[GIT PULL] nvme fixes for Linux 4.15

[GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

Hi Jens,

a couple nvme fixes for 4.15:

 - expand the queue ready fix that we only had for RDMA to also cover FC and
   loop by moving it to common code (Sagi)
 - fix an array out of bounds in the PCIe HMB code (Minwoo Im)
 - two new device quirks (Jeff Lien and Kai-Heng Feng)
 - static checkers fixes (Keith Busch)
 - FC target refcount fix (James Smart)
 - A trivial spelling fix in new code (Colin Ian King)


The following changes since commit 1690102de5651bb85b23d5eeaff682a6b96d705b:

  blktrace: Use blk_trace_bio_get_cgid inside blk_add_trace_bio (2017-11-19 12:37:26 -0700)

are available in the git repository at:

  git://git.infradead.org/nvme.git nvme-4.15

for you to fetch changes up to 8c97eeccf0ad8783c057830119467b877bdfced7:

  nvme-pci: add quirk for delay before CHK RDY for WDC SN200 (2017-11-23 09:12:08 +0100)

----------------------------------------------------------------
Colin Ian King (1):
      nvme: fix spelling mistake: "requeing" -> "requeuing"

James Smart (1):
      nvmet-fc: correct ref counting error when deferred rcv used

Jeff Lien (1):
      nvme-pci: add quirk for delay before CHK RDY for WDC SN200

Kai-Heng Feng (1):
      nvme-pci: disable APST on Samsung SSD 960 EVO + ASUS PRIME B350M-A

Keith Busch (2):
      nvme: Fix NULL dereference on reservation request
      nvme: Suppress static analyis warning

Minwoo Im (1):
      nvme-pci: avoid hmb desc array idx out-of-bound when hmmaxd set.

Sagi Grimberg (3):
      nvme-fabrics: introduce init command check for a queue that is not alive
      nvme-fc: check if queue is ready in queue_rq
      nvme-loop: check if queue is ready in queue_rq

 drivers/nvme/host/core.c      | 19 ++++++++-----------
 drivers/nvme/host/fabrics.h   | 30 ++++++++++++++++++++++++++++++
 drivers/nvme/host/fc.c        | 19 ++++++++++++++++++-
 drivers/nvme/host/multipath.c |  2 +-
 drivers/nvme/host/nvme.h      |  2 +-
 drivers/nvme/host/pci.c       | 16 +++++++++++++---
 drivers/nvme/host/rdma.c      | 32 ++++++--------------------------
 drivers/nvme/target/fc.c      |  9 ++++++---
 drivers/nvme/target/loop.c    | 25 ++++++++++++++++++++++++-
 9 files changed, 107 insertions(+), 47 deletions(-)

[GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

Hi Jens,

a couple nvme fixes for 4.15:

 - expand the queue ready fix that we only had for RDMA to also cover FC and
   loop by moving it to common code (Sagi)
 - fix an array out of bounds in the PCIe HMB code (Minwoo Im)
 - two new device quirks (Jeff Lien and Kai-Heng Feng)
 - static checkers fixes (Keith Busch)
 - FC target refcount fix (James Smart)
 - A trivial spelling fix in new code (Colin Ian King)


The following changes since commit 1690102de5651bb85b23d5eeaff682a6b96d705b:

  blktrace: Use blk_trace_bio_get_cgid inside blk_add_trace_bio (2017-11-19 12:37:26 -0700)

are available in the git repository at:

  git://git.infradead.org/nvme.git nvme-4.15

for you to fetch changes up to 8c97eeccf0ad8783c057830119467b877bdfced7:

  nvme-pci: add quirk for delay before CHK RDY for WDC SN200 (2017-11-23 09:12:08 +0100)

----------------------------------------------------------------
Colin Ian King (1):
      nvme: fix spelling mistake: "requeing" -> "requeuing"

James Smart (1):
      nvmet-fc: correct ref counting error when deferred rcv used

Jeff Lien (1):
      nvme-pci: add quirk for delay before CHK RDY for WDC SN200

Kai-Heng Feng (1):
      nvme-pci: disable APST on Samsung SSD 960 EVO + ASUS PRIME B350M-A

Keith Busch (2):
      nvme: Fix NULL dereference on reservation request
      nvme: Suppress static analyis warning

Minwoo Im (1):
      nvme-pci: avoid hmb desc array idx out-of-bound when hmmaxd set.

