Linux Device Mapper development
 help / color / mirror / Atom feed
From: "Benjamin Marzinski" <bmarzins@redhat.com>
To: Keith Busch <keith.busch@intel.com>
Cc: dm-devel@redhat.com
Subject: Re: [PATCH 2/2] Fill NVMe specific path info
Date: Mon, 20 Feb 2017 11:57:59 -0600	[thread overview]
Message-ID: <20170220175759.GA22981@octiron.msp.redhat.com> (raw)
In-Reply-To: <1487107154-24883-2-git-send-email-keith.busch@intel.com>

On Tue, Feb 14, 2017 at 04:19:14PM -0500, Keith Busch wrote:
> Signed-off-by: Keith Busch <keith.busch@intel.com>
> ---
> Pat of this is dependent on udev updates. Pull request sent to systemd here:
> 
>   https://github.com/systemd/systemd/pull/5348
> 
> Can always add that line by hand in the mean time.
> 
>  libmultipath/discovery.c | 36 ++++++++++++++++++++++++++++++++++++
>  libmultipath/hwtable.c   |  9 +++++++++
>  libmultipath/structs.h   |  1 +
>  3 files changed, 46 insertions(+)
> 
> diff --git a/libmultipath/discovery.c b/libmultipath/discovery.c
> index d1aec31..db7b04a 100644
> --- a/libmultipath/discovery.c
> +++ b/libmultipath/discovery.c
> @@ -1187,6 +1187,37 @@ scsi_sysfs_pathinfo (struct path * pp, vector hwtable)
>  }
>  
>  static int
> +nvme_sysfs_pathinfo (struct path * pp, vector hwtable)
> +{
> +	struct udev_device *parent, *nvme = NULL;
> +
> +	parent = pp->udev;
> +	while (parent) {
> +		const char *subsys = udev_device_get_subsystem(parent);
> +
> +		if (subsys && !strncmp(subsys, "nvme", 4)) {
> +			nvme = parent;
> +			break;
> +		}
> +		parent = udev_device_get_parent(parent);
> +	}
> +	if (!nvme)
> +		return 1;
> +
> +	snprintf(pp->vendor_id, SCSI_VENDOR_SIZE, "NVME");
> +	snprintf(pp->product_id, SCSI_PRODUCT_SIZE, "%s", udev_device_get_sysattr_value(nvme, "model"));
> +	snprintf(pp->serial, SERIAL_SIZE, "%s", udev_device_get_sysattr_value(nvme, "serial"));
> +	snprintf(pp->rev, SCSI_REV_SIZE, "%s", udev_device_get_sysattr_value(nvme, "firmware_rev"));
> +	snprintf(pp->wwid, WWID_SIZE, "%s", udev_device_get_sysattr_value(pp->udev, "wwid"));

With your udev rules change, you shouldn't need to set the wwid here. It
should just get dealt with by the get_uid call in pathinfo (or with
the new uevent merging code, before multipathd ever calls pathinfo).

Otherwise, it looks good to me.

