linux-toradex.git/include/linux/math64.h, branch v6.4-rc1

sched/clock/x86: Mark sched_clock() noinstr

2023-01-31T14:01:47+00:00

In order to use sched_clock() from noinstr code, mark it and all it's
implenentations noinstr.

The whole pvclock thing (used by KVM/Xen) is a bit of a pain,
since it calls out to watchdogs, create a
pvclock_clocksource_read_nowd() variant doesn't do that and can be
noinstr.

Signed-off-by: Peter Zijlstra (Intel) 
Signed-off-by: Ingo Molnar 
Link: https://lore.kernel.org/r/20230126151323.702003578@infradead.org

math64: fix kernel-doc return value warnings

2022-11-21T21:30:54+00:00

Fix the following kernel-doc warnings by adding a description for return
values of div_[us]64.

	math64.h:126: warning: No description found for return value of 'div_u64'
	math64.h:139: warning: No description found for return value of 'div_s64'

Signed-off-by: Liam Beguin 
Link: https://lore.kernel.org/r/20221118182309.3824530-3-liambeguin@gmail.com
Signed-off-by: Jonathan Corbet

math64: add kernel-doc for DIV64_U64_ROUND_UP

2022-11-21T21:30:54+00:00

Add kernel-doc for DIV64_U64_ROUND_UP so that it appears in the
documentation.

Signed-off-by: Liam Beguin 
Acked-by: Randy Dunlap 
Tested-by: Randy Dunlap 
Link: https://lore.kernel.org/r/20221118182309.3824530-2-liambeguin@gmail.com
Signed-off-by: Jonathan Corbet

math64: favor kernel-doc from header files

2022-11-21T21:30:53+00:00

Fix the kernel-doc markings for div64 functions to point to the header
file instead of the lib/ directory.  This avoids having implementation
specific comments in generic documentation.  Furthermore, given that
some kernel-doc comments are identical, drop them from lib/math64 and
only keep there comments that add implementation details.

Signed-off-by: Liam Beguin 
Acked-by: Randy Dunlap 
Tested-by: Randy Dunlap 
Link: https://lore.kernel.org/r/20221118182309.3824530-1-liambeguin@gmail.com
Signed-off-by: Jonathan Corbet

math64: New DIV_U64_ROUND_CLOSEST helper

2022-02-25T09:27:57+00:00

Provide DIV_U64_ROUND_CLOSEST helper which uses div_u64 to perform
division rounded to the closest integer using unsigned 64bit
dividend and unsigned 32bit divisor.

Reviewed-by: Marek Behún 
Signed-off-by: Pali Rohár 
Signed-off-by: Marek Behún 
Link: https://lore.kernel.org/r/20220219152818.4319-2-kabel@kernel.org
Signed-off-by: Greg Kroah-Hartman

math64.h: Add mul_s64_u64_shr()

2021-06-17T17:09:27+00:00

This function is needed for KVM's nested virtualization. The nested TSC
scaling implementation requires multiplying the signed TSC offset with
the unsigned TSC multiplier.

Signed-off-by: Ilias Stamatis 
Reviewed-by: Maxim Levitsky 
Message-Id: <20210526184418.28881-2-ilstam@amazon.com>
Signed-off-by: Paolo Bonzini

math64.h: kernel-docs: Convert some markups into normal comments

2020-10-15T05:49:46+00:00

There are several functions at math64.h that are also
defined at div64.c. As both are included at kernel-api.rst,
Sphinx 3.x complains about symbol duplication:

	./lib/math/div64.c:73: WARNING: Duplicate C declaration, also defined in 'core-api/kernel-api'.
	Declaration is 'div_s64_rem'.
	./lib/math/div64.c:104: WARNING: Duplicate C declaration, also defined in 'core-api/kernel-api'.
	Declaration is 'div64_u64_rem'.
	./lib/math/div64.c:144: WARNING: Duplicate C declaration, also defined in 'core-api/kernel-api'.
	Declaration is 'div64_u64'.
	./lib/math/div64.c:172: WARNING: Duplicate C declaration, also defined in 'core-api/kernel-api'.
	Declaration is 'div64_s64'.

