| Age | Commit message (Collapse) | Author |
|---|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse
Pull fuse fix from Miklos Szeredi:
"This fixes a deadlock when fuse, direct I/O and loop device are
combined"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
fuse: direct-io: don't dirty ITER_BVEC pages
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/vfs
Pull overlayfs fix from Miklos Szeredi:
"This fixes a regression caused by the last pull request"
* 'overlayfs-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/vfs:
ovl: fix workdir creation
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs
Pull btrfs fixes from Chris Mason:
"I'm not proud of how long it took me to track down that one liner in
btrfs_sync_log(), but the good news is the patches I was trying to
blame for these problems were actually fine (sorry Filipe)"
* 'for-linus-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
btrfs: introduce tickets_id to determine whether asynchronous metadata reclaim work makes progress
btrfs: remove root_log_ctx from ctx list before btrfs_sync_log returns
btrfs: do not decrease bytes_may_use when replaying extents
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"We've got quite a few fixes at this time, and all are stable patches.
syzkaller strikes back again (episode 19 or so), and we had to plug
some holes in ALSA core part (mostly timer).
In addition, a couple of FireWire audio fixes for the invalid copy
user calls in locks, and a few quirks for HD-audio and USB-audio as
usual are included"
* tag 'sound-4.8-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ALSA: rawmidi: Fix possible deadlock with virmidi registration
ALSA: timer: Fix zero-division by continue of uninitialized instance
ALSA: timer: fix NULL pointer dereference in read()/ioctl() race
ALSA: fireworks: accessing to user space outside spinlock
ALSA: firewire-tascam: accessing to user space outside spinlock
ALSA: hda - Enable subwoofer on Dell Inspiron 7559
ALSA: hda - Add headset mic quirk for Dell Inspiron 5468
ALSA: usb-audio: Add sample rate inquiry quirk for B850V3 CP2114
ALSA: timer: fix NULL pointer dereference on memory allocation failure
ALSA: timer: fix division by zero after SNDRV_TIMER_IOCTL_CONTINUE
|
|
virtio_console uses a small DMA buffer for control requests. Move
that buffer into heap memory.
Doing virtio DMA on the stack is normally okay on non-DMA-API virtio
systems (which is currently most of them), but it breaks completely
if the stack is virtually mapped.
Tested by typing both directions using picocom aimed at /dev/hvc0.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Amit Shah <amit.shah@redhat.com>
|
|
We get 1 warning when building kernel with W=1:
drivers/virtio/virtio_ring.c:170:16: warning: no previous prototype for 'vring_dma_dev' [-Wmissing-prototypes]
In fact, this function is only used in the file in which it is
declared and don't need a declaration, but can be made static.
so this patch marks this function with 'static'.
Signed-off-by: Baoyou Xie <baoyou.xie@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Pull arm64 fixes from Catalin Marinas:
- smp_mb__before_spinlock() changed to smp_mb() on arm64 since the
generic definition to smp_wmb() is not sufficient
- avoid a recursive loop with the graph tracer by using using
preempt_(enable|disable)_notrace in _percpu_(read|write)
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: use preempt_disable_notrace in _percpu_read/write
arm64: spinlocks: implement smp_mb__before_spinlock() as smp_mb()
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc fixes from Michael Ellerman:
"Fixes marked for stable:
- Don't alias user region to other regions below PAGE_OFFSET from
Paul Mackerras
- Fix again csum_partial_copy_generic() on 32-bit from Christophe
Leroy
- Fix corrupted PE allocation bitmap on releasing PE from Gavin Shan
Fixes for code merged this cycle:
- Fix crash on releasing compound PE from Gavin Shan
- Fix processor numbers in OPAL ICP from Benjamin Herrenschmidt
- Fix little endian build with CONFIG_KEXEC=n from Thiago Jung
Bauermann"
* tag 'powerpc-4.8-5' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/mm: Don't alias user region to other regions below PAGE_OFFSET
powerpc/32: Fix again csum_partial_copy_generic()
powerpc/powernv: Fix corrupted PE allocation bitmap on releasing PE
powerpc/powernv: Fix crash on releasing compound PE
powerpc/xics/opal: Fix processor numbers in OPAL ICP
powerpc/pseries: Fix little endian build with CONFIG_KEXEC=n
|
|
Pull ARM fixes from Russell King:
"A few ARM fixes:
- Robin Murphy noticed that the non-secure privileged entry was
relying on undefined behaviour, which needed to be fixed.
