From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 B10EB3CB578 for ; Wed, 9 Sep 2026 10:42:00 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1788950522; cv=none; b=fJ8ydh9UXuNZNLI20jmUo/JxzsSWp96mGdYKUBQfzlCrNFXpqMKFqGFGfL800wG/T6M2GZkk35o+OuybxHVT8ucI8YIsmZ2PS4IXYrAwDpztzP4XiJkuNcqprWOsVE4zMx8vI3nXrgyHPn8ekGsT1Bwa0v5x8/C4ujryG2VxnTc= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1788950522; c=relaxed/simple; bh=Go36hZXQxLa+oUcmd10dC+zzrj3xr68QqBjnywOPaLM=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=jFsrOUJ1JBmAsEw+rV9Tp+nlhYCj6z5mFKeBCD+1whBZaUssgSGuzvi6B56nrbgPyYygKLwhjdM7yWvemfRujo7xbQ9sDJr6s1h6dJbL4gczhpNLxk395kO/LhzXAfJ7h5xOnfe1GRlgSOWY0HEdbuPIYmRx22emIIITd2GjoVE= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=fjgVa1sT; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="fjgVa1sT" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 1A7191F00A3D; Wed, 9 Sep 2026 10:42:00 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1788950520; bh=vMBqJ3rymERsmjRjfu7njfLzzu37MScAoDr6rlHOp80=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=fjgVa1sTq40PtnCXUcVyvKeAYiAAD7amLrkEIrJFmrN98eVlzCHwV4GP130jmFM7B V8SPgdvzdHZH8f7ZME0kUzmVDxX+ASfOI+wVGbhpoIGq6Gcd/yxLzCdVNUZrn0TPJ+ FojKyqaj3aTIrL9V0Ws70esQFf4NYRESOHvDLQ76gj7dks8qnSYiMRMGxUkgME4nvt CsnnuasAXvRi57POAjG3ae9DHuy/qTkl4Vp3L7JQCJZpr3u95GrphQQrgoJAKLtedZ R6CjlIOw5e5/E9Xp6rQHnqt/ristb5/P6wTlCSJMbJiWONEvZaXI9zeE9SKiVnEcnb LM/dK3eBJ/bdA== From: Sasha Levin To: stable@vger.kernel.org Cc: Chuck Lever , Sasha Levin Subject: [PATCH 6.1.y 1/5] svcrdma: Update synopsis of svc_rdma_read_multiple_chunks() Date: Wed, 9 Sep 2026 06:41:54 -0400 Message-ID: <20260909104158.1554268-1-sashal@kernel.org> X-Mailer: git-send-email 2.53.0 In-Reply-To: <2026090310-onto-impotence-7944@gregkh> References: <2026090310-onto-impotence-7944@gregkh> Precedence: bulk X-Mailing-List: stable@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: 8bit From: Chuck Lever [ Upstream commit 740a3c895d94fc72494f2a4abf1ab31b6be482e8 ] Since the RDMA Read I/O state is now contained in the recv_ctxt, svc_rdma_read_multiple_chunks() can use that recv_ctxt to derive the read_info rather than the other way around. This removes another usage of the ri_readctxt field, enabling its removal in a subsequent patch. Signed-off-by: Chuck Lever [Stable 6.1 dependency adaptation] The stable tree still stores Read I/O state in svc_rdma_read_info and reconstructs rq_arg inside the Read builders. Fold the necessary context and synopsis conversions into this dependency so that upstream commit 3779b7b9e7d1 ("svcrdma: Validate Read chunk positions before reconstruction") can be cherry-picked without editing its patch. Move the existing chunk context and Read counters into svc_rdma_recv_ctxt, convert the existing Read helpers to rqstp/head arguments, and remove the now-unused read_info allocation and free helpers. Preserve the stable transport pointer, synchronous completion wait, and RDMA context cleanup. Save the decoded inline rq_arg before constructing Reads. Move the existing single-chunk and whole-message reconstruction into svc_rdma_process_read_list, after the builders return successfully, leaving the target's validation sites in the same form as upstream without adding any functions. Retain the upstream inline-copy page-offset correction and page-array bound, using rq_page_end for the latter on 6.1. Set the reconstructed whole-message length from rc_readbytes, which already includes copied inline bytes, so that those bytes are