Sagi Grimberg (3):
      nvme-fabrics: introduce init command check for a queue that is not alive
      nvme-fc: check if queue is ready in queue_rq
      nvme-loop: check if queue is ready in queue_rq

 drivers/nvme/host/core.c      | 19 ++++++++-----------
 drivers/nvme/host/fabrics.h   | 30 ++++++++++++++++++++++++++++++
 drivers/nvme/host/fc.c        | 19 ++++++++++++++++++-
 drivers/nvme/host/multipath.c |  2 +-
 drivers/nvme/host/nvme.h      |  2 +-
 drivers/nvme/host/pci.c       | 16 +++++++++++++---
 drivers/nvme/host/rdma.c      | 32 ++++++--------------------------
 drivers/nvme/target/fc.c      |  9 ++++++---
 drivers/nvme/target/loop.c    | 25 ++++++++++++++++++++++++-
 9 files changed, 107 insertions(+), 47 deletions(-)

Re: [GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/23/2017 07:44 AM, Christoph Hellwig wrote:
> Sagi Grimberg (3):
>       nvme-fc: check if queue is ready in queue_rq
>       nvme-loop: check if queue is ready in queue_rq

The nvme-loop part looks fine, but why is the nvme-fc part using:

enum nvme_fc_queue_flags {                                                     
        NVME_FC_Q_CONNECTED = (1 << 0),                                         
+       NVME_FC_Q_LIVE = (1 << 1),                                              
}; 

for flags that are used with set_bit() and friends? That's just
misleading, should be 0, 1, etc, not a shift.

The rest looks pretty straight forward, but the above is an eye sore.

-- 
Jens Axboe

[GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/23/2017 07:44 AM, Christoph Hellwig wrote:
> Sagi Grimberg (3):
>       nvme-fc: check if queue is ready in queue_rq
>       nvme-loop: check if queue is ready in queue_rq

The nvme-loop part looks fine, but why is the nvme-fc part using:

enum nvme_fc_queue_flags {                                                     
        NVME_FC_Q_CONNECTED = (1 << 0),                                         
+       NVME_FC_Q_LIVE = (1 << 1),                                              
}; 

for flags that are used with set_bit() and friends? That's just
misleading, should be 0, 1, etc, not a shift.

The rest looks pretty straight forward, but the above is an eye sore.

-- 
Jens Axboe

Re: [GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/23/2017 09:54 PM, Jens Axboe wrote:
> On 11/23/2017 07:44 AM, Christoph Hellwig wrote:
>> Sagi Grimberg (3):
>>       nvme-fc: check if queue is ready in queue_rq
>>       nvme-loop: check if queue is ready in queue_rq
> 
> The nvme-loop part looks fine, but why is the nvme-fc part using:
> 
> enum nvme_fc_queue_flags {                                                     
>         NVME_FC_Q_CONNECTED = (1 << 0),                                         
> +       NVME_FC_Q_LIVE = (1 << 1),                                              
> }; 
> 
> for flags that are used with set_bit() and friends? That's just
> misleading, should be 0, 1, etc, not a shift.
> 
> The rest looks pretty straight forward, but the above is an eye sore.

diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
index e0577bf33f45..0a8af4daef89 100644
--- a/drivers/nvme/host/fc.c
+++ b/drivers/nvme/host/fc.c
@@ -31,8 +31,8 @@
 
 
 enum nvme_fc_queue_flags {
-	NVME_FC_Q_CONNECTED = (1 << 0),
-	NVME_FC_Q_LIVE = (1 << 1),
+	NVME_FC_Q_CONNECTED = 0,
+	NVME_FC_Q_LIVE,
 };
 
 #define NVMEFC_QUEUE_DELAY	3		/* ms units */

-- 
Jens Axboe

[GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/23/2017 09:54 PM, Jens Axboe wrote:
> On 11/23/2017 07:44 AM, Christoph Hellwig wrote:
>> Sagi Grimberg (3):
>>       nvme-fc: check if queue is ready in queue_rq
>>       nvme-loop: check if queue is ready in queue_rq
> 
> The nvme-loop part looks fine, but why is the nvme-fc part using:
> 
> enum nvme_fc_queue_flags {                                                     
>         NVME_FC_Q_CONNECTED = (1 << 0),                                         
> +       NVME_FC_Q_LIVE = (1 << 1),                                              
> }; 
> 
> for flags that are used with set_bit() and friends? That's just
> misleading, should be 0, 1, etc, not a shift.
> 
> The rest looks pretty straight forward, but the above is an eye sore.

diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
index e0577bf33f45..0a8af4daef89 100644
--- a/drivers/nvme/host/fc.c
+++ b/drivers/nvme/host/fc.c
@@ -31,8 +31,8 @@
 
 
 enum nvme_fc_queue_flags {
-	NVME_FC_Q_CONNECTED = (1 << 0),
-	NVME_FC_Q_LIVE = (1 << 1),
+	NVME_FC_Q_CONNECTED = 0,
+	NVME_FC_Q_LIVE,
 };
 
 #define NVMEFC_QUEUE_DELAY	3		/* ms units */

-- 
Jens Axboe

Re: [GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

On Thu, Nov 23, 2017 at 10:04:55PM -0700, Jens Axboe wrote:
> > +       NVME_FC_Q_LIVE = (1 << 1),                                              
> > }; 
> > 
> > for flags that are used with set_bit() and friends? That's just
> > misleading, should be 0, 1, etc, not a shift.
> > 
> > The rest looks pretty straight forward, but the above is an eye sore.

Yes, it is misleading, but so far harmless.

Your fixup looks good to me:

Reviewed-by: Christoph Hellwig <hch@lst.de>

[GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

On Thu, Nov 23, 2017@10:04:55PM -0700, Jens Axboe wrote:
> > +       NVME_FC_Q_LIVE = (1 << 1),                                              
> > }; 
> > 
> > for flags that are used with set_bit() and friends? That's just
> > misleading, should be 0, 1, etc, not a shift.
> > 
> > The rest looks pretty straight forward, but the above is an eye sore.

Yes, it is misleading, but so far harmless.

Your fixup looks good to me:

Reviewed-by: Christoph Hellwig <hch at lst.de>

Re: [GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/24/2017 01:32 AM, Christoph Hellwig wrote:
> On Thu, Nov 23, 2017 at 10:04:55PM -0700, Jens Axboe wrote:
>>> +       NVME_FC_Q_LIVE = (1 << 1),                                              
>>> }; 
>>>
>>> for flags that are used with set_bit() and friends? That's just
>>> misleading, should be 0, 1, etc, not a shift.
>>>
>>> The rest looks pretty straight forward, but the above is an eye sore.
> 
> Yes, it is misleading, but so far harmless.

Yeah, but it just takes a few more bits...

> Your fixup looks good to me:
> 
> Reviewed-by: Christoph Hellwig <hch@lst.de>

Thanks, I've added it.

-- 
Jens Axboe

[GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/24/2017 01:32 AM, Christoph Hellwig wrote:
> On Thu, Nov 23, 2017@10:04:55PM -0700, Jens Axboe wrote:
>>> +       NVME_FC_Q_LIVE = (1 << 1),                                              
>>> }; 
>>>
>>> for flags that are used with set_bit() and friends? That's just
>>> misleading, should be 0, 1, etc, not a shift.
>>>
>>> The rest looks pretty straight forward, but the above is an eye sore.
> 
> Yes, it is misleading, but so far harmless.

Yeah, but it just takes a few more bits...

> Your fixup looks good to me:
> 
> Reviewed-by: Christoph Hellwig <hch at lst.de>

Thanks, I've added it.

-- 
Jens Axboe

Re: [GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 1/5/18 10:15 AM, Christoph Hellwig wrote:
> Hi Jens,
> 
> I have a few more small nvme fixes for you for Linux 4.15:
> 
> The following changes since commit 6d0e4827b72afc71349784336d5eb6df4df106e6:
> 
>   Revert "bdi: add error handle for bdi_debug_register" (2017-12-21 10:01:30 -0700)
> 
> are available in the git repository at:
> 
>   git://git.infradead.org/nvme.git nvme-4.15

Looks fine to me, pulled.

4.16 is coming up, would be nice to get at least the preliminary nvme
stuff pulled in for that.

-- 
Jens Axboe

[GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 1/5/18 10:15 AM, Christoph Hellwig wrote:
> Hi Jens,
> 
> I have a few more small nvme fixes for you for Linux 4.15:
> 
> The following changes since commit 6d0e4827b72afc71349784336d5eb6df4df106e6:
> 
>   Revert "bdi: add error handle for bdi_debug_register" (2017-12-21 10:01:30 -0700)
> 
> are available in the git repository at:
> 
>   git://git.infradead.org/nvme.git nvme-4.15

Looks fine to me, pulled.

4.16 is coming up, would be nice to get at least the preliminary nvme
stuff pulled in for that.

-- 
Jens Axboe

[GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

Hi Jens,

I have a few more small nvme fixes for you for Linux 4.15:

The following changes since commit 6d0e4827b72afc71349784336d5eb6df4df106e6:

  Revert "bdi: add error handle for bdi_debug_register" (2017-12-21 10:01:30 -0700)

are available in the git repository at:

  git://git.infradead.org/nvme.git nvme-4.15

for you to fetch changes up to 254beb84faccbe2f4eda0b51924857bdfb679969:

  nvme-fcloop: avoid possible uninitialized variable warning (2017-12-29 10:37:21 +0100)

----------------------------------------------------------------
James Smart (1):
      nvme-fcloop: avoid possible uninitialized variable warning

Jeff Lien (1):
      nvme: fix sector units when going between formats

Minwoo Im (1):
      nvme-pci: move use_sgl initialization to nvme_init_iod()

Sagi Grimberg (2):
      nvme-rdma: fix concurrent reset and reconnect
      nvme-mpath: fix last path removal during traffic

 drivers/nvme/host/core.c     |  7 ++++++-
 drivers/nvme/host/nvme.h     | 12 ++++++++++++
 drivers/nvme/host/pci.c      | 42 ++++++++++++++++++++----------------------
 drivers/nvme/host/rdma.c     | 14 +++++++++++++-
 drivers/nvme/target/fcloop.c |  2 +-
 5 files changed, 52 insertions(+), 25 deletions(-)

[GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

Hi Jens,

I have a few more small nvme fixes for you for Linux 4.15:

The following changes since commit 6d0e4827b72afc71349784336d5eb6df4df106e6:

  Revert "bdi: add error handle for bdi_debug_register" (2017-12-21 10:01:30 -0700)

are available in the git repository at:

  git://git.infradead.org/nvme.git nvme-4.15

for you to fetch changes up to 254beb84faccbe2f4eda0b51924857bdfb679969:

  nvme-fcloop: avoid possible uninitialized variable warning (2017-12-29 10:37:21 +0100)

----------------------------------------------------------------
James Smart (1):
      nvme-fcloop: avoid possible uninitialized variable warning

Jeff Lien (1):
      nvme: fix sector units when going between formats

Minwoo Im (1):
      nvme-pci: move use_sgl initialization to nvme_init_iod()

Sagi Grimberg (2):
      nvme-rdma: fix concurrent reset and reconnect
      nvme-mpath: fix last path removal during traffic

 drivers/nvme/host/core.c     |  7 ++++++-
 drivers/nvme/host/nvme.h     | 12 ++++++++++++
 drivers/nvme/host/pci.c      | 42 ++++++++++++++++++++----------------------
 drivers/nvme/host/rdma.c     | 14 +++++++++++++-
 drivers/nvme/target/fcloop.c |  2 +-
 5 files changed, 52 insertions(+), 25 deletions(-)

Re: [GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/29/2017 09:16 AM, Christoph Hellwig wrote:
> Hi Jens,
> 
> a few more nvme updates for 4.15.  A single small PCIe fix, and a number
> of patches for RDMA that are a little larger than what I'd like to see
> for -rc2, but they fix important issues seen in the wild.