- Vladimir Murzin noticed that prov-v7 fails to build for MMUless
configurations because a required header file wasn't included.
- A bunch of fixes for StrongARM regressions found while testing
4.8-rc on such platforms"
* 'fixes' of git://git.armlinux.org.uk/~rmk/linux-arm:
ARM: sa1100: clear reset status prior to reboot
ARM: 8600/1: Enforce some NS-SVC initialisation
ARM: 8599/1: mm: pull asm/memory.h explicitly
ARM: sa1100: register clocks early
ARM: sa1100: fix 3.6864MHz clock
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb into usb-linus
Felipe writes:
usb: fixes for v4.8-rc6
Unfortunately we have a bogus dwc3 patch leaked through the cracks and
got merged into Linus' HEAD. That patch ended up causing off-by-1 error
in our TRB accounting logic. Thankfully John Youn found out the problem
and we provided a revert to the bogus dwc3 patch in no time.
Apart from this off-by-1 error, we have two fixes to the Renesas drivers,
a small fix to our generic phy driver, a NULL pointer dereference fix for
f_eem and a build warning fix in dwc3.
|
|
Commit 3c97f5de1f28 ("ath10k: implement NAPI support") conflicts with
ath-current. To avoid any merge problems merge ath-current to ath-next already now.
Conflicts:
drivers/net/wireless/ath/ath10k/htt_rx.c
|
|
carl9170_usb_stop() is used from several places to flush and cleanup any
pending work. The normal pattern is to send a request and wait for the
irq handler to call complete(). The completion is not reinitialized
during normal operation and as the old comment indicates it is important
to keep calls to wait_for_completion_timeout() and complete() balanced.
Calling complete_all() brings this equilibirum out of balance and needs
to be fixed by a reinit_completion(). But that opens a small race
window. It is possible that the sequence of complete_all(),
reinit_completion() is faster than the wait_for_completion_timeout() can
do its work. The wake up is not lost but the done counter test is after
reinit_completion() has been executed. The only reason we don't see
carl9170_exec_cmd() hang forever is we use the timeout version of
wait_for_copletion().
Let's fix this by reinitializing the completion (that is just setting
done counter to 0) just before we send out an request. Now,
carl9170_usb_stop() can be sure a complete() call is enough to make
progess since there is only one waiter at max. This is a common pattern
also seen in various drivers which use completion.
Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
The ath6kl driver attempts to get the txpower value from the radio by first
clearing the existing stored value and then watching the value to become
non-zero.
APs allow setting client power to values from -127..127, but this radio
is not capable of setting values less then 0 and so will report txpower
as 0dbm for both negative and 0 client power values.
When the radio has txpower set to 0dbm txpower (equivalent to 1mw) the
ath6kl_cfg80211_get_txpower() function will remain in the
wait_event_interruptible_timeout() loop waiting for the value to be
non-zero, and will eventually timeout. This results in a 5 second delay in
response. However, the correct value of zero is eventually returned.
The 6004 defaults to 63dbm which is then limited by regulatory and
hardware limits with max of 18dbm (6003 max is 16dbm), therefore we can
use values larger then these to be able to determine when the value has
been updated.
To correct the issue, set the value to a nonsensical value (255) and wait
for it to change to the valid value.
Tested on both 6003 and 6004 based radios. Return value of zero is
correctly returned in an expected amount of time (similar to when
returning non-zero values) when AP client power is set to both 0 and
negative values.
Signed-off-by: Eric Bentley <eric.bentley@lairdtech.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
The firmware crash dumps on the 6004 are the same as the 6003. Remove the
statement guarding it from dumping on the 6004. Renamed the
REG_DUMP_COUNT_AR6003 to reflect support on both chips.
Signed-off-by: Dan Kephart <dan.kephart@lairdtech.com>
Reviewed-by: Steve deRosier <steve.derosier@lairdtech.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
The constant "123", which is the number of elements in
mask_m / mask_p, is repeated several times in this function.
Replace memsets with array initialization, and replace a loop
conditional with ARRAY_SIZE() so that we don't repeat ourselves.
Signed-off-by: Bob Copeland <me@bobcopeland.com>
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
When user requests for survey dump data, driver is providing wrong survey
information. This information we sent is the survey data that we have
collected during previous user request.
This issue occurs because we request survey dump for wrong channel. With
this change, we correctly display the correct and current survey
information to userspace.
Fixes: fa7937e3d5c2 ("ath10k: update bss channel survey information")
Signed-off-by: Ashok Raj Nagarajan <arnagara@qti.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
The error assigned does not seems to be used anywhere,
fixes nothing just a small cleanup
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
For non-aggregated frames, ath9k was leaving handling of powersave
filtered packets to mac80211. This can be too slow if the intermediate
queue is already filled with packets and mac80211 does not immediately
send a new packet via drv_tx().
Improve response time with filtered frames by triggering clearing the
powersave filter internally.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
It removes the need for undoing the padding changes to skb->data and it
improves performance by eliminating one tx status lookup per MPDU in the
status path. It is also useful for preparing a follow-up fix to better
handle powersave filtering.
A side effect is that these counters, available via debugfs, become now invalid:
* dot11TransmittedFragmentCount
* dot11FrameDuplicateCount,
* dot11ReceivedFragmentCount
* dot11MulticastReceivedFrameCount
Signed-off-by: Felix Fietkau <nbd@nbd.name>
[kvalo@qca.qualcomm.com: add a note about counters, thanks to Zefir Kurtisi]
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
commit 7a0adc83f34d ("ath10k: improve tx scheduling") is causing
severe throughput drop in multi client mode. This issue is originally
reported in veriwave setup with 50 clients with TCP downlink traffic.
While increasing number of clients, the average throughput drops
gradually. With 50 clients, the combined peak throughput is decreased
to 98 Mbps whereas reverting given commit restored it to 550 Mbps.
Processing txqs for every tx completion is causing overhead. Ideally for
management frame tx completion, pending txqs processing can be avoided.
The change partly reverts the commit "ath10k: improve tx scheduling".
Processing pending txqs after all skbs tx completion will yeild enough
room to burst tx frames.
Fixes: 7a0adc83f34d ("ath10k: improve tx scheduling")
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
Add NAPI support for rx and tx completion. NAPI poll is scheduled
from interrupt handler. The design is as below
- on interrupt
- schedule napi and mask interrupts
- on poll
- process all pipes (no actual Tx/Rx)
- process Rx within budget
- if quota exceeds budget reschedule napi poll by returning budget
- process Tx completions and update budget if necessary
- process Tx fetch indications (pull-push)
- push any other pending Tx (if possible)
- before resched or napi completion replenish htt rx ring buffer
- if work done < budget, complete napi poll and unmask interrupts
This change also get rid of two tasklets (intr_tq and txrx_compl_task).
Measured peak throughput with NAPI on IPQ4019 platform in controlled
environment. No noticeable reduction in throughput is seen and also
observed improvements in CPU usage. Approx. 15% CPU usage got reduced
in UDP uplink case.
DL: AP DUT Tx
UL: AP DUT Rx
IPQ4019 (avg. cpu usage %)
========
TOT +NAPI
=========== =============
TCP DL 644 Mbps (42%) 645 Mbps (36%)
TCP UL 673 Mbps (30%) 675 Mbps (26%)
UDP DL 682 Mbps (49%) 680 Mbps (49%)
UDP UL 720 Mbps (28%) 717 Mbps (11%)
Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/peter.chen/usb into usb-linus
Peter writes:
Fix the possible kernel panic when the hardware signal is bad for chipidea udc.
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio into staging-linus
Jonathan writes:
Second set of IIO fixes for the 4.8 cycle.