not counted twice. Stable-dep-of: 3779b7b9e7d1 ("svcrdma: Validate Read chunk positions before reconstruction") Signed-off-by: Sasha Levin --- include/linux/sunrpc/svc_rdma.h | 22 +++ net/sunrpc/xprtrdma/svc_rdma_rw.c | 318 ++++++++++++------------------ 2 files changed, 144 insertions(+), 196 deletions(-) diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h index fbc4bd423b355..536b95e28a8e2 100644 --- a/include/linux/sunrpc/svc_rdma.h +++ b/include/linux/sunrpc/svc_rdma.h @@ -126,6 +126,21 @@ enum { #define RPCSVC_MAXPAYLOAD_RDMA RPCSVC_MAXPAYLOAD +/* A chunk context tracks all I/O for moving one Read or Write + * chunk. This is a set of rdma_rw's that handle data movement + * for all segments of one chunk. + */ +struct svc_rdma_chunk_ctxt { + struct rpc_rdma_cid cc_cid; + struct ib_cqe cc_cqe; + struct svcxprt_rdma *cc_rdma; + struct list_head cc_rwctxts; + ktime_t cc_posttime; + int cc_sqecount; + enum ib_wc_status cc_status; + struct completion cc_done; +}; + struct svc_rdma_recv_ctxt { struct llist_node rc_node; struct list_head rc_list; @@ -141,6 +156,13 @@ struct svc_rdma_recv_ctxt { u32 rc_inv_rkey; __be32 rc_msgtype; + /* State for pulling Read chunks */ + unsigned int rc_pageoff; + unsigned int rc_curpage; + unsigned int rc_readbytes; + struct xdr_buf rc_saved_arg; + struct svc_rdma_chunk_ctxt rc_cc; + struct svc_rdma_pcl rc_call_pcl; struct svc_rdma_pcl rc_read_pcl; diff --git a/net/sunrpc/xprtrdma/svc_rdma_rw.c b/net/sunrpc/xprtrdma/svc_rdma_rw.c index ed581c183f7c8..f32635d0f5216 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_rw.c +++ b/net/sunrpc/xprtrdma/svc_rdma_rw.c @@ -142,25 +142,6 @@ static int svc_rdma_rw_ctx_init(struct svcxprt_rdma *rdma, return ret; } -/* A chunk context tracks all I/O for moving one Read or Write - * chunk. This is a set of rdma_rw's that handle data movement - * for all segments of one chunk. - * - * These are small, acquired with a single allocator call, and - * no more than one is needed per chunk. They are allocated on - * demand, and not cached. - */ -struct svc_rdma_chunk_ctxt { - struct rpc_rdma_cid cc_cid; - struct ib_cqe cc_cqe; - struct svcxprt_rdma *cc_rdma; - struct list_head cc_rwctxts; - ktime_t cc_posttime; - int cc_sqecount; - enum ib_wc_status cc_status; - struct completion cc_done; -}; - static void svc_rdma_cc_cid_init(struct svcxprt_rdma *rdma, struct rpc_rdma_cid *cid) { @@ -287,38 +268,6 @@ static void svc_rdma_write_done(struct ib_cq *cq, struct ib_wc *wc) svc_rdma_write_info_free(info); } -/* State for pulling a Read chunk. - */ -struct svc_rdma_read_info { - struct svc_rqst *ri_rqst; - struct svc_rdma_recv_ctxt *ri_readctxt; - unsigned int ri_pageno; - unsigned int ri_pageoff; - unsigned int ri_totalbytes; - - struct svc_rdma_chunk_ctxt ri_cc; -}; - -static struct svc_rdma_read_info * -svc_rdma_read_info_alloc(struct svcxprt_rdma *rdma) -{ - struct svc_rdma_read_info *info; - - info = kmalloc(sizeof(*info), GFP_KERNEL); - if (!info) - return info; - - svc_rdma_cc_init(rdma, &info->ri_cc); - info->ri_cc.cc_cqe.done = svc_rdma_wc_read_done; - return info; -} - -static void svc_rdma_read_info_free(struct svc_rdma_read_info *info) -{ - svc_rdma_cc_release(&info->ri_cc, DMA_FROM_DEVICE); - kfree(info); -} - /** * svc_rdma_wc_read_done - Handle completion of an RDMA Read ctx * @cq: controlling Completion Queue @@ -330,12 +279,12 @@ static void svc_rdma_wc_read_done(struct ib_cq *cq, struct ib_wc *wc) struct ib_cqe *cqe = wc->wr_cqe; struct svc_rdma_chunk_ctxt *cc = container_of(cqe, struct