Commit graph

982,081 commits

Author SHA1 Message Date
Vandita Kulkarni
7b864f9588 drm/i915/tgl/dsi: Do not override TA_SURE
Do not override TA_SURE timing parameter to
zero for DSI 8X frequency 800MHz or below on
TGL.

Signed-off-by: Vandita Kulkarni <vandita.kulkarni@intel.com>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Signed-off-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190730073648.5157-4-vandita.kulkarni@intel.com
2019-08-08 18:37:50 +05:30
Vandita Kulkarni
960e9836f7 drm/i915/tgl/dsi: Set latency PCS_DW1 for tgl
Latency programming remains same as that of ICL and
setting latency otimization for PCS_DW1 lanes is same as
that of EHL, hence extending it to TGL.

Signed-off-by: Vandita Kulkarni <vandita.kulkarni@intel.com>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Signed-off-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190730073648.5157-3-vandita.kulkarni@intel.com
2019-08-08 18:37:50 +05:30
Vandita Kulkarni
3522a33a27 drm/i915/tgl/dsi: Program TRANS_VBLANK register
Program vblank register for mipi dsi in video mode
on TGL.

Signed-off-by: Vandita Kulkarni <vandita.kulkarni@intel.com>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
Signed-off-by: Uma Shankar <uma.shankar@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190730073648.5157-2-vandita.kulkarni@intel.com
2019-08-08 18:37:49 +05:30
Vasily Gorbik
739bacbf7a s390/build: use size command to perform empty .bss check
Currently empty .bss checks performed do not pay attention to "common
objects" in object files which end up in .bss section eventually.

The "size" tool is a part of binutils and since version 2.18 provides
"--common" command line option, which allows to account "common objects"
sizes in .bss section size. Utilize "size --common" to perform accurate
check that .bss section is unused. Besides that the size tool handles
object files without .bss section gracefully and doesn't require
additional objdump run.

The linux kernel requires binutils 2.20 since 4.13.

Kbuild exports OBJSIZE to reference the right size tool.

Link: http://lkml.kernel.org/r/patch-2.thread-2257a1.git-2257a1c53d4a.your-ad-here.call-01565088755-ext-5120@work.hours
Reported-and-tested-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
2019-08-08 14:36:53 +02:00
Vasily Gorbik
7bac98707f kbuild: add OBJSIZE variable for the size tool
Define and export OBJSIZE variable for "size" tool from binutils to be
used in architecture specific Makefiles (naming the variable just "SIZE"
would be too risky). In particular this tool is useful to perform checks
that early boot code is not using bss section (which might have not been
zeroed yet or intersects with initrd or other files boot loader might
have put right after the linux kernel).

Link: http://lkml.kernel.org/r/patch-1.thread-2257a1.git-188f5a3d81d5.your-ad-here.call-01565088755-ext-5120@work.hours
Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
2019-08-08 14:36:51 +02:00
Will Deacon
2af2e72b18 iommu/arm-smmu-v3: Defer TLB invalidation until ->iotlb_sync()
Update the iommu_iotlb_gather structure passed to ->tlb_add_page() and
use this information to defer all TLB invalidation until ->iotlb_sync().
This drastically reduces contention on the command queue, since we can
insert our commands in batches rather than one-by-one.

Tested-by: Ganapatrao Kulkarni  <gkulkarni@marvell.com>
Signed-off-by: Will Deacon <will@kernel.org>
2019-08-08 13:31:54 +01:00
Will Deacon
587e6c10a7 iommu/arm-smmu-v3: Reduce contention during command-queue insertion
The SMMU command queue is a bottleneck in large systems, thanks to the
spin_lock which serialises accesses from all CPUs to the single queue
supported by the hardware.

Attempt to improve this situation by moving to a new algorithm for
inserting commands into the queue, which is lock-free on the fast-path.

Tested-by: Ganapatrao Kulkarni  <gkulkarni@marvell.com>
Signed-off-by: Will Deacon <will@kernel.org>
2019-08-08 13:31:54 +01:00
Nishka Dasgupta
057b05d5ac dmaengine: qcom: hidma_mgmt: Add of_node_put() before goto
Each iteration of for_each_available_child_of_node puts the previous
node, but in the case of a goto from the middle of the loop, there is
no put, thus causing a memory leak.
Hence add an of_node_put under the label that the gotos point to.
In order to avoid decrementing an already-decremented refcount, copy the
original contents of the label (including the return statement) to just
above the label, so that the code under the label is executed only when
a goto exit from the loop occurs.
Additionally, remove an unnecessary get/put pair from the loop, as the
loop itself already keeps track of refcount.
Issue found with Coccinelle.