We have a big rework of the kxsd9 driver queued up behind the fix below and
a fix for a recent fix that was marked for stable.
Hence this fix series is perhaps a little more urgent than average for IIO.
* core
- a fix for a fix in the last set. The recent fix for blocking ops when
! task running left a path (unlikely one) in which the function return
value was not set - so initialise it to 0.
- The IIO_TYPE_FRACTIONAL code previously didn't cope with negative
fractions. Turned out a fix for this was in Analog's tree but hadn't made
it upstream.
* bmc150
- reset chip at init time. At least one board out there ends up coming up
in an unstable state due to noise during power up. The reset does no
harm on other boards.
* kxsd9
- Fix a bug in the reported scaling due to failing to set the integer
part to 0.
* hid-sensors-pressure
- Output was in the wrong units to comply with the IIO ABI.
* tools
- iio_generic_buffer: Fix the trigger-less mode by ensuring we don't fault
out for having no trigger when we explicitly said we didn't want to have
one.
|
|
When debug preempt or preempt tracer is enabled, preempt_count_add/sub()
can be traced by function and function graph tracing, and
preempt_disable/enable() would call preempt_count_add/sub(), so in Ftrace
subsystem we should use preempt_disable/enable_notrace instead.
In the commit 345ddcc882d8 ("ftrace: Have set_ftrace_pid use the bitmap
like events do") the function this_cpu_read() was added to
trace_graph_entry(), and if this_cpu_read() calls preempt_disable(), graph
tracer will go into a recursive loop, even if the tracing_on is
disabled.
So this patch change to use preempt_enable/disable_notrace instead in
this_cpu_read().
Since Yonghui Yang helped a lot to find the root cause of this problem,
so also add his SOB.
Signed-off-by: Yonghui Yang <mark.yang@spreadtrum.com>
Signed-off-by: Chunyan Zhang <zhang.chunyan@linaro.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
|
|
smp_mb__before_spinlock() is intended to upgrade a spin_lock() operation
to a full barrier, such that prior stores are ordered with respect to
loads and stores occuring inside the critical section.
Unfortunately, the core code defines the barrier as smp_wmb(), which
is insufficient to provide the required ordering guarantees when used in
conjunction with our load-acquire-based spinlock implementation.
This patch overrides the arm64 definition of smp_mb__before_spinlock()
to map to a full smp_mb().
Cc: <stable@vger.kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Reported-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
|
|
The patch corrects the revision id register and uses it along with
magic value and chip version registers to download appropriate firmware
image.
PCIe8997 Z chipset variant code has been removed, as it won't be used in
production.
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
This patch:
- improves code layout
- removes a useless memset(0) for some memory allocated with kzalloc
- removes a useless if. We know that 'if (chan_band_tlv)' will succeed
because it has been tested a few lines above
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
Problems with the signal integrity of the high speed USB data lines or
noise on reference ground lines can cause the i.MX6 USB controller to
violate USB specs and exhibit unexpected behavior.
It was observed that USBi_UI interrupts were triggered first and when
isr_setup_status_phase was called, ci->status was NULL, which lead to a
NULL pointer dereference kernel panic.
This patch fixes the kernel panic, emits a warning once and returns
-EPIPE to halt the device and let the host get stalled.
It also adds a comment to point people, who are experiencing this issue,
to their USB hardware design.
Cc: <stable@vger.kernel.org> #4.1+
Signed-off-by: Clemens Gruber <clemens.gruber@pqgruber.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
|
|
This patch implements pre and post FLR handlers to support PCIe FLR
functionality. Software cleanup is performed in pre-FLR whereas
firmware is downloaded and software is re-initialised in
post-FLR handler.
Following command triggers FLR.
echo "1" > /sys/bus/pci/devices/$NUMBER/reset
This feature can be used as a recovery mechanism when firmware gets
hang.
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
In brcmf_sdio_kso_control() there is a retry loop as hardware may take
time to settle. However, when the call to brcmf_sdiod_regrb() returns
an error it is due to SDIO access failure and it makes no sense to wait
for hardware to settle. This patch aborts the loop after a number of
subsequent access errors.
