If the controller is degraded, the driver should stay out of the way so
the user can recover the drive. This patch skips driver initiated async
event requests when the drive is in this state.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

This moves nvme_setup_{flush,discard,rw} calls into a common
nvme_setup_cmd() helper. So we can eventually hide all the command
setup in the core module and don't even need to update the fabrics
drivers for any specific command type.
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

This rewrites nvme_setup_discard() with blk_add_request_payload().
It allocates only the necessary amount(16 bytes) for the payload.
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

The helper returns the number of bytes that need to be mapped
using PRPs/SGL entries.
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

When unloading driver, nvme_disable_io_queues() calls nvme_delete_queue()
that sends nvme_admin_delete_cq command to admin sq. So when the command
completed, the lock acquired by nvme_irq() actually belongs to admin queue.

While the lock that nvme_del_cq_end() trying to acquire belongs to io queue.
So it will not deadlock.

This patch adds lock nesting notation to fix following report.

[  109.840952] =============================================
[  109.846379] [ INFO: possible recursive locking detected ]
[  109.851806] 4.5.0+ #180 Tainted: G            E
[  109.856533] ---------------------------------------------
[  109.861958] swapper/0/0 is trying to acquire lock:
[  109.866771]  (&(&nvmeq->q_lock)->rlock){-.....}, at: [<ffffffffc0820bc6>] nvme_del_cq_end+0x26/0x70 [nvme]
[  109.876535]
[  109.876535] but task is already holding lock:
[  109.882398]  (&(&nvmeq->q_lock)->rlock){-.....}, at: [<ffffffffc0820c2b>] nvme_irq+0x1b/0x50 [nvme]
[  109.891547]
[  109.891547] other info that might help us debug this:
[  109.898107]  Possible unsafe locking scenario:
[  109.898107]
[  109.904056]        CPU0
[  109.906515]        ----
[  109.908974]   lock(&(&nvmeq->q_lock)->rlock);
[  109.913381]   lock(&(&nvmeq->q_lock)->rlock);
[  109.917787]
[  109.917787]  *** DEADLOCK ***
[  109.917787]
[  109.923738]  May be due to missing lock nesting notation
[  109.923738]
[  109.930558] 1 lock held by swapper/0/0:
[  109.934413]  #0:  (&(&nvmeq->q_lock)->rlock){-.....}, at: [<ffffffffc0820c2b>] nvme_irq+0x1b/0x50 [nvme]
[  109.944010]
[  109.944010] stack backtrace:
[  109.948389] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G            E   4.5.0+ #180
[  109.955734] Hardware name: Dell Inc. OptiPlex 7010/0YXT71, BIOS A15 08/12/2013
[  109.962989]  0000000000000000 ffff88011e203c38 ffffffff81383d9c ffffffff81c13540
[  109.970478]  ffffffff826711d0 ffff88011e203ce8 ffffffff810bb429 0000000000000046
[  109.977964]  0000000000000046 0000000000000000 0000000000b2e597 ffffffff81f4cb00
[  109.985453] Call Trace:
[  109.987911]  <IRQ>  [<ffffffff81383d9c>] dump_stack+0x85/0xc9
[  109.993711]  [<ffffffff810bb429>] __lock_acquire+0x19b9/0x1c60
[  109.999575]  [<ffffffff810b6d1d>] ? trace_hardirqs_off+0xd/0x10
[  110.005524]  [<ffffffff810b386d>] ? complete+0x3d/0x50
[  110.010688]  [<ffffffff810bb760>] lock_acquire+0x90/0xf0
[  110.016029]  [<ffffffffc0820bc6>] ? nvme_del_cq_end+0x26/0x70 [nvme]
[  110.022418]  [<ffffffff81772afb>] _raw_spin_lock_irqsave+0x4b/0x60
[  110.028632]  [<ffffffffc0820bc6>] ? nvme_del_cq_end+0x26/0x70 [nvme]
[  110.035019]  [<ffffffffc0820bc6>] nvme_del_cq_end+0x26/0x70 [nvme]
[  110.041232]  [<ffffffff8135b485>] blk_mq_end_request+0x35/0x60
[  110.047095]  [<ffffffffc0821ad8>] nvme_complete_rq+0x68/0x190 [nvme]
[  110.053481]  [<ffffffff8135b53f>] __blk_mq_complete_request+0x8f/0x130
[  110.060043]  [<ffffffff8135b611>] blk_mq_complete_request+0x31/0x40
[  110.066343]  [<ffffffffc08209e3>] __nvme_process_cq+0x83/0x240 [nvme]
[  110.072818]  [<ffffffffc0820c35>] nvme_irq+0x25/0x50 [nvme]
[  110.078419]  [<ffffffff810cdb66>] handle_irq_event_percpu+0x36/0x110
[  110.084804]  [<ffffffff810cdc77>] handle_irq_event+0x37/0x60
[  110.090491]  [<ffffffff810d0ea3>] handle_edge_irq+0x93/0x150
[  110.096180]  [<ffffffff81012306>] handle_irq+0xa6/0x130
[  110.101431]  [<ffffffff81011abe>] do_IRQ+0x5e/0x120
[  110.106333]  [<ffffffff8177384c>] common_interrupt+0x8c/0x8c
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Signed-off-by: NJens Axboe <axboe@fb.com>

Multiple users have reported device initialization failure due the driver
not receiving legacy PCI interrupts. This is not unique to any particular
controller, but has been observed on multiple platforms.

There have been no issues reported or observed when with message signaled
interrupts, so this patch attempts to use MSI-x during initialization,
falling back to MSI. If that fails, legacy would become the default.

The setup_io_queues error handling had to change as a result: the admin
queue's msix_entry used to be initialized to the legacy IRQ. The case
where nr_io_queues is 0 would fail request_irq when setting up the admin
queue's interrupt since re-enabling MSI-x fails with 0 vectors, leaving
the admin queue's msix_entry invalid. Instead, return success immediately.
Reported-by: NTim Muhlemmer <muhlemmer@gmail.com>
Reported-by: NJon Derrick <jonathan.derrick@intel.com>
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Make sure the CQE phase (validity) is read before the rest of the
structure. The phase bit is the highest address and the CQE
read will happen on most platforms from lower to upper addresses
and will be done by multiple non-atomic loads. If the structure
is updated by PCI during the reads from the processor, the
processor may get a corrupted copy.

The addition of the new nvme_cqe_valid function that verifies
the validity bit also allows refactoring of the other CQE read
sequences.
Signed-off-by: NMarta Rybczynska <marta.rybczynska@kalray.eu>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

PPAs sent to device is separately acknowledge in a 64bit status
variable. The status is stored in DW0 and DW1 of the completion queue
entry. Store this status inside the nvm_rq for further processing.

This can later be used to implement retry techniques for failed writes
and reads.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NMatias Bjørling <m@bjorling.me>
Signed-off-by: NJens Axboe <axboe@fb.com>

The method to uniquely identify a namespace depends on the controller's
specification revision level and implemented capabilities. This patch
has the driver figure this out and exports the unique string through a
single 'wwid' attribute so the user doesn't have this burden.

The longest namespace unique identifier is used if available. If not
available, the driver will concat the controller's vendor, serial,
and model with the namespace ID. The specification provides this as a
unique indentifier.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Signed-off-by: NJon Derrick <jonathan.derrick@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The NVMe specification does not require discarded blocks return zeroes on
read, but provides that behavior as a possibility. Some applications more
efficiently use an SSD if reads on discarded blocks were deterministically
zero, based on the "discard_zeroes_data" queue attribute.

There is no specification defined way to determine device behavior on
discarded blocks, so the driver always left the queue setting disabled. We
can only know behavior based on individual device models, so this patch
adds a flag to the NVMe "quirk" list that vendors may set if they know
their controller works that way. The patch also sets the new flag for one
such known device.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Suggested-by: NArtur Paszkiewicz <artur.paszkiewicz@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

When the media manager runs in dual or quad plane mode, lightnvm
abstracts away plane specific commands. This poses a problem for
get bad block table, as it reports bad blocks per plane, making the
table either two or four times bigger than expected. Fold the bad block
list before returning.
Signed-off-by: NMatias Bjørling <m@bjorling.me>
Signed-off-by: NJens Axboe <axboe@fb.com>

The block layer uses an unsigned short for max_segments.  The way we
calculate the value for NVMe tends to generate very large 32-bit values,
which after integer truncation may lead to a zero value instead of
the desired outcome.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reported-by: NJeff Lien <Jeff.Lien@hgst.com>
Tested-by: NJeff Lien <Jeff.Lien@hgst.com>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Factor out a helper to set all the device specific queue limits and apply
them to the admin queue in addition to the I/O queues.  Without this the
command size on the admin queue is arbitrarily low, and the missing
other limitations are just minefields waiting for victims.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reported-by: NJeff Lien <Jeff.Lien@hgst.com>
Tested-by: NJeff Lien <Jeff.Lien@hgst.com>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

A user could send a passthrough IO command with a metadata pointer to a
namespace without metadata. With metadata length of 0, kmalloc returns
ZERO_SIZE_PTR. Since that is not NULL, the driver would have set this as
the bio's integrity payload, which causes an access fault on completion.

This patch ignores the users metadata buffer if the namespace format
does not support separate metadata.
Reported-by: NStephen Bates <stephen.bates@microsemi.com>
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

The command flags can change the meaning of other fields in the command
that the driver is not prepared to handle. Specifically, the user could
passthrough an SGL flag, causing the controller to misinterpret the PRP
list the driver created, potentially corrupting memory or data.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NJon Derrick <jonathan.derrick@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

This moves failed queue handling out of the namespace removal path and
into the reset failure path, fixing a hanging condition if the controller
fails or link down during del_gendisk. Previously the driver had to see
the controller as degraded prior to calling del_gendisk to setup the
queues to fail. But, if the controller happened to fail after this,
there was no task to end outstanding requests.