Signed-off-by: Nishka Dasgupta <nishkadg.linux@gmail.com>
Acked-by: Sinan Kaya <okaya@kernel.org>
Link: https://lore.kernel.org/r/20190724081609.9724-1-nishkadg.linux@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
2019-08-08 17:56:50 +05:30
Robin Gong
232a7f18cf dmaengine: fsl-edma: add i.mx7ulp edma2 version support
Add edma2 for i.mx7ulp by version v3, since v2 has already
been used by mcf-edma.
The big changes based on v1 are belows:
1. only one dmamux.
2. another clock dma_clk except dmamux clk.
3. 16 independent interrupts instead of only one interrupt for
all channels.

Signed-off-by: Robin Gong <yibin.gong@nxp.com>
Link: https://lore.kernel.org/r/1563952834-7731-1-git-send-email-yibin.gong@nxp.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
2019-08-08 17:54:36 +05:30
Fuqian Huang
719e25dba4 dmaengine: qcom_hidma: Remove call to memset after dmam_alloc_coherent
In commit 518a2f1925
("dma-mapping: zero memory returned from dma_alloc_*"),
dma_alloc_coherent has already zeroed the memory.
So memset is not needed.

Signed-off-by: Fuqian Huang <huangfq.daxian@gmail.com>
Acked-by: Sinan Kaya <okaya@kernel.org>
Link: https://lore.kernel.org/r/20190715031723.6375-1-huangfq.daxian@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
2019-08-08 17:51:56 +05:30
Fuqian Huang
9603a7ab60 dmaengine: imx-sdma: Remove call to memset after dma_alloc_coherent
In commit 518a2f1925
("dma-mapping: zero memory returned from dma_alloc_*"),
dma_alloc_coherent has already zeroed the memory.
So memset is not needed.

Signed-off-by: Fuqian Huang <huangfq.daxian@gmail.com>
Link: https://lore.kernel.org/r/20190715031716.6328-1-huangfq.daxian@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
2019-08-08 17:51:43 +05:30
Hans de Goede
6cf9481b44 pwm: Fallback to the static lookup-list when acpi_pwm_get fails
Commit 4a6ef8e37c ("pwm: Add support referencing PWMs from ACPI")
made pwm_get unconditionally return the acpi_pwm_get return value if
the device passed to pwm_get has an ACPI fwnode.

But even if the passed in device has an ACPI fwnode, it does not
necessarily have the necessary ACPI package defining its pwm bindings,
especially since the binding / API of this ACPI package has only been
introduced very recently.

Up until now X86/ACPI devices which use a separate pwm controller for
controlling their LCD screen's backlight brightness have been relying
on the static lookup-list to get their pwm.

pwm_get unconditionally returning the acpi_pwm_get return value breaks
this, breaking backlight control on these devices.

This commit fixes this by making pwm_get fall back to the static
lookup-list if acpi_pwm_get returns -ENOENT.

BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=96571
Reported-by: youling257@gmail.com
Fixes: 4a6ef8e37c ("pwm: Add support referencing PWMs from ACPI")
Cc: Nikolaus Voss <nikolaus.voss@loewensteinmedical.de>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Nikolaus Voss <nikolaus.voss@loewensteinmedical.de>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
2019-08-08 13:17:38 +02:00
Oliver Neukum
c468a8aa79 usb: iowarrior: fix deadlock on disconnect
We have to drop the mutex before we close() upon disconnect()
as close() needs the lock. This is safe to do by dropping the
mutex as intfdata is already set to NULL, so open() will fail.

Fixes: 03f36e885f ("USB: open disconnect race in iowarrior")
Reported-by: syzbot+a64a382964bf6c71a9c0@syzkaller.appspotmail.com
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Oliver Neukum <oneukum@suse.com>
Link: https://lore.kernel.org/r/20190808092728.23417-1-oneukum@suse.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-08-08 12:43:18 +02:00
Oliver Neukum
2ca359f4f8 Revert "USB: rio500: simplify locking"
This reverts commit d710734b06.
This simplification causes a deadlock.

