From patchwork Thu Oct 10 07:11:04 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Leon Romanovsky X-Patchwork-Id: 11182779 X-Patchwork-Delegate: jgg@ziepe.ca 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 129691709 for ; Thu, 10 Oct 2019 07:11:16 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id E7D9C20B7C for ; Thu, 10 Oct 2019 07:11:15 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1570691476; bh=gJiegQZSfx7jtdSVQan5bfm1ct1XNfMwiuXombuFRaM=; h=From:To:Cc:Subject:Date:In-Reply-To:References:List-ID:From; b=ohETsa7/0vIu2kpSJDkFSLW/gR2WKWOb+neumMbJe/lIyI7+722cGxlYoOnvxwCPH ZGG/9DCPxoJ1nMuaTy+3eJYzjNqufYWR5+Ve9Qtfrp2k5l6TEJVv5hXCe2eiKcKx30 U/j/qekBwpnxXNmGJicjjIHUVMpmCaCXdWE1Z7EE= Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1732928AbfJJHLP (ORCPT ); Thu, 10 Oct 2019 03:11:15 -0400 Received: from mail.kernel.org ([198.145.29.99]:53386 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726864AbfJJHLP (ORCPT ); Thu, 10 Oct 2019 03:11:15 -0400 Received: from localhost (unknown [193.47.165.251]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id E8A952190F; Thu, 10 Oct 2019 07:11:13 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1570691474; bh=gJiegQZSfx7jtdSVQan5bfm1ct1XNfMwiuXombuFRaM=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=uOx5XtRcHmT2tDJ+qhjN09Py0Lw7b8V9AewQD3cgc6JWBVC6IiX7a1uZ43M+8bgKZ 5WTxLvBiJNzedKDIxLmoH37Kg2be41KofFT7E75lrd0Wcu+9QrzsfN1xs172NSoB6a f2SDBaiez7SUie9TybjatD5vS54YVFsiu489EmTg= From: Leon Romanovsky To: Doug Ledford , Jason Gunthorpe Cc: Leon Romanovsky , RDMA mailing list Subject: [PATCH rdma-next v1 1/2] RDMA/restrack: Remove PID namespace support Date: Thu, 10 Oct 2019 10:11:04 +0300 Message-Id: <20191010071105.25538-2-leon@kernel.org> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20191010071105.25538-1-leon@kernel.org> References: <20191010071105.25538-1-leon@kernel.org> MIME-Version: 1.0 Sender: linux-rdma-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org From: Leon Romanovsky IB resources are bounded to IB device and file descriptors, both entities are unaware to PID namespaces and to task lifetime. The difference in model caused to unpredictable behavior for the following scenario: 1. Create FD and context 2. Share it with ephemeral child 3. Create any object and exit that child The end result of this flow, that those newly created objects will be tracked by restrack, but won't be visible for users because task_struct associated with them already exited. The right thing is to rely on net namespace only for any filtering purposes and drop PID namespace. Signed-off-by: Leon Romanovsky --- drivers/infiniband/core/counters.c | 14 -------------- drivers/infiniband/core/nldev.c | 13 +------------ drivers/infiniband/core/restrack.c | 20 +------------------- drivers/infiniband/core/restrack.h | 1 - 4 files changed, 2 insertions(+), 46 deletions(-) diff --git a/drivers/infiniband/core/counters.c b/drivers/infiniband/core/counters.c index 736ab760025d..12ba2685abcf 100644 --- a/drivers/infiniband/core/counters.c +++ b/drivers/infiniband/core/counters.c @@ -149,9 +149,6 @@ static bool auto_mode_match(struct ib_qp *qp, struct rdma_counter *counter, struct auto_mode_param *param = &counter->mode.param; bool match = true; - if (!rdma_is_visible_in_pid_ns(&qp->res)) - return false; - /* Ensure that counter belongs to the right PID */ if (task_pid_nr(counter->res.task) != task_pid_nr(qp->res.task)) return false; @@ -229,9 +226,6 @@ static struct rdma_counter *rdma_get_counter_auto_mode(struct ib_qp *qp, rt = &dev->res[RDMA_RESTRACK_COUNTER]; xa_lock(&rt->xa); xa_for_each(&rt->xa, id, res) { - if (!rdma_is_visible_in_pid_ns(res)) - continue; - counter = container_of(res, struct rdma_counter, res); if ((counter->device != qp->device) || (counter->port != port)) goto next; @@ -412,9 +406,6 @@ static struct ib_qp *rdma_counter_get_qp(struct ib_device *dev, u32 qp_num) if (IS_ERR(res)) return NULL; - if (!rdma_is_visible_in_pid_ns(res)) - goto err; - qp = container_of(res, struct ib_qp, res); if (qp->qp_type == IB_QPT_RAW_PACKET && !capable(CAP_NET_RAW)) goto err; @@ -445,11 +436,6 @@ static struct rdma_counter *rdma_get_counter_by_id(struct ib_device *dev, if (IS_ERR(res)) return NULL; - if (!rdma_is_visible_in_pid_ns(res)) { - rdma_restrack_put(res); - return NULL; - } - counter = container_of(res, struct rdma_counter, res); kref_get(&counter->kref); rdma_restrack_put(res); diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c index 7a7474000100..71bc08510064 100644 --- a/drivers/infiniband/core/nldev.c +++ b/drivers/infiniband/core/nldev.c @@ -698,9 +698,6 @@ static int fill_stat_counter_qps(struct sk_buff *msg, rt = &counter->device->res[RDMA_RESTRACK_QP]; xa_lock(&rt->xa); xa_for_each(&rt->xa, id, res) { - if (!rdma_is_visible_in_pid_ns(res)) - continue; - qp = container_of(res, struct ib_qp, res); if (qp->qp_type == IB_QPT_RAW_PACKET && !capable(CAP_NET_RAW)) continue; @@ -1222,15 +1219,10 @@ static int res_get_common_doit(struct sk_buff *skb, struct nlmsghdr *nlh, goto err; } - if (!rdma_is_visible_in_pid_ns(res)) { - ret = -ENOENT; - goto err_get; - } - msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL); if (!msg) { ret = -ENOMEM; - goto err; + goto err_get; } nlh = nlmsg_put(msg, NETLINK_CB(skb).portid, nlh->nlmsg_seq, @@ -1334,9 +1326,6 @@ static int res_get_common_dumpit(struct sk_buff *skb, * objects. */ xa_for_each(&rt->xa, id, res) { - if (!rdma_is_visible_in_pid_ns(res)) - continue; - if (idx < start || !rdma_restrack_get(res)) goto next; diff --git a/drivers/infiniband/core/restrack.c b/drivers/infiniband/core/restrack.c index a07665f7ef8c..62fbb0ae9cb4 100644 --- a/drivers/infiniband/core/restrack.c +++ b/drivers/infiniband/core/restrack.c @@ -116,11 +116,8 @@ int rdma_restrack_count(struct ib_device *dev, enum rdma_restrack_type type) u32 cnt = 0; xa_lock(&rt->xa); - xas_for_each(&xas, e, U32_MAX) { - if (!rdma_is_visible_in_pid_ns(e)) - continue; + xas_for_each(&xas, e, U32_MAX) cnt++; - } xa_unlock(&rt->xa); return cnt; } @@ -346,18 +343,3 @@ void rdma_restrack_del(struct rdma_restrack_entry *res) } } EXPORT_SYMBOL(rdma_restrack_del); - -bool rdma_is_visible_in_pid_ns(struct rdma_restrack_entry *res) -{ - /* - * 1. Kern resources should be visible in init - * namespace only - * 2. Present only resources visible in the current - * namespace - */ - if (rdma_is_kernel_res(res)) - return task_active_pid_ns(current) == &init_pid_ns; - - /* PID 0 means that resource is not found in current namespace */ - return task_pid_vnr(res->task); -} diff --git a/drivers/infiniband/core/restrack.h b/drivers/infiniband/core/restrack.h index 7bd177cc0a61..d084e5f89849 100644 --- a/drivers/infiniband/core/restrack.h +++ b/drivers/infiniband/core/restrack.h @@ -27,5 +27,4 @@ int rdma_restrack_init(struct ib_device *dev); void rdma_restrack_clean(struct ib_device *dev); void rdma_restrack_attach_task(struct rdma_restrack_entry *res, struct task_struct *task); -bool rdma_is_visible_in_pid_ns(struct rdma_restrack_entry *res); #endif /* _RDMA_CORE_RESTRACK_H_ */ From patchwork Thu Oct 10 07:11:05 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Leon Romanovsky X-Patchwork-Id: 11182781 X-Patchwork-Delegate: jgg@ziepe.ca 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 B860D1709 for ; Thu, 10 Oct 2019 07:11:19 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 9982021929 for ; Thu, 10 Oct 2019 07:11:19 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1570691479; bh=nnYLH1uG+9nJcpWo+o1Jw9z0y1gV6PAnxk7tE/HKVG8=; h=From:To:Cc:Subject:Date:In-Reply-To:References:List-ID:From; b=yE9ZRIw8CKU82h1rTbjVVx6mr8jNVvF4GvxOeEWhysOGxPxNZarTTJywVC/yy9nsy cwPDsGwBRypZYnHF7YAUzf5LA0ksdVbqFESjnnLGB8zgVT9dkZS2DJo2yHRbNL0Z7o L7Jt35fWtC2tw1jsFwb2S45MEt0da/aZj41WyrOE= Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1732993AbfJJHLT (ORCPT ); Thu, 10 Oct 2019 03:11:19 -0400 Received: from mail.kernel.org ([198.145.29.99]:53410 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726864AbfJJHLS (ORCPT ); Thu, 10 Oct 2019 03:11:18 -0400 Received: from localhost (unknown [193.47.165.251]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id 1D43420B7C; Thu, 10 Oct 2019 07:11:16 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1570691477; bh=nnYLH1uG+9nJcpWo+o1Jw9z0y1gV6PAnxk7tE/HKVG8=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=YklC/xhmVB6choJA62r3cHHjhTR11/dJbsUeD1mcNeqB0XHTiH1MLvGKh0fvkwdAe 1VyWW8qHC7KXf+Bpl9M6TJ8iKUbg/Qu6gOuHFmltxgDcFt+ZcjzAadqPbNeXNX29rN Mmky9zzOU1gv4VEUm/e/U6hm2rfXQ35jhOdp+Fjs= From: Leon Romanovsky To: Doug Ledford , Jason Gunthorpe Cc: Leon Romanovsky , RDMA mailing list Subject: [PATCH rdma-next v1 2/2] RDMA/core: Check that process is still alive before sending it to the users Date: Thu, 10 Oct 2019 10:11:05 +0300 Message-Id: <20191010071105.25538-3-leon@kernel.org> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20191010071105.25538-1-leon@kernel.org> References: <20191010071105.25538-1-leon@kernel.org> MIME-Version: 1.0 Sender: linux-rdma-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org From: Leon Romanovsky The PID information can disappear asynchronically because task can be killed and moved to zombie state. In such case, PID will be zero in similar way to the kernel tasks. Recognize such situation where we are asking to return orphaned object and simply skip filling PID attribute. As part of this change, document the same scenario in counter.c code. Signed-off-by: Leon Romanovsky --- drivers/infiniband/core/counters.c | 14 ++++++++++++-- drivers/infiniband/core/nldev.c | 28 +++++++++++++++++++++------- 2 files changed, 33 insertions(+), 9 deletions(-) diff --git a/drivers/infiniband/core/counters.c b/drivers/infiniband/core/counters.c index 12ba2685abcf..47c551a0bcb0 100644 --- a/drivers/infiniband/core/counters.c +++ b/drivers/infiniband/core/counters.c @@ -149,8 +149,18 @@ static bool auto_mode_match(struct ib_qp *qp, struct rdma_counter *counter, struct auto_mode_param *param = &counter->mode.param; bool match = true; - /* Ensure that counter belongs to the right PID */ - if (task_pid_nr(counter->res.task) != task_pid_nr(qp->res.task)) + /* + * Ensure that counter belongs to the right PID. + * This operation can race with user space which kills + * the process and leaves QP and counters orphans. + * + * It is not a big deal because exitted task will leave both + * QP and counter in the same bucket of zombie process. Just ensure + * that process is still alive before procedding. + * + */ + if (task_pid_nr(counter->res.task) != task_pid_nr(qp->res.task) || + !task_pid_nr(qp->res.task)) return false; if (auto_mask & RDMA_COUNTER_MASK_QP_TYPE) diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c index 71bc08510064..0ebe95c79ae0 100644 --- a/drivers/infiniband/core/nldev.c +++ b/drivers/infiniband/core/nldev.c @@ -399,20 +399,34 @@ static int fill_res_info(struct sk_buff *msg, struct ib_device *device) static int fill_res_name_pid(struct sk_buff *msg, struct rdma_restrack_entry *res) { + int err = 0; + /* * For user resources, user is should read /proc/PID/comm to get the * name of the task file. */ if (rdma_is_kernel_res(res)) { - if (nla_put_string(msg, RDMA_NLDEV_ATTR_RES_KERN_NAME, - res->kern_name)) - return -EMSGSIZE; + err = nla_put_string(msg, RDMA_NLDEV_ATTR_RES_KERN_NAME, + res->kern_name); } else { - if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_PID, - task_pid_vnr(res->task))) - return -EMSGSIZE; + pid_t pid; + + pid = task_pid_vnr(res->task); + /* + * Task is dead and in zombie state. + * There is no need to print PID anymore. + */ + if (pid) + /* + * This part is racy, task can be killed and PID will + * be zero right here but it is ok, next query won't + * return PID. We don't promise real-time reflection + * of SW objects. + */ + err = nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_PID, pid); } - return 0; + + return err ? -EMSGSIZE : 0; } static bool fill_res_entry(struct ib_device *dev, struct sk_buff *msg,