From nobody Sun Dec 28 17:28:55 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 372A2C4167B for ; Wed, 6 Dec 2023 08:12:57 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1376990AbjLFIMs (ORCPT ); Wed, 6 Dec 2023 03:12:48 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59524 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229505AbjLFIMo (ORCPT ); Wed, 6 Dec 2023 03:12:44 -0500 Received: from smtp-out2.suse.de (smtp-out2.suse.de [IPv6:2a07:de40:b251:101:10:150:64:2]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 46CEB18D for ; Wed, 6 Dec 2023 00:12:51 -0800 (PST) Received: from imap2.dmz-prg2.suse.org (imap2.dmz-prg2.suse.org [IPv6:2a07:de40:b281:104:10:150:64:98]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by smtp-out2.suse.de (Postfix) with ESMTPS id A215A1FBF9; Wed, 6 Dec 2023 08:12:49 +0000 (UTC) Received: from imap2.dmz-prg2.suse.org (localhost [127.0.0.1]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by imap2.dmz-prg2.suse.org (Postfix) with ESMTPS id 8E794133DD; Wed, 6 Dec 2023 08:12:49 +0000 (UTC) Received: from dovecot-director2.suse.de ([10.150.64.162]) by imap2.dmz-prg2.suse.org with ESMTPSA id NcSnIQEtcGUTHgAAn2gu4w (envelope-from ); Wed, 06 Dec 2023 08:12:49 +0000 From: Daniel Wagner To: linux-nvme@lists.infradead.org Cc: linux-kernel@vger.kernel.org, Keith Busch , Christoph Hellwig , Sagi Grimberg , Hannes Reinecke , Daniel Wagner Subject: [PATCH v3 1/4] nvme: lookup ctrl from request instead from namespace Date: Wed, 6 Dec 2023 09:12:41 +0100 Message-ID: <20231206081244.32733-2-dwagner@suse.de> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20231206081244.32733-1-dwagner@suse.de> References: <20231206081244.32733-1-dwagner@suse.de> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Spamd-Bar: +++++++++++ Authentication-Results: smtp-out2.suse.de; dkim=none; dmarc=fail reason="No valid SPF, No valid DKIM" header.from=suse.de (policy=none); spf=softfail (smtp-out2.suse.de: 2a07:de40:b281:104:10:150:64:98 is neither permitted nor denied by domain of dwagner@suse.de) smtp.mailfrom=dwagner@suse.de X-Rspamd-Server: rspamd2 X-Spamd-Result: default: False [11.73 / 50.00]; ARC_NA(0.00)[]; RCVD_VIA_SMTP_AUTH(0.00)[]; SPAMHAUS_XBL(0.00)[2a07:de40:b281:104:10:150:64:98:from]; FROM_HAS_DN(0.00)[]; TO_DN_SOME(0.00)[]; R_MISSING_CHARSET(2.50)[]; TO_MATCH_ENVRCPT_ALL(0.00)[]; MIME_GOOD(-0.10)[text/plain]; NEURAL_SPAM_SHORT(2.94)[0.980]; BROKEN_CONTENT_TYPE(1.50)[]; R_SPF_SOFTFAIL(4.60)[~all:c]; RCVD_COUNT_THREE(0.00)[3]; MX_GOOD(-0.01)[]; RCPT_COUNT_SEVEN(0.00)[7]; MID_CONTAINS_FROM(1.00)[]; DBL_BLOCKED_OPENRESOLVER(0.00)[suse.de:email]; FUZZY_BLOCKED(0.00)[rspamd.com]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(2.20)[]; MIME_TRACE(0.00)[0:+]; RCVD_TLS_ALL(0.00)[]; BAYES_HAM(-3.00)[100.00%]; DMARC_POLICY_SOFTFAIL(0.10)[suse.de : No valid SPF, No valid DKIM,none] X-Rspamd-Queue-Id: A215A1FBF9 Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" In preparation to use struct nvme_ns_head pointers instead of a struct nvme_ns pointers, get the ctrl pointer from a request. The reason is that there is no easy and fast way to lookup the ctrl from struct nvme_ns_head. Reviewed-by: Christoph Hellwig Reviewed-by: Sagi Grimberg Signed-off-by: Daniel Wagner --- drivers/nvme/host/core.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index d699f0c8b13e..900c045fcae0 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -769,6 +769,7 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *= ns, struct request *req, struct nvme_command *cmnd) { unsigned short segments =3D blk_rq_nr_discard_segments(req), n =3D 0; + struct nvme_ctrl *ctrl =3D nvme_req(req)->ctrl; struct nvme_dsm_range *range; struct bio *bio; =20 @@ -786,10 +787,10 @@ static blk_status_t nvme_setup_discard(struct nvme_ns= *ns, struct request *req, * discard page. If that's also busy, it's safe to return * busy, as we know we can make progress once that's freed. */ - if (test_and_set_bit_lock(0, &ns->ctrl->discard_page_busy)) + if (test_and_set_bit_lock(0, &ctrl->discard_page_busy)) return BLK_STS_RESOURCE; =20 - range =3D page_address(ns->ctrl->discard_page); + range =3D page_address(ctrl->discard_page); } =20 if (queue_max_discard_segments(req->q) =3D=3D 1) { @@ -815,8 +816,8 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *= ns, struct request *req, } =20 if (WARN_ON_ONCE(n !=3D segments)) { - if (virt_to_page(range) =3D=3D ns->ctrl->discard_page) - clear_bit_unlock(0, &ns->ctrl->discard_page_busy); + if (virt_to_page(range) =3D=3D ctrl->discard_page) + clear_bit_unlock(0, &ctrl->discard_page_busy); else kfree(range); return BLK_STS_IOERR; @@ -861,9 +862,11 @@ static void nvme_set_ref_tag(struct nvme_ns *ns, struc= t nvme_command *cmnd, static inline blk_status_t nvme_setup_write_zeroes(struct nvme_ns *ns, struct request *req, struct nvme_command *cmnd) { + struct nvme_ctrl *ctrl =3D nvme_req(req)->ctrl; + memset(cmnd, 0, sizeof(*cmnd)); =20 - if (ns->ctrl->quirks & NVME_QUIRK_DEALLOCATE_ZEROES) + if (ctrl->quirks & NVME_QUIRK_DEALLOCATE_ZEROES) return nvme_setup_discard(ns, req, cmnd); =20 cmnd->write_zeroes.opcode =3D nvme_cmd_write_zeroes; --=20 2.43.0 From nobody Sun Dec 28 17:28:55 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1A46DC46CA0 for ; Wed, 6 Dec 2023 08:13:01 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1377132AbjLFIMw (ORCPT ); Wed, 6 Dec 2023 03:12:52 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59540 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230412AbjLFIMq (ORCPT ); Wed, 6 Dec 2023 03:12:46 -0500 Received: from smtp-out2.suse.de (smtp-out2.suse.de [195.135.223.131]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 203ADD3 for ; Wed, 6 Dec 2023 00:12:52 -0800 (PST) Received: from imap2.dmz-prg2.suse.org (imap2.dmz-prg2.suse.org [IPv6:2a07:de40:b281:104:10:150:64:98]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by smtp-out2.suse.de (Postfix) with ESMTPS id 4ADB21FCFA; Wed, 6 Dec 2023 08:12:50 +0000 (UTC) Received: from imap2.dmz-prg2.suse.org (localhost [127.0.0.1]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by imap2.dmz-prg2.suse.org (Postfix) with ESMTPS id 378F1133DD; Wed, 6 Dec 2023 08:12:50 +0000 (UTC) Received: from dovecot-director2.suse.de ([10.150.64.162]) by imap2.dmz-prg2.suse.org with ESMTPSA id EekhDAItcGUVHgAAn2gu4w (envelope-from ); Wed, 06 Dec 2023 08:12:50 +0000 From: Daniel Wagner To: linux-nvme@lists.infradead.org Cc: linux-kernel@vger.kernel.org, Keith Busch , Christoph Hellwig , Sagi Grimberg , Hannes Reinecke , Daniel Wagner Subject: [PATCH v3 2/4] nvme: initialize head before namespace Date: Wed, 6 Dec 2023 09:12:42 +0100 Message-ID: <20231206081244.32733-3-dwagner@suse.de> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20231206081244.32733-1-dwagner@suse.de> References: <20231206081244.32733-1-dwagner@suse.de> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Spamd-Bar: +++++++++++ Authentication-Results: smtp-out2.suse.de; dkim=none; dmarc=fail reason="No valid SPF, No valid DKIM" header.from=suse.de (policy=none); spf=softfail (smtp-out2.suse.de: 2a07:de40:b281:104:10:150:64:98 is neither permitted nor denied by domain of dwagner@suse.de) smtp.mailfrom=dwagner@suse.de X-Rspamd-Server: rspamd2 X-Spamd-Result: default: False [11.76 / 50.00]; ARC_NA(0.00)[]; RCVD_VIA_SMTP_AUTH(0.00)[]; SPAMHAUS_XBL(0.00)[2a07:de40:b281:104:10:150:64:98:from]; FROM_HAS_DN(0.00)[]; TO_DN_SOME(0.00)[]; R_MISSING_CHARSET(2.50)[]; TO_MATCH_ENVRCPT_ALL(0.00)[]; MIME_GOOD(-0.10)[text/plain]; NEURAL_SPAM_SHORT(2.97)[0.989]; BROKEN_CONTENT_TYPE(1.50)[]; R_SPF_SOFTFAIL(4.60)[~all:c]; RCVD_COUNT_THREE(0.00)[3]; MX_GOOD(-0.01)[]; RCPT_COUNT_SEVEN(0.00)[7]; MID_CONTAINS_FROM(1.00)[]; DBL_BLOCKED_OPENRESOLVER(0.00)[suse.de:email]; FUZZY_BLOCKED(0.00)[rspamd.com]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(2.20)[]; MIME_TRACE(0.00)[0:+]; RCVD_TLS_ALL(0.00)[]; BAYES_HAM(-3.00)[100.00%]; DMARC_POLICY_SOFTFAIL(0.10)[suse.de : No valid SPF, No valid DKIM,none] X-Rspamd-Queue-Id: 4ADB21FCFA Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" In preparation to use struct nvme_ns_head