From e5d574ab37f5f2e7937405613d9b1a724811e5ad Mon Sep 17 00:00:00 2001 From: Chunguang Xu Date: Tue, 11 Jun 2024 18:02:08 +0800 Subject: [PATCH 1/4] nvme: avoid double free special payload If a discard request needs to be retried, and that retry may fail before a new special payload is added, a double free will result. Clear the RQF_SPECIAL_LOAD when the request is cleaned. Signed-off-by: Chunguang Xu Reviewed-by: Sagi Grimberg Reviewed-by: Max Gurtovoy Signed-off-by: Keith Busch --- drivers/nvme/host/core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index f5d150c62955..c40930d10bd3 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -998,6 +998,7 @@ void nvme_cleanup_cmd(struct request *req) clear_bit_unlock(0, &ctrl->discard_page_busy); else kfree(bvec_virt(&req->special_vec)); + req->rq_flags &= ~RQF_SPECIAL_PAYLOAD; } } EXPORT_SYMBOL_GPL(nvme_cleanup_cmd); From d76584e53f4244dbc154bec447c3852600acc914 Mon Sep 17 00:00:00 2001 From: Daniel Wagner Date: Wed, 12 Jun 2024 16:02:40 +0200 Subject: [PATCH 2/4] nvmet-passthru: propagate status from id override functions The id override functions return a status which is not propagated to the caller. Fixes: c1fef73f793b ("nvmet: add passthru code to process commands") Signed-off-by: Daniel Wagner Reviewed-by: Chaitanya Kulkarni Reviewed-by: Christoph Hellwig Signed-off-by: Keith Busch --- drivers/nvme/target/passthru.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/nvme/target/passthru.c b/drivers/nvme/target/passthru.c index bb4a69d538fd..f003782d4ecf 100644 --- a/drivers/nvme/target/passthru.c +++ b/drivers/nvme/target/passthru.c @@ -226,13 +226,13 @@ static void nvmet_passthru_execute_cmd_work(struct work_struct *w) req->cmd->common.opcode == nvme_admin_identify) { switch (req->cmd->identify.cns) { case NVME_ID_CNS_CTRL: - nvmet_passthru_override_id_ctrl(req); + status = nvmet_passthru_override_id_ctrl(req); break; case NVME_ID_CNS_NS: - nvmet_passthru_override_id_ns(req); + status = nvmet_passthru_override_id_ns(req); break; case NVME_ID_CNS_NS_DESC_LIST: - nvmet_passthru_override_id_descs(req); + status = nvmet_passthru_override_id_descs(req); break; } } else if (status < 0) From cd0c1b8e045a8d2785342b385cb2684d9b48e426 Mon Sep 17 00:00:00 2001 From: Daniel Wagner Date: Wed, 12 Jun 2024 16:11:59 +0200 Subject: [PATCH 3/4] nvmet: always initialize cqe.result The spec doesn't mandate that the first two double words (aka results) for the command queue entry need to be set to 0 when they are not used (not specified). Though, the target implemention returns 0 for TCP and FC but not for RDMA. Let's make RDMA behave the same and thus explicitly initializing the result field. This prevents leaking any data from the stack. Signed-off-by: Daniel Wagner Reviewed-by: Christoph Hellwig Signed-off-by: Keith Busch --- drivers/nvme/target/core.c | 1 + drivers/nvme/target/fabrics-cmd-auth.c | 3 --- drivers/nvme/target/fabrics-cmd.c | 6 ------ 3 files changed, 1 insertion(+), 9 deletions(-) diff --git a/drivers/nvme/target/core.c b/drivers/nvme/target/core.c index 06f0c587f343..4ff460ba2826 100644 --- a/drivers/nvme/target/core.c +++ b/drivers/nvme/target/core.c @@ -957,6 +957,7 @@ bool nvmet_req_init(struct