| Age | Commit message (Collapse) | Author |
|---|
|
Commit fd76863 (RAID1: a new I/O barrier implementation to remove resync
window) introduces a user-after-free bug.
Signed-off-by: Shaohua Li <shli@fb.com>
|
|
When I run a parallel reading performan testing on a md raid1 device with
two NVMe SSDs, I observe very bad throughput in supprise: by fio with 64KB
block size, 40 seq read I/O jobs, 128 iodepth, overall throughput is
only 2.7GB/s, this is around 50% of the idea performance number.
The perf reports locking contention happens at allow_barrier() and
wait_barrier() code,
- 41.41% fio [kernel.kallsyms] [k] _raw_spin_lock_irqsave
- _raw_spin_lock_irqsave
+ 89.92% allow_barrier
+ 9.34% __wake_up
- 37.30% fio [kernel.kallsyms] [k] _raw_spin_lock_irq
- _raw_spin_lock_irq
- 100.00% wait_barrier
The reason is, in these I/O barrier related functions,
- raise_barrier()
- lower_barrier()
- wait_barrier()
- allow_barrier()
They always hold conf->resync_lock firstly, even there are only regular
reading I/Os and no resync I/O at all. This is a huge performance penalty.
The solution is a lockless-like algorithm in I/O barrier code, and only
holding conf->resync_lock when it has to.
The original idea is from Hannes Reinecke, and Neil Brown provides
comments to improve it. I continue to work on it, and make the patch into
current form.
In the new simpler raid1 I/O barrier implementation, there are two
wait barrier functions,
- wait_barrier()
Which calls _wait_barrier(), is used for regular write I/O. If there is
resync I/O happening on the same I/O barrier bucket, or the whole
array is frozen, task will wait until no barrier on same barrier bucket,
or the whold array is unfreezed.
- wait_read_barrier()
Since regular read I/O won't interfere with resync I/O (read_balance()
will make sure only uptodate data will be read out), it is unnecessary
to wait for barrier in regular read I/Os, waiting in only necessary
when the whole array is frozen.
The operations on conf->nr_pending[idx], conf->nr_waiting[idx], conf->
barrier[idx] are very carefully designed in raise_barrier(),
lower_barrier(), _wait_barrier() and wait_read_barrier(), in order to
avoid unnecessary spin locks in these functions. Once conf->
nr_pengding[idx] is increased, a resync I/O with same barrier bucket index
has to wait in raise_barrier(). Then in _wait_barrier() if no barrier
raised in same barrier bucket index and array is not frozen, the regular
I/O doesn't need to hold conf->resync_lock, it can just increase
conf->nr_pending[idx], and return to its caller. wait_read_barrier() is
very similar to _wait_barrier(), the only difference is it only waits when
array is frozen. For heavy parallel reading I/Os, the lockless I/O barrier
code almostly gets rid of all spin lock cost.
This patch significantly improves raid1 reading peroformance. From my
testing, a raid1 device built by two NVMe SSD, runs fio with 64KB
blocksize, 40 seq read I/O jobs, 128 iodepth, overall throughput
increases from 2.7GB/s to 4.6GB/s (+70%).
Changelog
V4:
- Change conf->nr_queued[] to atomic_t.
- Define BARRIER_BUCKETS_NR_BITS by (PAGE_SHIFT - ilog2(sizeof(atomic_t)))
V3:
- Add smp_mb__after_atomic() as Shaohua and Neil suggested.
- Change conf->nr_queued[] from atomic_t to int.
- Change conf->array_frozen from atomic_t back to int, and use
READ_ONCE(conf->array_frozen) to check value of conf->array_frozen
in _wait_barrier() and wait_read_barrier().
- In _wait_barrier() and wait_read_barrier(), add a call to
wake_up(&conf->wait_barrier) after atomic_dec(&conf->nr_pending[idx]),
to fix a deadlock between _wait_barrier()/wait_read_barrier and
freeze_array().
V2:
- Remove a spin_lock/unlock pair in raid1d().
- Add more code comments to explain why there is no racy when checking two
atomic_t variables at same time.
V1:
- Original RFC patch for comments.
Signed-off-by: Coly Li <colyli@suse.de>
Cc: Shaohua Li <shli@fb.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Guoqing Jiang <gqjiang@suse.com>
Reviewed-by: Neil Brown <neilb@suse.de>
Signed-off-by: Shaohua Li <shli@fb.com>
|
|
'Commit 79ef3a8aa1cb ("raid1: Rewrite the implementation of iobarrier.")'
introduces a sliding resync window for raid1 I/O barrier, this idea limits
I/O barriers to happen only inside a slidingresync window, for regular
I/Os out of this resync window they don't need to wait for barrier any
more. On large raid1 device, it helps a lot to improve parallel writing
I/O throughput when there are background resync I/Os performing at
same time.
The idea of sliding resync widow is awesome, but code complexity is a
challenge. Sliding resync window requires several variables to work
collectively, this is complexed and very hard to make it work correctly.
Just grep "Fixes: 79ef3a8aa1" in kernel git log, there are 8 more patches
to fix the original resync window patch. This is not the end, any further
related modification may easily introduce more regreassion.
Therefore I decide to implement a much simpler raid1 I/O barrier, by
removing resync window code, I believe life will be much easier.
The brief idea of the simpler barrier is,
- Do not maintain a global unique resync window
- Use multiple hash buckets to reduce I/O barrier conflicts, regular
I/O only has to wait for a resync I/O when both them have same barrier
bucket index, vice versa.
- I/O barrier can be reduced to an acceptable number if there are enough
barrier buckets
Here I explain how the barrier buckets are designed,
- BARRIER_UNIT_SECTOR_SIZE
The whole LBA address space of a raid1 device is divided into multiple
barrier units, by the size of BARRIER_UNIT_SECTOR_SIZE.
Bio requests won't go across border of barrier unit size, that means
maximum bio size is BARRIER_UNIT_SECTOR_SIZE<<9 (64MB) in bytes.
For random I/O 64MB is large enough for both read and write requests,
for sequential I/O considering underlying block layer may merge them
into larger requests, 64MB is still good enough.
Neil also points out that for resync operation, "we want the resync to
move from region to region fairly quickly so that the slowness caused
by having to synchronize with the resync is averaged out over a fairly
small time frame". For full speed resync, 64MB should take less then 1
second. When resync is competing with other I/O, it could take up a few
minutes. Therefore 64MB size is fairly good range for resync.
- BARRIER_BUCKETS_NR
There are BARRIER_BUCKETS_NR buckets in total, which is defined by,
#define BARRIER_BUCKETS_NR_BITS (PAGE_SHIFT - 2)
#define BARRIER_BUCKETS_NR (1<<BARRIER_BUCKETS_NR_BITS)
this patch makes the bellowed members of struct r1conf from integer
to array of integers,
- int nr_pending;
- int nr_waiting;
- int nr_queued;
- int barrier;
+ int *nr_pending;
+ int *nr_waiting;
+ int *nr_queued;
+ int *barrier;
number of the array elements is defined as BARRIER_BUCKETS_NR. For 4KB
kernel space page size, (PAGE_SHIFT - 2) indecates there are 1024 I/O
barrier buckets, and each array of integers occupies single memory page.
1024 means for a request which is smaller than the I/O barrier unit size
has ~0.1% chance to wait for resync to pause, which is quite a small
enough fraction. Also requesting single memory page is more friendly to
kernel page allocator than larger memory size.
- I/O barrier bucket is indexed by bio start sector
If multiple I/O requests hit different I/O barrier units, they only need
to compete I/O barrier with other I/Os which hit the same I/O barrier
bucket index with each other. The index of a barrier bucket which a
bio should look for is calculated by sector_to_idx() which is defined
in raid1.h as an inline function,
static inline int sector_to_idx(sector_t sector)
{
return hash_long(sector >> BARRIER_UNIT_SECTOR_BITS,
BARRIER_BUCKETS_NR_BITS);
}
Here sector_nr is the start sector number of a bio.
- Single bio won't go across boundary of a I/O barrier unit
If a request goes across boundary of barrier unit, it will be split. A
bio may be split in raid1_make_request() or raid1_sync_request(), if
sectors returned by align_to_barrier_unit_end() is smaller than
original bio size.
Comparing to single sliding resync window,
- Currently resync I/O grows linearly, therefore regular and resync I/O
will conflict within a single barrier units. So the I/O behavior is
similar to single sliding resync window.
- But a barrier unit bucket is shared by all barrier units with identical
barrier uinit index, the probability of conflict might be higher
than single sliding resync window, in condition that writing I/Os
always hit barrier units which have identical barrier bucket indexs with
the resync I/Os. This is a very rare condition in real I/O work loads,
I cannot imagine how it could happen in practice.
- Therefore we can achieve a good enough low conflict rate with much
simpler barrier algorithm and implementation.
There are two changes should be noticed,
- In raid1d(), I change the code to decrease conf->nr_pending[idx] into
single loop, it looks like this,
spin_lock_irqsave(&conf->device_lock, flags);
conf->nr_queued[idx]--;
spin_unlock_irqrestore(&conf->device_lock, flags);
This change generates more spin lock operations, but in next patch of
this patch set, it will be replaced by a single line code,
atomic_dec(&conf->nr_queueud[idx]);
So we don't need to worry about spin lock cost here.
- Mainline raid1 code split original raid1_make_request() into
raid1_read_request() and raid1_write_request(). If the original bio
goes across an I/O barrier unit size, this bio will be split before
calling raid1_read_request() or raid1_write_request(), this change
the code logic more simple and clear.
- In this patch wait_barrier() is moved from raid1_make_request() to
raid1_write_request(). In raid_read_request(), original wait_barrier()
is replaced by raid1_read_request().
The differnece is wait_read_barrier() only waits if array is frozen,
using different barrier function in different code path makes the code
more clean and easy to read.
Changelog
V4:
- Add alloc_r1bio() to remove redundant r1bio memory allocation code.
- Fix many typos in patch comments.
- Use (PAGE_SHIFT - ilog2(sizeof(int))) to define BARRIER_BUCKETS_NR_BITS.
V3:
- Rebase the patch against latest upstream kernel code.
- Many fixes by review comments from Neil,
- Back to use pointers to replace arraries in struct r1conf
- Remove total_barriers from struct r1conf
- Add more patch comments to explain how/why the values of
BARRIER_UNIT_SECTOR_SIZE and BARRIER_BUCKETS_NR are decided.
- Use get_unqueued_pending() to replace get_all_pendings() and
get_all_queued()
- Increase bucket number from 512 to 1024
- Change code comments format by review from Shaohua.
V2:
- Use bio_split() to split the orignal bio if it goes across barrier unit
bounday, to make the code more simple, by suggestion from Shaohua and
Neil.
- Use hash_long() to replace original linear hash, to avoid a possible
confilict between resync I/O and sequential write I/O, by suggestion from
Shaohua.
- Add conf->total_barriers to record barrier depth, which is used to
control number of parallel sync I/O barriers, by suggestion from Shaohua.
- In V1 patch the bellowed barrier buckets related members in r1conf are
allocated in memory page. To make the code more simple, V2 patch moves
the memory space into struct r1conf, like this,
- int nr_pending;
- int nr_waiting;
- int nr_queued;
- int barrier;
+ int nr_pending[BARRIER_BUCKETS_NR];
+ int nr_waiting[BARRIER_BUCKETS_NR];
+ int nr_queued[BARRIER_BUCKETS_NR];
+ int barrier[BARRIER_BUCKETS_NR];
This change is by the suggestion from Shaohua.
- Remove some inrelavent code comments, by suggestion from Guoqing.
- Add a missing wait_barrier() before jumping to retry_write, in
raid1_make_write_request().
V1:
- Original RFC patch for comments
Signed-off-by: Coly Li <colyli@suse.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Guoqing Jiang <gqjiang@suse.com>
Reviewed-by: Neil Brown <neilb@suse.de>
Signed-off-by: Shaohua Li <shli@fb.com>
|
|
Some Cavium dev boards have firmware which doesn't supply a proper
ethernet-phy-ieee802.3-c22" compatible property. Restore these boards
to working order by whitelisting this compatible value.
Signed-off-by: David Daney <david.daney@cavium.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add constants and callback functions for the dwmac on rk3328 socs.
As can be seen, the base structure is the same, only registers and the
bits in them moved slightly.
Signed-off-by: david.wu <david.wu@rock-chips.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We already have counters for sent/recv packets and sent/recv bytes.
Doing a batched update to reduce the number of
u64_stats_update_begin/end().
Take care not to bother with stats update when called
speculatively.
Cc: Willem de Bruijn <willemb@google.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
1) In the case where rate == priv->pkt_rate_low == priv->pkt_rate_high,
mlx4_en_auto_moderation() does a divide by zero.
2) We want to properly change the moderation parameters if rx_frames
was changed (like in ethtool -C eth0 rx-frames 16)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
After sending device capability queries and requests to the vNIC Server,
an interrupt is triggered and the responses are written to the driver's
CRQ response buffer. Since the interrupt can be triggered before all
responses are written and visible to the partition, there is a danger
that the interrupt handler or tasklet can terminate before all responses
are read, resulting in a failure to initialize the device.
To avoid this scenario, when capability commands are sent, we set
a flag that will be checked in the following interrupt tasklet that
will handle the capability responses from the server. Once all
responses have been handled, the flag is disabled; and the tasklet
is allowed to terminate.
Signed-off-by: Thomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Two different counters were being used for capabilities
requests and queries. These commands are not called
at the same time so there is no reason a single counter
cannot be used.
Signed-off-by: Thomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Create a tasklet to process queued commands or messages received from
firmware instead of processing them in the interrupt handler. Note that
this handler does not process network traffic, but communications related
to resource allocation and device settings.
Signed-off-by: Thomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This adds the backbone required for the various HW initalizations
which are necessary for the FCoE driver (qedf) for QLogic FastLinQ
4xxxx line of adapters - FW notification, resource initializations, etc.
Signed-off-by: Arun Easi <arun.easi@cavium.com>
Signed-off-by: Yuval Mintz <yuval.mintz@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
What happens is that a write to /dev/sg is given a request with non-zero
->iovec_count combined with zero ->dxfer_len. Or with ->dxferp pointing
to an array full of empty iovecs.
Having write permission to /dev/sg shouldn't be equivalent to the
ability to trigger BUG_ON() while holding spinlocks...
Found by Dmitry Vyukov and syzkaller.
[ The BUG_ON() got changed to a WARN_ON_ONCE(), but this fixes the
underlying issue. - Linus ]
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Don't crash the machine just because of an empty transfer. Use WARN_ON()
combined with returning an error.
Found by Dmitry Vyukov and syzkaller.
[ Changed to "WARN_ON_ONCE()". Al has a patch that should fix the root
cause, but a BUG_ON() is not acceptable in any case, and a WARN_ON()
might still be a cause of excessive log spamming.
NOTE! If this warning ever triggers, we may end up leaking resources,
since this doesn't bother to try to clean the command up. So this
WARN_ON_ONCE() triggering does imply real problems. But BUG_ON() is
much worse.
People really need to stop using BUG_ON() for "this shouldn't ever
happen". It makes pretty much any bug worse. - Linus ]
Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: James Bottomley <jejb@linux.vnet.ibm.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
'spi/topic/topcliff-pch' into spi-next
|
|
'spi/topic/s3c64xx', 'spi/topic/sh-msiof' and 'spi/topic/slave' into spi-next
|
|
'spi/topic/mpc52xx', 'spi/topic/ppc4xx' and 'spi/topic/pxa2xx' into spi-next
|
|
'spi/topic/falcon' and 'spi/topic/fsl-lpspi' into spi-next
|
|
'spi/topic/bcm-qspi' and 'spi/topic/bcm53xx' into spi-next
|
|
|
|
|
|
'spi/fix/s3c64xx' into spi-linus
|
|
'regulator/topic/supplies' and 'regulator/topic/tps65217' into regulator-next
|
|
'regulator/topic/pv88090', 'regulator/topic/qcom-smd', 'regulator/topic/rc5t583' and 'regulator/topic/rn5t618' into regulator-next
|
|
'regulator/topic/pcap', 'regulator/topic/pcf50633', 'regulator/topic/pfuze100' and 'regulator/topic/pv88060' into regulator-next
|
|
'regulator/topic/max8907', 'regulator/topic/max8925', 'regulator/topic/max8952' and 'regulator/topic/palmas' into regulator-next
|
|
'regulator/topic/max14577', 'regulator/topic/max77620', 'regulator/topic/max77686' and 'regulator/topic/max77693' into regulator-next
|
|
'regulator/topic/fan53555', 'regulator/topic/gpio', 'regulator/topic/hi655x' and 'regulator/topic/lp8755' into regulator-next
|
|
'regulator/topic/arizona', 'regulator/topic/as3711' and 'regulator/topic/bcm590xx' into regulator-next
|
|
'regulator/topic/88pm8607', 'regulator/topic/aat2870', 'regulator/topic/act8945a' and 'regulator/topic/ad5938' into regulator-next
|
|
|
|
'regulator/fix/tps65086' into regulator-linus
|
|
|
|
'asoc/topic/nau8540', 'asoc/topic/nau8825' and 'asoc/topic/omap' into asoc-next
|
|
'asoc/topic/cq93vc' and 'asoc/topic/da7218' into asoc-next
|
|
'dma_request_chan_by_mask()' can not return NULL.
Try to keep the logic in 'no_dma:' by resetting 'qspi->rx_chan' in case
of error.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
|
|
'dma_request_chan_by_mask()' can not return NULL.
Try to keep the logic in 'no_dma:' by resetting 'qspi->rx_chan' in case
of error.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
This driver should compile on all platforms, activate it under compile
test. The Lantiq specific parts are under ifdef and should be removed
when Lantiq platform supports common clock framework.
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
The call to spi_master_put() in a3700_spi_remove() is redundant since
the master is registered using devm_spi_register_master() and no
reference hold by using spi_master_get() in a3700_spi_remove().
This is detected by Coccinelle semantic patch.
Fixes: 5762ab71eb24 ("spi: Add support for Armada 3700 SPI Controller")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
A quote from the IB spec:
However, if the Consumer does not wait for the Affiliated Asynchronous
Last WQE Reached Event, then WQE and Data Segment leakage may occur.
Therefore, it is good programming practice to tear down a QP that is
associated with an SRQ by using the following process:
* Put the QP in the Error State;
* wait for the Affiliated Asynchronous Last WQE Reached Event;
* either:
* drain the CQ by invoking the Poll CQ verb and either wait for CQ
to be empty or the number of Poll CQ operations has exceeded CQ
capacity size; or
* post another WR that completes on the same CQ and wait for this WR to return as a WC;
* and then invoke a Destroy QP or Reset QP.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Laurence Oberman <loberman@redhat.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Steve Wise <swise@opengridcomputing.com>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Max Gurtovoy <maxg@mellanox.com>
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
Avoid that the following message is printed if login fails:
scsi host0: ib_srp: Sending CM DREQ failed
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Laurence Oberman <loberman@redhat.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
Report the destination port GID if connecting fails.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Laurence Oberman <loberman@redhat.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
Use lockdep_assert_held() statements to verify at run-time
whether the proper locks are held.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Laurence Oberman <loberman@redhat.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
Avoid that srp_process_rsp() overwrites the status information
in ch if the SRP target response timed out and processing of
another task management function has already started. Avoid that
issuing multiple task management functions concurrently triggers
list corruption. This patch prevents that the following stack
trace appears in the system log:
WARNING: CPU: 8 PID: 9269 at lib/list_debug.c:52 __list_del_entry_valid+0xbc/0xc0
list_del corruption. prev->next should be ffffc90004bb7b00, but was ffff8804052ecc68
CPU: 8 PID: 9269 Comm: sg_reset Tainted: G W 4.10.0-rc7-dbg+ #3
Call Trace:
dump_stack+0x68/0x93
__warn+0xc6/0xe0
warn_slowpath_fmt+0x4a/0x50
__list_del_entry_valid+0xbc/0xc0
wait_for_completion_timeout+0x12e/0x170
srp_send_tsk_mgmt+0x1ef/0x2d0 [ib_srp]
srp_reset_device+0x5b/0x110 [ib_srp]
scsi_ioctl_reset+0x1c7/0x290
scsi_ioctl+0x12a/0x420
sd_ioctl+0x9d/0x100
blkdev_ioctl+0x51e/0x9f0
block_ioctl+0x38/0x40
do_vfs_ioctl+0x8f/0x700
SyS_ioctl+0x3c/0x70
entry_SYSCALL_64_fastpath+0x18/0xad
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Steve Feeley <Steve.Feeley@sandisk.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
After srp_process_rsp() returns there is a short time during which
the scsi_host_find_tag() call will return a pointer to the SCSI
command that is being completed. If during that time a duplicate
response is received, avoid that the following call stack appears:
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: srp_recv_done+0x450/0x6b0 [ib_srp]
Oops: 0000 [#1] SMP
CPU: 10 PID: 0 Comm: swapper/10 Not tainted 4.10.0-rc7-dbg+ #1
Call Trace:
<IRQ>
__ib_process_cq+0x4b/0xd0 [ib_core]
ib_poll_handler+0x1d/0x70 [ib_core]
irq_poll_softirq+0xba/0x120
__do_softirq+0xba/0x4c0
irq_exit+0xbe/0xd0
smp_apic_timer_interrupt+0x38/0x50
apic_timer_interrupt+0x90/0xa0
</IRQ>
RIP: srp_recv_done+0x450/0x6b0 [ib_srp] RSP: ffff88046f483e20
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Steve Feeley <Steve.Feeley@sandisk.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
Tests have shown that the following error message is reported when
using SG-GAPS registration with an mlx5 adapter:
scsi host1: ib_srp: failed RECV status WR flushed (5) for CQE ffff880bd4270eb0
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 0f007806 2500002a ad9fafd1
scsi host1: ib_srp: reconnect succeeded
mlx5_0:dump_cqe:262:(pid 7369): dump error cqe
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 0f007806 25000032 00105dd0
scsi host1: ib_srp: failed FAST REG status memory management operation error (6) for CQE ffff880b92860138
Hence avoid using SG-GAPS memory registrations. Additionally,
always configure the blk_queue_virt_boundary() to avoid to trigger
a mapping failure when using adapters that support SG-GAPS (e.g.
mlx5).
Fixes: commit ad8e66b4a801 ("IB/srp: fix mr allocation when the device supports sg gaps")
Fixes: commit 509c5f33f4f6 ("IB/srp: Prevent mapping failures")
Reported-by: Laurence Oberman <loberman@redhat.com>
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mark Bloch <markb@mellanox.com>
Cc: Yuval Shaia <yuval.shaia@oracle.com>
Cc: <stable@vger.kernel.org> # 4.7+
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
'qedr_alloc_pbl_tbl()' can not return NULL.
In qedr_init_user_queue():
- simplify the test for the return value, no need to test for NULL
- propagate the error pointer if needed, otherwise 0 (success) is returned.
This is spurious.
In init_mr_info():
- test the return value with IS_ERR
- propagate the error pointer if needed instead of an exlictit -ENOMEM.
This is a no-op as the only error pointer that we can have here is
already -ENOMEM
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Acked-by: Ram Amrani <Ram.Amrani@cavium.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
When CONFIG_DCB is disabled, we get a link error:
drivers/infiniband/built-in.o: In function `bnxt_re_setup_qos':
trace.c:(.text+0x155774): undefined reference to `dcb_ieee_getapp_mask'
trace.c:(.text+0x155774): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `dcb_ieee_getapp_mask'
trace.c:(.text+0x155794): undefined reference to `dcb_ieee_getapp_mask'
trace.c:(.text+0x155794): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `dcb_ieee_getapp_mask'
Like the other drivers that use this function, a Kconfig dependency is
the correct fix.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
|
I ran into a build error on arm64 randconfig testing:
infiniband/hw/hns/hns_roce_main.c:539:1: error: data definition has no type or storage class [-Werror]
infiniband/hw/hns/hns_roce_main.c:539:1: error: type defaults to 'int' in declaration of 'MODULE_DEVICE_TABLE' [-Werror=implicit-int]
infiniband/hw/hns/hns_roce_main.c:539:1: error: parameter names (without types) in function declaration [-Werror]
infiniband/hw/hns/hns_roce_main.c:979:226: error: data definition has no type or storage class [-Werror]
infiniband/hw/hns/hns_roce_main.c:979:226: error: type defaults to 'int' in declaration of 'module_init' [-Werror=implicit-int]
infiniband/hw/hns/hns_roce_main.c:979:1: error: parameter names (without types) in function declaration [-Werror]
Including the module.h makes it build again.
Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Doug Ledford <dledford@redhat.com>
|
