This repairs problem with compile library in userspace (libnl).
Signed-off-by: Jiří Župka <jzupka@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Allows ss command (iproute2) to display "ecnseen" if at least one packet
with ECT(0) or ECT(1) or ECN was received by this socket.
"ecn" means ECN was negotiated at session establishment (TCP level)
"ecnseen" means we received at least one packet with ECT fields set (IP
level)
ss -i
...
ESTAB 0 0 192.168.20.110:22 192.168.20.144:38016
ino:5950 sk:f178e400
mem:(r0,w0,f0,t0) ts sack ecn ecnseen bic wscale:7,8 rto:210
rtt:12.5/7.5 cwnd:10 send 9.3Mbps rcv_space:14480
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The following commit removed some including headers:
"net: sh_eth: move the asm/sh_eth.h to include/linux/"
(commit id: d4fa0e35fd)
Then, the build failure happened on the linux-next:
drivers/net/ethernet/renesas/sh_eth.c:601: error: 'THIS_MODULE' undeclared here (not in a function)
drivers/net/ethernet/renesas/sh_eth.c:1970: error: expected declaration specifiers or '...' before string constant
drivers/net/ethernet/renesas/sh_eth.c:1970: warning: data definition has no type or storage class
drivers/net/ethernet/renesas/sh_eth.c:1970: warning: type defaults to 'int' in declaration of 'MODULE_AUTHOR'
drivers/net/ethernet/renesas/sh_eth.c:1970: warning: function declaration isn't a prototype
drivers/net/ethernet/renesas/sh_eth.c:1971: error: expected declaration specifiers or '...' before string constant
drivers/net/ethernet/renesas/sh_eth.c:1971: warning: data definition has no type or storage class
drivers/net/ethernet/renesas/sh_eth.c:1971: warning: type defaults to 'int' in declaration of 'MODULE_DESCRIPTION'
drivers/net/ethernet/renesas/sh_eth.c:1971: warning: function declaration isn't a prototype
drivers/net/ethernet/renesas/sh_eth.c:1972: error: expected declaration specifiers or '...' before string constant
drivers/net/ethernet/renesas/sh_eth.c:1972: warning: data definition has no type or storage class
drivers/net/ethernet/renesas/sh_eth.c:1972: warning: type defaults to 'int' in declaration of 'MODULE_LICENSE'
drivers/net/ethernet/renesas/sh_eth.c:1972: warning: function declaration isn't a prototype
This patch fixes the issue. This patch also get back include/kernel.h
and linux/spinlock.h.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
During a test where a pair of bonding interfaces using ARP monitoring
were both brought up and torn down (with an rmmod) repeatedly, a panic
in the timer code was noticed. I tracked this down and determined that
any of the bonding functions that ran as workqueue handlers and requeued
more work might not properly exit when the module was removed.
There was a flag protected by the bond lock called kill_timers that is
set when the interface goes down or the module is removed, but many of
the functions that monitor link status now unlock the bond lock to take
rtnl first. There is a chance that another CPU running the rmmod could
get the lock and set kill_timers after the first check has passed.
This patch does not allow any function to queue work that will make
itself run unless kill_timers is not set. I also noticed while doing
this work that bond_resend_igmp_join_requests did not have a check for
kill_timers, so I added the needed call there as well.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net>
Reported-by: Liang Zheng <lzheng@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The allocation of "phyinfo" wasn't checked, and also the allocation
wasn't freed on error paths. Sjur Brændeland pointed out as well
that "phy_driver" should be freed on the error path too.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Sjur Brændeland <sjur.brandeland@stericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Apart from the obvious cleanup, this should make the line
cursor_end = x - xorigin + w;
correct now.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Fixes cursor disappearing prematurely when moving off a top/left edge which
is not located at the desktop top/left edge.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The mouse cursor hotspot calculation when the cursor is partially off the
top or left side of the screen was off by one.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=41158
Signed-off-by: Nicholas Miell <nmiell@gmail.com>
Reviewed-by: Michel Dänzer <michel@daenzer.net>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Only disable the pipe if the monitor is physically
disconnected. The previous logic also disabled the
pipe if the link was trained.
Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=41248
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
The previous code could potentially loop forever. Limit
the number of DP aux defer retries to 4 for native aux
transactions, same as i2c over aux transactions.