Reviewed-by: Hante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: Franky Lin <franky.lin@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
The driver already supports the bcm4339 chipset but only for the variant
that shares the same modalias as the bcm4335, ie. sdio:c00v02D0d4335.
It turns out that there are also bcm4339 devices out there that have a
more distiguishable modalias sdio:c00v02D0d4339.
Reported-by: Steve deRosier <derosier@gmail.com>
Reviewed-by: Hante Meuleman <hante.meuleman@broadcom.com>
Reviewed-by: Pieter-Paul Giesberts <pieter-paul.giesberts@broadcom.com>
Reviewed-by: Franky Lin <franky.lin@broadcom.com>
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
Use the newly added linux/bitfield.h.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
There is no need to include linux/version.h in a in-tree
driver.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
Remove redefinition of GENMASK which should not be there
in the upstream version of the code.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
Common approach to accessing register fields is to define
structures or sets of macros containing mask and shift pair.
Operations on the register are then performed as follows:
field = (reg >> shift) & mask;
reg &= ~(mask << shift);
reg |= (field & mask) << shift;
Defining shift and mask separately is tedious. Ivo van Doorn
came up with an idea of computing them at compilation time
based on a single shifted mask (later refined by Felix) which
can be used like this:
#define REG_FIELD 0x000ff000
field = FIELD_GET(REG_FIELD, reg);
reg &= ~REG_FIELD;
reg |= FIELD_PREP(REG_FIELD, field);
FIELD_{GET,PREP} macros take care of finding out what the
appropriate shift is based on compilation time ffs operation.
GENMASK can be used to define registers (which is usually
less error-prone and easier to match with datasheets).
This approach is the most convenient I've seen so to limit code
multiplication let's move the macros to a global header file.
Attempts to use static inlines instead of macros failed due
to false positive triggering of BUILD_BUG_ON()s, especially with
GCC < 6.0.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
This patch adds cfg80211 testmode support so that userspace tools can
download necessary commands to firmware during manufacturing mode tests.
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
By default normal mode is chosen when driver is loaded. This
patch adds a provision to choose manufacturing mode via module
parameters.
Below command loads driver in manufacturing mode
insmod mwifiex.ko mfg_mode=1.
Tested-by: chunfan chen <jeffc@marvell.com>
Signed-off-by: Xinming Hu <huxm@marvell.com>
Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
While fixing another bug, I noticed that bcma manually sets up
a dma_mask pointer for its child devices. We have a generic
helper for that now, which should be able to cope better with
any variations that might be needed to deal with cache coherency,
unusual DMA address offsets, iommus, or limited DMA masks, none
of which are currently handled here.
This changes the core to use the of_dma_configure(), like
we do for platform devices that are probed directly from
DT.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
|
|
In commit f02db315b8d8 ("ipv4: IP_TOS and IP_TTL can be specified as
ancillary data") Francesco added IP_TOS values specified as integer.
However, kernel sends to userspace (at recvmsg() time) an IP_TOS value
in a single byte, when IP_RECVTOS is set on the socket.
It can be very useful to reflect all ancillary options as given by the
kernel in a subsequent sendmsg(), instead of aborting the sendmsg() with
EINVAL after Francesco patch.
So this patch extends IP_TOS ancillary to accept an u8, so that an UDP
server can simply reuse same ancillary block without having to mangle
it.
Jesper can then augment
https://github.com/netoptimizer/network-testing/blob/master/src/udp_example02.c
to add TOS reflection ;)
Fixes: f02db315b8d8 ("ipv4: IP_TOS and IP_TTL can be specified as ancillary data")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Francesco Fusco <ffusco@redhat.com>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
LLVM can generate code that tests for direct packet access via
skb->data/data_end in a way that currently gets rejected by the
verifier, example:
[...]
7: (61) r3 = *(u32 *)(r6 +80)
8: (61) r9 = *(u32 *)(r6 +76)
9: (bf) r2 = r9
10: (07) r2 += 54
11: (3d) if r3 >= r2 goto pc+12
R1=inv R2=pkt(id=0,off=54,r=0) R3=pkt_end R4=inv R6=ctx
R9=pkt(id=0,off=0,r=0) R10=fp
12: (18) r4 = 0xffffff7a
14: (05) goto pc+430
[...]
from 11 to 24: R1=inv R2=pkt(id=0,off=54,r=0) R3=pkt_end R4=inv
R6=ctx R9=pkt(id=0,off=0,r=0) R10=fp
24: (7b) *(u64 *)(r10 -40) = r1
25: (b7) r1 = 0
26: (63) *(u32 *)(r6 +56) = r1
27: (b7) r2 = 40
28: (71) r8 = *(u8 *)(r9 +20)
invalid access to packet, off=20 size=1, R9(id=0,off=0,r=0)
The reason why this gets rejected despite a proper test is that we
currently call find_good_pkt_pointers() only in case where we detect
tests like rX > pkt_end, where rX is of type pkt(id=Y,off=Z,r=0) and
derived, for example, from a register of type pkt(id=Y,off=0,r=0)
pointing to skb->data. find_good_pkt_pointers() then fills the range
in the current branch to pkt(id=Y,off=0,r=Z) on success.
For above case, we need to extend that to recognize pkt_end >= rX
pattern and mark the other branch that is taken on success with the
appropriate pkt(id=Y,off=0,r=Z) type via find_good_pkt_pointers().
Since eBPF operates on BPF_JGT (>) and BPF_JGE (>=), these are the
only two practical options to test for from what LLVM could have
generated, since there's no such thing as BPF_JLT (<) or BPF_JLE (<=)
that we would need to take into account as well.
After the fix:
[...]
7: (61) r3 = *(u32 *)(r6 +80)
8: (61) r9 = *(u32 *)(r6 +76)
9: (bf) r2 = r9
10: (07) r2 += 54
11: (3d) if r3 >= r2 goto pc+12
R1=inv R2=pkt(id=0,off=54,r=0) R3=pkt_end R4=inv R6=ctx
R9=pkt(id=0,off=0,r=0) R10=fp
12: (18) r4 = 0xffffff7a
14: (05) goto pc+430
[...]
from 11 to 24: R1=inv R2=pkt(id=0,off=54,r=54) R3=pkt_end R4=inv
R6=ctx R9=pkt(id=0,off=0,r=54) R10=fp
24: (7b) *(u64 *)(r10 -40) = r1
25: (b7) r1 = 0
26: (63) *(u32 *)(r6 +56) = r1
27: (b7) r2 = 40
28: (71) r8 = *(u8 *)(r9 +20)
29: (bf) r1 = r8
30: (25) if r8 > 0x3c goto pc+47
R1=inv56 R2=imm40 R3=pkt_end R4=inv R6=ctx R8=inv56
R9=pkt(id=0,off=0,r=54) R10=fp
31: (b7) r1 = 1
[...]
Verifier test cases are also added in this work, one that demonstrates
the mentioned example here and one that tries a bad packet access for
the current/fall-through branch (the one with types pkt(id=X,off=Y,r=0),
pkt(id=X,off=0,r=0)), then a case with good and bad accesses, and two
with both test variants (>, >=).
Fixes: 969bf05eb3ce ("bpf: direct packet access")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Over the years, TCP BDP has increased by several orders of magnitude,
and some people are considering to reach the 2 Gbytes limit.
Even with current window scale limit of 14, ~1 Gbytes maps to ~740,000
MSS.
In presence of packet losses (or reorders), TCP stores incoming packets
into an out of order queue, and number of skbs sitting there waiting for
the missing packets to be received can be in the 10^5 range.
Most packets are appended to the tail of this queue, and when
packets can finally be transferred to receive queue, we scan the queue
from its head.
However, in presence of heavy losses, we might have to find an arbitrary
point in this queue, involving a linear scan for every incoming packet,
throwing away cpu caches.
This patch converts it to a RB tree, to get bounded latencies.
Yaogong wrote a preliminary patch about 2 years ago.
Eric did the rebase, added ofo_last_skb cache, polishing and tests.
Tested with network dropping between 1 and 10 % packets, with good
success (about 30 % increase of throughput in stress tests)
Next step would be to also use an RB tree for the write queue at sender
side ;)
Signed-off-by: Yaogong Wang <wygivan@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Acked-By: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Jakub Kicinski says:
====================
nfp: fixes and trivial cleanup
First patch drops unnecessary version.h includes. Second one
drops support for pre-release versions of FW ABI. Removing
FW ABI 0.0 from supported set is particularly good since 0
could just be uninitialized memory. Last but not least I drop
unnecessary padding of frames on RX which makes us count bytes
incorrectly for the VF2VF traffic.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
There is no need to pad frames to ETH_ZLEN on RX.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Be more strict about FW versions. Drop support for old
transitional revisions which were never used in production.
Dropping support for FW ABI version 0.0.0.0 is particularly
useful because 0 could just be uninitialized memory.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Remove unnecessary version.h includes.
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Commit 76174004a0f19785a328f40388e87e982bbf69b9
(tcp: do not slow start when cwnd equals ssthresh )
introduced regression in TCP YeAH. Using 100ms delay 1% loss virtual
ethernet link kernel 4.2 shows bandwidth ~500KB/s for single TCP
connection and kernel 4.3 and above (including 4.8-rc4) shows bandwidth
~100KB/s.
That is caused by stalled cwnd when cwnd equals ssthresh. This patch
fixes it by proper increasing cwnd in this case.
Signed-off-by: Artem Germanov <agermanov@anchorfree.com>
Acked-by: Dmitry Adamushko <d.adamushko@anchorfree.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Eric Garver says:
====================
openvswitch: add 802.1ad support
This series adds 802.1ad support to openvswitch. It is a continuation of the
work originally started by Thomas F Herbert - hence the large rev number.
The extra VLAN is implemented by using an additional level of the
OVS_KEY_ATTR_ENCAP netlink attribute.
In OVS flow speak, this looks like
eth_type(0x88a8),vlan(vid=100),encap(eth_type(0x8100), vlan(vid=200),
encap(eth_type(0x0800), ...))
The userspace counterpart has also seen recent activity on the ovs-dev mailing
lists. There are some new 802.1ad OVS tests being added - also on the ovs-dev
list. This patch series has been tested using the most recent version of
userspace (v3) and tests (v2).
v22 changes:
- merge patch 4 into patch 3
- fix checkpatch.pl errors
- Still some 80 char warnings for long string literals
- refresh pointer after pskb_may_pull()
- refactor vlan nlattr parsing to remove some double checks
- introduce ovs_nla_put_vlan()
- move triple VLAN check to after ethertype serialization
- WARN_ON_ONCE() on triple VLAN and unexpected encap values
v21 changes:
- Fix (and simplify) netlink attribute parsing
- re-add handling of truncated VLAN tags
- fix if/else dangling assignment in {push,pop}_vlan()
- simplify parse_vlan()
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add support for 802.1ad including the ability to push and pop double
tagged vlans. Add support for 802.1ad to netlink parsing and flow
conversion. Uses double nested encap attributes to represent double
tagged vlan. Inner TPID encoded along with ctci in nested attributes.
This is based on Thomas F Herbert's original v20 patch. I made some
small clean ups and bug fixes.
Signed-off-by: Thomas F Herbert <thomasfherbert@gmail.com>
Signed-off-by: Eric Garver <e@erig.me>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This is to simplify using double tagged vlans. This function allows all
valid vlan ethertypes to be checked in a single function call.
Also replace some instances that check for both ETH_P_8021Q and
ETH_P_8021AD.
Patch based on one originally by Thomas F Herbert.
Signed-off-by: Thomas F Herbert <thomasfherbert@gmail.com>
Signed-off-by: Eric Garver <e@erig.me>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
openvswitch: Add support for 8021.AD
Change the description of the VLAN tpid field.
Signed-off-by: Thomas F Herbert <thomasfherbert@gmail.com>
Acked-by: Pravin B Shelar <pshelar@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
