From nobody Mon Apr 6 18:55:59 2026 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=quarantine dis=none) header.from=redhat.com ARC-Seal: i=1; a=rsa-sha256; t=1773847993; cv=none; d=zohomail.com; s=zohoarc; b=NnA0Madako7UqQryLUmnCrbkse04eai4nqrwSXDax/Ge1oAqFCVV+O5uYdSres0lPEMJXjhGvqrHV9Q557JFmgUjxc67Wo6d9O4xHrSrRAKP8UjzNs35/Cpv3PuN6sxAY4S+uPAjkFxOzMaH0aDa9lxkE00ma8jMw6XUFNv7aBY= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1773847993; h=Content-Type:Content-Transfer-Encoding:Cc:Cc:Date:Date:From:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:Subject:To:To:Message-Id:Reply-To; bh=V2/+0Og3S4wOSV0lrvsiMpNrEkJuUBoXaBIw/93KO9Q=; b=MQKpo34ibRI8PSlTDEG8Rveyv9PQpLP24RPKd9kkDjXkIKT6JDSEdnwzclKwqI2S4UsJ7xoQwU25ViL1dSpvwT/rcedb+D8iaanixoO5t+hUK3BQVxThc/NzjlFjPNqrEWyGuj2ijjewF3OOon3Wgga/E0AhDtm+d2kUH/kcciE= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=quarantine dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1773847993519253.19548665531795; Wed, 18 Mar 2026 08:33:13 -0700 (PDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1w2ssw-0002ZF-IS; Wed, 18 Mar 2026 11:32:38 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1w2sst-0002Y3-OU for qemu-devel@nongnu.org; Wed, 18 Mar 2026 11:32:35 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.133.124]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1w2ssr-0000Q8-Mo for qemu-devel@nongnu.org; Wed, 18 Mar 2026 11:32:35 -0400 Received: from mail-wr1-f71.google.com (mail-wr1-f71.google.com [209.85.221.71]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-618-4vVGwpL8OWaD-VSdYdDgNg-1; Wed, 18 Mar 2026 11:32:25 -0400 Received: by mail-wr1-f71.google.com with SMTP id ffacd0b85a97d-43b3c9568a3so337669f8f.0 for ; Wed, 18 Mar 2026 08:32:24 -0700 (PDT) Received: from localhost (p200300cfd737d03304b9bf783d26f60b.dip0.t-ipconnect.de. [2003:cf:d737:d033:4b9:bf78:3d26:f60b]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-43b518522d7sm7490639f8f.13.2026.03.18.08.32.21 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 18 Mar 2026 08:32:21 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1773847952; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=V2/+0Og3S4wOSV0lrvsiMpNrEkJuUBoXaBIw/93KO9Q=; b=enRTSXT3lD/ty5i+iWEFnRDw9wfbpZe5qUbWADrJgYg5/hmjh7PIZb08x6M9D4LD2nHAwf iMokqC/E4cOzRxgXNPMiWE6qn/qEQTBoQBmdKAVJ0/KnB4NK9yq0HUB10oni7xqKLR+q2N RvCdqdpAldynnfTQ22A8HF+0qXtXI5k= X-MC-Unique: 4vVGwpL8OWaD-VSdYdDgNg-1 X-Mimecast-MFC-AGG-ID: 4vVGwpL8OWaD-VSdYdDgNg_1773847944 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=google; t=1773847944; x=1774452744; darn=nongnu.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=V2/+0Og3S4wOSV0lrvsiMpNrEkJuUBoXaBIw/93KO9Q=; b=rc5mBJ21dTQTwJma7ohMFRiylnJJ+aKypbpMSzRLFIqu+tgxa0Bko4L49i4KnwwHid gfciNtbCAMOLZSjBU2wa1NGzmJhUMKHOMYlBD3rKoQZcZ+eMuFpmZnJ1tt8/jHrIz1WY KcKRRlCECzXRHq14uaMhOJ7/lua8ZeerAApMcxhdUhGS87pvCIYeqW2oOc8bnYiFRPDv wkNwFdal6I32IEWjxt6DjTSUfcRCaTURes+PQ3Qd7oyS+nU2zZB/yY21alpZ4KS1lOEq 1Yc2fNj/vQcK/nJOMqdO6x5zfbpla0FDOl/gyfIhANRHLC4hIuVOqMNHoLZy+K0LOXDk 9thw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1773847944; x=1774452744; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-gg:x-gm-message-state:from :to:cc:subject:date:message-id:reply-to; bh=V2/+0Og3S4wOSV0lrvsiMpNrEkJuUBoXaBIw/93KO9Q=; b=J43tHJm4DBxB2+U1eMAqljLui/GMYxL0KZe6DFTsN8nWhZ1OCY2NmG5rB342zJTPno 3TbTAnlqohxHcI7tKTZLG9ee2ifO5L6wFTT/MoSIBvQtxeYjKxhEKUbXryYvHu+v6ROJ 5HnDjUmGLWvZD1Mio21cdSmxjJ3I2Tyr+4HVXTdlQy81lGAI1J2l0fDBD159A0M4LAMr 3Ik/otaT9xleQ2pcs/OScnCyDFM38306hzPrUGtzDTN8u4kXNnKijpnY1P2ISMp22VFI kTC8yQu5khR4Px63zOCSjsttmR59/IGBSlxkXcCu9wR+IMjfR3wX+9k6rkElQmAOmh7O NsUA== X-Gm-Message-State: AOJu0Yy7600U4Yah15WrqRdDV8qePTD1rTFjl5H/VLNG1Eg7kmnbrrgA vctxlnUqfOtNIVG4Rt3IYwgi2FuAV49ZoX7RgFeQmdgC7pcazsvBMlTlXmkqTT6ziM82kAWr94k DlALlD26ji6LRVmUMkwQCQU9rJYfx91g4SM2uolfm6vVa3ttAJ8/RPQtF X-Gm-Gg: ATEYQzw3C7f6C5ZWQ2vWAlXIB8t4pHU7Mua2QYx75s0AjOPqsm5YPvJnW0CjLQ8N4TE +9hEA9NJJIfnhmQ/au8TDu+ieqpzQ+NwiQaFmTrqCCYSDNJH4kk3LL7PAg3HZR4GJPxYy+tzSMC 65hY4F3DSFpf/eBVhb0wFM7ow8Tkg+RvPiValeSuJ+V7o5Ua4J8PEg9ZurMC6aPgpREYSsPH+F+ 4fB4rwFE/lRVLcNQjAcuy/07j62w7j72e5EzSUtMKRE3sCcKoIBTFMkPErHjgijvXJ4SrimMYaV wNJdAR3bJ6Iy4S9hgwdnxCiloSRfFvducNVm2Q2/vWQw28ea2l9ZYDYr8NaOXQABAxiMtf+xrBf 3KepFQkBGPltIWJj61OGO0McSV6CtTK/RJqqm7/ZzjBNNNWeyDtSEh3Cg7kYJiiD8XzprG1qyN0 94jfM= X-Received: by 2002:a05:6000:4387:b0:439:b744:c601 with SMTP id ffacd0b85a97d-43b527c55cfmr6073007f8f.40.1773847943459; Wed, 18 Mar 2026 08:32:23 -0700 (PDT) X-Received: by 2002:a05:6000:4387:b0:439:b744:c601 with SMTP id ffacd0b85a97d-43b527c55cfmr6072937f8f.40.1773847942840; Wed, 18 Mar 2026 08:32:22 -0700 (PDT) From: Hanna Czenczek To: qemu-block@nongnu.org Cc: qemu-devel@nongnu.org, Hanna Czenczek , Kevin Wolf , Julia Suvorova , Aarushi Mehta , Stefan Hajnoczi , Stefano Garzarella Subject: [PATCH for-11.0 3/3] io-uring: Resubmit tails of short writes Date: Wed, 18 Mar 2026 16:32:05 +0100 Message-ID: <20260318153206.171494-4-hreitz@redhat.com> X-Mailer: git-send-email 2.53.0 In-Reply-To: <20260318153206.171494-1-hreitz@redhat.com> References: <20260318153206.171494-1-hreitz@redhat.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=170.10.133.124; envelope-from=hreitz@redhat.com; helo=us-smtp-delivery-124.mimecast.com X-Spam_score_int: -3 X-Spam_score: -0.4 X-Spam_bar: / X-Spam_report: (-0.4 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H5=0.001, RCVD_IN_MSPIKE_WL=0.001, RCVD_IN_VALIDITY_RPBL_BLOCKED=0.819, RCVD_IN_VALIDITY_SAFE_BLOCKED=0.903, SPF_HELO_PASS=-0.001, SPF_PASS=-0.001 autolearn=no autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: qemu development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: qemu-devel-bounces+importer=patchew.org@nongnu.org X-ZohoMail-DKIM: pass (identity @redhat.com) X-ZM-MESSAGEID: 1773847996174154100 Content-Type: text/plain; charset="utf-8" Short writes can happen, too, not just short reads. The difference to aio=3Dnative is that the kernel will actually retry the tail of short requests internally already -- so it is harder to reproduce. But if