nvmet_req *req, struct nvmet_cq *cq, req->metadata_sg_cnt = 0; req->transfer_len = 0; req->metadata_len = 0; + req->cqe->result.u64 = 0; req->cqe->status = 0; req->cqe->sq_head = 0; req->ns = NULL; diff --git a/drivers/nvme/target/fabrics-cmd-auth.c b/drivers/nvme/target/fabrics-cmd-auth.c index d61b8c6ff3b2..cb34d644ed08 100644 --- a/drivers/nvme/target/fabrics-cmd-auth.c +++ b/drivers/nvme/target/fabrics-cmd-auth.c @@ -333,7 +333,6 @@ done: pr_debug("%s: ctrl %d qid %d nvme status %x error loc %d\n", __func__, ctrl->cntlid, req->sq->qid, status, req->error_loc); - req->cqe->result.u64 = 0; if (req->sq->dhchap_step != NVME_AUTH_DHCHAP_MESSAGE_SUCCESS2 && req->sq->dhchap_step != NVME_AUTH_DHCHAP_MESSAGE_FAILURE2) { unsigned long auth_expire_secs = ctrl->kato ? ctrl->kato : 120; @@ -516,8 +515,6 @@ void nvmet_execute_auth_receive(struct nvmet_req *req) status = nvmet_copy_to_sgl(req, 0, d, al); kfree(d); done: - req->cqe->result.u64 = 0; - if (req->sq->dhchap_step == NVME_AUTH_DHCHAP_MESSAGE_SUCCESS2) nvmet_auth_sq_free(req->sq); else if (req->sq->dhchap_step == NVME_AUTH_DHCHAP_MESSAGE_FAILURE1) { diff --git a/drivers/nvme/target/fabrics-cmd.c b/drivers/nvme/target/fabrics-cmd.c index 042b379cbb36..69d77d34bec1 100644 --- a/drivers/nvme/target/fabrics-cmd.c +++ b/drivers/nvme/target/fabrics-cmd.c @@ -226,9 +226,6 @@ static void nvmet_execute_admin_connect(struct nvmet_req *req) if (status) goto out; - /* zero out initial completion result, assign values as needed */ - req->cqe->result.u32 = 0; - if (c->recfmt != 0) { pr_warn("invalid connect version (%d).\n", le16_to_cpu(c->recfmt)); @@ -305,9 +302,6 @@ static void nvmet_execute_io_connect(struct nvmet_req *req) if (status) goto out; - /* zero out initial completion result, assign values as needed */ - req->cqe->result.u32 = 0; - if (c->recfmt != 0) { pr_warn("invalid connect version (%d).\n", le16_to_cpu(c->recfmt)); From ff0ffe5b7c3c12c6e0cca16652905963ae817b44 Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Thu, 13 Jun 2024 09:36:50 -0700 Subject: [PATCH 4/4] nvme: fix namespace removal list This function wants to move a subset of a list from one element to the tail into another list. It also needs to use the srcu synchronize instead of the regular rcu version. Do this one element at a time because that's the only to do it. Fixes: be647e2c76b27f4 ("nvme: use srcu for iterating namespace list") Reported-by: Venkat Rao Bagalkote Tested-by: Venkat Rao Bagalkote Signed-off-by: Keith Busch --- drivers/nvme/host/core.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index c40930d10bd3..782090ce0bc1 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -3960,12 +3960,13 @@ static void nvme_remove_invalid_namespaces(struct nvme_ctrl *ctrl, mutex_lock(&ctrl->namespaces_lock); list_for_each_entry_safe(ns, next, &ctrl->namespaces, list) { - if (ns->head->ns_id > nsid) - list_splice_init_rcu(&ns->list, &rm_list, - synchronize_rcu); + if (ns->head->ns_id > nsid) { + list_del_rcu(&ns->list); + synchronize_srcu(&ctrl->srcu); + list_add_tail_rcu(&ns->list, &rm_list); + } } mutex_unlock(&ctrl->namespaces_lock); - synchronize_srcu(&ctrl->srcu); list_for_each_entry_safe(ns, next, &rm_list, list) nvme_ns_remove(ns);