pointers instead of a struct nvme_ns pointers, initialize the head pointer before we create the disk. This allows us to attach the head as private data to the disk object. Signed-off-by: Daniel Wagner --- drivers/nvme/host/core.c | 46 ++++++++++++++++++++++------------------ 1 file changed, 25 insertions(+), 21 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 900c045fcae0..1fabe1b81de0 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -3474,10 +3474,11 @@ static int nvme_global_check_duplicate_ids(struct n= vme_subsystem *this, return ret; } =20 -static int nvme_init_ns_head(struct nvme_ns *ns, struct nvme_ns_info *info) +static int nvme_init_ns_head(struct nvme_ctrl *ctrl, + struct nvme_ns_info *info, + struct nvme_ns_head **head) { - struct nvme_ctrl *ctrl =3D ns->ctrl; - struct nvme_ns_head *head =3D NULL; + struct nvme_ns_head *h =3D NULL; int ret; =20 ret =3D nvme_global_check_duplicate_ids(ctrl->subsys, &info->ids); @@ -3499,8 +3500,8 @@ static int nvme_init_ns_head(struct nvme_ns *ns, stru= ct nvme_ns_info *info) * up at any time. */ nvme_print_device_info(ctrl); - if ((ns->ctrl->ops->flags & NVME_F_FABRICS) || /* !PCIe */ - ((ns->ctrl->subsys->cmic & NVME_CTRL_CMIC_MULTI_CTRL) && + if ((ctrl->ops->flags & NVME_F_FABRICS) || /* !PCIe */ + ((ctrl->subsys->cmic & NVME_CTRL_CMIC_MULTI_CTRL) && info->is_shared)) { dev_err(ctrl->device, "ignoring nsid %d because of duplicate IDs\n", @@ -3519,8 +3520,8 @@ static int nvme_init_ns_head(struct nvme_ns *ns, stru= ct nvme_ns_info *info) } =20 mutex_lock(&ctrl->subsys->lock); - head =3D nvme_find_ns_head(ctrl, info->nsid); - if (!head) { + h =3D nvme_find_ns_head(ctrl, info->nsid); + if (!h) { ret =3D nvme_subsys_check_duplicate_ids(ctrl->subsys, &info->ids); if (ret) { dev_err(ctrl->device, @@ -3528,20 +3529,20 @@ static int nvme_init_ns_head(struct nvme_ns *ns, st= ruct nvme_ns_info *info) info->nsid); goto out_unlock; } - head =3D nvme_alloc_ns_head(ctrl, info); - if (IS_ERR(head)) { - ret =3D PTR_ERR(head); + h =3D nvme_alloc_ns_head(ctrl, info); + if (IS_ERR(h)) { + ret =3D PTR_ERR(h); goto out_unlock; } } else { ret =3D -EINVAL; - if (!info->is_shared || !head->shared) { + if (!info->is_shared || !h->shared) { dev_err(ctrl->device, "Duplicate unshared namespace %d\n", info->nsid); goto out_put_ns_head; } - if (!nvme_ns_ids_equal(&head->ids, &info->ids)) { + if (!nvme_ns_ids_equal(&h->ids, &info->ids)) { dev_err(ctrl->device, "IDs don't match for shared namespace %d\n", info->nsid); @@ -3557,13 +3558,12 @@ static int nvme_init_ns_head(struct nvme_ns *ns, st= ruct nvme_ns_info *info) } } =20 - list_add_tail_rcu(&ns->siblings, &head->list); - ns->head =3D head; + *head =3D h; mutex_unlock(&ctrl->subsys->lock); return 0; =20 out_put_ns_head: - nvme_put_ns_head(head); + nvme_put_ns_head(h); out_unlock: mutex_unlock(&ctrl->subsys->lock); return ret; @@ -3615,15 +3615,22 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, s= truct nvme_ns_info *info) if (!ns) return; =20 + if (nvme_init_ns_head(ctrl, info, &ns->head)) + goto out_free_ns; + disk =3D blk_mq_alloc_disk(ctrl->tagset, ns); if (IS_ERR(disk)) - goto out_free_ns; + goto out_free_head; disk->fops =3D &nvme_bdev_ops; disk->private_data =3D ns; =20 ns->disk =3D disk; ns->queue =3D disk->queue; =20 + mutex_lock(&ctrl->subsys->lock); + list_add_tail_rcu(&ns->siblings, &ns->head->list); + mutex_unlock(&ctrl->subsys->lock); + if (ctrl->opts && ctrl->opts->data_digest) blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES, ns->queue); =20 @@ -3635,9 +3642,6 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, str= uct nvme_ns_info *info) ns->ctrl =3D ctrl; kref_init(&ns->kref); =20 - if (nvme_init_ns_head(ns, info)) - goto out_cleanup_disk; - /* * If multipathing is enabled, the device name for all disks and not * just those that represent shared namespaces needs to be based on the @@ -3691,9 +3695,9 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, str= uct nvme_ns_info *info) if (list_empty(&ns->head->list)) list_del_init(&ns->head->entry); mutex_unlock(&ctrl->subsys->lock); - nvme_put_ns_head(ns->head); - out_cleanup_disk: put_disk(disk); + out_free_head: + nvme_put_ns_head(ns->head); out_free_ns: kfree(ns); } --=20 2.43.0 From nobody Sun Dec 28 17:28:55 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1436CC10DC1 for ; Wed, 6 Dec 2023 08:13:07 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1377162AbjLFIM5 (ORCPT ); Wed, 6 Dec 2023 03:12:57 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59570 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1377078AbjLFIMt (ORCPT ); Wed, 6 Dec 2023 03:12:49 -0500 Received: from smtp-out1.suse.de (smtp-out1.suse.de [IPv6:2a07:de40:b251:101:10:150:64:1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 707681A4 for ; Wed, 6 Dec 2023 00:12:52 -0800 (PST) Received: from imap2.dmz-prg2.suse.org (imap2.dmz-prg2.suse.org [IPv6:2a07:de40:b281:104:10:150:64:98]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by smtp-out1.suse.de (Postfix) with ESMTPS id E5FBA2206F; Wed, 6 Dec 2023 08:12:50 +0000 (UTC) Received: from imap2.dmz-prg2.suse.org (localhost [127.0.0.1]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by imap2.dmz-prg2.suse.org (Postfix) with ESMTPS id CCB57133DD; Wed, 6 Dec 2023 08:12:50 +0000 (UTC) Received: from dovecot-director2.suse.de ([10.150.64.162]) by imap2.dmz-prg2.suse.org with ESMTPSA id clNbMAItcGUXHgAAn2gu4w (envelope-from ); Wed, 06 Dec 2023 08:12:50 +0000 From: Daniel Wagner To: linux-nvme@lists.infradead.org Cc: linux-kernel@vger.kernel.org, Keith Busch , Christoph Hellwig , Sagi Grimberg , Hannes Reinecke , Daniel Wagner Subject: [PATCH v3 3/4] nvme: move ns id info to struct nvme_ns_head Date: Wed, 6 Dec 2023 09:12:43 +0100 Message-ID: <20231206081244.32733-4-dwagner@suse.de> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20231206081244.32733-1-dwagner@suse.de> References: <20231206081244.32733-1-dwagner@suse.de> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Spamd-Result: default: False [1.98 / 50.00]; ARC_NA(0.00)[]; RCVD_VIA_SMTP_AUTH(0.00)[]; RBL_SPAMHAUS_BLOCKED_OPENRESOLVER(0.00)[2a07:de40:b281:104:10:150:64:98:from]; FROM_HAS_DN(0.00)[]; TO_DN_SOME(0.00)[]; R_MISSING_CHARSET(2.50)[]; TO_MATCH_ENVRCPT_ALL(0.00)[]; MIME_GOOD(-0.10)[text/plain]; BROKEN_CONTENT_TYPE(1.50)[]; R_SPF_SOFTFAIL(0.00)[~all:c]; RCVD_COUNT_THREE(0.00)[3]; MX_GOOD(-0.01)[]; NEURAL_HAM_SHORT(-0.01)[-0.046]; RCPT_COUNT_SEVEN(0.00)[7]; MID_CONTAINS_FROM(1.00)[]; DBL_BLOCKED_OPENRESOLVER(0.00)[suse.de:email]; FUZZY_BLOCKED(0.00)[rspamd.com]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(0.00)[]; MIME_TRACE(0.00)[0:+]; RCVD_TLS_ALL(0.00)[]; BAYES_HAM(-3.00)[100.00%]; DMARC_POLICY_SOFTFAIL(0.10)[suse.de : No valid SPF, No valid DKIM,none] X-Spamd-Bar: + X-Rspamd-Server: rspamd1 Authentication-Results: smtp-out1.suse.de; dkim=none; dmarc=fail reason="No valid SPF, No valid DKIM" header.from=suse.de (policy=none); spf=softfail (smtp-out1.suse.de: 2a07:de40:b281:104:10:150:64:98 is neither permitted nor denied by domain of dwagner@suse.de) smtp.mailfrom=dwagner@suse.de X-Rspamd-Queue-Id: E5FBA2206F Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Move the namesapce info to struct nvme_ns_head, because it's the same for all associated namespaces. The head pointer is accessible from the ns pointer so we could just update all places with ns->x to ns->head->x. While this is okay for the slow path, it's not for the fast path. Thus we store the head pointer as private data into request_queue. Signed-off-by: Daniel Wagner --- drivers/nvme/host/apple.c | 4 +- drivers/nvme/host/core.c | 192 +++++++++++++++++---------------- drivers/nvme/host/fc.c | 4 +- drivers/nvme/host/ioctl.c | 20 ++-- drivers/nvme/host/multipath.c | 31 +++--- drivers/nvme/host/nvme.h | 60 ++++++----- drivers/nvme/host/rdma.c | 10 +- drivers/nvme/host/tcp.c | 8 +- drivers/nvme/host/zns.c | 34 +++--- drivers/nvme/target/loop.c | 4 +- drivers/nvme/target/passthru.c | 8 +- 11 files changed, 197 insertions(+), 178 deletions(-) diff --git a/drivers/nvme/host/apple.c b/drivers/nvme/host/apple.c index 596bb11eeba5..04c1c5fbd4d1 