From patchwork Fri Nov 15 20:57:05 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Vivek Goyal X-Patchwork-Id: 11247113 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 1CEE76C1 for ; Fri, 15 Nov 2019 20:57:39 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id E9D2E2073A for ; Fri, 15 Nov 2019 20:57:38 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="J5eTPN9H" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727078AbfKOU5d (ORCPT ); Fri, 15 Nov 2019 15:57:33 -0500 Received: from us-smtp-delivery-1.mimecast.com ([205.139.110.120]:45062 "EHLO us-smtp-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1726980AbfKOU5a (ORCPT ); Fri, 15 Nov 2019 15:57:30 -0500 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1573851448; 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=VY0uoXYg3UA2Mb0lTBtQ4l9rtvW+i7vfhQnMJFuX8IM=; b=J5eTPN9HJzlJOfKqSovo+y861eIH4YznTCt7rnSg7Sz4l5nFqbfGptUxhOlPA3M0fVaHBT P4W8PX0T7UgJGEV0IMOKeLJA7HvdqX7uSI1t55R55lzOMoV0Sy6Kz41hPxvSu5QWP11/Oq aTo1vT0jFR7Oi5MhMFcVPUT+XicZtVY= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-244-4aQALp9CONW1FVP-2qIZJg-1; Fri, 15 Nov 2019 15:57:22 -0500 Received: from smtp.corp.redhat.com (int-mx04.intmail.prod.int.phx2.redhat.com [10.5.11.14]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 680E418B9FAB; Fri, 15 Nov 2019 20:57:21 +0000 (UTC) Received: from horse.redhat.com (unknown [10.18.25.35]) by smtp.corp.redhat.com (Postfix) with ESMTP id 291CD5E263; Fri, 15 Nov 2019 20:57:15 +0000 (UTC) Received: by horse.redhat.com (Postfix, from userid 10451) id B5A6B224778; Fri, 15 Nov 2019 15:57:14 -0500 (EST) From: Vivek Goyal To: virtio-fs@redhat.com, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org Cc: vgoyal@redhat.com, stefanha@redhat.com, dgilbert@redhat.com, miklos@szeredi.hu Subject: [PATCH 4/4] virtiofs: Support blocking posix locks (fcntl(F_SETLKW)) Date: Fri, 15 Nov 2019 15:57:05 -0500 Message-Id: <20191115205705.2046-5-vgoyal@redhat.com> In-Reply-To: <20191115205705.2046-1-vgoyal@redhat.com> References: <20191115205705.2046-1-vgoyal@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.14 X-MC-Unique: 4aQALp9CONW1FVP-2qIZJg-1 X-Mimecast-Spam-Score: 0 Sender: linux-fsdevel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-fsdevel@vger.kernel.org As of now we don't support blocking variant of posix locks and daemon returns -EOPNOTSUPP. Reason being that it can lead to deadlocks. Virtqueue size is limited and it is possible we fill virtqueue with all the requests of fcntl(F_SETLKW) and wait for reply. And later a subsequent unlock request can't make progress because virtqueue is full. And that means F_SETLKW can't make progress and we are deadlocked. Use notification queue to solve this problem. After submitting lock request device will send a reply asking requester to wait. Once lock is available, requester will get a notification saying locking is available. That way we don't keep the request virtueue busy while we are waiting for lock and further unlock requests can make progress. When we get a reply in response to lock request, we need a way to know if we need to wait for notification or not. I have overloaded the fuse_out_header->error field. If value is ->error is 1, that's a signal to caller to wait for lock notification. Signed-off-by: Vivek Goyal --- fs/fuse/virtio_fs.c | 78 ++++++++++++++++++++++++++++++++++++++- include/uapi/linux/fuse.h | 7 ++++ 2 files changed, 84 insertions(+), 1 deletion(-) diff --git a/fs/fuse/virtio_fs.c b/fs/fuse/virtio_fs.c index 21d8d9d7d317..8aa9fc996556 100644 --- a/fs/fuse/virtio_fs.c +++ b/fs/fuse/virtio_fs.c @@ -35,6 +35,7 @@ struct virtio_fs_vq { struct work_struct done_work; struct list_head queued_reqs; struct list_head end_reqs; /* End these requests */ + struct list_head wait_reqs; /* requests waiting for notification */ struct virtio_fs_notify_node *notify_nodes; struct list_head notify_reqs; /* List for queuing notify requests */ struct delayed_work dispatch_work; @@ -85,7 +86,6 @@ struct virtio_fs_notify_node { static int virtio_fs_enqueue_all_notify(struct virtio_fs_vq *fsvq); - static inline struct virtio_fs_vq *vq_to_fsvq(struct virtqueue *vq) { struct virtio_fs *fs = vq->vdev->priv; @@ -513,13 +513,75 @@ static int virtio_fs_enqueue_all_notify(struct virtio_fs_vq *fsvq) return 0; } +static int notify_complete_waiting_req(struct virtio_fs *vfs, + struct fuse_notify_lock_out *out_args) +{ + struct virtio_fs_vq *fsvq = &vfs->vqs[VQ_REQUEST]; + struct fuse_req *req, *next; + bool found = false; + struct fuse_conn *fc = fsvq->fud->fc; + + /* Find waiting request with the unique number and end it */ + spin_lock(&fsvq->lock); + list_for_each_entry_safe(req, next, &fsvq->wait_reqs, list) { + if (req->in.h.unique == out_args->id) { + list_del_init(&req->list); + clear_bit(FR_SENT, &req->flags); + /* Transfer error code from notify */ + req->out.h.error = out_args->error; + found = true; + break; + } + } + spin_unlock(&fsvq->lock); + + /* + * TODO: It is possible that some re-ordering happens in notify + * comes before request is complete. Deal with it. + */ + if (found) { + fuse_request_end(fc, req); + spin_lock(&fsvq->lock); + dec_in_flight_req(fsvq); + spin_unlock(&fsvq->lock); + } else + pr_debug("virtio-fs: Did not find waiting request with" + " unique=0x%llx\n", out_args->id); + + return 0; +} + +static int virtio_fs_handle_notify(struct virtio_fs *vfs, + struct virtio_fs_notify *notify) +{ + int ret = 0; + struct fuse_out_header *oh = ¬ify->out_hdr; + struct fuse_notify_lock_out *lo; + + /* + * For notifications, oh.unique is 0 and oh->error contains code + * for which notification as arrived. + */ + switch(oh->error) { + case FUSE_NOTIFY_LOCK: + lo = (struct fuse_notify_lock_out *) ¬ify->outarg; + notify_complete_waiting_req(vfs, lo); + break; + default: + printk("virtio-fs: Unexpected notification %d\n", oh->error); + } + return ret; +} + static void virtio_fs_notify_done_work(struct work_struct *work) { struct virtio_fs_vq *fsvq = container_of(work, struct virtio_fs_vq, done_work); struct virtqueue *vq = fsvq->vq; + struct virtio_fs *vfs = vq->vdev->priv; LIST_HEAD(reqs); struct virtio_fs_notify_node *notify, *next; + struct fuse_out_header *oh; spin_lock(&fsvq->lock); do { @@ -535,6 +597,10 @@ static void virtio_fs_notify_done_work(struct work_struct *work) /* Process notify */ list_for_each_entry_safe(notify, next, &reqs, list) { + oh = ¬ify->notify.out_hdr; + WARN_ON(oh->unique); + /* Handle notification */ + virtio_fs_handle_notify(vfs, ¬ify->notify); spin_lock(&fsvq->lock); dec_in_flight_req(fsvq); list_del_init(¬ify->list); @@ -656,6 +722,15 @@ static void virtio_fs_requests_done_work(struct work_struct *work) * TODO verify that server properly follows FUSE protocol * (oh.uniq, oh.len) */ + if (req->out.h.error == 1) { + /* Wait for notification to complete request */ + list_del_init(&req->list); + spin_lock(&fsvq->lock); + list_add_tail(&req->list, &fsvq->wait_reqs); + spin_unlock(&fsvq->lock); + continue; + } + args = req->args; copy_args_from_argbuf(args, req); @@ -705,6 +780,7 @@ static int virtio_fs_init_vq(struct virtio_fs *fs, struct virtio_fs_vq *fsvq, strncpy(fsvq->name, name, VQ_NAME_LEN); spin_lock_init(&fsvq->lock); INIT_LIST_HEAD(&fsvq->queued_reqs); + INIT_LIST_HEAD(&fsvq->wait_reqs); INIT_LIST_HEAD(&fsvq->end_reqs); INIT_LIST_HEAD(&fsvq->notify_reqs); init_completion(&fsvq->in_flight_zero); diff --git a/include/uapi/linux/fuse.h b/include/uapi/linux/fuse.h index 373cada89815..45f0c4efec8e 100644 --- a/include/uapi/linux/fuse.h +++ b/include/uapi/linux/fuse.h @@ -481,6 +481,7 @@ enum fuse_notify_code { FUSE_NOTIFY_STORE = 4, FUSE_NOTIFY_RETRIEVE = 5, FUSE_NOTIFY_DELETE = 6, + FUSE_NOTIFY_LOCK = 7, FUSE_NOTIFY_CODE_MAX, }; @@ -868,6 +869,12 @@ struct fuse_notify_retrieve_in { uint64_t dummy4; }; +struct fuse_notify_lock_out { + uint64_t id; + int32_t error; + int32_t padding; +}; + /* Device ioctls: */ #define FUSE_DEV_IOC_CLONE _IOR(229, 0, uint32_t)