @@ -949,12 +949,10 @@ static ssize_t nfs_direct_write_schedule_iovec(struct= nfs_direct_req *dreq, spin_unlock(&dreq->lock); nfs_unlock_request(req); nfs_mark_request_commit(req, NULL, &cinfo, 0); + requested_bytes +=3D req_len; desc.pg_error =3D 0; defer =3D true; } - if (i < npages) - nfs_direct_release_pages(pagevec + i, npages - i, pinned); - kvfree(pagevec); if (result < 0) break; } --=20 2.55.0.691.gc56d675ccc-goog