On failure, all namespace states are set to dead. This has capacity
revalidate to 0, and ends all new requests with error status.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

A reset failure schedules the device to unbind from the driver through
the pci driver's remove. This cleans up all intialization, so there is
no need to duplicate the potentially racy cleanup.

To help understand why a reset failed, the status is logged with the
existing warning message.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

This patch makes nvme namespace removal lockless. It is up to the caller
to ensure no active namespace scanning is occuring. To ensure no scan
work occurs, the nvme pci driver adds a removing state to the controller
device to avoid queueing scan work during removal. The work is flushed
after setting the state, so no new scan work can be queued.

The lockless removal allows the driver to cleanup a namespace
request_queue if the controller fails during removal. Previously this
could deadlock trying to acquire the namespace mutex in order to handle
such events.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

A namespace may be detached from a controller, but a user may be holding
a reference to it. Attaching a new namespace with the same NSID will create
duplicate names when using the NSID to name the disk.

This patch uses an IDA that is released only when the last reference is
released instead of using the namespace ID.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

Unmapping the registers on reset or shutdown is not necessary. Keeping
the mapping simplifies reset handling.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

For NVMe over Fabrics, the cntlid will be used by systemd/udev to
create link to the device, for example,

/dev/disk/by-path/<fabrics-info>-<cntlid>-<namespace> -> /dev/nvme0n1
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Both LighNVM and NVMe over Fabrics need to look at more than just the
status and result field.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMatias Bj?rling <m@bjorling.me>
Reviewed-by: NJay Freyensee <james.p.freyensee@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The only work left in the kthread is the periodic health check for each
controller.  There is no need to run this from process context or keep
a thread context around for it, so replace it with a simpler timer.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

There is no reason to do unconditional polling of CQs per the NVMe
spec.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

Use a dedicated work item to submit async event requests instead of the
global kthread.  This simplifies the code and reduces the latencies to
resubmit a request once an even notification happened.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

We don't need to spam the kernel logs with thousands of IO cancelling
messages. We can infer all IO's are being cancelled with fewer, or
even none at all. This patch rate limits the message and uses the debug
log level as it is mainly used for testing purposes.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

A device failure or link down wouldn't have been detected during namespace
removal. This patch keeps the device in the list for polling so that the
thread may see such failure and initiate a reset. The device is removed
from the list after disable, so we can safely flush the reset work as
it can't be requeued when disable completes.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

It's possible a request may get to the driver after the nvme queue was
disabled. This has the request requeue if that happens.

Note the request is still "started" by the driver, but requeuing will
clear the start state for timeout handling.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

It is generally more efficient to submit larger IO.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The function returns true when the controller can't handle IO.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Acked-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

NVMe over Fabrics drivers are going to reuse the core,
so splits nvme.ko into 2 modules:

nvme-core.ko: the core part
nvme.ko: the PCI driver

Export symbols from nvme-core.ko.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Split dev_list_lock into one in the core and one in the PCI driver.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

These variables are used by PCI driver and will also be used in the
forthcoming NVMe over Fabrics drivers.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We don't want to be able to unload the fabric driver when we have
openened referenced to our namespaces. Thus, for each nvme_open we
take a reference on the fabric driver and put it in nvme_release.
This behavior is consistent with the scsi model.

This resolves the panic when unloading a fabric module with
mpath holders.
Signed-off-by: NSagi Grimberg <sagig@mellanox.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NIan Bakshan <ianb@mellanox.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Having the ctrl name "nvmeX" seems much more friendly than
the underlying device name. Also, with other nvme transports
such as the soon to come nvme-loop we don't have an underlying
device so it doesn't makes sense to make up one.

In order to help matching an instance name to a pci function,
we add a info print in nvme_probe.
Signed-off-by: NSagi Grimberg <sagig@mellanox.com>
Acked-by: NKeith Busch <keith.busch@intel.com>

Manually fixed up the hunk in nvme_cancel_queue_ios().
Signed-off-by: NJens Axboe <axboe@fb.com>

Pass the right private data to device_create_with_groups from the
beginning, and remove the superflous call to dev_set_drvdata.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJon Derrick <jonathan.derrick@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

This notifies blk-mq when the tag set contains a different number of
queues prior to freeing unused ones that the request queue points to.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

The specification currently limits the number of MLC pairs to 886. Make
sure that a device is unable to be instantiate if more is configured.

Also, previously the patch had the wrong math for copying MLC pairs, as
it only copied half of the actual entries.

Fixes: ca5927e7 "lightnvm: introduce mlc lower page table mappings"
Signed-off-by: NMatias Bjørling <m@bjorling.me>
Signed-off-by: NJens Axboe <axboe@fb.com>