Noticed by: Brad Campbell <lists2009@fnarfbargle.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Brad Campbell <lists2009@fnarfbargle.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
An incorrect ordering in the error checking code lead
to DP aux defer being skipped in the aux native write
path. Move the bytes transferred check (ret == 0)
below the defer check.
Tracked down by: Brad Campbell <brad@fnarfbargle.com>
Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=41121
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Brad Campbell <brad@fnarfbargle.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
Need to allow application to update existing fdb entries that already
exist. This makes bridge netlink neighbor API have same flags and
semantics as ip neighbor table.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When port is added to a bridge, the old code would send the new neighbor
netlink message before the subsequent new link message. This bug makes
it difficult to use the monitoring API in an application.
This code changes the ordering to add the forwarding entry
after the port is setup. One of the error checks (for invalid address)
is moved earlier in the process to avoid having to do unwind.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Also changing it's frequency to once every 64s instead of existing 32s as
it was shown to affect performance
Signed-off-by: Somnath Kotur <somnath.kotur@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This was manifesting as a crash when FAT Dump extraction was attempted on a PPC machine.
Signed-off-by: Somnath Kotur <somnath.kotur@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This was showing up as junk value on PPC /Big endian machines since
it was marked as a byte.
Signed-off-by: Somnath Kotur <somnath.kotur@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add 60s delay before timeout on polling Bit 31 so that FAT dump can
complete when reset occurs.
Signed-off-by: Somnath Kotur <somnath.kotur@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Amir Vadai wrote:
> When a stream is paused, and its rule is expired while it is paused,
> no new rule will be configured to the HW when traffic resume.
[...]
> - When stream was resumed, traffic was steered again by RSS, and
> because current-cpu was equal to desired-cpu, ndo_rx_flow_steer
> wasn't called and no rule was configured to the HW.
Fix this by setting the flow's current CPU only in the table for the
newly selected RX queue.
Reported-and-tested-by: Amir Vadai <amirv@dev.mellanox.co.il>
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Removes some bitrot from the au1100fb driver and fix it up so it works again.
Signed-off-by: Manuel Lauss <manuel.lauss@googlemail.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
When blanking an already blanked framebuffer, a kernel NULL pointer
dereference occurred, because mx3fb driver handles all kinds of screen
blanking (normal, vsync suspend, powerdown) in the same way.
Certain programs (Xorg X11 server) first do a normal blank, followed by
a powerdown blank, which triggered the bug.
Add an additional safeguard and make sdc_disable_channel() safe against
multiple calls independent of other logic.
Signed-off-by: Michael Thalmeier <michael.thalmeier@hale.at>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
Since commit [e58aa3d2: genirq: Run irq handlers with interrupts disabled],
We run all interrupt handlers with interrupts disabled
and we even check and yell when an interrupt handler
returns with interrupts enabled (see commit [b738a50a:
genirq: Warn when handler enables interrupts]).
So now this flag is a NOOP and can be removed.
Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Acked-by: David Brown <davidb@codeaurora.org>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
Having "edid" as char caused a problem in ufx_read_edid() where we
compared "edid[i] != 0xFF". Because of the type difference, the
condition was never true and the error checking failed.
Also I added a __user notation to silence a sparse complaint.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
The Sigmatel/IDT parser should have the same naming convention
for input jacks as the other codecs have.
BugLink: http://bugs.launchpad.net/bugs/859704
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The struct ore_striping_info will be used later in other
structures. And ore_calc_stripe_info as well. Rename them
make struct ore_striping_info public. ore_calc_stripe_info
is still static, will be made public on first use.
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
The struct pnfs_osd_data_map data_map member of exofs_sb_info was
never used after mount. In fact all it's members were duplicated
by the ore_layout structure. So just remove the duplicated information.
Also removed some stupid, but perfectly supported, restrictions on
layout parameters. The case where num_devices is not divisible by
mirror_count+1 is perfectly fine since the rotating device view
will eventually use all the devices it can get.
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
Signed-off-by: Benny Halevy <bhalevy@tonian.com>
ore_components already has a comps member so this leads
to things like comps->comps which is annoying. the name oc
was already used in new code. So rename all old usage of
ore_components comps => ore_components oc.
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
This quiets the following sparse noise:
warning: symbol 'exofs_sync_fs' was not declared. Should it be static?
warning: symbol 'exofs_free_sbi' was not declared. Should it be static?
warning: symbol 'exofs_get_parent' was not declared. Should it be static?