Looks good to me, pulled. Thanks.

-- 
Jens Axboe

[GIT PULL] nvme fixes for Linux 4.15

[Jens Axboe]

On 11/29/2017 09:16 AM, Christoph Hellwig wrote:
> Hi Jens,
> 
> a few more nvme updates for 4.15.  A single small PCIe fix, and a number
> of patches for RDMA that are a little larger than what I'd like to see
> for -rc2, but they fix important issues seen in the wild.

Looks good to me, pulled. Thanks.

-- 
Jens Axboe

[GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

Hi Jens,

a few more nvme updates for 4.15.  A single small PCIe fix, and a number
of patches for RDMA that are a little larger than what I'd like to see
for -rc2, but they fix important issues seen in the wild.

The following changes since commit 6c4ca1e36cdc1a0a7a84797804b87920ccbebf51:

  bcache: check return value of register_shrinker (2017-11-24 16:23:01 -0700)

are available in the git repository at:

  git://git.infradead.org/nvme.git nvme-4.15

for you to fetch changes up to 7e5dd57ef3081ff6c03908d786ed5087f6fbb7ae:

  nvme-pci: fix NULL pointer dereference in nvme_free_host_mem() (2017-11-28 08:49:26 -0800)

----------------------------------------------------------------
Israel Rukshin (1):
      nvme-rdma: Use mr pool

Max Gurtovoy (1):
      nvme-rdma: fix memory leak during queue allocation

Minwoo Im (1):
      nvme-pci: fix NULL pointer dereference in nvme_free_host_mem()

Sagi Grimberg (4):
      nvme-rdma: don't suppress send completions
      nvme-rdma: don't complete requests before a send work request has completed
      nvme-rdma: wait for local invalidation before completing a request
      nvme-rdma: Check remotely invalidated rkey matches our expected rkey

 drivers/nvme/host/pci.c  |   1 +
 drivers/nvme/host/rdma.c | 234 ++++++++++++++++++++++++-----------------------
 2 files changed, 119 insertions(+), 116 deletions(-)

[GIT PULL] nvme fixes for Linux 4.15

[Christoph Hellwig]

Hi Jens,

a few more nvme updates for 4.15.  A single small PCIe fix, and a number
of patches for RDMA that are a little larger than what I'd like to see
for -rc2, but they fix important issues seen in the wild.

The following changes since commit 6c4ca1e36cdc1a0a7a84797804b87920ccbebf51:

  bcache: check return value of register_shrinker (2017-11-24 16:23:01 -0700)

are available in the git repository at:

  git://git.infradead.org/nvme.git nvme-4.15

for you to fetch changes up to 7e5dd57ef3081ff6c03908d786ed5087f6fbb7ae:

  nvme-pci: fix NULL pointer dereference in nvme_free_host_mem() (2017-11-28 08:49:26 -0800)

----------------------------------------------------------------
Israel Rukshin (1):
      nvme-rdma: Use mr pool

Max Gurtovoy (1):
      nvme-rdma: fix memory leak during queue allocation

Minwoo Im (1):
      nvme-pci: fix NULL pointer dereference in nvme_free_host_mem()

Sagi Grimberg (4):
      nvme-rdma: don't suppress send completions
      nvme-rdma: don't complete requests before a send work request has completed
      nvme-rdma: wait for local invalidation before completing a request
      nvme-rdma: Check remotely invalidated rkey matches our expected rkey

 drivers/nvme/host/pci.c  |   1 +
 drivers/nvme/host/rdma.c | 234 ++++++++++++++++++++++++-----------------------
 2 files changed, 119 insertions(+), 116 deletions(-)