the tail of a short request returns an error to the kernel, we will see it in userspace still. To reproduce this, apply the following patch on top of the one shown in HEAD^ (again %s/escaped // to apply): escaped diff --git a/block/export/fuse.c b/block/export/fuse.c escaped index 67dc50a412..2b98489a32 100644 escaped --- a/block/export/fuse.c escaped +++ b/block/export/fuse.c @@ -1059,8 +1059,15 @@ fuse_co_read(FuseExport *exp, void **bufptr, uint64_= t offset, uint32_t size) int64_t blk_len; void *buf; int ret; + static uint32_t error_size; - size =3D MIN(size, 4096); + if (error_size =3D=3D size) { + error_size =3D 0; + return -EIO; + } else if (size > 4096) { + error_size =3D size - 4096; + size =3D 4096; + } /* Limited by max_read, should not happen */ if (size > FUSE_MAX_READ_BYTES) { @@ -1111,8 +1118,15 @@ fuse_co_write(FuseExport *exp, struct fuse_write_out= *out, { int64_t blk_len; int ret; + static uint32_t error_size; - size =3D MIN(size, 4096); + if (error_size =3D=3D size) { + error_size =3D 0; + return -EIO; + } else if (size > 4096) { + error_size =3D size - 4096; + size =3D 4096; + } QEMU_BUILD_BUG_ON(FUSE_MAX_WRITE_BYTES > BDRV_REQUEST_MAX_BYTES); /* Limited by max_write, should not happen */ I know this is a bit artificial because to produce this, there must be an I/O error somewhere anyway, but if it does happen, qemu will understand it to mean ENOSPC for short writes, which is incorrect. So I believe we need to resubmit the tail to maybe have it succeed now, or at least get the correct error code. Reproducer as before: $ ./qemu-img create -f raw test.raw 8k Formatting 'test.raw', fmt=3Draw size=3D8192 $ ./qemu-io -f raw -c 'write -P 42 0 8k' test.raw wrote 8192/8192 bytes at offset 0 8 KiB, 1 ops; 00.00 sec (64.804 MiB/sec and 8294.9003 ops/sec) $ hexdump -C test.raw 00000000 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a |**************= **| * 00002000 $ storage-daemon/qemu-storage-daemon \ --blockdev file,node-name=3Dtest,filename=3Dtest.raw \ --export fuse,id=3Dexp,node-name=3Dtest,mountpoint=3Dtest.raw,writable= =3Dtrue $ ./qemu-io --image-opts -c 'read -P 23 0 8k' \ driver=3Dfile,filename=3Dtest.raw,cache.direct=3Don,aio=3Dio_uring read 8192/8192 bytes at offset 0 8 KiB, 1 ops; 00.00 sec (58.481 MiB/sec and 7485.5342 ops/sec) $ ./qemu-io --image-opts -c 'write -P 23 0 8k' \ driver=3Dfile,filename=3Dtest.raw,cache.direct=3Don,aio=3Dio_uring write failed: No space left on device $ hexdump -C test.raw 00000000 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 |..............= ..| * 00001000 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a 2a |**************= **| * 00002000 So short reads already work (because there is code for that), but short writes incorrectly produce ENOSPC. This patch fixes that by resubmitting not only the tail of short reads but short writes also. Signed-off-by: Hanna Czenczek Reviewed-by: Kevin Wolf --- block/io_uring.c | 83 ++++++++++++++++++++++++++-------------------- block/trace-events | 2 +- 2 files