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
This quiets the sparse noise:
warning: symbol '_calc_trunk_info' was not declared. Should it be static?
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
Any driver that selects SND_SOC_WM8994 should also make sure that
MFD_WM8994 is set, since the codec relies on the mfd code:
sound/built-in.o: In function `wm8994_read':
last.c:(.text+0x20160): undefined reference to `wm8994_reg_read'
sound/built-in.o: In function `wm8994_write':
last.c:(.text+0x20e68): undefined reference to `wm8994_reg_write'
This solves the problem by selecting the MFD driver directly
and adding extra 'depends on' statements to make sure that we
respect the dependencies of that driver.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Add zorder support on OMAP4, this feature allows deciding the visibility order
of the overlays based on the zorder value provided as an overlay info parameter
or a sysfs attribute of the overlay object.
Use the overlay cap OMAP_DSS_OVL_CAP_ZORDER to determine whether zorder is
supported for the overlay or not. Use dss feature FEAT_ALPHA_FREE_ZORDER
if the caps are not available.
Ensure that all overlays that are enabled and connected to the same manager
have different zorders. Swapping zorders of 2 enabled overlays currently
requires disabling one of the overlays.
Signed-off-by: Archit Taneja <archit@ti.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Add support for VIDEO3 pipeline on OMAP4:
- Add VIDEO3 pipeline information in dss_features and omapdss.h
- Add VIDEO3 pipeline register coefficients in dispc.h
- Create a new overlay structure corresponding to VIDEO3.
- Make changes in dispc.c for VIDEO3
Signed-off-by: Archit Taneja <archit@ti.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
On OMAP3, in order to enable alpha blending for LCD and TV managers, we needed
to set LCDALPHABLENDERENABLE/TVALPHABLENDERENABLE bits in DISPC_CONFIG. On
OMAP4, alpha blending is always enabled by default, if the above bits are set,
we switch to an OMAP3 compatibility mode where the zorder values in the pipeline
attribute registers are ignored and a fixed priority is configured.
Rename the manager_info member "alpha_enabled" to "partial_alpha_enabled" for
more clarity. Introduce two dss_features FEAT_ALPHA_FIXED_ZORDER and
FEAT_ALPHA_FREE_ZORDER which represent OMAP3-alpha compatibility mode and OMAP4
alpha mode respectively. Introduce an overlay cap for ZORDER. The DSS2 user is
expected to check for the ZORDER cap, if an overlay doesn't have this cap, the
user is expected to set the parameter partial_alpha_enabled. If the overlay has
ZORDER cap, the DSS2 user can assume that alpha blending is already enabled.
Don't support OMAP3 compatibility mode for now. Trying to read/write to
alpha_blending_enabled sysfs attribute issues a warning for OMAP4 and does not
set the LCDALPHABLENDERENABLE/TVALPHABLENDERENABLE bits.
Change alpha_enabled to partial_alpha_enabled in the omap_vout driver. Use
overlay cap "OMAP_DSS_OVL_CAP_GLOBAL_ALPHA" to check if overlay supports alpha
blending or not. Replace this with checks for VIDEO1 pipeline.
Cc: linux-media@vger.kernel.org
Cc: Lajos Molnar <molnar@ti.com>
Signed-off-by: Archit Taneja <archit@ti.com>
Acked-by: Vaibhav Hiremath <hvaibhav@ti.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Applications may want to read ELD information to
understand what codecs are supported on the HDMI
receiver and handle the a-v delay for better lip-sync.
ELD information is exposed in a device-specific
IFACE_PCM kcontrol. Tested both with amixer and
PulseAudio; with a corresponding patch passthrough modes
are enabled automagically.
ELD control size is set to zero in case of errors or
wrong configurations. No notifications are implemented
for now, it is expected that jack detection is used to
reconfigure the audio outputs.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
As request_percpu_irq() doesn't allow for a percpu interrupt to have
its type configured (it is generally impossible to configure it on all
CPUs at once), add a 'type' argument to enable_percpu_irq().
This allows some low-level, board specific init code to be switched to
a generic API.
[ tglx: Added WARN_ON argument ]
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The ARM GIC interrupt controller offers per CPU interrupts (PPIs),
which are usually used to connect local timers to each core. Each CPU
has its own private interface to the GIC, and only sees the PPIs that
are directly connect to it.
While these timers are separate devices and have a separate interrupt
line to a core, they all use the same IRQ number.