100644 --- a/drivers/nvme/host/apple.c +++ b/drivers/nvme/host/apple.c @@ -732,7 +732,7 @@ static int apple_nvme_remove_sq(struct apple_nvme *anv) static blk_status_t apple_nvme_queue_rq(struct blk_mq_hw_ctx *hctx, const struct blk_mq_queue_data *bd) { - struct nvme_ns *ns =3D hctx->queue->queuedata; + struct nvme_ns_head *head =3D hctx->queue->queuedata; struct apple_nvme_queue *q =3D hctx->driver_data; struct apple_nvme *anv =3D queue_to_apple_nvme(q); struct request *req =3D bd->rq; @@ -753,7 +753,7 @@ static blk_status_t apple_nvme_queue_rq(struct blk_mq_h= w_ctx *hctx, if (!nvme_check_ready(&anv->ctrl, req, true)) return nvme_fail_nonready_command(&anv->ctrl, req); =20 - ret =3D nvme_setup_cmd(ns, req); + ret =3D nvme_setup_cmd(head, req); if (ret) return ret; =20 diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 1fabe1b81de0..caa52c2f57c8 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -308,16 +308,16 @@ static void nvme_retry_req(struct request *req) =20 static void nvme_log_error(struct request *req) { - struct nvme_ns *ns =3D req->q->queuedata; + struct nvme_ns_head *head =3D req->q->queuedata; struct nvme_request *nr =3D nvme_req(req); =20 - if (ns) { - pr_err_ratelimited("%s: %s(0x%x) @ LBA %llu, %llu blocks, %s (sct 0x%x /= sc 0x%x) %s%s\n", - ns->disk ? ns->disk->disk_name : "?", + if (head) { + pr_err_ratelimited("%s: %s(0x%x) @ LBA %llu, %u blocks, %s (sct 0x%x / s= c 0x%x) %s%s\n", + head->disk ? head->disk->disk_name : "?", nvme_get_opcode_str(nr->cmd->common.opcode), nr->cmd->common.opcode, - (unsigned long long)nvme_sect_to_lba(ns, blk_rq_pos(req)), - (unsigned long long)blk_rq_bytes(req) >> ns->lba_shift, + nvme_sect_to_lba(head, blk_rq_pos(req)), + blk_rq_bytes(req) >> head->lba_shift, nvme_get_error_status_str(nr->status), nr->status >> 8 & 7, /* Status Code Type */ nr->status & 0xff, /* Status Code */ @@ -757,16 +757,16 @@ bool __nvme_check_ready(struct nvme_ctrl *ctrl, struc= t request *rq, } EXPORT_SYMBOL_GPL(__nvme_check_ready); =20 -static inline void nvme_setup_flush(struct nvme_ns *ns, +static inline void nvme_setup_flush(struct nvme_ns_head *head, struct nvme_command *cmnd) { memset(cmnd, 0, sizeof(*cmnd)); cmnd->common.opcode =3D nvme_cmd_flush; - cmnd->common.nsid =3D cpu_to_le32(ns->head->ns_id); + cmnd->common.nsid =3D cpu_to_le32(head->ns_id); } =20 -static blk_status_t nvme_setup_discard(struct nvme_ns *ns, struct request = *req, - struct nvme_command *cmnd) +static blk_status_t nvme_setup_discard(struct nvme_ns_head *head, + struct request *req, struct nvme_command *cmnd) { unsigned short segments =3D blk_rq_nr_discard_segments(req), n =3D 0; struct nvme_ctrl *ctrl =3D nvme_req(req)->ctrl; @@ -794,8 +794,8 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *= ns, struct request *req, } =20 if (queue_max_discard_segments(req->q) =3D=3D 1) { - u64 slba =3D nvme_sect_to_lba(ns, blk_rq_pos(req)); - u32 nlb =3D blk_rq_sectors(req) >> (ns->lba_shift - 9); + u64 slba =3D nvme_sect_to_lba(head, blk_rq_pos(req)); + u32 nlb =3D blk_rq_sectors(req) >> (head->lba_shift - 9); =20 range[0].cattr =3D cpu_to_le32(0); range[0].nlb =3D cpu_to_le32(nlb); @@ -803,8 +803,9 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *= ns, struct request *req, n =3D 1; } else { __rq_for_each_bio(bio, req) { - u64 slba =3D nvme_sect_to_lba(ns, bio->bi_iter.bi_sector); - u32 nlb =3D bio->bi_iter.bi_size >> ns->lba_shift; + u64 slba =3D nvme_sect_to_lba(head, + bio->bi_iter.bi_sector); + u32 nlb =3D bio->bi_iter.bi_size >> head->lba_shift; =20 if (n < segments) { range[n].cattr =3D cpu_to_le32(0); @@ -825,7 +826,7 @@ static blk_status_t nvme_setup_discard(struct nvme_ns *= ns, struct request *req, =20 memset(cmnd, 0, sizeof(*cmnd)); cmnd->dsm.opcode =3D nvme_cmd_dsm; - cmnd->dsm.nsid =3D cpu_to_le32(ns->head->ns_id); + cmnd->dsm.nsid =3D cpu_to_le32(head->ns_id); cmnd->dsm.nr =3D cpu_to_le32(segments - 1); cmnd->dsm.attributes =3D cpu_to_le32(NVME_DSMGMT_AD); =20 @@ -835,14 +836,14 @@ static blk_status_t nvme_setup_discard(struct nvme_ns= *ns, struct request *req, return BLK_STS_OK; } =20 -static void nvme_set_ref_tag(struct nvme_ns *ns, struct nvme_command *cmnd, - struct request *req) +static void nvme_set_ref_tag(struct nvme_ns_head *head, + struct nvme_command *cmnd, struct request *req) { u32 upper, lower; u64 ref48; =20 /* both rw and write zeroes share the same reftag format */ - switch (ns->guard_type) { + switch (head->guard_type) { case NVME_NVM_NS_16B_GUARD: cmnd->rw.reftag =3D cpu_to_le32(t10_pi_ref_tag(req)); break; @@ -859,7 +860,7 @@ static void nvme_set_ref_tag(struct nvme_ns *ns, struct= nvme_command *cmnd, } } =20 -static inline blk_status_t nvme_setup_write_zeroes(struct nvme_ns *ns, +static inline blk_status_t nvme_setup_write_zeroes(struct nvme_ns_head *he= ad, struct request *req, struct nvme_command *cmnd) { struct nvme_ctrl *ctrl =3D nvme_req(req)->ctrl; @@ -867,25 +868,25 @@ static inline blk_status_t nvme_setup_write_zeroes(st= ruct nvme_ns *ns, memset(cmnd, 0, sizeof(*cmnd)); =20 if (ctrl->quirks & NVME_QUIRK_DEALLOCATE_ZEROES) - return nvme_setup_discard(ns, req, cmnd); + return nvme_setup_discard(head, req, cmnd); =20 cmnd->write_zeroes.opcode =3D nvme_cmd_write_zeroes; - cmnd->write_zeroes.nsid =3D cpu_to_le32(ns->head->ns_id); + cmnd->write_zeroes.nsid =3D cpu_to_le32(head->ns_id); cmnd->write_zeroes.slba =3D - cpu_to_le64(nvme_sect_to_lba(ns, blk_rq_pos(req))); + cpu_to_le64(nvme_sect_to_lba(head, blk_rq_pos(req))); cmnd->write_zeroes.length =3D - cpu_to_le16((blk_rq_bytes(req) >> ns->lba_shift) - 1); + cpu_to_le16((blk_rq_bytes(req) >> head->lba_shift) - 1); =20 - if (!(req->cmd_flags & REQ_NOUNMAP) && (ns->features & NVME_NS_DEAC)) + if (!(req->cmd_flags & REQ_NOUNMAP) && (head->features & NVME_NS_DEAC)) cmnd->write_zeroes.control |=3D cpu_to_le16(NVME_WZ_DEAC); =20 - if (nvme_ns_has_pi(ns)) { + if (nvme_ns_has_pi(head)) { cmnd->write_zeroes.control |=3D cpu_to_le16(NVME_RW_PRINFO_PRACT); =20 - switch (ns->pi_type) { + switch (head->pi_type) { case NVME_NS_DPS_PI_TYPE1: case NVME_NS_DPS_PI_TYPE2: - nvme_set_ref_tag(ns, cmnd, req); + nvme_set_ref_tag(head, cmnd, req); break; } } @@ -893,7 +894,7 @@ static inline blk_status_t nvme_setup_write_zeroes(stru= ct nvme_ns *ns, return BLK_STS_OK; } =20 -static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns, +static inline blk_status_t nvme_setup_rw(struct nvme_ns_head *head, struct request *req, struct nvme_command *cmnd, enum nvme_opcode op) { @@ -910,17 +911,19 @@ static inline blk_status_t nvme_setup_rw(struct nvme_= ns *ns, =20 cmnd->rw.opcode =3D op; cmnd->rw.flags =3D 0; - cmnd->rw.nsid =3D cpu_to_le32(ns->head->ns_id); + cmnd->rw.nsid =3D cpu_to_le32(head->ns_id); cmnd->rw.cdw2 =3D 0; cmnd->rw.cdw3 =3D 0; cmnd->rw.metadata =3D 0; - cmnd->rw.slba =3D cpu_to_le64(nvme_sect_to_lba(ns, blk_rq_pos(req))); - cmnd->rw.length =3D cpu_to_le16((blk_rq_bytes(req) >> ns->lba_shift) - 1); + cmnd->rw.slba =3D + cpu_to_le64(nvme_sect_to_lba(head, blk_rq_pos(req))); + cmnd->rw.length =3D + cpu_to_le16((blk_rq_bytes(req) >> head->lba_shift) - 1); cmnd->rw.reftag =3D 0; cmnd->rw.apptag =3D 0; cmnd->rw.appmask =3D 0; =20 - if (ns->ms) { + if (head->ms) { /* * If formated with metadata, the block layer always provides a * metadata buffer if CONFIG_BLK_DEV_INTEGRITY is enabled. Else @@ -928,12 +931,12 @@ static inline blk_status_t nvme_setup_rw(struct nvme_= ns *ns, * namespace capacity to zero to prevent any I/O. */ if (!blk_integrity_rq(req)) { - if (WARN_ON_ONCE(!nvme_ns_has_pi(ns))) + if (WARN_ON_ONCE(!nvme_ns_has_pi(head))) return BLK_STS_NOTSUPP; control |=3D NVME_RW_PRINFO_PRACT; } =20 - switch (ns->pi_type) { + switch (head->pi_type) { case NVME_NS_DPS_PI_TYPE3: control |=3D NVME_RW_PRINFO_PRCHK_GUARD; break; @@ -943,7 +946,7 @@ static inline blk_status_t nvme_setup_rw(struct nvme_ns= *ns, NVME_RW_PRINFO_PRCHK_REF; if (op =3D=3D nvme_cmd_zone_append) control |=3D NVME_RW_APPEND_PIREMAP; - nvme_set_ref_tag(ns, cmnd, req); + nvme_set_ref_tag(head, cmnd, req); break; } } @@ -966,7 +969,7 @@ void nvme_cleanup_cmd(struct request *req) } EXPORT_SYMBOL_GPL(nvme_cleanup_cmd); =20 -blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req) +blk_status_t nvme_setup_cmd(struct nvme_ns_head *head, struct request *req) { struct nvme_command *cmd =3D nvme_req(req)->cmd; blk_status_t ret =3D BLK_STS_OK; @@ -980,35 +983,39 @@ blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struc= t request *req) /* these are setup prior to execution in nvme_init_request() */ break; case REQ_OP_FLUSH: - nvme_setup_flush(ns, cmd); + nvme_setup_flush(head, cmd); break; case REQ_OP_ZONE_RESET_ALL: case REQ_OP_ZONE_RESET: - ret =3D nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_RESET); + ret =3D nvme_setup_zone_mgmt_send(head, req, + cmd, NVME_ZONE_RESET); break; case REQ_OP_ZONE_OPEN: - ret =3D nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_OPEN); + ret =3D nvme_setup_zone_mgmt_send(head, req, + cmd, NVME_ZONE_OPEN); break; case REQ_OP_ZONE_CLOSE: - ret =3D nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_CLOSE); + ret =3D nvme_setup_zone_mgmt_send(head, req, + cmd, NVME_ZONE_CLOSE); break; case REQ_OP_ZONE_FINISH: - ret =3D nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_FINISH); + ret =3D nvme_setup_zone_mgmt_send(head, req, + cmd, NVME_ZONE_FINISH); break; case REQ_OP_WRITE_ZEROES: - ret =3D nvme_setup_write_zeroes(ns, req, cmd); + ret =3D nvme_setup_write_zeroes(head, req, cmd); break; case REQ_OP_DISCARD: - ret =3D nvme_setup_discard(ns, req, cmd); + ret =3D nvme_setup_discard(head, req, cmd); break; case REQ_OP_READ: - ret =3D nvme_setup_rw(ns, req, cmd, nvme_cmd_read); + ret =3D nvme_setup_rw(head, req, cmd, nvme_cmd_read); break; case REQ_OP_WRITE: - ret =3D nvme_setup_rw(ns, req, cmd, nvme_cmd_write); + ret =3D nvme_setup_rw(head, req, cmd, nvme_cmd_write); break; case REQ_OP_ZONE_APPEND: - ret =3D nvme_setup_rw(ns, req, cmd, nvme_cmd_zone_append); + ret =3D nvme_setup_rw(head, req, cmd, nvme_cmd_zone_append); break; default: WARN_ON_ONCE(1); @@ -1084,12 +1091,13 @@ int nvme_submit_sync_cmd(struct request_queue *q, s= truct nvme_command *cmd, } EXPORT_SYMBOL_GPL(nvme_submit_sync_cmd); =20 -u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u8 op= code) +u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, + u8 opcode) { u32 effects =3D 0; =20 - if (ns) { - effects =3D le32_to_cpu(ns->head->effects->iocs[opcode]); + if (head) { + effects =3D le32_to_cpu(head->effects->iocs[opcode]); if (effects & ~(NVME_CMD_EFFECTS_CSUPP | NVME_CMD_EFFECTS_LBCC)) dev_warn_once(ctrl->device, "IO command:%02x has unusual effects:%08x\n", @@ -1109,9 +1117,10 @@ u32 nvme_command_effects(struct nvme_ctrl *ctrl, str= uct nvme_ns *ns, u8 opcode) } EXPORT_SYMBOL_NS_GPL(nvme_command_effects, NVME_TARGET_PASSTHRU); =20 -u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u8 opc= ode) +u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, + u8 opcode) { - u32 effects =3D nvme_command_effects(ctrl, ns, opcode); + u32 effects =3D nvme_command_effects(ctrl, head, opcode); =20 /* * For simplicity, IO to all namespaces is quiesced even if the command @@ -1129,8 +1138,8 @@ u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struc= t nvme_ns *ns, u8 opcode) } EXPORT_SYMBOL_NS_GPL(nvme_passthru_start, NVME_TARGET_PASSTHRU); =20 -void nvme_passthru_end(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u32 eff= ects, - struct nvme_command *cmd, int status) +void nvme_passthru_end(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, + u32 effects, struct nvme_command *cmd, int status) { if (effects & NVME_CMD_EFFECTS_CSE_MASK) { nvme_unfreeze(ctrl); @@ -1149,7 +1158,7 @@ void nvme_passthru_end(struct nvme_ctrl *ctrl, struct= nvme_ns *ns, u32 effects, nvme_queue_scan(ctrl); flush_work(&ctrl->scan_work); } - if (ns) + if (head) return; =20 switch (cmd->common.opcode) { @@ -1679,9 +1688,9 @@ static void nvme_init_integrity(struct gendisk *disk,= struct nvme_ns *ns, { struct blk_integrity integrity =3D { }; =20 - switch (ns->pi_type) { + switch (ns->head->pi_type) { case NVME_NS_DPS_PI_TYPE3: - switch (ns->guard_type) { + switch (ns->head->guard_type) { case NVME_NVM_NS_16B_GUARD: integrity.profile =3D &t10_pi_type3_crc; integrity.tag_size =3D sizeof(u16) + sizeof(u32); @@ -1699,7 +1708,7 @@ static void nvme_init_integrity(struct gendisk *disk,= struct nvme_ns *ns, break; case NVME_NS_DPS_PI_TYPE1: case NVME_NS_DPS_PI_TYPE2: - switch (ns->guard_type) { + switch (ns->head->guard_type) { case NVME_NVM_NS_16B_GUARD: integrity.profile =3D &t10_pi_type1_crc; integrity.tag_size =3D sizeof(u16); @@ -1720,7 +1729,7 @@ static void nvme_init_integrity(struct gendisk *disk,= struct nvme_ns *ns, break; } =20 - integrity.tuple_size =3D ns->ms; + integrity.tuple_size =3D ns->head->ms; blk_integrity_register(disk, &integrity); blk_queue_max_integrity_segments(disk->queue, max_integrity_segments); } @@ -1737,8 +1746,10 @@ static void nvme_config_discard(struct gendisk *disk= , struct nvme_ns *ns) struct request_queue *queue =3D disk->queue; u32 size =3D queue_logical_block_size(queue); =20 - if (ctrl->dmrsl && ctrl->dmrsl <=3D nvme_sect_to_lba(ns, UINT_MAX)) - ctrl->max_discard_sectors =3D nvme_lba_to_sect(ns, ctrl->dmrsl); + if (ctrl->dmrsl && + ctrl->dmrsl <=3D nvme_sect_to_lba(ns->head, UINT_MAX)) + ctrl->max_discard_sectors =3D + nvme_lba_to_sect(ns->head, ctrl->dmrsl); =20 if (ctrl->max_discard_sectors =3D=3D 0) { blk_queue_max_discard_sectors(queue, 0); @@ -1779,11 +1790,11 @@ static int nvme_init_ms(struct nvme_ns *ns, struct = nvme_id_ns *id) int ret =3D 0; u32 elbaf; =20 - ns->pi_size =3D 0; - ns->ms =3D le16_to_cpu(id->lbaf[lbaf].ms); + ns->head->pi_size =3D 0; + ns->head->ms =3D le16_to_cpu(id->lbaf[lbaf].ms); if (!(ctrl->ctratt & NVME_CTRL_ATTR_ELBAS)) { - ns->pi_size =3D sizeof(struct t10_pi_tuple); - ns->guard_type =3D NVME_NVM_NS_16B_GUARD; + ns->head->pi_size =3D sizeof(struct t10_pi_tuple); + ns->head->guard_type =3D NVME_NVM_NS_16B_GUARD; goto set_pi; } =20 @@ -1806,13 +1817,13 @@ static int nvme_init_ms(struct nvme_ns *ns, struct = nvme_id_ns *id) if (nvme_elbaf_sts(elbaf)) goto free_data; =20 - ns->guard_type =3D nvme_elbaf_guard_type(elbaf); - switch (ns->guard_type) { + ns->head->guard_type =3D nvme_elbaf_guard_type(elbaf); + switch (ns->head->guard_type) { case NVME_NVM_NS_64B_GUARD: - ns->pi_size =3D sizeof(struct crc64_pi_tuple); + ns->head->pi_size =3D sizeof(struct crc64_pi_tuple); break; case NVME_NVM_NS_16B_GUARD: - ns->pi_size =3D sizeof(struct t10_pi_tuple); + ns->head->pi_size =3D sizeof(struct t10_pi_tuple); break; default: break; @@ -1821,10 +1832,10 @@ static int nvme_init_ms(struct nvme_ns *ns, struct = nvme_id_ns *id) free_data: kfree(nvm); set_pi: - if (ns->pi_size && (first || ns->ms =3D=3D ns->pi_size)) - ns->pi_type =3D id->dps & NVME_NS_DPS_PI_MASK; + if (ns->head->pi_size && (first || ns->head->ms =3D=3D ns->head->pi_size)) + ns->head->pi_type =3D id->dps & NVME_NS_DPS_PI_MASK; else - ns->pi_type =3D 0; + ns->head->pi_type =3D 0; =20 return ret; } @@ -1838,8 +1849,8 @@ static int nvme_configure_metadata(struct nvme_ns *ns= , struct nvme_id_ns *id) if (ret) return ret; =20 - ns->features &=3D ~(NVME_NS_METADATA_SUPPORTED | NVME_NS_EXT_LBAS); - if (!ns->ms || !(ctrl->ops->flags & NVME_F_METADATA_SUPPORTED)) + ns->head->features &=3D ~(NVME_NS_METADATA_SUPPORTED | NVME_NS_EXT_LBAS); + if (!ns->head->ms || !(ctrl->ops->flags & NVME_F_METADATA_SUPPORTED)) return 0; =20 if (ctrl->ops->flags & NVME_F_FABRICS) { @@ -1851,7 +1862,7 @@ static int nvme_configure_metadata(struct nvme_ns *ns= , struct nvme_id_ns *id) if (WARN_ON_ONCE(!