changed, 48 insertions(+), 37 deletions(-) diff --git a/block/io_uring.c b/block/io_uring.c index cb131d3b8b..61b54647ae 100644 --- a/block/io_uring.c +++ b/block/io_uring.c @@ -27,10 +27,10 @@ typedef struct { BdrvRequestFlags flags; =20 /* - * Buffered reads may require resubmission, see - * luring_resubmit_short_read(). + * Short reads/writes require resubmission, see + * luring_resubmit_short_io(). */ - int total_read; + int total_done; QEMUIOVector resubmit_qiov; =20 CqeHandler cqe_handler; @@ -44,6 +44,10 @@ static void luring_prep_sqe(struct io_uring_sqe *sqe, vo= id *opaque) int fd =3D req->fd; BdrvRequestFlags flags =3D req->flags; =20 + if (req->resubmit_qiov.iov !=3D NULL) { + qiov =3D &req->resubmit_qiov; + } + switch (req->type) { case QEMU_AIO_WRITE: { @@ -51,7 +55,8 @@ static void luring_prep_sqe(struct io_uring_sqe *sqe, voi= d *opaque) if (luring_flags !=3D 0 || qiov->niov > 1) { #ifdef HAVE_IO_URING_PREP_WRITEV2 io_uring_prep_writev2(sqe, fd, qiov->iov, - qiov->niov, offset, luring_flags); + qiov->niov, offset + req->total_done, + luring_flags); #else /* * FUA should only be enabled with HAVE_IO_URING_PREP_WRITEV2,= see @@ -59,12 +64,14 @@ static void luring_prep_sqe(struct io_uring_sqe *sqe, v= oid *opaque) */ assert(luring_flags =3D=3D 0); =20 - io_uring_prep_writev(sqe, fd, qiov->iov, qiov->niov, offset); + io_uring_prep_writev(sqe, fd, qiov->iov, qiov->niov, + offset + req->total_done); #endif } else { /* The man page says non-vectored is faster than vectored */ struct iovec *iov =3D qiov->iov; - io_uring_prep_write(sqe, fd, iov->iov_base, iov->iov_len, offs= et); + io_uring_prep_write(sqe, fd, iov->iov_base, iov->iov_len, + offset + req->total_done); } break; } @@ -73,17 +80,14 @@ static void luring_prep_sqe(struct io_uring_sqe *sqe, v= oid *opaque) break; case QEMU_AIO_READ: { - if (req->resubmit_qiov.iov !=3D NULL) { - qiov =3D &req->resubmit_qiov; - } if (qiov->niov > 1) { io_uring_prep_readv(sqe, fd, qiov->iov, qiov->niov, - offset + req->total_read); + offset + req->total_done); } else { /* The man page says non-vectored is faster than vectored */ struct iovec *iov =3D qiov->iov; io_uring_prep_read(sqe, fd, iov->iov_base, iov->iov_len, - offset + req->total_read); + offset + req->total_done); } break; } @@ -98,21 +102,26 @@ static void luring_prep_sqe(struct io_uring_sqe *sqe, = void *opaque) } =20 /** - * luring_resubmit_short_read: + * luring_resubmit_short_io: * - * Short reads are rare but may occur. The remaining read request needs to= be - * resubmitted. + * Short reads and writes are rare but may occur. The remaining request n= eeds + * to be resubmitted. + * + * For example, short reads can be reproduced by a FUSE export deliberately + * executing short reads. The tail of short writes is generally resubmitt= ed by + * io-uring in the kernel, but if that resubmission encounters an I/O erro= r, the + * already submitted portion will be returned as a short write. */ -static void luring_resubmit_short_read(LuringRequest *req, int nread) +static