For these devices, request_irq() is not the right API as it assumes
that an IRQ number is visible by a number of CPUs (through the
affinity setting), but makes it very awkward to express that an IRQ
number can be handled by all CPUs, and yet be a different interrupt
line on each CPU, requiring a different dev_id cookie to be passed
back to the handler.
The *_percpu_irq() functions is designed to overcome these
limitations, by providing a per-cpu dev_id vector:
int request_percpu_irq(unsigned int irq, irq_handler_t handler,
const char *devname, void __percpu *percpu_dev_id);
void free_percpu_irq(unsigned int, void __percpu *);
int setup_percpu_irq(unsigned int irq, struct irqaction *new);
void remove_percpu_irq(unsigned int irq, struct irqaction *act);
void enable_percpu_irq(unsigned int irq);
void disable_percpu_irq(unsigned int irq);
The API has a number of limitations:
- no interrupt sharing
- no threading
- common handler across all the CPUs
Once the interrupt is requested using setup_percpu_irq() or
request_percpu_irq(), it must be enabled by each core that wishes its
local interrupt to be delivered.
Based on an initial patch by Thomas Gleixner.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lkml.kernel.org/r/1316793788-14500-2-git-send-email-marc.zyngier@arm.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
After checking the code and datasheet, I think what we want in the second
snd_soc_update_bits call is to update WM8741_DACRMSB_ATTENUATION register
instead of WM8741_DACRLSB_ATTENUATION.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
We have __exit annotation for txx9aclc_generic_remove(),
thus add __devexit_p to wrap it.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Acked-by: Mika Westerberg <mika.westerberg@iki.fi>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
It is exported via resume callback of struct dev_pm_ops rather than referenced
directly and so should be staticised.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
This assignment is done by the snd_soc_register_codec so there is no need
to redo it in probe function of a codec driver.
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
One thing puzzled me is that in JBOD case, the per-disk writeout
performance is smaller than the corresponding single-disk case even
when they have comparable bdi_thresh. Tracing shows find that in single
disk case, bdi_writeback is always kept high while in JBOD case, it
could drop low from time to time and correspondingly bdi_reclaimable
could sometimes rush high.
The fix is to watch bdi_reclaimable and kick background writeback as
soon as it goes high. This resembles the global background threshold
but in per-bdi manner. The trick is, as long as bdi_reclaimable does
not go high, bdi_writeback naturally won't go low because
bdi_reclaimable+bdi_writeback ~= bdi_thresh.
With less fluctuated writeback pages, JBOD performance is observed to
increase noticeably in various cases.
vmstat:nr_written values before/after patch:
3.1.0-rc4-wo-underrun+ 3.1.0-rc4-bgthresh3+
------------------------ ------------------------
125596480 +25.9% 158179363 JBOD-10HDD-16G/ext4-100dd-1M-24p-16384M-20:10-X
61790815 +110.4% 130032231 JBOD-10HDD-16G/ext4-10dd-1M-24p-16384M-20:10-X
58853546 -0.1% 58823828 JBOD-10HDD-16G/ext4-1dd-1M-24p-16384M-20:10-X
110159811 +24.7% 137355377 JBOD-10HDD-16G/xfs-100dd-1M-24p-16384M-20:10-X
69544762 +10.8% 77080047 JBOD-10HDD-16G/xfs-10dd-1M-24p-16384M-20:10-X
50644862 +0.5% 50890006 JBOD-10HDD-16G/xfs-1dd-1M-24p-16384M-20:10-X
42677090 +28.0% 54643527 JBOD-10HDD-thresh=100M/ext4-100dd-1M-24p-16384M-100M:10-X
47491324 +13.3% 53785605 JBOD-10HDD-thresh=100M/ext4-10dd-1M-24p-16384M-100M:10-X
52548986 +0.9% 53001031 JBOD-10HDD-thresh=100M/ext4-1dd-1M-24p-16384M-100M:10-X
26783091 +36.8% 36650248 JBOD-10HDD-thresh=100M/xfs-100dd-1M-24p-16384M-100M:10-X
35526347 +14.0% 40492312 JBOD-10HDD-thresh=100M/xfs-10dd-1M-24p-16384M-100M:10-X
44670723 -1.1% 44177606 JBOD-10HDD-thresh=100M/xfs-1dd-1M-24p-16384M-100M:10-X
127996037 +22.4% 156719990 JBOD-10HDD-thresh=2G/ext4-100dd-1M-24p-16384M-2048M:10-X
57518856 +3.8% 59677625 JBOD-10HDD-thresh=2G/ext4-10dd-1M-24p-16384M-2048M:10-X
51919909 +12.2% 58269894 JBOD-10HDD-thresh=2G/ext4-1dd-1M-24p-16384M-2048M:10-X
86410514 +79.0% 154660433 JBOD-10HDD-thresh=2G/xfs-100dd-1M-24p-16384M-2048M:10-X
40132519 +38.6% 55617893 JBOD-10HDD-thresh=2G/xfs-10dd-1M-24p-16384M-2048M:10-X
48423248 +7.5% 52042927 JBOD-10HDD-thresh=2G/xfs-1dd-1M-24p-16384M-2048M:10-X
206041046 +44.1% 296846536 JBOD-10HDD-thresh=4G/xfs-100dd-1M-24p-16384M-4096M:10-X
72312903 -19.4% 58272885 JBOD-10HDD-thresh=4G/xfs-10dd-1M-24p-16384M-4096M:10-X
50635672 -0.5% 50384787 JBOD-10HDD-thresh=4G/xfs-1dd-1M-24p-16384M-4096M:10-X
68308534 +115.7% 147324758 JBOD-10HDD-thresh=800M/ext4-100dd-1M-24p-16384M-800M:10-X
57882933 +14.5% 66269621 JBOD-10HDD-thresh=800M/ext4-10dd-1M-24p-16384M-800M:10-X
52183472 +12.8% 58855181 JBOD-10HDD-thresh=800M/ext4-1dd-1M-24p-16384M-800M:10-X
53788956 +94.2% 104460352 JBOD-10HDD-thresh=800M/xfs-100dd-1M-24p-16384M-800M:10-X
44493342 +35.5% 60298210 JBOD-10HDD-thresh=800M/xfs-10dd-1M-24p-16384M-800M:10-X
42641209 +18.9% 50681038 JBOD-10HDD-thresh=800M/xfs-1dd-1M-24p-16384M-800M:10-X
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Keep a minimal pool of dirty pages for each bdi, so that the disk IO
queues won't underrun. Also gently increase a small bdi_thresh to avoid
it stuck in 0 for some light dirtied bdi.
It's particularly useful for JBOD and small memory system.
It may result in (pos_ratio > 1) at the setpoint and push the dirty
pages high. This is more or less intended because the bdi is in the
danger of IO queue underflow.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
The dirty pause time shall ultimately be controlled by adjusting
nr_dirtied_pause, since there is relationship
pause = pages_dirtied / task_ratelimit
Assuming
pages_dirtied ~= nr_dirtied_pause
task_ratelimit ~= dirty_ratelimit
We get
nr_dirtied_pause ~= dirty_ratelimit * desired_pause
Here dirty_ratelimit is preferred over task_ratelimit because it's
more stable.
It's also important to limit possible large transitional errors:
- bw is changing quickly
- pages_dirtied << nr_dirtied_pause on entering dirty exceeded area
- pages_dirtied >> nr_dirtied_pause on btrfs (to be improved by a
separate fix, but still expect non-trivial errors)
So we end up using the above formula inside clamp_val().
The best test case for this code is to run 100 "dd bs=4M" tasks on
btrfs and check its pause time distribution.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Apply two policies to scale down the max pause time for
1) small number of concurrent dirtiers
2) small memory system (comparing to storage bandwidth)
MAX_PAUSE=200ms may only be suitable for high end servers with lots of
concurrent dirtiers, where the large pause time can reduce much overheads.
Otherwise, smaller pause time is desirable whenever possible, so as to
get good responsiveness and smooth user experiences. It's actually
required for good disk utilization in the case when all the dirty pages
can be synced to disk within MAX_PAUSE=200ms.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
As proposed by Chris, Dave and Jan, don't start foreground writeback IO
inside balance_dirty_pages(). Instead, simply let it idle sleep for some
time to throttle the dirtying task. In the mean while, kick off the
per-bdi flusher thread to do background writeback IO.
RATIONALS
=========
- disk seeks on concurrent writeback of multiple inodes (Dave Chinner)
If every thread doing writes and being throttled start foreground
writeback, it leads to N IO submitters from at least N different
inodes at the same time, end up with N different sets of IO being
issued with potentially zero locality to each other, resulting in
much lower elevator sort/merge efficiency and hence we seek the disk
all over the place to service the different sets of IO.