svc_rdma_chunk_ctxt, cc_cqe); - struct svc_rdma_read_info *info; + struct svc_rdma_recv_ctxt *ctxt; switch (wc->status) { case IB_WC_SUCCESS: - info = container_of(cc, struct svc_rdma_read_info, ri_cc); - trace_svcrdma_wc_read(wc, &cc->cc_cid, info->ri_totalbytes, + ctxt = container_of(cc, struct svc_rdma_recv_ctxt, rc_cc); + trace_svcrdma_wc_read(wc, &cc->cc_cid, ctxt->rc_readbytes, cc->cc_posttime); break; case IB_WC_WR_FLUSH_ERR: @@ -689,7 +638,8 @@ int svc_rdma_send_reply_chunk(struct svcxprt_rdma *rdma, /** * svc_rdma_build_read_segment - Build RDMA Read WQEs to pull one RDMA segment - * @info: context for ongoing I/O + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * @segment: co-ordinates of remote memory to be read * * Returns: @@ -698,19 +648,18 @@ int svc_rdma_send_reply_chunk(struct svcxprt_rdma *rdma, * %-ENOMEM: allocating a local resources failed * %-EIO: a DMA mapping error occurred */ -static int svc_rdma_build_read_segment(struct svc_rdma_read_info *info, +static int svc_rdma_build_read_segment(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head, const struct svc_rdma_segment *segment) { - struct svc_rdma_recv_ctxt *head = info->ri_readctxt; - struct svc_rdma_chunk_ctxt *cc = &info->ri_cc; - struct svc_rqst *rqstp = info->ri_rqst; + struct svc_rdma_chunk_ctxt *cc = &head->rc_cc; unsigned int sge_no, seg_len, len; struct svc_rdma_rw_ctxt *ctxt; struct scatterlist *sg; int ret; len = segment->rs_length; - sge_no = PAGE_ALIGN(info->ri_pageoff + len) >> PAGE_SHIFT; + sge_no = PAGE_ALIGN(head->rc_pageoff + len) >> PAGE_SHIFT; ctxt = svc_rdma_get_rw_ctxt(cc->cc_rdma, sge_no); if (!ctxt) return -ENOMEM; @@ -719,25 +668,25 @@ static int svc_rdma_build_read_segment(struct svc_rdma_read_info *info, sg = ctxt->rw_sg_table.sgl; for (sge_no = 0; sge_no < ctxt->rw_nents; sge_no++) { seg_len = min_t(unsigned int, len, - PAGE_SIZE - info->ri_pageoff); + PAGE_SIZE - head->rc_pageoff); - if (!info->ri_pageoff) + if (!head->rc_pageoff) head->rc_page_count++; - sg_set_page(sg, rqstp->rq_pages[info->ri_pageno], - seg_len, info->ri_pageoff); + sg_set_page(sg, rqstp->rq_pages[head->rc_curpage], + seg_len, head->rc_pageoff); sg = sg_next(sg); - info->ri_pageoff += seg_len; - if (info->ri_pageoff == PAGE_SIZE) { - info->ri_pageno++; - info->ri_pageoff = 0; + head->rc_pageoff += seg_len; + if (head->rc_pageoff == PAGE_SIZE) { + head->rc_curpage++; + head->rc_pageoff = 0; } len -= seg_len; /* Safety check */ if (len && - &rqstp->rq_pages[info->ri_pageno + 1] > rqstp->rq_page_end) + &rqstp->rq_pages[head->rc_curpage + 1] > rqstp->rq_page_end) goto out_overrun; } @@ -752,13 +701,14 @@ static int svc_rdma_build_read_segment(struct svc_rdma_read_info *info, return 0; out_overrun: - trace_svcrdma_page_overrun_err(cc->cc_rdma, rqstp, info->ri_pageno); + trace_svcrdma_page_overrun_err(cc->cc_rdma, rqstp, head->rc_curpage); return -EINVAL; } /** * svc_rdma_build_read_chunk - Build RDMA Read WQEs to pull one RDMA chunk - * @info: context for ongoing I/O + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * @chunk: Read chunk to pull * * Return values: @@ -767,7 +717,8 @@ static int svc_rdma_build_read_segment(struct svc_rdma_read_info *info, * %-ENOMEM: allocating a local resources failed * %-EIO: a DMA mapping error occurred */ -static int svc_rdma_build_read_chunk(struct svc_rdma_read_info *info, +static int svc_rdma_build_read_chunk(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head, const struct svc_rdma_chunk *chunk) { const struct svc_rdma_segment *segment; @@ -775,56 +726,59 @@ static int svc_rdma_build_read_chunk(struct svc_rdma_read_info *info, ret = -EINVAL; pcl_for_each_segment(segment, chunk) { - ret = svc_rdma_build_read_segment(info, segment); + ret = svc_rdma_build_read_segment(rqstp, head, segment); if (ret < 0) break; - info->ri_totalbytes += segment->rs_length; + head->rc_readbytes += segment->rs_length; } return ret; } /** * svc_rdma_copy_inline_range - Copy part of the inline content into pages - * @info: context for RDMA Reads + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * @offset: offset into the Receive buffer of region to copy * @remaining: length of region to copy * * Take a page at a time from rqstp->rq_pages and copy the inline * content from the Receive buffer into that page. Update - * info->ri_pageno and info->ri_pageoff so that the next RDMA Read + * head->rc_curpage and head->rc_pageoff so that the next RDMA Read * result will land contiguously with the copied content. * * Return values: * %0: Inline content was successfully copied * %-EINVAL: offset or length was incorrect */ -static int svc_rdma_copy_inline_range(struct svc_rdma_read_info *info, +static int svc_rdma_copy_inline_range(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head, unsigned int offset, unsigned int remaining) { - struct svc_rdma_recv_ctxt *head = info->ri_readctxt; unsigned char *dst, *src = head->rc_recv_buf; - struct svc_rqst *rqstp = info->ri_rqst; unsigned int page_no, numpages; - numpages = PAGE_ALIGN(info->ri_pageoff + remaining) >> PAGE_SHIFT; + numpages = PAGE_ALIGN(head->rc_pageoff + remaining) >> PAGE_SHIFT; for (page_no = 0; page_no < numpages; page_no++) { unsigned int page_len; + if (head->rc_curpage >= rqstp->rq_page_end - rqstp->rq_pages) + return -EINVAL; + page_len = min_t(unsigned int, remaining, - PAGE_SIZE - info->ri_pageoff); + PAGE_SIZE - head->rc_pageoff); - if (!info->ri_pageoff) + if (!head->rc_pageoff) head->rc_page_count++; - dst = page_address(rqstp->rq_pages[info->ri_pageno]); - memcpy(dst + info->ri_pageno, src + offset, page_len); + dst = page_address(rqstp->rq_pages[head->rc_curpage]); + memcpy(dst + head->rc_pageoff, src + offset, page_len); - info->ri_totalbytes += page_len; - info->ri_pageoff += page_len; - if (info->ri_pageoff == PAGE_SIZE) { - info->ri_pageno++; - info->ri_pageoff = 0; + head->rc_readbytes += page_len; + head->rc_pageoff += page_len; + if (head->rc_pageoff == PAGE_SIZE) { + head->rc_curpage++; + head->rc_pageoff = 0; } remaining -= page_len; offset += page_len; @@ -835,7 +789,8 @@ static int svc_rdma_copy_inline_range(struct svc_rdma_read_info *info, /** * svc_rdma_read_multiple_chunks - Construct RDMA Reads to pull data item Read chunks - * @info: context for RDMA Reads + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * * The chunk data lands in rqstp->rq_arg as a series of contiguous pages, * like an incoming TCP call. @@ -847,11 +802,11 @@ static int svc_rdma_copy_inline_range(struct svc_rdma_read_info *info, * %-ENOTCONN: posting failed (connection is lost), * %-EIO: rdma_rw initialization failed (DMA mapping, etc). */ -static noinline int svc_rdma_read_multiple_chunks(struct svc_rdma_read_info *info) +static noinline int +svc_rdma_read_multiple_chunks(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head) { - struct svc_rdma_recv_ctxt *head = info->ri_readctxt; const struct svc_rdma_pcl *pcl = &head->rc_read_pcl; - struct xdr_buf *buf = &info->ri_rqst->rq_arg; struct svc_rdma_chunk *chunk, *next; unsigned int start, length; int