Reported-by: syzbot+7bbcbe9c9ff0cd49592a@syzkaller.appspotmail.com
Fixes: d710734b06 ("USB: rio500: simplify locking")
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Oliver Neukum <oneukum@suse.com>
Link: https://lore.kernel.org/r/20190808092854.23519-1-oneukum@suse.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-08-08 12:43:18 +02:00
Andy Shevchenko
bf5ab1bded pinctrl: denverton: Update pin names according to v1.08
Version 1.08 of pin list has some changes in pin names for Intel Denverton.

Update the driver accordingly.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
2019-08-08 12:57:01 +03:00
Greg Kroah-Hartman
b6fd948832 MAINTAINERS: mark wusbcore and UWB as obsolete
Joe rightly points out that we should be using the "Obsolete" status for
these two subsystems.

Also I got the path name wrong for the wusbcore tree.

Reported-by: Joe Perches <joe@perches.com>
Fixes: 71ed79b0e4 ("USB: Move wusbcore and UWB to staging as it is obsolete")
Link: https://lore.kernel.org/r/20190808092509.GA20173@kroah.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-08-08 11:42:18 +02:00
Rafael J. Wysocki
ee8193ee96 ACPI: PM: Print debug messages on device power state changes
Add an acpi_handle_debug() statement to acpi_device_set_power() to
allow ACPI device power state changes to be tracked.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2019-08-08 11:32:51 +02:00
Rafael J. Wysocki
ac9eafbe93 ACPI: PM: s2idle: Execute LPS0 _DSM functions with suspended devices
According to Section 3.5 of the "Intel Low Power S0 Idle" document [1],
Function 5 of the LPS0 _DSM is expected to be invoked when the system
configuration matches the criteria for entering the target low-power
state of the platform.  In particular, this means that all devices
should be suspended and in low-power states already when that function
is invoked.

This is not the case currently, however, because Function 5 of the
LPS0 _DSM is invoked by it before the "noirq" phase of device suspend,
which means that some devices may not have been put into low-power
states yet at that point.  That is a consequence of the previous
design of the suspend-to-idle flow that allowed the "noirq" phase of
device suspend and the "noirq" phase of device resume to be carried
out for multiple times while "suspended" (if any spurious wakeup
events were detected) and the point of the LPS0 _DSM Function 5
invocation was chosen so as to call it (and LPS0 _DSM Function 6
analogously) once per suspend-resume cycle (regardless of how many
times the "noirq" phases of device suspend and resume were carried
out while "suspended").

Now that the suspend-to-idle flow has been redesigned to carry out
the "noirq" phases of device suspend and resume once in each cycle,
the code can be reordered to follow the specification that it is
based on more closely.

For this purpose, add ->prepare_late and ->restore_early platform
callbacks for suspend-to-idle, to be executed, respectively, after
the "noirq" phase of suspending devices and before the "noirq"
phase of resuming them and make ACPI use them for the invocation
of LPS0 _DSM functions as appropriate.

While at it, move the LPS0 entry requirements check to be made
before invoking Functions 3 and 5 of the LPS0 _DSM (also once
per cycle) as follows from the specification [1].

Link: https://uefi.org/sites/default/files/resources/Intel_ACPI_Low_Power_S0_Idle.pdf # [1]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
2019-08-08 11:26:01 +02:00
Rafael J. Wysocki
29113f2f0a ACPI: EC: PM: Make acpi_ec_dispatch_gpe() print debug message
Add a pm_pr_dbg() debug statement to acpi_ec_dispatch_gpe() to print
a message when the EC GPE has been dispatched (because its status
was set).

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
2019-08-08 11:25:53 +02:00
Rafael J. Wysocki
d758940493 ACPI: EC: PM: Consolidate some code depending on PM_SLEEP
Move some routines, including acpi_ec_dispatch_gpe(), that are only
used if CONFIG_PM_SLEEP is set to the #ifdef block containing the EC
suspend and resume callbacks, to make the "full EC PM picture" easier
to follow.

While at it, move the header of acpi_ec_dispatch_gpe() in the
header file to a CONFIG_PM_SLEEP #ifdef block.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
2019-08-08 11:25:43 +02:00
Rafael J. Wysocki
6e86633a79 ACPI: PM: s2idle: Eliminate acpi_sleep_no_ec_events()
Change acpi_ec_suspend() to use pm_suspend_no_platform() instead of
acpi_sleep_no_ec_events(), which allows the latter to be eliminated
along with the s2idle_in_progress variable which is only used by it.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
2019-08-08 11:25:37 +02:00
Rafael J. Wysocki
fcd0a04267 ACPI: PM: s2idle: Switch EC over to polling during "noirq" suspend
Since the ACPI SCI is set up for system wakeup before the "noirq"
suspend of devices, it is better to make suspend-to-idle follow
suspend-to-RAM (S3) and switch over the EC to polling during "noirq"
suspend (and back to interrupt-based flow during "noirq" resume).

The frequency of spurious wakeup interrupts from the EC may be
reduced this way.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
2019-08-08 11:25:30 +02:00
Rafael J. Wysocki
068b47d098 ACPI: PM: s2idle: Add acpi.sleep_no_lps0 module parameter
Add a module parameter to prevent the ACPI LPS0 _DSM functions
from being invoked (if need be) and rework the suspend-to-idle
blacklist entries in acpisleep_dmi_table[] to make them simply
prevent suspend-to-idle from being used by default on the systems
in question (which really is the original purpose of those entries).

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
2019-08-08 11:25:22 +02:00
Rafael J. Wysocki
2e2c2fdc53 ACPI: PM: s2idle: Rearrange lps0_device_attach()
To allow a subsequent change to be simpler, rearrange the code in
lps0_device_attach() to reduce the indentation level and (while
at it) make it avoid calling lpi_device_get_constraints() when
lps0_device_handle is not going to be set.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
2019-08-08 11:24:47 +02:00
Colin Ian King
6b7c3b86f0 drm/vmwgfx: fix memory leak when too many retries have occurred
Currently when too many retries have occurred there is a memory
leak on the allocation for reply on the error return path. Fix
this by kfree'ing reply before returning.

Addresses-Coverity: ("Resource leak")
Fixes: a9cd9c044a ("drm/vmwgfx: Add a check to handle host message failure")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Deepak Rawat <drawat@vmware.com>
Signed-off-by: Deepak Rawat <drawat@vmware.com>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
2019-08-08 11:22:54 +02:00
Wenwen Wang
1be3c1fae6 ALSA: firewire: fix a memory leak bug
In iso_packets_buffer_init(), 'b->packets' is allocated through
kmalloc_array(). Then, the aligned packet size is checked. If it is
larger than PAGE_SIZE, -EINVAL will be returned to indicate the error.
However, the allocated 'b->packets' is not deallocated on this path,
leading to a memory leak.

To fix the above issue, free 'b->packets' before returning the error code.

Fixes: 31ef9134eb ("ALSA: add LaCie FireWire Speakers/Griffin FireWave Surround driver")
Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Cc: <stable@vger.kernel.org> # v2.6.39+
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2019-08-08 11:12:26 +02:00
Niklas Söderlund
ddc8cd7d92 dt-bindings: rcar-{csi2,vin}: Rename bindings documentation files
Renesas media binding documentation files use a naming schema of
'renesas,<module>.txt'. Rename the VIN and CSI-2 files to match this
pattern.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Ulrich Hecht <uli+renesas@fpond.eu>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
2019-08-08 10:40:35 +02:00
Simon Horman
72a83ded7f dt-bindings: rcar-imr: Rename bindings documentation file
Renesas media binding documentation files use a naming schema of
'renesas,<module>.txt'. Rename the IMR file to match this pattern.

Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
2019-08-08 10:40:19 +02:00
Simon Horman
f32d0c6363 dt-bindings: Rename file of DT bindings for Renesas memory controllers
For consistency with the naming of (most) other documentation files for DT
bindings for Renesas IP blocks rename the Renesas R-Mobile and SH-Mobile
memory controllers documentation file from renesas-memory-controllers.txt
to renesas,dbsc.txt.

Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
2019-08-08 10:39:48 +02:00
Jani Nikula
707d26dcc1 drm/i915: split out intel_pch.[ch] from i915_drv.[ch]
Abstract the rather self-contained piece of code from i915_drv.[ch]. No
functional changes.

Cc: José Roberto de Souza <jose.souza@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190807120415.17917-1-jani.nikula@intel.com
2019-08-08 11:38:22 +03:00
Geert Uytterhoeven
e96a926104 clk: renesas: rcar-usb2-clock-sel: Use devm_platform_ioremap_resource() helper
Use the devm_platform_ioremap_resource() helper instead of open-coding
the same operation.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
2019-08-08 10:22:59 +02:00
Geert Uytterhoeven
6da53cfa0f pinctrl: sh-pfc: Use dev_notice_once() instead of open-coding
At the time of commit 9a643c9a11 ("sh-pfc: Convert message
printing from pr_* to dev_*"), the dev_*_once() variants didn't exist
yet, so the once behavior was open-coded.

Since commit e135303bd5 ("device: Add dev_<level>_once variants")
they do, so "revert" to the good practice of using a helper.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Ulrich Hecht <uli+renesas@fpond.eu>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
2019-08-08 10:18:57 +02:00
Narendra K
1c5fecb612 efi: Export Runtime Configuration Interface table to sysfs
System firmware advertises the address of the 'Runtime
Configuration Interface table version 2 (RCI2)' via
an EFI Configuration Table entry. This code retrieves the RCI2
table from the address and exports it to sysfs as a binary
attribute 'rci2' under /sys/firmware/efi/tables directory.
The approach adopted is similar to the attribute 'DMI' under
/sys/firmware/dmi/tables.

RCI2 table contains BIOS HII in XML format and is used to populate
BIOS setup page in Dell EMC OpenManage Server Administrator tool.
The BIOS setup page contains BIOS tokens which can be configured.

Signed-off-by: Narendra K <Narendra.K@dell.com>
Reviewed-by: Mario Limonciello <mario.limonciello@dell.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-08-08 11:10:25 +03:00
Ard Biesheuvel
5828efb95b efi: ia64: move SAL systab handling out of generic EFI code
The SAL systab is an Itanium specific EFI configuration table, so
move its handling into arch/ia64 where it belongs.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-08-08 11:01:48 +03:00
Ard Biesheuvel
ec7e1605d7 efi/x86: move UV_SYSTAB handling into arch/x86
The SGI UV UEFI machines are tightly coupled to the x86 architecture
so there is no need to keep any awareness of its existence in the
generic EFI layer, especially since we already have the infrastructure
to handle arch-specific configuration tables, and were even already
using it to some extent.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-08-08 11:01:48 +03:00
Ard Biesheuvel
e55f31a599 efi: x86: move efi_is_table_address() into arch/x86
The function efi_is_table_address() and the associated array of table
pointers is specific to x86. Since we will be adding some more x86
specific tables, let's move this code out of the generic code first.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
2019-08-08 11:01:48 +03:00
Qais Yousef
5c3ceef9ad cpufreq: schedutil: fix equation in comment
scale_irq_capacity() call in schedutil_cpu_util() does

	util *= (max - irq)
	util /= max

But the comment says

	util *= (1 - irq)
	util /= max

Fix the comment to match what the scaling function does.

Signed-off-by: Qais Yousef <qais.yousef@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: "Rafael J . Wysocki" <rjw@rjwysocki.net>
Link: https://lkml.kernel.org/r/20190802104628.8410-1-qais.yousef@arm.com
2019-08-08 09:09:31 +02:00
Peter Zijlstra
67692435c4 sched: Rework pick_next_task() slow-path
Avoid the RETRY_TASK case in the pick_next_task() slow path.

By doing the put_prev_task() early, we get the rt/deadline pull done,
and by testing rq->nr_running we know if we need newidle_balance().

This then gives a stable state to pick a task from.

Since the fast-path is fair only; it means the other classes will
always have pick_next_task(.prev=NULL, .rf=NULL) and we can simplify.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
Link: https://lkml.kernel.org/r/aa34d24b36547139248f32a30138791ac6c02bd6.1559129225.git.vpillai@digitalocean.com
2019-08-08 09:09:31 +02:00
Peter Zijlstra
5f2a45fc9e sched: Allow put_prev_task() to drop rq->lock
Currently the pick_next_task() loop is convoluted and ugly because of
how it can drop the rq->lock and needs to restart the picking.

For the RT/Deadline classes, it is put_prev_task() where we do
balancing, and we could do this before the picking loop. Make this
possible.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
Link: https://lkml.kernel.org/r/e4519f6850477ab7f3d257062796e6425ee4ba7c.1559129225.git.vpillai@digitalocean.com
2019-08-08 09:09:31 +02:00
Peter Zijlstra
5ba553eff0 sched/fair: Expose newidle_balance()
For pick_next_task_fair() it is the newidle balance that requires
dropping the rq->lock; provided we do put_prev_task() early, we can
also detect the condition for doing newidle early.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
Link: https://lkml.kernel.org/r/9e3eb1859b946f03d7e500453a885725b68957ba.1559129225.git.vpillai@digitalocean.com
2019-08-08 09:09:31 +02:00
Peter Zijlstra
03b7fad167 sched: Add task_struct pointer to sched_class::set_curr_task
In preparation of further separating pick_next_task() and
set_curr_task() we have to pass the actual task into it, while there,
rename the thing to better pair with put_prev_task().

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
Link: https://lkml.kernel.org/r/a96d1bcdd716db4a4c5da2fece647a1456c0ed78.1559129225.git.vpillai@digitalocean.com
2019-08-08 09:09:31 +02:00
Peter Zijlstra
10e7071b2f sched: Rework CPU hotplug task selection
The CPU hotplug task selection is the only place where we used
put_prev_task() on a task that is not current. While looking at that,
it occured to me that we can simplify all that by by using a custom
pick loop.

Since we don't need to put current, we can do away with the fake task
too.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
2019-08-08 09:09:30 +02:00
Peter Zijlstra
f95d4eaee6 sched/{rt,deadline}: Fix set_next_task vs pick_next_task
Because pick_next_task() implies set_curr_task() and some of the
details haven't mattered too much, some of what _should_ be in
set_curr_task() ended up in pick_next_task, correct this.

This prepares the way for a pick_next_task() variant that does not
affect the current state; allowing remote picking.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
Link: https://lkml.kernel.org/r/38c61d5240553e043c27c5e00b9dd0d184dd6081.1559129225.git.vpillai@digitalocean.com
2019-08-08 09:09:30 +02:00
Peter Zijlstra
5feeb7837a sched: Fix kerneldoc comment for ia64_set_curr_task
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
Link: https://lkml.kernel.org/r/fde3a65ea3091ec6b84dac3c19639f85f452c5d1.1559129225.git.vpillai@digitalocean.com
2019-08-08 09:09:30 +02:00
Peter Zijlstra
99d84bf8c6 stop_machine: Fix stop_cpus_in_progress ordering
Make sure the entire for loop has stop_cpus_in_progress set.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lwe@gmail.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: mingo@kernel.org
Cc: Phil Auld <pauld@redhat.com>
Cc: Julien Desfossez <jdesfossez@digitalocean.com>
Cc: Nishanth Aravamudan <naravamudan@digitalocean.com>
Link: https://lkml.kernel.org/r/0fd8fd4b99b9b9aa88d8b2dff897f7fd0d88f72c.1559129225.git.vpillai@digitalocean.com
2019-08-08 09:09:30 +02:00
Dave Chiluk
de53fd7aed sched/fair: Fix low cpu usage with high throttling by removing expiration of cpu-local slices
It has been observed, that highly-threaded, non-cpu-bound applications
running under cpu.cfs_quota_us constraints can hit a high percentage of
periods throttled while simultaneously not consuming the allocated
amount of quota. This use case is typical of user-interactive non-cpu
bound applications, such as those running in kubernetes or mesos when
run on multiple cpu cores.

This has been root caused to cpu-local run queue being allocated per cpu
bandwidth slices, and then not fully using that slice within the period.
At which point the slice and quota expires. This expiration of unused
slice results in applications not being able to utilize the quota for
which they are allocated.

The non-expiration of per-cpu slices was recently fixed by
'commit 512ac999d2 ("sched/fair: Fix bandwidth timer clock drift
condition")'. Prior to that it appears that this had been broken since
at least 'commit 51f2176d74 ("sched/fair: Fix unlocked reads of some
cfs_b->quota/period")' which was introduced in v3.16-rc1 in 2014. That
added the following conditional which resulted in slices never being
expired.

if (cfs_rq->runtime_expires != cfs_b->runtime_expires) {
	/* extend local deadline, drift is bounded above by 2 ticks */
	cfs_rq->runtime_expires += TICK_NSEC;

Because this was broken for nearly 5 years, and has recently been fixed
and is now being noticed by many users running kubernetes
(https://github.com/kubernetes/kubernetes/issues/67577) it is my opinion
that the mechanisms around expiring runtime should be removed
altogether.

This allows quota already allocated to per-cpu run-queues to live longer
than the period boundary. This allows threads on runqueues that do not
use much CPU to continue to use their remaining slice over a longer
period of time than cpu.cfs_period_us. However, this helps prevent the
above condition of hitting throttling while also not fully utilizing
your cpu quota.