OTOH, if there is only one submission thread, it doesn't jump between
inodes in the same way when congestion clears - it keeps writing to
the same inode, resulting in large related chunks of sequential IOs
being issued to the disk. This is more efficient than the above
foreground writeback because the elevator works better and the disk
seeks less.
- lock contention and cache bouncing on concurrent IO submitters (Dave Chinner)
With this patchset, the fs_mark benchmark on a 12-drive software RAID0 goes
from CPU bound to IO bound, freeing "3-4 CPUs worth of spinlock contention".
* "CPU usage has dropped by ~55%", "it certainly appears that most of
the CPU time saving comes from the removal of contention on the
inode_wb_list_lock" (IMHO at least 10% comes from the reduction of
cacheline bouncing, because the new code is able to call much less
frequently into balance_dirty_pages() and hence access the global
page states)
* the user space "App overhead" is reduced by 20%, by avoiding the
cacheline pollution by the complex writeback code path
* "for a ~5% throughput reduction", "the number of write IOs have
dropped by ~25%", and the elapsed time reduced from 41:42.17 to
40:53.23.
* On a simple test of 100 dd, it reduces the CPU %system time from 30% to 3%,
and improves IO throughput from 38MB/s to 42MB/s.
- IO size too small for fast arrays and too large for slow USB sticks
The write_chunk used by current balance_dirty_pages() cannot be
directly set to some large value (eg. 128MB) for better IO efficiency.
Because it could lead to more than 1 second user perceivable stalls.
Even the current 4MB write size may be too large for slow USB sticks.
The fact that balance_dirty_pages() starts IO on itself couples the
IO size to wait time, which makes it hard to do suitable IO size while
keeping the wait time under control.
Now it's possible to increase writeback chunk size proportional to the
disk bandwidth. In a simple test of 50 dd's on XFS, 1-HDD, 3GB ram,
the larger writeback size dramatically reduces the seek count to 1/10
(far beyond my expectation) and improves the write throughput by 24%.
- long block time in balance_dirty_pages() hurts desktop responsiveness
Many of us may have the experience: it often takes a couple of seconds
or even long time to stop a heavy writing dd/cp/tar command with
Ctrl-C or "kill -9".
- IO pipeline broken by bumpy write() progress
There are a broad class of "loop {read(buf); write(buf);}" applications
whose read() pipeline will be under-utilized or even come to a stop if
the write()s have long latencies _or_ don't progress in a constant rate.
The current threshold based throttling inherently transfers the large
low level IO completion fluctuations to bumpy application write()s,
and further deteriorates with increasing number of dirtiers and/or bdi's.
For example, when doing 50 dd's + 1 remote rsync to an XFS partition,
the rsync progresses very bumpy in legacy kernel, and throughput is
improved by 67% by this patchset. (plus the larger write chunk size,
it will be 93% speedup).
The new rate based throttling can support 1000+ dd's with excellent
smoothness, low latency and low overheads.
For the above reasons, it's much better to do IO-less and low latency
pauses in balance_dirty_pages().
Jan Kara, Dave Chinner and me explored the scheme to let
balance_dirty_pages() wait for enough writeback IO completions to
safeguard the dirty limit. However it's found to have two problems:
- in large NUMA systems, the per-cpu counters may have big accounting
errors, leading to big throttle wait time and jitters.
- NFS may kill large amount of unstable pages with one single COMMIT.
Because NFS server serves COMMIT with expensive fsync() IOs, it is
desirable to delay and reduce the number of COMMITs. So it's not
likely to optimize away such kind of bursty IO completions, and the
resulted large (and tiny) stall times in IO completion based throttling.
So here is a pause time oriented approach, which tries to control the
pause time in each balance_dirty_pages() invocations, by controlling
the number of pages dirtied before calling balance_dirty_pages(), for
smooth and efficient dirty throttling:
- avoid useless (eg. zero pause time) balance_dirty_pages() calls
- avoid too small pause time (less than 4ms, which burns CPU power)
- avoid too large pause time (more than 200ms, which hurts responsiveness)
- avoid big fluctuations of pause times
It can control pause times at will. The default policy (in a followup
patch) will be to do ~10ms pauses in 1-dd case, and increase to ~100ms
in 1000-dd case.
