Re: [Xen-devel] [BUG] blkback reporting incorrect number of sectors, unable to boot

[Mike Reardon <mule@xxxxxxxx>]

So am I correct in reading this that for at least the foreseeable future storage using 4k sector sizes is not gonna happen?  I'm just trying to figure out if I need to get some different hardware.

Thank you!

On Tue, Nov 7, 2017 at 5:41 AM, Roger Pau Monné <roger.pau@xxxxxxxxxx> wrote:

On Tue, Nov 07, 2017 at 04:31:06AM -0700, Jan Beulich wrote:
> >>> On 07.11.17 at 11:30, <roger.pau@xxxxxxxxxx> wrote:
> > On Mon, Nov 06, 2017 at 05:33:37AM -0700, Jan Beulich wrote:
> >> >>> On 04.11.17 at 05:48, <mule@xxxxxxxx> wrote:
> >> > I added some additional storage to my server with some native 4k sector
> >> > size disks.  The LVM volumes on that array seem to work fine when mounted
> >> > by the host, and when passed through to any of the Linux guests, but
> >> > Windows guests aren't able to use them when using PV drivers.  The work
> >> > fine to install when I first install Windows (Windows 10, latest build) but
> >> > once I install the PV drivers it will no longer boot and give an
> >> > inaccessible boot device error.  If I assign the storage to a different
> >> > Windows guest that already has the drivers installed (as secondary storage,
> >> > not as the boot device) I see the disk listed in disk management, but the
> >> > size of the disk is 8x larger than it should be.  After looking into it a
> >> > bit, the disk is reporting 8x the number of sectors it should have when I
> >> > run xenstore-ls.  Here is the info from xenstore-ls for the relevant volume:
> >> >
> >> >       51712 = ""
> >> >        frontend = "/local/domain/8/device/vbd/51712"
> >> >        params = "/dev/tv_storage/main-storage"
> >> >        script = "/etc/xen/scripts/block"
> >> >        frontend-id = "8"
> >> >        _online_ = "1"
> >> >        removable = "0"
> >> >        bootable = "1"
> >> >        state = "2"
> >> >        dev = "xvda"
> >> >        type = "phy"
> >> >        mode = "w"
> >> >        device-type = "disk"
> >> >        discard-enable = "1"
> >> >        feature-max-indirect-segments = "256"
> >> >        multi-queue-max-queues = "12"
> >> >        max-ring-page-order = "4"
> >> >        physical-device = "fe:0"
> >> >        physical-device-path = "/dev/dm-0"
> >> >        hotplug-status = "connected"
> >> >        feature-flush-cache = "1"
> >> >        feature-discard = "0"
> >> >        feature-barrier = "1"
> >> >        feature-persistent = "1"
> >> >        sectors = "34359738368"
> >> >        info = "0"
> >> >        sector-size = "4096"
> >> >        physical-sector-size = "4096"
> >> >
> >> >
> >> > Here are the numbers for the volume as reported by fdisk:
> >> >
> >> > Disk /dev/tv_storage/main-storage: 16 TiB, 17592186044416 bytes, 4294967296
> >> > sectors
> >> > Units: sectors of 1 * 4096 = 4096 bytes
> >> > Sector size (logical/physical): 4096 bytes / 4096 bytes
> >> > I/O size (minimum/optimal): 4096 bytes / 4096 bytes
> >> > Disklabel type: dos
> >> > Disk identifier: 0x00000000
> >> >
> >> > Device                        Boot Start        End    Sectors Size Id Type
> >> > /dev/tv_storage/main-storage1          1 4294967295 4294967295  16T ee GPT
> >> >
> >> >
> >> > As with the size reported in Windows disk management, the number of sectors
> >> > from xenstore seems is 8x higher than what it should be.  The disks aren't
> >> > using 512b sector emulation, they are natively 4k, so I have no idea where
> >> > the 8x increase is coming from.
> >>
> >> Hmm, looks like a backend problem indeed: struct hd_struct's
> >> nr_sects (which get_capacity() returns) looks to be in 512-byte
> >> units, regardless of actual sector size. Hence the plain
> >> get_capacity() use as well the (wrongly open coded) use of
> >> part_nr_sects_read() looks insufficient in vbd_sz(). Roger,
> >> Konrad?
> >
> > Hm, AFAICT sector-size should always be set to 512.
>
> Which would mean that bdev_logical_block_size() can't be used by
> blkback to set this value. Yet then - what's the point of the xenstore
> setting if it's always the same value anyway?

Some frontends (at least FreeBSD) will choke if sector-size is not
set. So we have the following scenario:

 - Windows: acknowledges sector-size * sectors in order to set disk
   capacity.
 - Linux: sets disk capacity to sectors * 512.
 - FreeBSD: sets disk capacity to sector-size * sectors, will choke if
   sector-size is not set.

In order to keep compatibility with all of them AFAICT the only option
is to hardcode sector-size to 512 in xenstore.

Roger.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel