DFS offload support implemented:
- DFS_OFFLOAD feature is advertised depending on HW capabilities
- CAC_STARTED event forwarding from HW implemented
- start_radar_detection() callback now returning -ENOTSUPP
if DFS_OFFLOAD is enabled
Signed-off-by: Dmitry Lebed <dlebed@quantenna.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
NeilBrown says:
====================
A few rhashtables cleanups
2 patches fixes documentation
1 fixes a bit in rhashtable_walk_start()
1 improves rhashtable_walk stability.
All reviewed and Acked.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When a walk of an rhashtable is interrupted with rhastable_walk_stop()
and then rhashtable_walk_start(), the location to restart from is based
on a 'skip' count in the current hash chain, and this can be incorrect
if insertions or deletions have happened. This does not happen when
the walk is not stopped and started as iter->p is a placeholder which
is safe to use while holding the RCU read lock.
In rhashtable_walk_start() we can revalidate that 'p' is still in the
same hash chain. If it isn't then the current method is still used.
With this patch, if a rhashtable walker ensures that the current
object remains in the table over a stop/start period (possibly by
elevating the reference count if that is sufficient), it can be sure
that a walk will not miss objects that were in the hashtable for the
whole time of the walk.
rhashtable_walk_start() may not find the object even though it is
still in the hashtable if a rehash has moved it to a new table. In
this case it will (eventually) get -EAGAIN and will need to proceed
through the whole table again to be sure to see everything at least
once.
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The documentation claims that when rhashtable_walk_start_check()
detects a resize event, it will rewind back to the beginning
of the table. This is not true. We need to set ->slot and
->skip to be zero for it to be true.
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Neither rhashtable_walk_enter() or rhltable_walk_enter() sleep, though
they do take a spinlock without irq protection.
So revise the comments to accurately state the contexts in which
these functions can be called.
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
grow_decision and shink_decision no longer exist, so remove
the remaining references to them.
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RETPOLINE made calls to tp->af_specific->md5_lookup() quite expensive,
given they have no result.
We can omit the calls for sockets that have no md5 keys.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Updates to the bitfields in struct packet_sock are not atomic.
Serialize these read-modify-write cycles.
Move po->running into a separate variable. Its writes are protected by
po->bind_lock (except for one startup case at packet_create). Also
replace a textual precondition warning with lockdep annotation.
All others are set only in packet_setsockopt. Serialize these
updates by holding the socket lock. Analogous to other field updates,
also hold the lock when testing whether a ring is active (pg_vec).
Fixes: 8dc4194474 ("[PACKET]: Add optional checksum computation for recvmsg")
Reported-by: DaeRyong Jeong <threeearcat@gmail.com>
Reported-by: Byoungyoung Lee <byoungyoung@purdue.edu>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Similar to the readx_poll_timeout() macro calling ktime_* and using
ktime_t type, which is declared in <linux/ktime.h>. So, make
include/linux/regmap.h explicitly include <linux/ktime.h>, like
include/linux/iopoll.h does. Otherwise, users of the macro will see
below errors.
error: implicit declaration of function ‘ktime_add_us’
[-Werror=implicit-function-declaration]
error: implicit declaration of function ‘ktime_get’
[-Werror=implicit-function-declaration]
error: implicit declaration of function ‘ktime_compare’
[-Werror=implicit-function-declaration]
include/linux/regmap.h:128:2: error: unknown type name ‘ktime_t’
ktime_t __timeout = ktime_add_us(ktime_get(), __timeout_us); \
Signed-off-by: Sean Wang <sean.wang@mediatek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The method struct vga_switcheroo_handler::get_client_id() is defined
as returning an 'enum vga_switcheroo_client_id' but the implementation
in this driver, gmux_get_client_id(), returns an 'int'.
Fix this by returning 'enum vga_switcheroo_client_id' in this driver too.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Instead of synchronously cancelling the timer and re-enabling it inside
the reset callbacks, keep the timer enabled and let it die on its next
wakeup if no longer required. This allows
intel_engine_reset_breadcrumbs() to be used from an atomic
(timer/softirq) context such as required for resetting an engine.
It also allows us to react better to the user poking around debugfs for
testing missed irqs.
v2: Tighten the order of del_timer_sync as the fake_irq timer
may trigger the hangcheck timer, and so we should cancel it first and
then cancel the hangcheck (Mika)
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424142945.6787-1-chris@chris-wilson.co.uk
Since commit a521422ea4 ("ARM: shmobile: mackerel: Remove Legacy C
board code"), the only remaining platforms using this driver are SuperH
SH-Mobile SoCs (sh7723). As both SUPERH and ARCH_SHMOBILE are set for
these platforms, the SUPERH dependency can be dropped.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
We should get drvdata from struct device directly. Going via
platform_device is an unneeded step back and forth.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
We should get drvdata from struct device directly. Going via
platform_device is an unneeded step back and forth.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Use new return type vm_fault_t for fault handler
and page_mkwrite handler.
Signed-off-by: Souptick Joarder <jrdr.linux@gmail.com>
Reviewed-by: Matthew Wilcox <mawilcox@microsoft.com>
Cc: Jaya Kumar <jayakumar.lkml@gmail.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
radeonfb_pci_suspend() is never called in atomic context.
radeonfb_pci_suspend() is only set as ".suspend" in struct pci_driver.
This function is not called in atomic context.
Despite never getting called from atomic context, radeonfb_pci_suspend()
calls mdelay() to busily wait.
This is not necessary and can be replaced with msleep()
and usleep_range() to avoid busy waiting.
This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
aty128_set_suspend() is never called in atomic context.
The call chains ending up at aty128_set_suspend() are:
[1] aty128_set_suspend() <- aty128_pci_suspend()
[2] aty128_set_suspend() <- aty128_do_resume() <- aty128_pci_resume()
[3] aty128_set_suspend() <- aty128_do_resume() <- aty128_early_resume()
aty128_pci_suspend() is set as ".suspend" in struct pci_driver.
aty128_pci_resume() is set as ".resume" in struct pci_driver.
aty128_early_resume() is never called.
These functions are not called in atomic context.
Despite never getting called from atomic context, aty128_set_suspend()
calls mdelay() to busily wait.
This is not necessary and can be replaced with msleep() to
avoid busy waiting.
This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
savage_init_hw() is never called in atomic context.
The call chains ending up at savage_init_hw() are:
[1] savage_init_hw() <- savagefb_probe()
[2] savage_init_hw() <- savagefb_resume()
savagefb_probe() is only set as ".probe" in struct pci_driver.
savagefb_resume) is only set as ".resume" in struct pci_driver.
These functions are not called in atomic context.
Despite never getting called from atomic context, savage_init_hw()
calls mdelay() to busily wait.
This is not necessary and can be replaced with usleep_range to
avoid busy waiting.
This is found by a static analysis tool named DCNS written by myself.
And I also manually check it.
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
Cc: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
After moving all nodes under "soc" node in commit 5d99cc59a3 ("ARM:
dts: exynos: Move Exynos5250 and Exynos5420 nodes under soc"), the i2c20
alias in Peach Pit and Peach Pi stopped pointing to proper node:
arch/arm/boot/dts/exynos5420-peach-pit.dtb: Warning (alias_paths):
/aliases:i2c20: aliases property is not a valid node (/spi@12d40000/cros-ec@0/i2c-tunnel)
arch/arm/boot/dts/exynos5800-peach-pi.dtb: Warning (alias_paths):
/aliases:i2c20: aliases property is not a valid node (/spi@12d40000/cros-ec@0/i2c-tunnel)
Fixes: 5d99cc59a3 ("ARM: dts: exynos: Move Exynos5250 and Exynos5420 nodes under soc")
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Use the correct functions to allow a name prefix assigned through
codec_conf to be taken into consideration whilst enabling and disabling
the preloader widget.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
According to the hardware spec, checking the INTEVENT bit isn't a
reliable way to detect if an OICR interrupt has occurred. This is
because this bit can be cleared by the hardware/firmware before the
interrupt service routine has run. So instead, just check for OICR
events every time.
Fixes: 940b61af02 ("ice: Initialize PF and setup miscellaneous interrupt")
Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
We can do a sleeping allocation from an irq context when CONFIG_NUMA
is enabled. Fix this by initializing the NUMA crng instances in a
workqueue.
Reported-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reported-by: syzbot+9de458f6a5e713ee8c1a@syzkaller.appspotmail.com
Fixes: 8ef35c866f ("random: set up the NUMA crng instances...")
Cc: stable@vger.kernel.org
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Action type 5 defines large action generic values. Fix comment to
reflect that better.
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
ice_sched_add_nodes_to_layer is used recursively, and so we start
with num_nodes_added being 0. This way, in case of an error or if
num_nodes is NULL, the function just returns 0 to indicate that no
nodes were added.
Fixes: 5513b920a4 ("ice: Update Tx scheduler tree for VSI multi-Tx queue support")
Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
When Qav mode is enabled, queue 0 should be kept on Stream Reservation
mode. From the i210 datasheet, section 8.12.19:
"Note: Queue0 QueueMode must be set to 1b when TransmitMode is set to
Qav." ("QueueMode 1b" represents the Stream Reservation mode)
The solution is to give queue 0 the all the credits it might need, so
it has priority over queue 1.
A situation where this can happen is when cbs is "installed" only on
queue 1, leaving queue 0 alone. For example:
$ tc qdisc replace dev enp2s0 handle 100: parent root mqprio num_tc 3 \
map 2 2 1 0 2 2 2 2 2 2 2 2 2 2 2 2 queues 1@0 1@1 2@2 hw 0
$ tc qdisc replace dev enp2s0 parent 100:2 cbs locredit -1470 \
hicredit 30 sendslope -980000 idleslope 20000 offload 1
Signed-off-by: Vinicius Costa Gomes <vinicius.gomes@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
There is a potential execution path in which variable err is
returned without being properly initialized previously.
Fix this by initializing variable err to 0.
Addresses-Coverity-ID: 1468362 ("Uninitialized scalar variable")
Fixes: f4ecfbfc32 ("drm/i915: Check whitelist registers across resets")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180424131545.GA4053@embeddedor.com
Currently it is possible to read and/or write to suspend EB's.
Writing /dev/mtdX or /dev/mtdblockX from several processes may
break the flash state machine.
Taken from cfi_cmdset_0001 driver.
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: Richard Weinberger <richard@nod.at>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Some Micron chips does not work well wrt Erase suspend for
boot blocks. This avoids the issue by not allowing Erase suspend
for the boot blocks for the 28F00AP30(1GBit) chip.
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: Richard Weinberger <richard@nod.at>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Currently it is possible to read and/or write to suspend EB's.
Writing /dev/mtdX or /dev/mtdblockX from several processes may
break the flash state machine.
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@infinera.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: Richard Weinberger <richard@nod.at>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
The property of the legacy mode for the eMMC PHY turned out to
be wrong. Some eMMC devices are unstable due to the set-up/hold
timing violation. Correct the delay value.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Currently in ext4_valid_block_bitmap() we expect the bitmap to be
positioned anywhere between 0 and s_blocksize clusters, but that's
wrong because the bitmap can be placed anywhere in the block group. This
causes false positives when validating bitmaps on perfectly valid file
system layouts. Fix it by checking whether the bitmap is within the group
boundary.
The problem can be reproduced using the following
mkfs -t ext3 -E stride=256 /dev/vdb1
mount /dev/vdb1 /mnt/test
cd /mnt/test
wget https://cdn.kernel.org/pub/linux/kernel/v4.x/linux-4.16.3.tar.xz
tar xf linux-4.16.3.tar.xz
This will result in the warnings in the logs
EXT4-fs error (device vdb1): ext4_validate_block_bitmap:399: comm tar: bg 84: block 2774529: invalid block bitmap
[ Changed slightly for clarity and to not drop a overflow test -- TYT ]
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reported-by: Ilya Dryomov <idryomov@gmail.com>
Fixes: 7dac4a1726 ("ext4: add validity checks for bitmap block numbers")
Cc: stable@vger.kernel.org
Add syscon-phy-mode property specifying a phandle of system controller
to each ethernet node.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
The GIO clock/reset, Another MAC clock, and the PHY clock are required
for the ethernet of Pro4 SoC.
And add clock-names and reset-names to the ethernet node of PXs2 since
we need to distinguish clocks and resets now.
Suggested-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Add syscon-phy-mode property specifying a phandle of system controller
to each ethernet node.
In addition, LD11 SoC has a built-in ethernet PHY. When we set "internal"
to phy-mode property, this built-in PHY is available.
This patch changes phy-mode property for LD11 to "internal", as default.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Add clock-names and reset-names because this node recognizes multiple
clocks and resets. ("ether", and so on, for each)
Suggested-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
The current error handling for failed resource setup for xdp_ring
data is a break out of the loop and returning 0 indicated everything
was OK, when in fact it is not. Fix this by exiting via the
error exit label err_setup_tx that will clean up the resources
correctly and return and error status.
Detected by CoverityScan, CID#1466879 ("Logically dead code")
Fixes: 21092e9ce8 ("ixgbevf: Add support for XDP_TX action")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Put WARN_ON() and bail out if the given index is over the allocated
array of the given SPDIF controls. It's merely a sanity check to
catch any potential issues (if any).
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In early stage of firmware SDK, DICE seems to lose its backward
compatibility due to some registers on global address section. I found
this with Alesis Multimix 12 FireWire with ancient firmware (approx.
shipped version).
According to retrieved log from the unit, global section has 96 byte
space. On the other hand, current version of ALSA dice driver assumes
that all of supported unit has at least 100 byte space.
$ ./firewire-request /dev/fw1 read 0xffffe0000000 28
result: 000: 00 00 00 0a 00 00 00 18 00 00 00 22 00 00 00 8a
result: 010: 00 00 00 ac 00 00 01 12 00 00 00 00 00 00 00 00
result: 020: 00 00 00 00 00 00 00 00
This commit adds support for the ancient firmware. Check of global section
is loosened to accept the smaller space. The lack of information is
already compensated by hard-coded parameters.
I experienced that the latest version of Windows driver for this model
can't handle this unit, too. This means that TCAT releases firmware SDK
without backward compatibility for the ancient firmware.
Below list is a early history of driver/firmware package released by
Alesis. I investigated on wayback machine on Internet Archive:
* Unknown: PAL v1.0.41.2, firmware v1.0.3
* Mar 2006: PAL v1.54.0, firmware v1.0.4
* Dec 2006: PAL v2.0.0.2, firmware v2.0
* Jun 2007: PAL v3.0.41.5, firmware v2.0
* Jul 2007: PAL v3.0.56.2. firmware v2.0
* Jan 2008: PAL v3.0.81.1080, firmware v2.0
If I can assume that firmware version is the same as DICE version, DICE
version for the issued firmware may be v1.0.3. According to code base of
userspace driver project (FFADO), I can read DICE v1.0.4 supports global
space larger than 100 byte. I guess the smaller space of global section is
a feature of DICE v1.0.3.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Current implementation of auditing by executable name only implements
the 'equal' operator. This patch extends it to also support the 'not
equal' operator.
See: https://github.com/linux-audit/audit-kernel/issues/53
Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com>
Reviewed-by: Richard Guy Briggs <rgb@redhat.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
This reverts commit <c6849a3ac1> ("net: init sk_cookie for inet socket")
Per discussion with Eric, when update sock_net(sk)->cookie_gen, the
whole cache cache line will be invalidated, as this cache line is shared
with all cpus, that may cause great performace hit.
Bellow is the data form Eric.
"Performance is reduced from ~5 Mpps to ~3.8 Mpps with 16 RX queues on
my host" when running synflood test.
Have to revert it to prevent from cache line false sharing.
Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The HPS EMAC0 drive strength is changed to 4mA because the initial 8mA
drive strength has caused CE test to fail. This requires changes on the
pad skew for EMAC0 PHY driver. Based on several measurements done, Tx
clock does not require the extra 0.96ns delay.
Signed-off-by: Ooi, Joyce <joyce.ooi@intel.com>
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Currently the 32-bit device address only is supported for DMA. However,
starting from Intel Sunrisepoint PCH the DMA address of the device FIFO
can be 64-bit.
Change the respective variable to be compatible with DMA engine
expectations, i.e. to phys_addr_t.
Fixes: 34cadd9c1b ("spi: pxa2xx: Add support for Intel Sunrisepoint")
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org
Using 'at' or 'at24' as the <manufacturer> part of the compatible
string is now deprecated. Use a correct value: 'atmel,<model>'.
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Temporarily disable AES-GCM, as AES-CCM is only currently
enabled mechanism on client side. This fixes SMB3.11
encrypted mounts to Windows.
Also the tree connect request itself should be encrypted if
requested encryption ("seal" on mount), in addition we should be
enabling encryption in 3.11 based on whether we got any valid
encryption ciphers back in negprot (the corresponding session flag is
not set as it is in 3.0 and 3.02)
Signed-off-by: Steve French <smfrench@gmail.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
CC: Stable <stable@vger.kernel.org>
Dan Carpenter had pointed this out a while ago, but the code around
this had changed so wasn't causing any problems since that field
was not used in this error path.
Still, it is cleaner to always initialize this field, so changing
the error path to set it.
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
CC: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steve French <smfrench@gmail.com>
The datasheet has been expanded with more registers and the DT files
have been updated with the new size. This change updates the example so
writing new DT files can use the enhanced driver which uses the new
registers.
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
[robh: s/have/has/, s/enchanted/enhanced/]
Signed-off-by: Rob Herring <robh@kernel.org>
Tal Gilboa says:
====================
Introduce adaptive TX interrupt moderation to net DIM
Net DIM is a library designed for dynamic interrupt moderation. It was
implemented and optimized with receive side interrupts in mind, since these
are usually the CPU expensive ones. This patch-set introduces adaptive transmit
interrupt moderation to net DIM, complete with a usage in the mlx5e driver.
Using adaptive TX behavior would reduce interrupt rate for multiple scenarios.
Furthermore, it is essential for increasing bandwidth on cases where payload
aggregation is required.
v3: Remove "inline" from functions in .c files (requested by DaveM). Revert
adding "enabled" field from struct net_dim and applied mlx5e structural
suggestions (suggested by SaeedM).
v2: Rebase over proper tree.
v1: Fix compilation issues due to missed function renaming.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for adaptive TX moderation. This greatly reduces TX interrupt
rate and increases bandwidth, mostly for TCP bandwidth over ARM
architecture (below). There is a slight single stream TCP with very large
message sizes degradation (x86). In this case if there's any moderation on
transmitted packets the bandwidth would reduce due to hitting TCP output limit.
Since this is a synthetic case, this is still worth doing.
Performance improvement (ConnectX-4Lx 40GbE, ARM)
TCP 64B bandwidth with 1-50 streams increased 6-35%.
TCP 64B bandwidth with 100-500 streams increased 20-70%.
Performance improvement (ConnectX-5 100GbE, x86)
Bandwidth: increased up to 40% (1024B with 10s of streams).
Interrupt rate: reduced up to 50% (1024B with 1000s of streams).
Performance degradation (ConnectX-5 100GbE, x86)
Bandwidth: up to 10% decrease single stream TCP (1MB message size from
51Gb/s to 47Gb/s).
Signed-off-by: Tal Gilboa <talgi@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Acked-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Interrupt moderation for TX traffic requires different profiles than RX
interrupt moderation. The main goal here is to reduce interrupt rate and
allow better payload aggregation by keeping SKBs in the TX queue a bit
longer. Ping-pong behavior would get a profile with a short timer, so
latency wouldn't increase for these scenarios. There might be a slight
degradation in bandwidth for single stream with large message sizes, since
net.ipv4.tcp_limit_output_bytes is limiting the allowed TX traffic, but
with many streams performance is always improved.
Signed-off-by: Tal Gilboa <talgi@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>