BEHAVIOR CHANGE
===============
(1) dirty threshold
Users will notice that the applications will get throttled once crossing
the global (background + dirty)/2=15% threshold, and then balanced around
17.5%. Before patch, the behavior is to just throttle it at 20% dirtyable
memory in 1-dd case.
Since the task will be soft throttled earlier than before, it may be
perceived by end users as performance "slow down" if his application
happens to dirty more than 15% dirtyable memory.
(2) smoothness/responsiveness
Users will notice a more responsive system during heavy writeback.
"killall dd" will take effect instantly.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Add two fields to task_struct.
1) account dirtied pages in the individual tasks, for accuracy
2) per-task balance_dirty_pages() call intervals, for flexibility
The balance_dirty_pages() call interval (ie. nr_dirtied_pause) will
scale near-sqrt to the safety gap between dirty pages and threshold.
The main problem of per-task nr_dirtied is, if 1k+ tasks start dirtying
pages at exactly the same time, each task will be assigned a large
initial nr_dirtied_pause, so that the dirty threshold will be exceeded
long before each task reached its nr_dirtied_pause and hence call
balance_dirty_pages().
The solution is to watch for the number of pages dirtied on each CPU in
between the calls into balance_dirty_pages(). If it exceeds ratelimit_pages
(3% dirty threshold), force call balance_dirty_pages() for a chance to
set bdi->dirty_exceeded. In normal situations, this safeguarding
condition is not expected to trigger at all.
On the sqrt in dirty_poll_interval():
It will serve as an initial guess when dirty pages are still in the
freerun area.
When dirty pages are floating inside the dirty control scope [freerun,
limit], a followup patch will use some refined dirty poll interval to
get the desired pause time.
thresh-dirty (MB) sqrt
1 16
2 22
4 32
8 45
16 64
32 90
64 128
128 181
256 256
512 362
1024 512
The above table means, given 1MB (or 1GB) gap and the dd tasks polling
balance_dirty_pages() on every 16 (or 512) pages, the dirty limit won't
be exceeded as long as there are less than 16 (or 512) concurrent dd's.
So sqrt naturally leads to less overheads and more safe concurrent tasks
for large memory servers, which have large (thresh-freerun) gaps.
peter: keep the per-CPU ratelimit for safeguarding the 1k+ tasks case
CC: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reviewed-by: Andrea Righi <andrea@betterlinux.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
There are some imperfections in balanced_dirty_ratelimit.
1) large fluctuations
The dirty_rate used for computing balanced_dirty_ratelimit is merely
averaged in the past 200ms (very small comparing to the 3s estimation
period for write_bw), which makes rather dispersed distribution of
balanced_dirty_ratelimit.
It's pretty hard to average out the singular points by increasing the
estimation period. Considering that the averaging technique will
introduce very undesirable time lags, I give it up totally. (btw, the 3s
write_bw averaging time lag is much more acceptable because its impact
is one-way and therefore won't lead to oscillations.)
The more practical way is filtering -- most singular
balanced_dirty_ratelimit points can be filtered out by remembering some
prev_balanced_rate and prev_prev_balanced_rate. However the more
reliable way is to guard balanced_dirty_ratelimit with task_ratelimit.
2) due to truncates and fs redirties, the (write_bw <=> dirty_rate)
match could become unbalanced, which may lead to large systematical
errors in balanced_dirty_ratelimit. The truncates, due to its possibly
bumpy nature, can hardly be compensated smoothly. So let's face it. When
some over-estimated balanced_dirty_ratelimit brings dirty_ratelimit
high, dirty pages will go higher than the setpoint. task_ratelimit will
in turn become lower than dirty_ratelimit. So if we consider both
balanced_dirty_ratelimit and task_ratelimit and update dirty_ratelimit
only when they are on the same side of dirty_ratelimit, the systematical
errors in balanced_dirty_ratelimit won't be able to bring
dirty_ratelimit far away.
The balanced_dirty_ratelimit estimation may also be inaccurate near
@limit or @freerun, however is less an issue.
3) since we ultimately want to
- keep the fluctuations of task ratelimit as small as possible
- keep the dirty pages around the setpoint as long time as possible
the update policy used for (2) also serves the above goals nicely:
if for some reason the dirty pages are high (task_ratelimit < dirty_ratelimit),
and dirty_ratelimit is low (dirty_ratelimit < balanced_dirty_ratelimit),
there is no point to bring up dirty_ratelimit in a hurry only to hurt
both the above two goals.