This theoretically allows a machine to use slightly more than its
allotted quota in some periods. This overflow would be bounded by the
remaining quota left on each per-cpu runqueueu. This is typically no
more than min_cfs_rq_runtime=1ms per cpu. For CPU bound tasks this will
change nothing, as they should theoretically fully utilize all of their
quota in each period. For user-interactive tasks as described above this
provides a much better user/application experience as their cpu
utilization will more closely match the amount they requested when they
hit throttling. This means that cpu limits no longer strictly apply per
period for non-cpu bound applications, but that they are still accurate
over longer timeframes.

This greatly improves performance of high-thread-count, non-cpu bound
applications with low cfs_quota_us allocation on high-core-count
machines. In the case of an artificial testcase (10ms/100ms of quota on
80 CPU machine), this commit resulted in almost 30x performance
improvement, while still maintaining correct cpu quota restrictions.
That testcase is available at https://github.com/indeedeng/fibtest.

Fixes: 512ac999d2 ("sched/fair: Fix bandwidth timer clock drift condition")
Signed-off-by: Dave Chiluk <chiluk+linux@indeed.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: John Hammond <jhammond@indeed.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kyle Anderson <kwa@yelp.com>
Cc: Gabriel Munos <gmunoz@netflix.com>
Cc: Peter Oskolkov <posk@posk.io>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Brendan Gregg <bgregg@netflix.com>
Link: https://lkml.kernel.org/r/1563900266-19734-2-git-send-email-chiluk+linux@indeed.com
2019-08-08 09:09:30 +02:00
Peter Zijlstra
139d025cda sched: Clean up active_mm reference counting
The current active_mm reference counting is confusing and sub-optimal.

Rewrite the code to explicitly consider the 4 separate cases:

    user -> user

	When switching between two user tasks, all we need to consider
	is switch_mm().

    user -> kernel

	When switching from a user task to a kernel task (which
	doesn't have an associated mm) we retain the last mm in our
	active_mm. Increment a reference count on active_mm.

  kernel -> kernel

	When switching between kernel threads, all we need to do is
	pass along the active_mm reference.

  kernel -> user

	When switching between a kernel and user task, we must switch
	from the last active_mm to the next mm, hoping of course that
	these are the same. Decrement a reference on the active_mm.

The code keeps a different order, because as you'll note, both 'to
user' cases require switch_mm().

And where the old code would increment/decrement for the 'kernel ->
kernel' case, the new code observes this is a neutral operation and
avoids touching the reference count.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Rik van Riel <riel@surriel.com>
Reviewed-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: luto@kernel.org
2019-08-08 09:09:30 +02:00
Peter Zijlstra
130d9c331b rcu/tree: Fix SCHED_FIFO params
A rather embarrasing mistake had us call sched_setscheduler() before
initializing the parameters passed to it.

Fixes: 1a763fd7c6 ("rcu/tree: Call setschedule() gp ktread to SCHED_FIFO outside of atomic region")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paul E. McKenney <paulmck@linux.ibm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
2019-08-08 09:09:30 +02:00
Peter Zijlstra
e57d143091 mutex: Fix up mutex_waiter usage
The patch moving bits into mutex.c was a little too much; by also
moving struct mutex_waiter a few less common CONFIGs would no longer
build.

Fixes: 5f35d5a66b ("locking/mutex: Make __mutex_owner static to mutex.c")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
2019-08-08 09:09:25 +02:00
Ming Lei
491beed3b1 genirq/affinity: Create affinity mask for single vector
Since commit c66d4bd110 ("genirq/affinity: Add new callback for
(re)calculating interrupt sets"), irq_create_affinity_masks() returns
NULL in case of single vector. This change has caused regression on some
drivers, such as lpfc.

The problem is that single vector requests can happen in some generic cases:

  1) kdump kernel

  2) irq vectors resource is close to exhaustion.

If in that situation the affinity mask for a single vector is not created,
every caller has to handle the special case.

There is no reason why the mask cannot be created, so remove the check for
a single vector and create the mask.

Fixes: c66d4bd110 ("genirq/affinity: Add new callback for (re)calculating interrupt sets")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20190805011906.5020-1-ming.lei@redhat.com
2019-08-08 08:47:55 +02:00