In order to avoid Sphinx warnings about duplication, change the kernel-doc
markups to just comments at math64.h.

Reviewed-by: Vincenzo Frascino 
Signed-off-by: Mauro Carvalho Chehab

Merge tag 'for-v5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supply

2020-08-08T04:27:37+00:00

Pull power supply and reset updates from Sebastian Reichel:
 "Power-supply core:
   - add COOL/WARM/HOT state from JEITA JISC8712:2015 specification
   - convert simple-battery DT binding to YAML
   - add long-life charging mode

 Battery/charger drivers:
   - bq25150: new charger driver
   - bq27xxx: add support for BQ27z561 and BQ28z610
   - max17040: support CAPACITY_ALERT_MIN
   - sbs-battery: add PEC support
   - wilco-ec: support long-life charging mode
   - bq25890: fix DT binding
   - misc. fixes and cleanups

 Reset drivers:
   - linkstation: new reset driver"

* tag 'for-v5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supply: (32 commits)
  power: supply: wilco_ec: Add long life charging mode
  power: supply: bq27xxx_battery: Add the BQ28z610 Battery monitor
  dt-bindings: power: Add BQ28z610 compatible
  power: supply: bq27xxx_battery: Add the BQ27Z561 Battery monitor
  dt-bindings: power: Add BQ27Z561 compatible
  power: supply: test_power: Fix battery_current initial value
  power: supply: Fix kerneldoc of power_supply_temp2resist_simple()
  power: supply: cpcap-battery: Fix kerneldoc of cpcap_battery_read_accumulated()
  dt-bindings: power: Convert battery.txt to battery.yaml
  power: supply: rt5033_battery: Fix error code in rt5033_battery_probe()
  power: supply: max17040: Add POWER_SUPPLY_PROP_CAPACITY_ALERT_MIN
  power: supply: check if calc_soc succeeded in pm860x_init_battery
  power: supply: bq2xxxx: Replace HTTP links with HTTPS ones
  power: reset: add driver for LinkStation power off
  power: supply: sc27xx: prevent adc * 1000 from overflow
  math64: New DIV_S64_ROUND_CLOSEST helper
  power: fix duplicated words in bq2415x_charger.h
  power: Convert to DEFINE_SHOW_ATTRIBUTE
  power: reset: keystone-reset: Replace HTTP links with HTTPS ones
  power: supply: bq25150 introduce the bq25150
  ...

math64: New DIV_S64_ROUND_CLOSEST helper

2020-07-27T22:47:18+00:00

Provide DIV_S64_ROUND_CLOSEST helper which uses div_s64 to perform
division rounded to the closest integer using signed 64bit
dividend and signed 32bit divisor.

Signed-off-by: Chunyan Zhang 
Signed-off-by: Sebastian Reichel

sched/cputime: Improve cputime_adjust()

2020-06-15T12:10:00+00:00

People report that utime and stime from /proc//stat become very
wrong when the numbers are big enough, especially if you watch these
counters incrementally.

Specifically, the current implementation of: stime*rtime/total,
results in a saw-tooth function on top of the desired line, where the
teeth grow in size the larger the values become. IOW, it has a
relative error.

The result is that, when watching incrementally as time progresses
(for large values), we'll see periods of pure stime or utime increase,
irrespective of the actual ratio we're striving for.

Replace scale_stime() with a math64.h helper: mul_u64_u64_div_u64()
that is far more accurate. This also allows architectures to override
the implementation -- for instance they can opt for the old algorithm
if this new one turns out to be too expensive for them.

Signed-off-by: Oleg Nesterov 
Signed-off-by: Peter Zijlstra (Intel) 
Link: https://lkml.kernel.org/r/20200519172506.GA317395@hirez.programming.kicks-ass.net