From mboxrd@z Thu Jan 1 00:00:00 1970 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 bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id ED2FFD15D8A for ; Mon, 21 Oct 2024 12:42:20 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender:List-Subscribe:List-Help :List-Post:List-Archive:List-Unsubscribe:List-Id:Content-Transfer-Encoding: Content-Type:In-Reply-To:From:References:Cc:To:Subject:MIME-Version:Date: Message-ID:Reply-To:Content-ID:Content-Description:Resent-Date:Resent-From: Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner; bh=kxO8TqmKzbOLGur78jYVmcXl1m76p1c5MxmBr7LrheM=; b=XrDdW1aZ9p87RJXPFBktfbyIuP FpMppJzLtDB0/sO4R2eKdC9RJAX2gfQxsQRCYOpCA5ycPhDen9/DBi4BgtL3mwC2IV4ngP5rTdYI0 2Vcpt45CAMDIQcJTLPxq5YCAB1HarA2rjxyMYyFZ17+jsSWLzgu6mHE4fOUuAxqaFkVBtmC3SbSYp cWIQ2miqrUSTIgzeIglhnq8fixBNMGpEhU3clc/Eqnv3KjOrS6ko5ypnwBw+bWSrdR5r3cHDdMZ2P Z5vnWaKZ2yw0BwwdPSIfXyK05AkOmkmTOQ0A4rcsdxi0qGJ53Jij8b3vAO8bveTJypafI+dXyWDUm GbYR90yg==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.98 #2 (Red Hat Linux)) id 1t2rjk-00000007IN9-0BvO; Mon, 21 Oct 2024 12:42:16 +0000 Received: from mx0b-001b2d01.pphosted.com ([148.163.158.5]) by bombadil.infradead.org with esmtps (Exim 4.98 #2 (Red Hat Linux)) id 1t2rSz-00000007EKh-1pLq for linux-nvme@lists.infradead.org; Mon, 21 Oct 2024 12:24:59 +0000 Received: from pps.filterd (m0360072.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.18.1.2/8.18.1.2) with ESMTP id 49L2Kwpc028018; Mon, 21 Oct 2024 12:24:48 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ibm.com; h=cc :content-transfer-encoding:content-type:date:from:in-reply-to :message-id:mime-version:references:subject:to; s=pp1; bh=kxO8Tq mKzbOLGur78jYVmcXl1m76p1c5MxmBr7LrheM=; b=EKodM9c0vPzpTq8qcyfEnd mlgFA8SoayuhsRlJNU1XE0sQocGoY9F+hWZGuakZlwBvd1xkkQAFwLiYtLud+4+W HrA+DrSGU2JlBEs3cUZcA2cWwkAOtiz6W7H7D6DOmfa1+QeiBcjI66WPgaQj8Y2l GmHTwP8kwaFsffyS5lGmDChyQ2j2rDEqwTGSNQHrDe2JC2ElXmwIzOAkZiPauueA iG0rukBglsvSxbx7QfTglZ57OZxK4wGiSHxL0j316WHATfcjMPNFNxoLv0FHSLHk w2QTAIQqEL1Ey2IIO2QAtTdqzP0SoL0TFKbrmpKMTmEy1dyRPuewx+b1fUf2qC1g == Received: from ppma13.dal12v.mail.ibm.com (dd.9e.1632.ip4.static.sl-reverse.com [50.22.158.221]) by mx0a-001b2d01.pphosted.com (PPS) with ESMTPS id 42c5fcgnyj-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 21 Oct 2024 12:24:48 +0000 (GMT) Received: from pps.filterd (ppma13.dal12v.mail.ibm.com [127.0.0.1]) by ppma13.dal12v.mail.ibm.com (8.18.1.2/8.18.1.2) with ESMTP id 49L9PxIL018590; Mon, 21 Oct 2024 12:24:47 GMT Received: from smtprelay01.wdc07v.mail.ibm.com ([172.16.1.68]) by ppma13.dal12v.mail.ibm.com (PPS) with ESMTPS id 42csaj61qk-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 21 Oct 2024 12:24:47 +0000 Received: from smtpav05.wdc07v.mail.ibm.com (smtpav05.wdc07v.mail.ibm.com [10.39.53.232]) by smtprelay01.wdc07v.mail.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 49LCOkbI35914374 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 21 Oct 2024 12:24:46 GMT Received: from smtpav05.wdc07v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id AC5885805D; Mon, 21 Oct 2024 12:24:46 +0000 (GMT) Received: from smtpav05.wdc07v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id C764E58059; Mon, 21 Oct 2024 12:24:43 +0000 (GMT) Received: from [9.171.30.180] (unknown [9.171.30.180]) by smtpav05.wdc07v.mail.ibm.com (Postfix) with ESMTP; Mon, 21 Oct 2024 12:24:43 +0000 (GMT) Message-ID: <62a1e828-cb14-4d70-88a6-04b1ebdd089d@linux.ibm.com> Date: Mon, 21 Oct 2024 17:54:40 +0530 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCHv4 RFC 1/1] nvme-multipath: Add sysfs attributes for showing multipath info To: Hannes Reinecke , linux-nvme@lists.infradead.org Cc: dwagner@suse.de, hch@lst.de, kbusch@kernel.org, sagi@grimberg.me, axboe@fb.com, gjoyce@linux.ibm.com References: <20240911062653.1060056-1-nilay@linux.ibm.com> <20240911062653.1060056-4-nilay@linux.ibm.com> <5050777a-2812-4fcf-bed9-00f0cb5706fc@suse.de> <93af0dc5-c988-423c-9788-e93ca8703fd5@linux.ibm.com> <0f4528f0-8ef3-449f-bb15-47878d93a002@suse.de> <48bf8a1d-3a47-4c94-8cd0-7d8c4ea5d3ab@linux.ibm.com> <2476d00b-35d3-4516-b76c-9b76a499e4fc@linux.ibm.com> <3c92b9ae-1223-4787-90e3-955f82161269@suse.de> Content-Language: en-US From: Nilay Shroff In-Reply-To: <3c92b9ae-1223-4787-90e3-955f82161269@suse.de> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-TM-AS-GCONF: 00 X-Proofpoint-ORIG-GUID: 8M8I41vd3YcMDT_y_f_rXX1Skff7LuU_ X-Proofpoint-GUID: 8M8I41vd3YcMDT_y_f_rXX1Skff7LuU_ X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.293,Aquarius:18.0.1051,Hydra:6.0.680,FMLib:17.12.62.30 definitions=2024-10-15_01,2024-10-11_01,2024-09-30_01 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 priorityscore=1501 bulkscore=0 mlxscore=0 clxscore=1015 mlxlogscore=999 phishscore=0 impostorscore=0 malwarescore=0 lowpriorityscore=0 adultscore=0 suspectscore=0 spamscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.19.0-2409260000 definitions=main-2410210087 X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20241021_052457_794564_60009FDE X-CRM114-Status: GOOD ( 27.98 ) X-BeenThere: linux-nvme@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "Linux-nvme" Errors-To: linux-nvme-bounces+linux-nvme=archiver.kernel.org@lists.infradead.org On 10/16/24 12:22, Hannes Reinecke wrote: > On 10/16/24 05:19, Nilay Shroff wrote: >> Hi Hannes, >> >> On 10/7/24 21:03, Nilay Shroff wrote: >>> >>> >>> On 10/7/24 19:34, Hannes Reinecke wrote: >>>> On 10/7/24 15:47, Nilay Shroff wrote: >>>>> >>>>> >>>>> On 10/7/24 15:44, Hannes Reinecke wrote: >>>>>> On 9/11/24 08:26, Nilay Shroff wrote: >>>>>>> NVMe native multipath supports different IO policies for selecting I/O >>>>>>> path, however we don't have any visibility about which path is being >>>>>>> selected by multipath code for forwarding I/O. >>>>>>> This patch helps add that visibility by adding new sysfs attribute files >>>>>>> named "numa_nodes" and "queue_depth" under each namespace block device >>>>>>> path /sys/block/nvmeXcYnZ/. We also create a "multipath" sysfs directory >>>>>>> under head disk node and then from this directory add a link to each >>>>>>> namespace path device this head disk node points to. >>>>>>> >>>>>>> For instance, /sys/block/nvmeXnY/multipath/ would create a soft link to >>>>>>> each path the head disk node points to: >>>>>>> >>>>>>> $ ls -1 /sys/block/nvme1n1/ >>>>>>> nvme1c1n1 -> ../../../../../pci052e:78/052e:78:00.0/nvme/nvme1/nvme1c1n1 >>>>>>> nvme1c3n1 -> ../../../../../pci058e:78/058e:78:00.0/nvme/nvme3/nvme1c3n1 >>>>>>> >>>>>>> For round-robin I/O policy, we could easily infer from the above output >>>>>>> that I/O workload targeted to nvme3n1 would toggle across paths nvme1c1n1 >>>>>>> and nvme1c3n1. >>>>>>> >>>>>>> For numa I/O policy, the "numa_nodes" attribute file shows the numa nodes >>>>>>> being preferred by the respective block device path. The numa nodes value >>>>>>> is comma delimited list of nodes or A-B range of nodes. >>>>>>> >>>>>>> For queue-depth I/O policy, the "queue_depth" attribute file shows the >>>>>>> number of active/in-flight I/O requests currently queued for each path. >>>>>>> >>>>>>> Signed-off-by: Nilay Shroff >>>>>>> --- >>>>>>>     drivers/nvme/host/core.c      |  3 ++ >>>>>>>     drivers/nvme/host/multipath.c | 71 +++++++++++++++++++++++++++++++++++ >>>>>>>     drivers/nvme/host/nvme.h      | 20 ++++++++-- >>>>>>>     drivers/nvme/host/sysfs.c     | 20 ++++++++++ >>>>>>>     4 files changed, 110 insertions(+), 4 deletions(-) >>>>>>> >>>>>>> diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c >>>>>>> index 983909a600ad..6be29fd64236 100644 >>>>>>> --- a/drivers/nvme/host/core.c >>>>>>> +++ b/drivers/nvme/host/core.c >>>>>>> @@ -3951,6 +3951,9 @@ static void nvme_ns_remove(struct nvme_ns *ns) >>>>>>>           if (!nvme_ns_head_multipath(ns->head)) >>>>>>>             nvme_cdev_del(&ns->cdev, &ns->cdev_device); >>>>>>> + >>>>>>> +    