void luring_resubmit_short_io(LuringRequest *req, int ndone) { QEMUIOVector *resubmit_qiov; size_t remaining; =20 - trace_luring_resubmit_short_read(req, nread); + trace_luring_resubmit_short_io(req, ndone); =20 - /* Update read position */ - req->total_read +=3D nread; - remaining =3D req->qiov->size - req->total_read; + /* Update I/O position */ + req->total_done +=3D ndone; + remaining =3D req->qiov->size - req->total_done; =20 /* Shorten qiov */ resubmit_qiov =3D &req->resubmit_qiov; @@ -121,7 +130,7 @@ static void luring_resubmit_short_read(LuringRequest *r= eq, int nread) } else { qemu_iovec_reset(resubmit_qiov); } - qemu_iovec_concat(resubmit_qiov, req->qiov, req->total_read, remaining= ); + qemu_iovec_concat(resubmit_qiov, req->qiov, req->total_done, remaining= ); =20 aio_add_sqe(luring_prep_sqe, req, &req->cqe_handler); } @@ -153,26 +162,28 @@ static void luring_cqe_handler(CqeHandler *cqe_handle= r) return; } } else if (req->qiov) { - /* total_read is non-zero only for resubmitted read requests */ - int total_bytes =3D ret + req->total_read; + /* total_done is non-zero only for resubmitted requests */ + int total_bytes =3D ret + req->total_done; =20 if (total_bytes =3D=3D req->qiov->size) { ret =3D 0; - } else { + } else if (ret > 0 && (req->type =3D=3D QEMU_AIO_READ || + req->type =3D=3D QEMU_AIO_WRITE)) { /* Short Read/Write */ - if (req->type =3D=3D QEMU_AIO_READ) { - if (ret > 0) { - luring_resubmit_short_read(req, ret); - return; - } - - /* Pad with zeroes */ - qemu_iovec_memset(req->qiov, total_bytes, 0, - req->qiov->size - total_bytes); - ret =3D 0; - } else { - ret =3D -ENOSPC; - } + luring_resubmit_short_io(req, ret); + return; + } else if (req->type =3D=3D QEMU_AIO_READ) { + /* Read ret =3D=3D 0: EOF, pad with zeroes */ + qemu_iovec_memset(req->qiov, total_bytes, 0, + req->qiov->size - total_bytes); + ret =3D 0; + } else { + /* + * Normal write ret =3D=3D 0 means ENOSPC. + * For zone-append, we treat any 0 <=3D ret < qiov->size as EN= OSPC, + * too, because resubmitting the tail seems a little unsafe. + */ + ret =3D -ENOSPC; } } =20 diff --git a/block/trace-events b/block/trace-events index d170fc96f1..950c82d4b8 100644 --- a/block/trace-events +++ b/block/trace-events @@ -64,7 +64,7 @@ file_paio_submit(void *acb, void *opaque, int64_t offset,= int count, int type) " # io_uring.c luring_cqe_handler(void *req, int ret) "req %p ret %d" luring_co_submit(void *bs, void *req, int fd, uint64_t offset, size_t nbyt= es, int type) "bs %p req %p fd %d offset %" PRId64 " nbytes %zd type %d" -luring_resubmit_short_read(void *req, int nread) "req %p nread %d" +luring_resubmit_short_io(void *req, int ndone) "req %p ndone %d" =20 # qcow2.c qcow2_add_task(void *co, void *bs, void *pool, const char *action, int clu= ster_type, uint64_t host_offset, uint64_t offset, uint64_t bytes, void *qio= v, size_t qiov_offset) "co %p bs %p pool %p: %s: cluster_type %d file_clust= er_offset %" PRIu64 " offset %" PRIu64 " bytes %" PRIu64 " qiov %p qiov_off= set %zu" --=20 2.53.0