(id->flbas & NVME_NS_FLBAS_META_EXT))) return 0; =20 - ns->features |=3D NVME_NS_EXT_LBAS; + ns->head->features |=3D NVME_NS_EXT_LBAS; =20 /* * The current fabrics transport drivers support namespace @@ -1862,8 +1873,8 @@ static int nvme_configure_metadata(struct nvme_ns *ns= , struct nvme_id_ns *id) * Note, this check will need to be modified if any drivers * gain the ability to use other metadata formats. */ - if (ctrl->max_integrity_segments && nvme_ns_has_pi(ns)) - ns->features |=3D NVME_NS_METADATA_SUPPORTED; + if (ctrl->max_integrity_segments && nvme_ns_has_pi(ns->head)) + ns->head->features |=3D NVME_NS_METADATA_SUPPORTED; } else { /* * For PCIe controllers, we can't easily remap the separate @@ -1872,9 +1883,9 @@ static int nvme_configure_metadata(struct nvme_ns *ns= , struct nvme_id_ns *id) * We allow extended LBAs for the passthrough interface, though. */ if (id->flbas & NVME_NS_FLBAS_META_EXT) - ns->features |=3D NVME_NS_EXT_LBAS; + ns->head->features |=3D NVME_NS_EXT_LBAS; else - ns->features |=3D NVME_NS_METADATA_SUPPORTED; + ns->head->features |=3D NVME_NS_METADATA_SUPPORTED; } return 0; } @@ -1900,8 +1911,8 @@ static void nvme_set_queue_limits(struct nvme_ctrl *c= trl, static void nvme_update_disk_info(struct gendisk *disk, struct nvme_ns *ns, struct nvme_id_ns *id) { - sector_t capacity =3D nvme_lba_to_sect(ns, le64_to_cpu(id->nsze)); - u32 bs =3D 1U << ns->lba_shift; + sector_t capacity =3D nvme_lba_to_sect(ns->head, le64_to_cpu(id->nsze)); + u32 bs =3D 1U << ns->head->lba_shift; u32 atomic_bs, phys_bs, io_opt =3D 0; =20 /* @@ -1909,7 +1920,8 @@ static void nvme_update_disk_info(struct gendisk *dis= k, * or smaller than a sector size yet, so catch this early and don't * allow block I/O. */ - if (ns->lba_shift > PAGE_SHIFT || ns->lba_shift < SECTOR_SHIFT) { + if (ns->head->lba_shift > PAGE_SHIFT || + ns->head->lba_shift < SECTOR_SHIFT) { capacity =3D 0; bs =3D (1 << 9); } @@ -1952,12 +1964,12 @@ static void nvme_update_disk_info(struct gendisk *d= isk, * I/O to namespaces with metadata except when the namespace supports * PI, as it can strip/insert in that case. */ - if (ns->ms) { + if (ns->head->ms) { if (IS_ENABLED(CONFIG_BLK_DEV_INTEGRITY) && - (ns->features & NVME_NS_METADATA_SUPPORTED)) + (ns->head->features & NVME_NS_METADATA_SUPPORTED)) nvme_init_integrity(disk, ns, ns->ctrl->max_integrity_segments); - else if (!nvme_ns_has_pi(ns)) + else if (!nvme_ns_has_pi(ns->head)) capacity =3D 0; } =20 @@ -1988,7 +2000,7 @@ static void nvme_set_chunk_sectors(struct nvme_ns *ns= , struct nvme_id_ns *id) is_power_of_2(ctrl->max_hw_sectors)) iob =3D ctrl->max_hw_sectors; else - iob =3D nvme_lba_to_sect(ns, le16_to_cpu(id->noiob)); + iob =3D nvme_lba_to_sect(ns->head, le16_to_cpu(id->noiob)); =20 if (!iob) return; @@ -2021,7 +2033,7 @@ static int nvme_update_ns_info_generic(struct nvme_ns= *ns, if (nvme_ns_head_multipath(ns->head)) { blk_mq_freeze_queue(ns->head->disk->queue); set_disk_ro(ns->head->disk, nvme_ns_is_readonly(ns, info)); - nvme_mpath_revalidate_paths(ns); + nvme_mpath_revalidate_paths(ns->head); blk_stack_limits(&ns->head->disk->queue->limits, &ns->queue->limits, 0); ns->head->disk->flags |=3D GENHD_FL_HIDDEN; @@ -2055,7 +2067,7 @@ static int nvme_update_ns_info_block(struct nvme_ns *= ns, =20 blk_mq_freeze_queue(ns->disk->queue); lbaf =3D nvme_lbaf_index(id->flbas); - ns->lba_shift =3D id->lbaf[lbaf].ds; + ns->head->lba_shift =3D id->lbaf[lbaf].ds; nvme_set_queue_limits(ns->ctrl, ns->queue); =20 ret =3D nvme_configure_metadata(ns, id); @@ -2081,7 +2093,7 @@ static int nvme_update_ns_info_block(struct nvme_ns *= ns, * do not return zeroes. */ if ((id->dlfeat & 0x7) =3D=3D 0x1 && (id->dlfeat & (1 << 3))) - ns->features |=3D NVME_NS_DEAC; + ns->head->features |=3D NVME_NS_DEAC; set_disk_ro(ns->disk, nvme_ns_is_readonly(ns, info)); set_bit(NVME_NS_READY, &ns->flags); blk_mq_unfreeze_queue(ns->disk->queue); @@ -2096,7 +2108,7 @@ static int nvme_update_ns_info_block(struct nvme_ns *= ns, blk_mq_freeze_queue(ns->head->disk->queue); nvme_update_disk_info(ns->head->disk, ns, id); set_disk_ro(ns->head->disk, nvme_ns_is_readonly(ns, info)); - nvme_mpath_revalidate_paths(ns); + nvme_mpath_revalidate_paths(ns->head); blk_stack_limits(&ns->head->disk->queue->limits, &ns->queue->limits, 0); disk_update_readahead(ns->head->disk); @@ -3618,7 +3630,7 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, str= uct nvme_ns_info *info) if (nvme_init_ns_head(ctrl, info, &ns->head)) goto out_free_ns; =20 - disk =3D blk_mq_alloc_disk(ctrl->tagset, ns); + disk =3D blk_mq_alloc_disk(ctrl->tagset, ns->head); if (IS_ERR(disk)) goto out_free_head; disk->fops =3D &nvme_bdev_ops; diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c index fb22976a36a8..41f9bacae6f4 100644 --- a/drivers/nvme/host/fc.c +++ b/drivers/nvme/host/fc.c @@ -2799,7 +2799,7 @@ static blk_status_t nvme_fc_queue_rq(struct blk_mq_hw_ctx *hctx, const struct blk_mq_queue_data *bd) { - struct nvme_ns *ns =3D hctx->queue->queuedata; + struct nvme_ns_head *head =3D hctx->queue->queuedata; struct nvme_fc_queue *queue =3D hctx->driver_data; struct nvme_fc_ctrl *ctrl =3D queue->ctrl; struct request *rq =3D bd->rq; @@ -2813,7 +2813,7 @@ nvme_fc_queue_rq(struct blk_mq_hw_ctx *hctx, !nvme_check_ready(&queue->ctrl->ctrl, rq, queue_ready)) return nvme_fail_nonready_command(&queue->ctrl->ctrl, rq); =20 - ret =3D nvme_setup_cmd(ns, rq); + ret =3D nvme_setup_cmd(head, rq); if (ret) return ret; =20 diff --git a/drivers/nvme/host/ioctl.c b/drivers/nvme/host/ioctl.c index 529b9954d2b8..29263009bb6e 100644 --- a/drivers/nvme/host/ioctl.c +++ b/drivers/nvme/host/ioctl.c @@ -61,7 +61,7 @@ static bool nvme_cmd_allowed(struct nvme_ns *ns, struct n= vme_command *c, * and marks this command as supported. If not reject unprivileged * passthrough. */ - effects =3D nvme_command_effects(ns->ctrl, ns, c->common.opcode); + effects =3D nvme_command_effects(ns->ctrl, ns->head, c->common.opcode); if (!(effects & NVME_CMD_EFFECTS_CSUPP)) return false; =20 @@ -168,8 +168,8 @@ static int nvme_map_user_request(struct request *req, u= 64 ubuffer, unsigned int flags) { struct request_queue *q =3D req->q; - struct nvme_ns *ns =3D q->queuedata; - struct block_device *bdev =3D ns ? ns->disk->part0 : NULL; + struct nvme_ns_head *head =3D q->queuedata; + struct block_device *bdev =3D head ? head->disk->part0 : NULL; struct bio *bio =3D NULL; void *meta =3D NULL; int ret; @@ -222,7 +222,7 @@ static int nvme_submit_user_cmd(struct request_queue *q, void __user *meta_buffer, unsigned meta_len, u32 meta_seed, u64 *result, unsigned timeout, unsigned int flags) { - struct nvme_ns *ns =3D q->queuedata; + struct nvme_ns_head *head =3D q->queuedata; struct nvme_ctrl *ctrl; struct request *req; void *meta =3D NULL; @@ -245,7 +245,7 @@ static int nvme_submit_user_cmd(struct request_queue *q, bio =3D req->bio; ctrl =3D nvme_req(req)->ctrl; =20 - effects =3D nvme_passthru_start(ctrl, ns, cmd->common.opcode); + effects =3D nvme_passthru_start(ctrl, head, cmd->common.opcode); ret =3D nvme_execute_rq(req, false); if (result) *result =3D le64_to_cpu(nvme_req(req)->result.u64); @@ -257,7 +257,7 @@ static int nvme_submit_user_cmd(struct request_queue *q, blk_mq_free_request(req); =20 if (effects) - nvme_passthru_end(ctrl, ns, effects, cmd, ret); + nvme_passthru_end(ctrl, head, effects, cmd, ret); =20 return ret; } @@ -283,10 +283,10 @@ static int nvme_submit_io(struct nvme_ns *ns, struct = nvme_user_io __user *uio) return -EINVAL; } =20 - length =3D (io.nblocks + 1) << ns->lba_shift; + length =3D (io.nblocks + 1) << ns->head->lba_shift; =20 if ((io.control & NVME_RW_PRINFO_PRACT) && - ns->ms =3D=3D sizeof(struct t10_pi_tuple)) { + ns->head->ms =3D=3D sizeof(struct t10_pi_tuple)) { /* * Protection information is stripped/inserted by the * controller. @@ -296,11 +296,11 @@ static int nvme_submit_io(struct nvme_ns *ns, struct = nvme_user_io __user *uio) meta_len =3D 0; metadata =3D NULL; } else { - meta_len =3D (io.nblocks + 1) * ns->ms; + meta_len =3D (io.nblocks + 1) * ns->head->ms; metadata =3D nvme_to_user_ptr(io.metadata); } =20 - if (ns->features & NVME_NS_EXT_LBAS) { + if (ns->head->features & NVME_NS_EXT_LBAS) { length +=3D meta_len; meta_len =3D 0; } else if (meta_len) { diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c index 0a88d7bdc5e3..24ccd5b9c350 100644 --- a/drivers/nvme/host/multipath.c +++ b/drivers/nvme/host/multipath.c @@ -82,11 +82,14 @@ void nvme_mpath_start_freeze(struct nvme_subsystem *sub= sys) =20 void nvme_failover_req(struct request *req) { - struct nvme_ns *ns =3D req->q->queuedata; + struct nvme_ns_head *head =3D req->q->queuedata; + struct nvme_ctrl *ctrl =3D nvme_req(req)->ctrl; + struct nvme_ns *ns; u16 status =3D nvme_req(req)->status & 0x7ff; unsigned long flags; struct bio *bio; =20 + ns =3D nvme_find_get_ns(ctrl, head->ns_id); nvme_mpath_clear_current_path(ns); =20 /* @@ -94,14 +97,14 @@ void nvme_failover_req(struct request *req) * ready to serve this namespace. Kick of a re-read of the ANA * information page, and just try any other available path for now. */ - if (nvme_is_ana_error(status) && ns->ctrl->ana_log_buf) { + if (nvme_is_ana_error(status) && ctrl->ana_log_buf) { set_bit(NVME_NS_ANA_PENDING, &ns->flags); - queue_work(nvme_wq, &ns->ctrl->ana_work); + queue_work(nvme_wq, &ctrl->ana_work); } =20 - spin_lock_irqsave(&ns->head->requeue_lock, flags); + spin_lock_irqsave(&head->requeue_lock, flags); for (bio =3D req->bio; bio; bio =3D bio->bi_next) { - bio_set_dev(bio, ns->head->disk->part0); + bio_set_dev(bio, head->disk->part0); if (bio->bi_opf & REQ_POLLED) { bio->bi_opf &=3D ~REQ_POLLED; bio->bi_cookie =3D BLK_QC_T_NONE; @@ -115,17 +118,17 @@ void nvme_failover_req(struct request *req) */ bio->bi_opf &=3D ~REQ_NOWAIT; } - blk_steal_bios(&ns->head->requeue_list, req); - spin_unlock_irqrestore(&ns->head->requeue_lock, flags); + blk_steal_bios(&head->requeue_list, req); + spin_unlock_irqrestore(&head->requeue_lock, flags); =20 blk_mq_end_request(req, 0); - kblockd_schedule_work(&ns->head->requeue_work); + kblockd_schedule_work(&head->requeue_work); } =20 void nvme_mpath_start_request(struct request *rq) { - struct nvme_ns *ns =3D rq->q->queuedata; - struct gendisk *disk =3D ns->head->disk; + struct nvme_ns_head *head =3D rq->q->queuedata; + struct gendisk *disk =3D head->disk; =20 if (!blk_queue_io_stat(disk->queue) || blk_rq_is_passthrough(rq)) return; @@ -138,11 +141,11 @@ EXPORT_SYMBOL_GPL(nvme_mpath_start_request); =20 void nvme_mpath_end_request(struct request *rq) { - struct nvme_ns *ns =3D rq->q->queuedata; + struct nvme_ns_head *head =3D rq->q->queuedata; =20 if (!(nvme_req(rq)->flags & NVME_MPATH_IO_STATS)) return; - bdev_end_io_acct(ns->head->disk->part0, req_op(rq), + bdev_end_io_acct(head->disk->part0, req_op(rq), blk_rq_bytes(rq) >> SECTOR_SHIFT, nvme_req(rq)->start_time); } @@ -202,10 +205,10 @@ void nvme_mpath_clear_ctrl_paths(struct nvme_ctrl *ct= rl) up_read(&ctrl->namespaces_rwsem); } =20 -void nvme_mpath_revalidate_paths(struct nvme_ns *ns) +void nvme_mpath_revalidate_paths(struct nvme_ns_head *head) { - struct nvme_ns_head *head =3D ns->head; sector_t capacity =3D get_capacity(head->disk); + struct nvme_ns *ns; int node; int srcu_idx; =20 diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h index 578e6d311bc9..e6c7890b14c9 100644 --- a/drivers/nvme/host/nvme.h +++ b/drivers/nvme/host/nvme.h @@ -451,6 +451,17 @@ struct nvme_ns_head { bool shared; int instance; struct nvme_effects_log *effects; + int lba_shift; + u16 ms; + u16 pi_size; + u16 sgs; + u32 sws; + u8 pi_type; + u8 guard_type; +#ifdef CONFIG_BLK_DEV_ZONED + u64 zsze; +#endif + unsigned long features; =20 struct cdev cdev; struct device cdev_device; @@ -492,17 +503,6 @@ struct nvme_ns { struct kref kref; struct nvme_ns_head *head; =20 - int lba_shift; - u16 ms; - u16 pi_size; - u16 sgs; - u32 sws; - u8 pi_type; - u8 guard_type; -#ifdef CONFIG_BLK_DEV_ZONED - u64 zsze; -#endif - unsigned long features; unsigned long flags; #define NVME_NS_REMOVING 0 #define NVME_NS_ANA_PENDING 2 @@ -517,9 +517,9 @@ struct nvme_ns { }; =20 /* NVMe ns supports metadata actions by the controller (generate/strip) */ -static inline bool nvme_ns_has_pi(struct nvme_ns *ns) +static inline bool nvme_ns_has_pi(struct nvme_ns_head *head) { - return ns->pi_type && ns->ms =3D=3D ns->pi_size; + return head->pi_type && head->ms =3D=3D head->pi_size; } =20 struct nvme_ctrl_ops { @@ -651,17 +651,17 @@ static inline int nvme_reset_subsystem(struct nvme_ct= rl *ctrl) /* * Convert a 512B sector number to a device logical block number. */ -static inline u64 nvme_sect_to_lba(struct nvme_ns *ns, sector_t sector) +static inline u64 nvme_sect_to_lba(struct nvme_ns_head *head, sector_t sec= tor) { - return sector >> (ns->lba_shift - SECTOR_SHIFT); + return sector >> (head->lba_shift - SECTOR_SHIFT); } =20 /* * Convert a device logical block number to a 512B sector number. */ -static inline sector_t nvme_lba_to_sect(struct nvme_ns *ns, u64 lba) +static inline sector_t nvme_lba_to_sect(struct nvme_ns_head *head, u64 lba) { - return lba << (ns->lba_shift - SECTOR_SHIFT); + return lba << (head->lba_shift - SECTOR_SHIFT); } =20 /* @@ -792,7 +792,7 @@ static inline enum req_op nvme_req_op(struct nvme_comma= nd *cmd) #define NVME_QID_ANY -1 void nvme_init_request(struct request *req, struct nvme_command *cmd); void nvme_cleanup_cmd(struct request *req); -blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req); +blk_status_t nvme_setup_cmd(struct nvme_ns_head *head, struct request *req= ); blk_status_t nvme_fail_nonready_command(struct nvme_ctrl *ctrl, struct request *req); bool __nvme_check_ready(struct nvme_ctrl *ctrl, struct request *rq, @@ -901,7 +901,7 @@ void nvme_mpath_update(struct nvme_ctrl *ctrl); void nvme_mpath_uninit(struct nvme_ctrl *ctrl); void nvme_mpath_stop(struct nvme_ctrl *ctrl); bool nvme_mpath_clear_current_path(struct nvme_ns *ns); -void nvme_mpath_revalidate_paths(struct nvme_ns *ns); +void nvme_mpath_revalidate_paths(struct nvme_ns_head *head); void nvme_mpath_clear_ctrl_paths(struct nvme_ctrl *ctrl); void nvme_mpath_shutdown_disk(struct nvme_ns_head *head); void nvme_mpath_start_request(struct request *rq); @@ -909,10 +909,10 @@ void nvme_mpath_end_request(struct request *rq); =20 static inline void nvme_trace_bio_complete(struct request *req) { - struct nvme_ns *ns =3D req->q->queuedata; + struct nvme_ns_head *head =3D req->q->queuedata; =20 if ((req->cmd_flags & REQ_NVME_MPATH) && req->bio) - trace_block_bio_complete(ns->head->disk->queue, req->bio); + trace_block_bio_complete(head->disk->queue, req->bio); } =20 extern bool multipath; @@ -1004,13 +1004,14 @@ int nvme_ns_report_zones(struct nvme_ns *ns, sector= _t sector, unsigned int nr_zones, report_zones_cb cb, void *data); #ifdef CONFIG_BLK_DEV_ZONED int nvme_update_zone_info(struct nvme_ns *ns, unsigned lbaf); -blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns *ns, struct request = *req, +blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns_head *head, + struct request *req, struct nvme_command *cmnd, enum nvme_zone_mgmt_action action); #else -static inline blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns *ns, - struct request *req, struct nvme_command *cmnd, - enum nvme_zone_mgmt_action action) +static inline blk_status_t nvme_setup_zone_mgmt_send( + struct nvme_ns_head *head, struct request *req, + struct nvme_command *cmnd, enum nvme_zone_mgmt_action action) { return BLK_STS_NOTSUPP; } @@ -1086,12 +1087,13 @@ static inline int nvme_auth_wait(struct nvme_ctrl *= ctrl, int qid) static inline void nvme_auth_free(struct nvme_ctrl *ctrl) {}; #endif =20 -u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns *ns, +u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, u8 opcode); -u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u8 opc= ode); +u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, + u8 opcode); int nvme_execute_rq(struct request *rq, bool at_head); -void nvme_passthru_end(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u32 eff= ects, - struct nvme_command *cmd, int status); +void nvme_passthru_end(struct nvme_ctrl *ctrl, struct nvme_ns_head *head, + u32 effects, struct nvme_command *cmd, int status); struct nvme_ctrl *nvme_ctrl_from_file(struct file *file); struct nvme_ns *nvme_find_get_ns(struct nvme_ctrl *ctrl, unsigned nsid); void nvme_put_ns(struct nvme_ns *ns); diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c index 81e2621169e5..0e967413ee43 100644 --- a/drivers/nvme/host/rdma.c +++ b/drivers/nvme/host/rdma.c @@ -1407,7 +1407,7 @@ static int nvme_rdma_map_sg_pi(struct nvme_rdma_queue= *queue, struct nvme_rdma_sgl *sgl =3D &req->data_sgl; struct ib_reg_wr *wr =3D &req->reg_wr; struct request *rq =3D blk_mq_rq_from_pdu(req); - struct nvme_ns *ns =3D rq->q->queuedata; + struct