ret; @@ -859,12 +814,12 @@ static noinline int svc_rdma_read_multiple_chunks(struct svc_rdma_read_info *inf start = 0; chunk = pcl_first_chunk(pcl); length = chunk->ch_position; - ret = svc_rdma_copy_inline_range(info, start, length); + ret = svc_rdma_copy_inline_range(rqstp, head, start, length); if (ret < 0) return ret; pcl_for_each_chunk(chunk, pcl) { - ret = svc_rdma_build_read_chunk(info, chunk); + ret = svc_rdma_build_read_chunk(rqstp, head, chunk); if (ret < 0) return ret; @@ -873,31 +828,21 @@ static noinline int svc_rdma_read_multiple_chunks(struct svc_rdma_read_info *inf break; start += length; - length = next->ch_position - info->ri_totalbytes; - ret = svc_rdma_copy_inline_range(info, start, length); + length = next->ch_position - head->rc_readbytes; + ret = svc_rdma_copy_inline_range(rqstp, head, start, length); if (ret < 0) return ret; } start += length; length = head->rc_byte_len - start; - ret = svc_rdma_copy_inline_range(info, start, length); - if (ret < 0) - return ret; - - buf->len += info->ri_totalbytes; - buf->buflen += info->ri_totalbytes; - - buf->head[0].iov_base = page_address(info->ri_rqst->rq_pages[0]); - buf->head[0].iov_len = min_t(size_t, PAGE_SIZE, info->ri_totalbytes); - buf->pages = &info->ri_rqst->rq_pages[1]; - buf->page_len = info->ri_totalbytes - buf->head[0].iov_len; - return 0; + return svc_rdma_copy_inline_range(rqstp, head, start, length); } /** * svc_rdma_read_data_item - Construct RDMA Reads to pull data item Read chunks - * @info: context for RDMA Reads + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * * The chunk data lands in the page list of rqstp->rq_arg.pages. * @@ -912,50 +857,17 @@ static noinline int svc_rdma_read_multiple_chunks(struct svc_rdma_read_info *inf * %-ENOTCONN: posting failed (connection is lost), * %-EIO: rdma_rw initialization failed (DMA mapping, etc). */ -static int svc_rdma_read_data_item(struct svc_rdma_read_info *info) +static int svc_rdma_read_data_item(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head) { - struct svc_rdma_recv_ctxt *head = info->ri_readctxt; - struct xdr_buf *buf = &info->ri_rqst->rq_arg; - struct svc_rdma_chunk *chunk; - unsigned int length; - int ret; - - chunk = pcl_first_chunk(&head->rc_read_pcl); - ret = svc_rdma_build_read_chunk(info, chunk); - if (ret < 0) - goto out; - - /* Split the Receive buffer between the head and tail - * buffers at Read chunk's position. XDR roundup of the - * chunk is not included in either the pagelist or in - * the tail. - */ - buf->tail[0].iov_base = buf->head[0].iov_base + chunk->ch_position; - buf->tail[0].iov_len = buf->head[0].iov_len - chunk->ch_position; - buf->head[0].iov_len = chunk->ch_position; - - /* Read chunk may need XDR roundup (see RFC 8166, s. 3.4.5.2). - * - * If the client already rounded up the chunk length, the - * length does not change. Otherwise, the length of the page - * list is increased to include XDR round-up. - * - * Currently these chunks always start at page offset 0, - * thus the rounded-up length never crosses a page boundary. - */ - buf->pages = &info->ri_rqst->rq_pages[0]; - length = xdr_align_size(chunk->ch_length); - buf->page_len = length; - buf->len += length; - buf->buflen += length; - -out: - return ret; + return svc_rdma_build_read_chunk(rqstp, head, + pcl_first_chunk(&head->rc_read_pcl)); } /** - * svc_rdma_read_chunk_range - Build RDMA Read WQEs for portion of a chunk - * @info: context for RDMA Reads + * svc_rdma_read_chunk_range - Build RDMA Read WRs for portion of a chunk + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * @chunk: parsed Call chunk to pull * @offset: offset of region to pull * @length: length of region to pull @@ -967,7 +879,8 @@ static int svc_rdma_read_data_item(struct svc_rdma_read_info *info) * %-ENOTCONN: posting failed (connection is lost), * %-EIO: rdma_rw initialization failed (DMA mapping, etc). */ -static int svc_rdma_read_chunk_range(struct svc_rdma_read_info *info, +static int svc_rdma_read_chunk_range(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head, const struct svc_rdma_chunk *chunk, unsigned int offset, unsigned int length) { @@ -987,11 +900,11 @@ static int svc_rdma_read_chunk_range(struct svc_rdma_read_info *info, dummy.rs_length = min_t(u32, length, segment->rs_length) - offset; dummy.rs_offset = segment->rs_offset + offset; - ret = svc_rdma_build_read_segment(info, &dummy); + ret = svc_rdma_build_read_segment(rqstp, head, &dummy); if (ret < 0) break; - info->ri_totalbytes += dummy.rs_length; + head->rc_readbytes += dummy.rs_length; length -= dummy.rs_length; offset = 0; } @@ -1000,7 +913,8 @@ static int svc_rdma_read_chunk_range(struct svc_rdma_read_info *info, /** * svc_rdma_read_call_chunk - Build RDMA Read WQEs to pull a Long Message - * @info: context for RDMA Reads + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * * Return values: * %0: RDMA Read WQEs were successfully built @@ -1009,9 +923,9 @@ static int svc_rdma_read_chunk_range(struct svc_rdma_read_info *info, * %-ENOTCONN: posting failed (connection is lost), * %-EIO: rdma_rw initialization failed (DMA mapping, etc). */ -static int svc_rdma_read_call_chunk(struct svc_rdma_read_info *info) +static int svc_rdma_read_call_chunk(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head) { - struct svc_rdma_recv_ctxt *head = info->ri_readctxt; const struct svc_rdma_chunk *call_chunk = pcl_first_chunk(&head->rc_call_pcl); const struct svc_rdma_pcl *pcl = &head->rc_read_pcl; @@ -1020,17 +934,18 @@ static int svc_rdma_read_call_chunk(struct svc_rdma_read_info *info) int ret; if (pcl_is_empty(pcl)) - return svc_rdma_build_read_chunk(info, call_chunk); + return svc_rdma_build_read_chunk(rqstp, head, call_chunk); start = 0; chunk = pcl_first_chunk(pcl); length = chunk->ch_position; - ret = svc_rdma_read_chunk_range(info, call_chunk, start, length); + ret = svc_rdma_read_chunk_range(rqstp, head, call_chunk, + start, length); if (ret < 0) return ret; pcl_for_each_chunk(chunk, pcl) { - ret = svc_rdma_build_read_chunk(info, chunk); + ret = svc_rdma_build_read_chunk(rqstp, head, chunk); if (ret < 0) return ret; @@ -1039,8 +954,8 @@ static int svc_rdma_read_call_chunk(struct svc_rdma_read_info *info) break; start += length; - length = next->ch_position - info->ri_totalbytes; - ret = svc_rdma_read_chunk_range(info, call_chunk, + length = next->ch_position - head->rc_readbytes; + ret = svc_rdma_read_chunk_range(rqstp, head, call_chunk, start, length); if (ret < 0) return ret; @@ -1048,12 +963,14 @@ static int svc_rdma_read_call_chunk(struct svc_rdma_read_info *info) start += length; length = call_chunk->ch_length - start; - return svc_rdma_read_chunk_range(info, call_chunk, start, length); + return svc_rdma_read_chunk_range(rqstp, head, call_chunk, + start, length); } /** * svc_rdma_read_special - Build RDMA Read WQEs to pull a Long Message - * @info: context for RDMA Reads + * @rqstp: RPC transaction context + * @head: context for ongoing I/O * * The start of the data lands in the first page just after the * Transport header, and