nvme_mpath_remove_sysfs_link(ns); >>>>>>> + >>>>>>>         del_gendisk(ns->disk); >>>>>>>           mutex_lock(&ns->ctrl->namespaces_lock); >>>>>>> diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c >>>>>>> index 518e22dd4f9b..7d9c36a7a261 100644 >>>>>>> --- a/drivers/nvme/host/multipath.c >>>>>>> +++ b/drivers/nvme/host/multipath.c >>>>>>> @@ -654,6 +654,8 @@ static void nvme_mpath_set_live(struct nvme_ns *ns) >>>>>>>             nvme_add_ns_head_cdev(head); >>>>>>>         } >>>>>>>     +    nvme_mpath_add_sysfs_link(ns); >>>>>>> + >>>>>>>         mutex_lock(&head->lock); >>>>>>>         if (nvme_path_is_optimized(ns)) { >>>>>>>             int node, srcu_idx; >>>>>> Nearly there. >>>>> Thank you for your review comments! >>>>> >>>>>> >>>>>> You can only call 'nvme_mpath_add_sysfs_link()' if the gendisk on the head had been created. >>>>>> >>>>>> And there is one branch in nvme_mpath_add_disk(): >>>>>> >>>>>>                   if (desc.state) { >>>>>>                           /* found the group desc: update */ >>>>>>                           nvme_update_ns_ana_state(&desc, ns); >>>>>> >>>>>> which does not go via nvme_mpath_set_live(), yet a device link would need to be create here, too. >>>>>> But you can't call nvme_mpath_add_sysfs_link() from nvme_mpath_add_disk(), as the actual gendisk >>>>>> might only be created later on during ANA log parsing.>> >>>>>> It is a tangle, and I haven't found a good way out of this. >>>>>> But I am _very much_ in favour of having these links, so please >>>>>> update your patch. >>>>>> In case disk supports ANA group then yes it would go through nvme_mpath_add_disk()->nvme_update_ns_ana_state(); >>>>>> and later nvme_update_ns_ana_state() would also fall through function nvme_mpath_set_live where we call >>>>>> nvme_mpath_add_sysfs_link(). >>>>> >>>>> So I think that in any case while multipath namespace is being created it has to go through >>>>> nvme_mpath_set_live function. And as we see in nvme_mpath_set_live function, we only create >>>>> sysfs link after the gendisk on the head is created. Do you agree with this? Or please let >>>>> me know if you have any further question. >>>>> >>>> But it doesn't: >>>> >>>>          if (nvme_state_is_live(ns->ana_state) && >>>>              nvme_ctrl_state(ns->ctrl) == NVME_CTRL_LIVE) >>>>                  nvme_mpath_set_live(ns); >>>> >>>> so if a namespace is being created for which nvme_state_is_live() returns 'false' nvme_mpath_set_live() is not called, and no links are being created. >>>> This can happen eg. if the first namespace to be encountered is in any other state than 'optimized' or 'non-optimized'. >>>> >>> >>> OK I got what you're suggesting here. So in this particular case when ANA state of a shared namespace >>> is neither "optimized" nor "non-optimized", we would have gendisk for shared namepspace (i.e. >>> nvmeXcYnZ) created but we don't have yet gendisk for corresponding head node (i.e. nvmeXnY) created. >>> So without the gendisk for head node created, how could we create a link from it to the namespace node? >>> >>> The link from gendisk head node would be eventually created when the ANA state of the namespace transition >>> to "optimized" or "non-optimized" state. I think, it's not anyways possible to have multipathing function >>> enabled until the gendisk for the head node created, isn't it? So I don't yet understand why do we really >>> need device link created if we don't have gendisk for head not ready? Am I missing anything here? >>> > You are right, we will create 'the' link for the gendisk node once 'nvme_mpath_set_live()' is called. But: > Then we would have _one_ link (for the active path) only. > All other paths (ie the non-active ones) might already been present, but won't get a link. > > I'd love to have links to all paths present, active or not. > OK, I think I got what you're asking here. I think we can make it the way you want. I will work it out in the next patch. Thanks, --Nilay