-Ben

> +
> +	condlog(3, "%s: vendor:%s product:%s serial:%s rev:%s wwid:%s", pp->dev,
> +		pp->vendor_id, pp->product_id, pp->serial, pp->rev, pp->wwid);
> +	pp->hwe = find_hwe(hwtable, pp->vendor_id, pp->product_id, NULL);
> +
> +	return 0;
> +}
> +
> +static int
>  rbd_sysfs_pathinfo (struct path * pp, vector hwtable)
>  {
>  	sprintf(pp->vendor_id, "Ceph");
> @@ -1405,6 +1436,8 @@ sysfs_pathinfo(struct path * pp, vector hwtable)
>  		pp->bus = SYSFS_BUS_SCSI;
>  	if (!strncmp(pp->dev,"rbd", 3))
>  		pp->bus = SYSFS_BUS_RBD;
> +	if (!strncmp(pp->dev,"nvme", 4))
> +		pp->bus = SYSFS_BUS_NVME;
>  
>  	if (pp->bus == SYSFS_BUS_UNDEF)
>  		return 0;
> @@ -1420,6 +1453,9 @@ sysfs_pathinfo(struct path * pp, vector hwtable)
>  	} else if (pp->bus == SYSFS_BUS_RBD) {
>  		if (rbd_sysfs_pathinfo(pp, hwtable))
>  			return 1;
> +	} else if (pp->bus == SYSFS_BUS_NVME) {
> +		if (nvme_sysfs_pathinfo(pp, hwtable))
> +			return 1;
>  	}
>  	return 0;
>  }
> diff --git a/libmultipath/hwtable.c b/libmultipath/hwtable.c
> index f5a5f7b..c55998a 100644
> --- a/libmultipath/hwtable.c
> +++ b/libmultipath/hwtable.c
> @@ -1073,6 +1073,15 @@ static struct hwentry default_hw[] = {
>  		.prio_name     = PRIO_ALUA,
>  		.no_path_retry = 30,
>  	},
> +	/*
> +	 * Generic NVMe devices
> +	 */
> +	{
> +		.vendor	       = "NVME",
> +		.product       = ".*",
> +		.uid_attribute = "ID_WWN",
> +		.checker_name  = DIRECTIO,
> +	},
>  #if 0
>  	/*
>  	 * Copy this TEMPLATE to add new hardware.
> diff --git a/libmultipath/structs.h b/libmultipath/structs.h
> index 6edd927..dfd65ae 100644
> --- a/libmultipath/structs.h
> +++ b/libmultipath/structs.h
> @@ -53,6 +53,7 @@ enum sysfs_buses {
>  	SYSFS_BUS_CCW,
>  	SYSFS_BUS_CCISS,
>  	SYSFS_BUS_RBD,
> +	SYSFS_BUS_NVME,
>  };
>  
>  enum pathstates {
> -- 
> 2.5.5
> 
> --
> dm-devel mailing list
> dm-devel@redhat.com
> https://www.redhat.com/mailman/listinfo/dm-devel

  reply	other threads:[~2017-02-20 17:57 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-02-14 21:19 [PATCH 1/2] Don't blacklist nvme Keith Busch
2017-02-14 21:19 ` [PATCH 2/2] Fill NVMe specific path info Keith Busch
2017-02-20 17:57   ` Benjamin Marzinski [this message]
2017-02-21 21:06     ` Keith Busch
2017-02-14 21:35 ` [PATCH 1/2] Don't blacklist nvme Bart Van Assche
2017-02-14 23:00   ` Keith Busch
2017-02-15 14:57     ` Christoph Hellwig
2017-02-15 17:24       ` Keith Busch
2017-02-16  1:58         ` Mike Snitzer
2017-02-16  2:01     ` Mike Snitzer
2017-02-16  2:35       ` Mike Snitzer
2017-02-15 14:56 ` Christoph Hellwig
2017-02-16  2:53   ` hch's native NVMe multipathing [was: Re: [PATCH 1/2] Don't blacklist nvme] Mike Snitzer
2017-02-16  5:00     ` Bart Van Assche
2017-02-16 12:37       ` Mike Snitzer
2017-02-16 19:46         ` Bart Van Assche
2017-02-16 20:23           ` Mike Snitzer
2017-02-16 20:58             ` Bart Van Assche
2017-02-16 14:26     ` Christoph Hellwig
2017-02-16 15:13       ` Mike Snitzer
2017-02-16 17:38         ` Keith Busch
2017-02-16 17:37           ` Bart Van Assche
2017-02-16 18:07             ` Keith Busch
2017-02-16 18:21               ` Mike Snitzer
2017-02-16 20:40                 ` Keith Busch
2017-02-17  9:04                 ` [dm-devel] " hch
2017-02-17 14:43                   ` Mike Snitzer
2017-02-16 18:05         ` Sagi Grimberg
2017-02-17  9:05           ` [dm-devel] " Christoph Hellwig
2017-02-17 14:37             ` Mike Snitzer
2017-02-17  9:33         ` [dm-devel] " Christoph Hellwig
2017-02-17 14:32           ` Mike Snitzer
2017-02-20 18:17   ` [PATCH 1/2] Don't blacklist nvme Benjamin Marzinski
2017-02-20 14:14 ` Mike Snitzer
2017-02-27  5:37 ` Christophe Varoqui

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170220175759.GA22981@octiron.msp.redhat.com \
    --to=bmarzins@redhat.com \
    --cc=dm-devel@redhat.com \
    --cc=keith.busch@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox