Allow the user to limit the number of requests serviced through a single
connection, to help prevent faster clients from starving slower clients.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Don't call svc_xprt_enqueue() if the XPT_DATA flag is already set.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Rather than code up our own versions of the socket callbacks, just
call the defaults.
This also allows us to merge svc_udp_data_ready() and svc_tcp_data_ready().
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Prevent callbacks from triggering while we're detaching the socket.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Dropping and/or deferring requests has an impact on performance. Let's
make sure we can trace those events.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Add a tracepoint to track when the processing of incoming RPC data gets
deferred due to out-of-space issues on the outgoing transport.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Silent a few smatch warnings about indentation
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Those are now defined in fs/nfsd/vfs.h
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Reviewed-by: Jeff Layton <jlayton@poochiereds.net>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Have a simple flex file server where the mds (NFSv4.1 or NFSv4.2)
is also the ds (NFSv3). I.e., the metadata and the data file are
the exact same file.
This will allow testing of the flex file client.
Simply add the "pnfs" export option to your export
in /etc/exports and mount from a client that supports
flex files.
Signed-off-by: Tom Haynes <loghyr@primarydata.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Tom Haynes <loghyr@primarydata.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jeff Layton <jlayton@poochiereds.net>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
GSS-Proxy doesn't produce very much debug logging at all. Printing out
the gss minor status will aid in troubleshooting if the
GSS_Accept_sec_context upcall fails.
Signed-off-by: Scott Mayhew <smayhew@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
This addresses the conundrum referenced in RFC5661 18.35.3,
and will allow clients to return state to the server using the
machine credentials.
The biggest part of the problem is that we need to allow the client
to send a compound op with integrity/privacy on mounts that don't
have it enabled.
Add server support for properly decoding and using spo_must_enforce
and spo_must_allow bits. Add support for machine credentials to be
used for CLOSE, OPEN_DOWNGRADE, LOCKU, DELEGRETURN,
and TEST/FREE STATEID.
Implement a check so as to not throw WRONGSEC errors when these
operations are used if integrity/privacy isn't turned on.
Without this, Linux clients with credentials that expired while holding
delegations were getting stuck in an endless loop.
Signed-off-by: Andrew Elble <aweits@rit.edu>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Rename mach_creds_match() to nfsd4_mach_creds_match() and un-staticify
Signed-off-by: Andrew Elble <aweits@rit.edu>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
So these may be used in nfsd as well
Signed-off-by: Andrew Elble <aweits@rit.edu>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Willem de Bruijn says:
====================
limit sk_filter trim to payload
Sockets can apply a filter to incoming packets to drop or trim them.
Fix two codepaths that call skb_pull/__skb_pull after sk_filter
without checking for packet length.
Reading beyond skb->tail after trimming happens in more codepaths, but
safety of reading in the linear segment is based on minimum allocation
size (MAX_HEADER, GRO_MAX_HEAD, ..).
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Dccp verifies packet integrity, including length, at initial rcv in
dccp_invalid_packet, later pulls headers in dccp_enqueue_skb.
A call to sk_filter in-between can cause __skb_pull to wrap skb->len.
skb_copy_datagram_msg interprets this as a negative value, so
(correctly) fails with EFAULT. The negative length is reported in
ioctl SIOCINQ or possibly in a DCCP_WARN in dccp_close.
Introduce an sk_receive_skb variant that caps how small a filter
program can trim packets, and call this in dccp with the header
length. Excessively trimmed packets are now processed normally and
queued for reception as 0B payloads.
Fixes: 7c657876b6 ("[DCCP]: Initial implementation")
Signed-off-by: Willem de Bruijn <willemb@google.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sockets can have a filter program attached that drops or trims
incoming packets based on the filter program return value.
Rose requires data packets to have at least ROSE_MIN_LEN bytes. It
verifies this on arrival in rose_route_frame and unconditionally pulls
the bytes in rose_recvmsg. The filter can trim packets to below this
value in-between, causing pull to fail, leaving the partial header at
the time of skb_copy_datagram_msg.
Place a lower bound on the size to which sk_filter may trim packets
by introducing sk_filter_trim_cap and call this for rose packets.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
If a system configures the Nuvoton chip to use the alternative
EFM IO address (CR_EFIR2) then after probing the primary EFM IO
address (CR_EFIR) this region is not released.
If a driver for another function of the Nuvoton Super I/O
chip uses the same probing mechanism then it will hang if
loaded after the nuvoton-cir driver.
This was reported for the nct6775 hwmon driver.
Fix this by properly releasing the region after probing CR_EFIR.
This regression was introduced with kernel 4.6 so cc it to stable.
Reported-by: Antti Seppälä <a.seppala@gmail.com>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Tested-by: Antti Seppälä <a.seppala@gmail.com>
Cc: <stable@vger.kernel.org> # 4.6.x-
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
During shutdown sequence, in case all modules are already removed,
codec state is not updated. Though it's not causing any harm for now,
but it's good to maintain proper codec state. Fix this.
Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Alex Elder <elder@linaro.org>
In case module is removed dynamically with ongoing playback, during
module cleanup codec state is mistakenly modified. State should be
modified for module only. Fix this.
Fixes: 76414cb499b7 ("audio: Use single codec driver
registration")
Signed-off-by: Vaibhav Agarwal <vaibhav.agarwal@linaro.org>
Reviewed-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Alex Elder <elder@linaro.org>
The wait_event() call in dvb_unregister_frontend() waits synchronously
for other tasks to free a file descriptor, but it does that while
holding several mutexes. That alone is a bad idea, but if one user
process happens to keep a (defunct) file descriptor open indefinitely,
the kernel will correctly detect a hung task:
INFO: task kworker/0:1:314 blocked for more than 30 seconds.
Not tainted 4.7.0-rc1-hosting+ #50
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
kworker/0:1 D ffff88003daf7a50 0 314 2 0x00000000
Workqueue: usb_hub_wq hub_event
ffff88003daf7a50 0000000000000296 ffff88003daf7a30 ffff88003fc13f98
ffff88003dadce00 ffff88003daf8000 ffff88003e3fc010 ffff88003d48d4f8
ffff88003e3b5030 ffff88003e3f8898 ffff88003daf7a68 ffffffff810cf860
Call Trace:
[<ffffffff810cf860>] schedule+0x30/0x80
[<ffffffff812f88d3>] dvb_unregister_frontend+0x93/0xc0
[<ffffffff8107a000>] ? __wake_up_common+0x80/0x80
[<ffffffff813019c7>] dvb_usb_adapter_frontend_exit+0x37/0x70
[<ffffffff81300614>] dvb_usb_exit+0x34/0xb0
[<ffffffff81300d4a>] dvb_usb_device_exit+0x3a/0x50
[<ffffffff81302dc2>] pctv452e_usb_disconnect+0x52/0x60
[<ffffffff81295a07>] usb_unbind_interface+0x67/0x1e0
[<ffffffff810609f3>] ? __blocking_notifier_call_chain+0x53/0x70
[<ffffffff8127ba67>] __device_release_driver+0x77/0x110
[<ffffffff8127c2d3>] device_release_driver+0x23/0x30
[<ffffffff8127ab5d>] bus_remove_device+0x10d/0x150
[<ffffffff8127879b>] device_del+0x13b/0x260
[<ffffffff81299dea>] ? usb_remove_ep_devs+0x1a/0x30
[<ffffffff8129468e>] usb_disable_device+0x9e/0x1e0
[<ffffffff8128bb09>] usb_disconnect+0x89/0x260
[<ffffffff8128db8d>] hub_event+0x30d/0xfc0
[<ffffffff81059475>] process_one_work+0x1c5/0x4a0
[<ffffffff8105940c>] ? process_one_work+0x15c/0x4a0
[<ffffffff81059799>] worker_thread+0x49/0x480
[<ffffffff81059750>] ? process_one_work+0x4a0/0x4a0
[<ffffffff81059750>] ? process_one_work+0x4a0/0x4a0
[<ffffffff8105f65e>] kthread+0xee/0x110
[<ffffffff810400bf>] ret_from_fork+0x1f/0x40
[<ffffffff8105f570>] ? __kthread_unpark+0x70/0x70
5 locks held by kworker/0:1/314:
#0: ("usb_hub_wq"){......}, at: [<ffffffff8105940c>] process_one_work+0x15c/0x4a0
#1: ((&hub->events)){......}, at: [<ffffffff8105940c>] process_one_work+0x15c/0x4a0
#2: (&dev->mutex){......}, at: [<ffffffff8128d8cb>] hub_event+0x4b/0xfc0
#3: (&dev->mutex){......}, at: [<ffffffff8128bad2>] usb_disconnect+0x52/0x260
#4: (&dev->mutex){......}, at: [<ffffffff8127c2cb>] device_release_driver+0x1b/0x30
This patch removes the blocking wait, and postpones the kfree() call
until all file handles have been closed by using struct kref.
Signed-off-by: Max Kellermann <max@duempel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Not all libelf implementations have this "Please, ELF_C_READ, but use
mmap if possible" elf_begin() cmd, so provide a fallback to plain old
ELF_C_READ.
Case in point: Alpine Linux 3.4.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-1fctuknrawgoi5xqon4mu9dv@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Saeed Mahameed says:
====================
mlx5 tx timeout watchdog fixes
This patch set provides two trivial fixes for the tx timeout series lately
applied into net 4.7.
From Daniel, detect stuck queues due to BQL
From Mohamad, fix tx timeout watchdog false alarm
Hopefully those two fixes will make it to -stable, assuming
3947ca1859 ('net/mlx5e: Implement ndo_tx_timeout callback') was also backported to -stable.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Start all tx queues (including inactive ones) when opening the netdev.
Stop all tx queues (including inactive ones) when closing the netdev.
This is a workaround for the tx timeout watchdog false alarm issue in
which the netdev watchdog is polling all the tx queues which may include
inactive queues and thus once lowering the real tx queues number
(ethtool -L) it will generate tx timeout watchdog false alarms.
Fixes: 3947ca1859 ('net/mlx5e: Implement ndo_tx_timeout callback')
Signed-off-by: Mohamad Haj Yahia <mohamad@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change netif_tx_queue_stopped to netif_xmit_stopped. This will show
when queues are stopped due to byte queue limits.
Fixes: 3947ca1859 ('net/mlx5e: Implement ndo_tx_timeout callback')
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update vrf documentation for changes made to 4.4 - 4.8 kernels
and iproute2 support for vrf keyword.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The LIRC_CAN_REC() returns a boolean "flag & LIRC_CAN_REC_MASK"
to check whether the device can receive data.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
When opening or closing a lirc character device, the framework
provides to the user the possibility to keep track of opening or
closing of the device by calling two functions:
- set_use_inc() when opening the device
- set_use_dec() when closing the device
if those are not set by the lirc user, the system segfaults.
Check the pointer value before calling the above functions.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
If ioctl is called, it cannot be a case of invalid system call
number (ENOSYS), that is a ENOTTY case which means that the
device doesn't support that specific ioctl command.
So, replace ENOSYS with ENOTTY.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
When comparing a variable with a constant, the comparison should
start from the variable and not from the constant. It's also
written in the human DNA.
Swap the terms of comparisons whenever the constant comes first
and fix the following checkpatch warning:
WARNING: Comparisons should place the constant on the right side of the test
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The three if statements check the same thing, merge them in only
one statement.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
There are two if ... else which check the same thing in different
part of the code, they can be merged in a single check.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The whole function is inside an 'if' statement
("if (ir->d.add_to_buf)").
Check the opposite of that statement at the beginning and exit,
this way we can have one level less of indentation.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
It was set based on CONFIG_64BIT, that is available only when using
Kconfig, which we're working towards but not to the point of having this
CONFIG variable set, so synthesize it from available compiler defined
defines, __SIZEOF_LONG__ or, lacking that, __WORDSIZE.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-og5fmkr17856lhupacihwxvb@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The code can be rearranged so that some goto paths can be removed
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
This patch mutes also all the checkpatch warnings related to
printk.
Reword all the printouts so that the string doesn't need to be
split, which fixes the following checkpatch warning:
WARNING: quoted string split across lines
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Transmitters don't necessarily need to have a FIFO managed buffer
for their transfers.
When registering the driver, before allocating the buffer, check
whether the device is a transmitter or receiver. Allocate the
buffer only for receivers.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
During the driver registration, move the buffer allocation on a
separate function.
Signed-off-by: Andi Shyti <andi.shyti@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The newly added FCP support in the vsp1 driver causes a link error
when CONFIG_RENESAS_FCP=m, since it's not reachable by built-in code:
drivers/media/built-in.o: In function `vsp1_remove':
:(.text+0xdeca0): undefined reference to `rcar_fcp_put'
drivers/media/built-in.o: In function `vsp1_probe':
:(.text+0xdef44): undefined reference to `rcar_fcp_get'
We already have a conditional dependency on FCP that requires
it for ARM64, so for all others we just have to prevent setting
RENESAS_VSP1=y when RENESAS_FCP=m by extending the FCP dependency.
Fixes: 94fcdf8297 ("[media] v4l: vsp1: Add FCP support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Building without CONFIG_PM results in a harmless warning from
slightly incorrect #ifdef guards:
drivers/media/platform/vsp1/vsp1_drv.c:525:12: error: 'vsp1_pm_runtime_resume' defined but not used [-Werror=unused-function]
drivers/media/platform/vsp1/vsp1_drv.c:516:12: error: 'vsp1_pm_runtime_suspend' defined but not used [-Werror=unused-function]
This removes the existing #ifdef and instead marks all four
PM functions as __maybe_unused.
Fixes: 1e6af546ee ("[media] v4l: vsp1: Implement runtime PM support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
* topic/vsp1: (36 commits)
[media] v4l: vsp1: wpf: Add flipping support
[media] v4l: vsp1: rwpf: Support runtime modification of controls
[media] v4l: vsp1: Simplify alpha propagation
[media] v4l: vsp1: clu: Support runtime modification of controls
[media] v4l: vsp1: lut: Support runtime modification of controls
[media] v4l: vsp1: Support runtime modification of controls
[media] v4l: vsp1: Add Cubic Look Up Table (CLU) support
[media] v4l: vsp1: lut: Expose configuration through a control
[media] v4l: vsp1: lut: Initialize the mutex
[media] v4l: vsp1: dl: Don't free fragments with interrupts disabled
[media] v4l: vsp1: Set entities functions
[media] v4l: vsp1: Don't create LIF entity when the userspace API is enabled
[media] v4l: vsp1: Don't register media device when userspace API is disabled
[media] v4l: vsp1: Base link creation on availability of entities
[media] media: Add video statistics computation functions
[media] media: Add video processing entity functions
[media] v4l: vsp1: sru: Fix intensity control ID
[media] v4l: vsp1: Stop the pipeline upon the first STREAMOFF
[media] v4l: vsp1: Constify operation structures
[media] v4l: vsp1: pipe: Fix typo in comment
...
Now that the frame rate can be properly set, this commit adds support
for S_PARM and G_PARM.
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Instead of using two tables to estimate the temporal decimation
factor, use a formula. This allows to get the closest fps, which
sounds better than the current tables.
Also, the current code doesn't store the real framerate.
This patch fixes the above issues.
[Ezequiel:
- introduce a TW686X_MAX_FPS macro for max_fps.
- use hweight_long instead of open coding the set bits count.]
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
For 4K LBA or very large disks, atari_partition can easily get tricked
into thinking it has found an Atari partition table. Depending on the
data in the disk, it ends up creating partitions with awkward lengths.
We saw logs like this while playing with fio.
[5.625867] nvme2n1: AHDI p2
[5.625872] nvme2n1: p2 size 2910030523 extends beyond EOD, truncated
People has had issues with misinterpreted AHDI partition tables for a long
time, see this BSD thread from 1995, for example.
https://mail-index.netbsd.org/port-atari/1995/11/19/0001.html
Since the atari partition, according to the spec, doesn't even support
sector sizes with more than 512, a quick sanity check is reasonable to
just bail out early, before even attempting to read sector 0.
Signed-off-by: Gabriel Krisman Bertazi <krisman@linux.vnet.ibm.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