the rest lands in rqstp->rq_arg.pages. @@ -1069,25 +986,10 @@ static int svc_rdma_read_call_chunk(struct svc_rdma_read_info *info) * %-ENOTCONN: posting failed (connection is lost), * %-EIO: rdma_rw initialization failed (DMA mapping, etc). */ -static noinline int svc_rdma_read_special(struct svc_rdma_read_info *info) +static noinline int svc_rdma_read_special(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *head) { - struct xdr_buf *buf = &info->ri_rqst->rq_arg; - int ret; - - ret = svc_rdma_read_call_chunk(info); - if (ret < 0) - goto out; - - buf->len += info->ri_totalbytes; - buf->buflen += info->ri_totalbytes; - - buf->head[0].iov_base = page_address(info->ri_rqst->rq_pages[0]); - buf->head[0].iov_len = min_t(size_t, PAGE_SIZE, info->ri_totalbytes); - buf->pages = &info->ri_rqst->rq_pages[1]; - buf->page_len = info->ri_totalbytes - buf->head[0].iov_len; - -out: - return ret; + return svc_rdma_read_call_chunk(rqstp, head); } /** @@ -1117,30 +1019,54 @@ int svc_rdma_process_read_list(struct svcxprt_rdma *rdma, struct svc_rqst *rqstp, struct svc_rdma_recv_ctxt *head) { - struct svc_rdma_read_info *info; - struct svc_rdma_chunk_ctxt *cc; + struct svc_rdma_chunk_ctxt *cc = &head->rc_cc; + struct xdr_buf *buf = &rqstp->rq_arg; int ret; - info = svc_rdma_read_info_alloc(rdma); - if (!info) - return -ENOMEM; - cc = &info->ri_cc; - info->ri_rqst = rqstp; - info->ri_readctxt = head; - info->ri_pageno = 0; - info->ri_pageoff = 0; - info->ri_totalbytes = 0; + svc_rdma_cc_init(rdma, cc); + cc->cc_cqe.done = svc_rdma_wc_read_done; + head->rc_curpage = 0; + head->rc_pageoff = 0; + head->rc_readbytes = 0; + /* Preserve the decoded inline body before reconstructing rq_arg. */ + head->rc_saved_arg = *buf; if (pcl_is_empty(&head->rc_call_pcl)) { if (head->rc_read_pcl.cl_count == 1) - ret = svc_rdma_read_data_item(info); + ret = svc_rdma_read_data_item(rqstp, head); else - ret = svc_rdma_read_multiple_chunks(info); + ret = svc_rdma_read_multiple_chunks(rqstp, head); } else - ret = svc_rdma_read_special(info); + ret = svc_rdma_read_special(rqstp, head); if (ret < 0) goto out_err; + /* All Read chunks have been constructed; now assemble rq_arg. */ + if (pcl_is_empty(&head->rc_call_pcl) && + head->rc_read_pcl.cl_count == 1) { + struct svc_rdma_chunk *chunk = pcl_first_chunk(&head->rc_read_pcl); + unsigned int length = xdr_align_size(chunk->ch_length); + + /* Split the inline body at the Read chunk's position. */ + buf->tail[0].iov_base = buf->head[0].iov_base + chunk->ch_position; + buf->tail[0].iov_len = buf->head[0].iov_len - chunk->ch_position; + buf->head[0].iov_len = chunk->ch_position; + + /* Include XDR roundup of the Read chunk in the page list. */ + buf->pages = &rqstp->rq_pages[0]; + buf->page_len = length; + buf->len += length; + buf->buflen += length; + } else { + /* rc_readbytes includes both inline content and Read chunks. */ + buf->len = head->rc_readbytes; + buf->buflen = head->rc_readbytes; + buf->head[0].iov_base = page_address(rqstp->rq_pages[0]); + buf->head[0].iov_len = min_t(size_t, PAGE_SIZE, head->rc_readbytes); + buf->pages = &rqstp->rq_pages[1]; + buf->page_len = head->rc_readbytes - buf->head[0].iov_len; + } + trace_svcrdma_post_read_chunk(&cc->cc_cid, cc->cc_sqecount); init_completion(&cc->cc_done); ret = svc_rdma_post_chunk_ctxt(cc); @@ -1160,6 +1086,6 @@ int svc_rdma_process_read_list(struct svcxprt_rdma *rdma, head->rc_page_count = 0; out_err: - svc_rdma_read_info_free(info); + svc_rdma_cc_release(cc, DMA_FROM_DEVICE); return ret; } -- 2.53.0