So, we make use of task_ratelimit to limit the update of dirty_ratelimit
in two ways:
1) avoid changing dirty rate when it's against the position control target
(the adjusted rate will slow down the progress of dirty pages going
back to setpoint).
2) limit the step size. task_ratelimit is changing values step by step,
leaving a consistent trace comparing to the randomly jumping
balanced_dirty_ratelimit. task_ratelimit also has the nice smaller
errors in stable state and typically larger errors when there are big
errors in rate. So it's a pretty good limiting factor for the step
size of dirty_ratelimit.
Note that bdi->dirty_ratelimit is always tracking balanced_dirty_ratelimit.
task_ratelimit is merely used as a limiting factor.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
It's all about bdi->dirty_ratelimit, which aims to be (write_bw / N)
when there are N dd tasks.
On write() syscall, use bdi->dirty_ratelimit
============================================
balance_dirty_pages(pages_dirtied)
{
task_ratelimit = bdi->dirty_ratelimit * bdi_position_ratio();
pause = pages_dirtied / task_ratelimit;
sleep(pause);
}
On every 200ms, update bdi->dirty_ratelimit
===========================================
bdi_update_dirty_ratelimit()
{
task_ratelimit = bdi->dirty_ratelimit * bdi_position_ratio();
balanced_dirty_ratelimit = task_ratelimit * write_bw / dirty_rate;
bdi->dirty_ratelimit = balanced_dirty_ratelimit
}
Estimation of balanced bdi->dirty_ratelimit
===========================================
balanced task_ratelimit
-----------------------
balance_dirty_pages() needs to throttle tasks dirtying pages such that
the total amount of dirty pages stays below the specified dirty limit in
order to avoid memory deadlocks. Furthermore we desire fairness in that
tasks get throttled proportionally to the amount of pages they dirty.
IOW we want to throttle tasks such that we match the dirty rate to the
writeout bandwidth, this yields a stable amount of dirty pages:
dirty_rate == write_bw (1)
The fairness requirement gives us:
task_ratelimit = balanced_dirty_ratelimit
== write_bw / N (2)
where N is the number of dd tasks. We don't know N beforehand, but
still can estimate balanced_dirty_ratelimit within 200ms.
Start by throttling each dd task at rate
task_ratelimit = task_ratelimit_0 (3)
(any non-zero initial value is OK)
After 200ms, we measured
dirty_rate = # of pages dirtied by all dd's / 200ms
write_bw = # of pages written to the disk / 200ms
For the aggressive dd dirtiers, the equality holds
dirty_rate == N * task_rate
== N * task_ratelimit_0 (4)
Or
task_ratelimit_0 == dirty_rate / N (5)
Now we conclude that the balanced task ratelimit can be estimated by
write_bw
balanced_dirty_ratelimit = task_ratelimit_0 * ---------- (6)
dirty_rate
Because with (4) and (5) we can get the desired equality (1):
write_bw
balanced_dirty_ratelimit == (dirty_rate / N) * ----------
dirty_rate
== write_bw / N
Then using the balanced task ratelimit we can compute task pause times like:
task_pause = task->nr_dirtied / task_ratelimit
task_ratelimit with position control
------------------------------------
However, while the above gives us means of matching the dirty rate to
the writeout bandwidth, it at best provides us with a stable dirty page
count (assuming a static system). In order to control the dirty page
count such that it is high enough to provide performance, but does not
exceed the specified limit we need another control.
The dirty position control works by extending (2) to
task_ratelimit = balanced_dirty_ratelimit * pos_ratio (7)
where pos_ratio is a negative feedback function that subjects to
1) f(setpoint) = 1.0
2) df/dx < 0
That is, if the dirty pages are ABOVE the setpoint, we throttle each
task a bit more HEAVY than balanced_dirty_ratelimit, so that the dirty
pages are created less fast than they are cleaned, thus DROP to the
setpoints (and the reverse).
Based on (7) and the assumption that both dirty_ratelimit and pos_ratio
remains CONSTANT for the past 200ms, we get
task_ratelimit_0 = balanced_dirty_ratelimit * pos_ratio (8)
Putting (8) into (6), we get the formula used in
bdi_update_dirty_ratelimit():
write_bw
balanced_dirty_ratelimit *= pos_ratio * ---------- (9)
dirty_rate
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