nvme_ns_head *head =3D rq->q->queuedata; struct bio *bio =3D rq->bio; struct nvme_keyed_sgl_desc *sg =3D &c->common.dptr.ksgl; int nr; @@ -1423,7 +1423,7 @@ static int nvme_rdma_map_sg_pi(struct nvme_rdma_queue= *queue, goto mr_put; =20 nvme_rdma_set_sig_attrs(blk_get_integrity(bio->bi_bdev->bd_disk), c, - req->mr->sig_attrs, ns->pi_type); + req->mr->sig_attrs, head->pi_type); nvme_rdma_set_prot_checks(c, &req->mr->sig_attrs->check_mask); =20 ib_update_fast_reg_key(req->mr, ib_inc_rkey(req->mr->rkey)); @@ -1979,7 +1979,7 @@ static enum blk_eh_timer_return nvme_rdma_timeout(str= uct request *rq) static blk_status_t nvme_rdma_queue_rq(struct blk_mq_hw_ctx *hctx, const struct blk_mq_queue_data *bd) { - struct nvme_ns *ns =3D hctx->queue->queuedata; + struct nvme_ns_head *head =3D hctx->queue->queuedata; struct nvme_rdma_queue *queue =3D hctx->driver_data; struct request *rq =3D bd->rq; struct nvme_rdma_request *req =3D blk_mq_rq_to_pdu(rq); @@ -2007,7 +2007,7 @@ static blk_status_t nvme_rdma_queue_rq(struct blk_mq_= hw_ctx *hctx, ib_dma_sync_single_for_cpu(dev, sqe->dma, sizeof(struct nvme_command), DMA_TO_DEVICE); =20 - ret =3D nvme_setup_cmd(ns, rq); + ret =3D nvme_setup_cmd(head, rq); if (ret) goto unmap_qe; =20 @@ -2017,7 +2017,7 @@ static blk_status_t nvme_rdma_queue_rq(struct blk_mq_= hw_ctx *hctx, queue->pi_support && (c->common.opcode =3D=3D nvme_cmd_write || c->common.opcode =3D=3D nvme_cmd_read) && - nvme_ns_has_pi(ns)) + nvme_ns_has_pi(head)) req->use_sig_mr =3D true; else req->use_sig_mr =3D false; diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c index 08805f027810..169462faad47 100644 --- a/drivers/nvme/host/tcp.c +++ b/drivers/nvme/host/tcp.c @@ -2483,7 +2483,7 @@ static blk_status_t nvme_tcp_map_data(struct nvme_tcp= _queue *queue, return 0; } =20 -static blk_status_t nvme_tcp_setup_cmd_pdu(struct nvme_ns *ns, +static blk_status_t nvme_tcp_setup_cmd_pdu(struct nvme_ns_head *head, struct request *rq) { struct nvme_tcp_request *req =3D blk_mq_rq_to_pdu(rq); @@ -2492,7 +2492,7 @@ static blk_status_t nvme_tcp_setup_cmd_pdu(struct nvm= e_ns *ns, u8 hdgst =3D nvme_tcp_hdgst_len(queue), ddgst =3D 0; blk_status_t ret; =20 - ret =3D nvme_setup_cmd(ns, rq); + ret =3D nvme_setup_cmd(head, rq); if (ret) return ret; =20 @@ -2548,7 +2548,7 @@ static void nvme_tcp_commit_rqs(struct blk_mq_hw_ctx = *hctx) static blk_status_t nvme_tcp_queue_rq(struct blk_mq_hw_ctx *hctx, const struct blk_mq_queue_data *bd) { - struct nvme_ns *ns =3D hctx->queue->queuedata; + struct nvme_ns_head *head =3D hctx->queue->queuedata; struct nvme_tcp_queue *queue =3D hctx->driver_data; struct request *rq =3D bd->rq; struct nvme_tcp_request *req =3D blk_mq_rq_to_pdu(rq); @@ -2558,7 +2558,7 @@ static blk_status_t nvme_tcp_queue_rq(struct blk_mq_h= w_ctx *hctx, if (!nvme_check_ready(&queue->ctrl->ctrl, rq, queue_ready)) return nvme_fail_nonready_command(&queue->ctrl->ctrl, rq); =20 - ret =3D nvme_tcp_setup_cmd_pdu(ns, rq); + ret =3D nvme_tcp_setup_cmd_pdu(head, rq); if (unlikely(ret)) return ret; =20 diff --git a/drivers/nvme/host/zns.c b/drivers/nvme/host/zns.c index ec8557810c21..51ac3a61ce3f 100644 --- a/drivers/nvme/host/zns.c +++ b/drivers/nvme/host/zns.c @@ -11,7 +11,7 @@ int nvme_revalidate_zones(struct nvme_ns *ns) { struct request_queue *q =3D ns->queue; =20 - blk_queue_chunk_sectors(q, ns->zsze); + blk_queue_chunk_sectors(q, ns->head->zsze); blk_queue_max_zone_append_sectors(q, ns->ctrl->max_zone_append); =20 return blk_revalidate_disk_zones(ns->disk, NULL); @@ -99,11 +99,12 @@ int nvme_update_zone_info(struct nvme_ns *ns, unsigned = lbaf) goto free_data; } =20 - ns->zsze =3D nvme_lba_to_sect(ns, le64_to_cpu(id->lbafe[lbaf].zsze)); - if (!is_power_of_2(ns->zsze)) { + ns->head->zsze =3D + nvme_lba_to_sect(ns->head, le64_to_cpu(id->lbafe[lbaf].zsze)); + if (!is_power_of_2(ns->head->zsze)) { dev_warn(ns->ctrl->device, "invalid zone size:%llu for namespace:%u\n", - ns->zsze, ns->head->ns_id); + ns->head->zsze, ns->head->ns_id); status =3D -ENODEV; goto free_data; } @@ -128,7 +129,7 @@ static void *nvme_zns_alloc_report_buffer(struct nvme_n= s *ns, sizeof(struct nvme_zone_descriptor); =20 nr_zones =3D min_t(unsigned int, nr_zones, - get_capacity(ns->disk) >> ilog2(ns->zsze)); + get_capacity(ns->head->disk) >> ilog2(ns->head->zsze)); =20 bufsize =3D sizeof(struct nvme_zone_report) + nr_zones * sizeof(struct nvme_zone_descriptor); @@ -162,13 +163,13 @@ static int nvme_zone_parse_entry(struct nvme_ns *ns, =20 zone.type =3D BLK_ZONE_TYPE_SEQWRITE_REQ; zone.cond =3D entry->zs >> 4; - zone.len =3D ns->zsze; - zone.capacity =3D nvme_lba_to_sect(ns, le64_to_cpu(entry->zcap)); - zone.start =3D nvme_lba_to_sect(ns, le64_to_cpu(entry->zslba)); + zone.len =3D ns->head->zsze; + zone.capacity =3D nvme_lba_to_sect(ns->head, le64_to_cpu(entry->zcap)); + zone.start =3D nvme_lba_to_sect(ns->head, le64_to_cpu(entry->zslba)); if (zone.cond =3D=3D BLK_ZONE_COND_FULL) zone.wp =3D zone.start + zone.len; else - zone.wp =3D nvme_lba_to_sect(ns, le64_to_cpu(entry->wp)); + zone.wp =3D nvme_lba_to_sect(ns->head, le64_to_cpu(entry->wp)); =20 return cb(&zone, idx, data); } @@ -196,11 +197,11 @@ int nvme_ns_report_zones(struct nvme_ns *ns, sector_t= sector, c.zmr.zrasf =3D NVME_ZRASF_ZONE_REPORT_ALL; c.zmr.pr =3D NVME_REPORT_ZONE_PARTIAL; =20 - sector &=3D ~(ns->zsze - 1); + sector &=3D ~(ns->head->zsze - 1); while (zone_idx < nr_zones && sector < get_capacity(ns->disk)) { memset(report, 0, buflen); =20 - c.zmr.slba =3D cpu_to_le64(nvme_sect_to_lba(ns, sector)); + c.zmr.slba =3D cpu_to_le64(nvme_sect_to_lba(ns->head, sector)); ret =3D nvme_submit_sync_cmd(ns->queue, &c, report, buflen); if (ret) { if (ret > 0) @@ -220,7 +221,7 @@ int nvme_ns_report_zones(struct nvme_ns *ns, sector_t s= ector, zone_idx++; } =20 - sector +=3D ns->zsze * nz; + sector +=3D ns->head->zsze * nz; } =20 if (zone_idx > 0) @@ -232,14 +233,15 @@ int nvme_ns_report_zones(struct nvme_ns *ns, sector_t= sector, return ret; } =20 -blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns *ns, struct request = *req, - struct nvme_command *c, enum nvme_zone_mgmt_action action) +blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns_head *head, + struct request *req, struct nvme_command *c, + enum nvme_zone_mgmt_action action) { memset(c, 0, sizeof(*c)); =20 c->zms.opcode =3D nvme_cmd_zone_mgmt_send; - c->zms.nsid =3D cpu_to_le32(ns->head->ns_id); - c->zms.slba =3D cpu_to_le64(nvme_sect_to_lba(ns, blk_rq_pos(req))); + c->zms.nsid =3D cpu_to_le32(head->ns_id); + c->zms.slba =3D cpu_to_le64(nvme_sect_to_lba(head, blk_rq_pos(req))); c->zms.zsa =3D action; =20 if (req_op(req) =3D=3D REQ_OP_ZONE_RESET_ALL) diff --git a/drivers/nvme/target/loop.c b/drivers/nvme/target/loop.c index 9cb434c58075..1c3abadecaa7 100644 --- a/drivers/nvme/target/loop.c +++ b/drivers/nvme/target/loop.c @@ -131,7 +131,7 @@ static void nvme_loop_execute_work(struct work_struct *= work) static blk_status_t nvme_loop_queue_rq(struct blk_mq_hw_ctx *hctx, const struct blk_mq_queue_data *bd) { - struct nvme_ns *ns =3D hctx->queue->queuedata; + struct nvme_ns_head *head =3D hctx->queue->queuedata; struct nvme_loop_queue *queue =3D hctx->driver_data; struct request *req =3D bd->rq; struct nvme_loop_iod *iod =3D blk_mq_rq_to_pdu(req); @@ -141,7 +141,7 @@ static blk_status_t nvme_loop_queue_rq(struct blk_mq_hw= _ctx *hctx, if (!nvme_check_ready(&queue->ctrl->ctrl, req, queue_ready)) return nvme_fail_nonready_command(&queue->ctrl->ctrl, req); =20 - ret =3D nvme_setup_cmd(ns, req); + ret =3D nvme_setup_cmd(head, req); if (ret) return ret; =20 diff --git a/drivers/nvme/target/passthru.c b/drivers/nvme/target/passthru.c index 9fe07d7efa96..7c13084a8f87 100644 --- a/drivers/nvme/target/passthru.c +++ b/drivers/nvme/target/passthru.c @@ -216,11 +216,11 @@ static void nvmet_passthru_execute_cmd_work(struct wo= rk_struct *w) struct nvmet_req *req =3D container_of(w, struct nvmet_req, p.work); struct request *rq =3D req->p.rq; struct nvme_ctrl *ctrl =3D nvme_req(rq)->ctrl; - struct nvme_ns *ns =3D rq->q->queuedata; + struct nvme_ns_head *head =3D rq->q->queuedata; u32 effects; int status; =20 - effects =3D nvme_passthru_start(ctrl, ns, req->cmd->common.opcode); + effects =3D nvme_passthru_start(ctrl, head, req->cmd->common.opcode); status =3D nvme_execute_rq(rq, false); if (status =3D=3D NVME_SC_SUCCESS && req->cmd->common.opcode =3D=3D nvme_admin_identify) { @@ -243,7 +243,7 @@ static void nvmet_passthru_execute_cmd_work(struct work= _struct *w) blk_mq_free_request(rq); =20 if (effects) - nvme_passthru_end(ctrl, ns, effects, req->cmd, status); + nvme_passthru_end(ctrl, head, effects, req->cmd, status); } =20 static enum rq_end_io_ret nvmet_passthru_req_done(struct request *rq, @@ -339,7 +339,7 @@ static void nvmet_passthru_execute_cmd(struct nvmet_req= *req) * non-trivial effects, make sure to execute the command synchronously * in a workqueue so that nvme_passthru_end gets called. */ - effects =3D nvme_command_effects(ctrl, ns, req->cmd->common.opcode); + effects =3D nvme_command_effects(ctrl, ns->head, req->cmd->common.opcode); if (req->p.use_workqueue || (effects & ~(NVME_CMD_EFFECTS_CSUPP | NVME_CMD_EFFECTS_LBCC))) { INIT_WORK(&req->p.work, nvmet_passthru_execute_cmd_work); --=20 2.43.0 From nobody Sun Dec 28 17:28:55 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id BE714C4167B for ; Wed, 6 Dec 2023 08:13:03 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1377126AbjLFIMy (ORCPT ); Wed, 6 Dec 2023 03:12:54 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59548 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231697AbjLFIMr (ORCPT ); Wed, 6 Dec 2023 03:12:47 -0500 Received: from smtp-out1.suse.de (smtp-out1.suse.de [IPv6:2a07:de40:b251:101:10:150:64:1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id F3AFD18D for ; Wed, 6 Dec 2023 00:12:52 -0800 (PST) Received: from imap2.dmz-prg2.suse.org (imap2.dmz-prg2.suse.org [IPv6:2a07:de40:b281:104:10:150:64:98]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by smtp-out1.suse.de (Postfix) with ESMTPS id 86C6222072; Wed, 6 Dec 2023 08:12:51 +0000 (UTC) Received: from imap2.dmz-prg2.suse.org (localhost [127.0.0.1]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by imap2.dmz-prg2.suse.org (Postfix) with ESMTPS id 74358133DD; Wed, 6 Dec 2023 08:12:51 +0000 (UTC) Received: from dovecot-director2.suse.de ([10.150.64.162]) by imap2.dmz-prg2.suse.org with ESMTPSA id k+DmGgMtcGUZHgAAn2gu4w (envelope-from ); Wed, 06 Dec 2023 08:12:51 +0000 From: Daniel Wagner To: linux-nvme@lists.infradead.org Cc: linux-kernel@vger.kernel.org, Keith Busch , Christoph Hellwig , Sagi Grimberg , Hannes Reinecke , Daniel Wagner Subject: [PATCH v3 4/4] nvme: add csi, ms and nuse to sysfs Date: Wed, 6 Dec 2023 09:12:44 +0100 Message-ID: <20231206081244.32733-5-dwagner@suse.de> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20231206081244.32733-1-dwagner@suse.de> References: <20231206081244.32733-1-dwagner@suse.de> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Spamd-Result: default: False [4.09 / 50.00]; ARC_NA(0.00)[]; RCVD_VIA_SMTP_AUTH(0.00)[]; RBL_SPAMHAUS_BLOCKED_OPENRESOLVER(0.00)[2a07:de40:b281:104:10:150:64:98:from]; FROM_HAS_DN(0.00)[]; TO_DN_SOME(0.00)[]; R_MISSING_CHARSET(2.50)[]; TO_MATCH_ENVRCPT_ALL(0.00)[]; MIME_GOOD(-0.10)[text/plain]; NEURAL_SPAM_SHORT(2.10)[0.700]; BROKEN_CONTENT_TYPE(1.50)[]; R_SPF_SOFTFAIL(0.00)[~all:c]; RCVD_COUNT_THREE(0.00)[3]; MX_GOOD(-0.01)[]; RCPT_COUNT_SEVEN(0.00)[7]; MID_CONTAINS_FROM(1.00)[]; DBL_BLOCKED_OPENRESOLVER(0.00)[suse.de:email]; FUZZY_BLOCKED(0.00)[rspamd.com]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(0.00)[]; MIME_TRACE(0.00)[0:+]; RCVD_TLS_ALL(0.00)[]; BAYES_HAM(-3.00)[100.00%]; DMARC_POLICY_SOFTFAIL(0.10)[suse.de : No valid SPF, No valid DKIM,none] X-Spamd-Bar: ++++ X-Rspamd-Server: rspamd1 Authentication-Results: smtp-out1.suse.de; dkim=none; dmarc=fail reason="No valid SPF, No valid DKIM" header.from=suse.de (policy=none); spf=softfail (smtp-out1.suse.de: 2a07:de40:b281:104:10:150:64:98 is neither permitted nor denied by domain of dwagner@suse.de) smtp.mailfrom=dwagner@suse.de X-Rspamd-Queue-Id: 86C6222072 Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" libnvme is using the sysfs for enumarating the nvme resources. Though there are few missing attritbutes in the sysfs. For these libnvme issues commands during discovering. As the kernel already knows all these attributes and we would like to avoid libnvme to issue commands all the time, expose these missing attributes. The nuse value is updated on request because the nuse is a volatile value. Since any user can read the sysfs attribute, a very simple rate limit is added (update max every 5 seconds). A more sophisticated update strategy can be added later if there is actually a need for it. Signed-off-by: Daniel Wagner --- drivers/nvme/host/core.c | 28 ++++++++++++++++++++++++++++ drivers/nvme/host/nvme.h | 2 ++ drivers/nvme/host/sysfs.c | 31 +++++++++++++++++++++++++++++++ 3 files changed, 61 insertions(+) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index caa52c2f57c8..e7dd64ee1653 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -1663,6 +1663,33 @@ static void nvme_ns_release(struct nvme_ns *ns) nvme_put_ns(ns); } =20 +int nvme_ns_update_nuse(struct nvme_ns_head *head) +{ + static DEFINE_RATELIMIT_STATE(_rs, 5 * HZ, 1); + struct nvme_id_ns *id; + struct nvme_ns *ns; + int srcu_idx, ret =3D -EWOULDBLOCK; + + if (__ratelimit(&_rs)) + return 0; + + srcu_idx =3D srcu_read_lock(&head->srcu); + ns =3D nvme_find_path(head); + if (!ns) + goto out_unlock; + + ret =3D nvme_identify_ns(ns->ctrl, head->ns_id, &id); + if (ret) + goto out_unlock; + + head->nuse =3D le64_to_cpu(id->nuse); + kfree(id); + +out_unlock: + srcu_read_unlock(&head->srcu, srcu_idx); + return ret; +} + static int nvme_open(struct gendisk *disk, blk_mode_t mode) { return nvme_ns_open(disk->private_data); @@ -2068,6 +2095,7 @@ static int nvme_update_ns_info_block(struct nvme_ns *= ns, blk_mq_freeze_queue(ns->disk->queue); lbaf =3D nvme_lbaf_index(id->flbas); ns->head->lba_shift =3D id->lbaf[lbaf].ds; + ns->head->nuse =3D le64_to_cpu(id->nuse); nvme_set_queue_limits(ns->ctrl, ns->queue); =20 ret =3D nvme_configure_metadata(ns, id); diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h index e6c7890b14c9..6a928646dc09 100644 --- a/drivers/nvme/host/nvme.h +++ b/drivers/nvme/host/nvme.h @@ -456,6 +456,7 @@ struct nvme_ns_head { u16 pi_size; u16 sgs; u32 sws; + u64 nuse; u8 pi_type; u8 guard_type; #ifdef CONFIG_BLK_DEV_ZONED @@ -867,6 +868,7 @@ int nvme_ns_chr_uring_cmd(struct io_uring_cmd *ioucmd, unsigned int issue_flags); int nvme_ns_head_chr_uring_cmd(struct io_uring_cmd *ioucmd, unsigned int issue_flags); +int nvme_ns_update_nuse(struct nvme_ns_head *head); int nvme_getgeo(struct block_device *bdev, struct hd_geometry *geo); int nvme_dev_uring_cmd(struct io_uring_cmd *ioucmd, unsigned int issue_fla= gs); =20 diff --git a/drivers/nvme/host/sysfs.c b/drivers/nvme/host/sysfs.c index c6b7fbd4d34d..c24854eca496 100644 --- a/drivers/nvme/host/sysfs.c +++ b/drivers/nvme/host/sysfs.c @@ -114,12 +114,43 @@ static ssize_t nsid_show(struct device *dev, struct d= evice_attribute *attr, } static DEVICE_ATTR_RO(nsid); =20 +static ssize_t csi_show(struct device *dev, struct device_attribute *attr, + char *buf) +{ + return sysfs_emit(buf, "%d\n", dev_to_ns_head(dev)->ids.csi); +} +static DEVICE_ATTR_RO(csi); + +static ssize_t metadata_bytes_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + return sysfs_emit(buf, "%d\n", dev_to_ns_head(dev)->ms); +} +static DEVICE_ATTR_RO(metadata_bytes); + +static ssize_t nuse_show(struct device *dev, struct device_attribute *attr, + char *buf) +{ + struct nvme_ns_head *head =3D dev_to_ns_head(dev); + int ret; + + ret =3D nvme_ns_update_nuse(head); + if (ret) + return ret; + + return sysfs_emit(buf, "%llu\n", head->nuse); +} +static DEVICE_ATTR_RO(nuse); + static struct attribute *nvme_ns_id_attrs[] =3D { &dev_attr_wwid.attr, &dev_attr_uuid.attr, &dev_attr_nguid.attr, &dev_attr_eui.attr, + &dev_attr_csi.attr, &dev_attr_nsid.attr, + &dev_attr_metadata_bytes.attr, + &dev_attr_nuse.attr, #ifdef CONFIG_NVME_MULTIPATH &dev_attr_ana_grpid.attr, &dev_attr_ana_state.attr, --=20 2.43.0