Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1409597

[PATCH 4.6 023/100] sched/loadavg: Fix loadavg artifacts on fully idle and on fully loaded systems

Show all headers | View raw

4.6-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Vik Heyndrickx <vik.heyndrickx@veribox.net>

commit 20878232c52329f92423d27a60e48b6a6389e0dd upstream.

Systems show a minimal load average of 0.00, 0.01, 0.05 even when they
have no load at all.

Uptime and /proc/loadavg on all systems with kernels released during the
last five years up until kernel version 4.6-rc5, show a 5- and 15-minute
minimum loadavg of 0.01 and 0.05 respectively. This should be 0.00 on
idle systems, but the way the kernel calculates this value prevents it
from getting lower than the mentioned values.

Likewise but not as obviously noticeable, a fully loaded system with no
processes waiting, shows a maximum 1/5/15 loadavg of 1.00, 0.99, 0.95
(multiplied by number of cores).

Once the (old) load becomes 93 or higher, it mathematically can never
get lower than 93, even when the active (load) remains 0 forever.
This results in the strange 0.00, 0.01, 0.05 uptime values on idle
systems.  Note: 93/2048 = 0.0454..., which rounds up to 0.05.

It is not correct to add a 0.5 rounding (=1024/2048) here, since the
result from this function is fed back into the next iteration again,
so the result of that +0.5 rounding value then gets multiplied by
(2048-2037), and then rounded again, so there is a virtual "ghost"
load created, next to the old and active load terms.

By changing the way the internally kept value is rounded, that internal
value equivalent now can reach 0.00 on idle, and 1.00 on full load. Upon
increasing load, the internally kept load value is rounded up, when the
load is decreasing, the load value is rounded down.

The modified code was tested on nohz=off and nohz kernels. It was tested
on vanilla kernel 4.6-rc5 and on centos 7.1 kernel 3.10.0-327. It was
tested on single, dual, and octal cores system. It was tested on virtual
hosts and bare hardware. No unwanted effects have been observed, and the
problems that the patch intended to fix were indeed gone.

Tested-by: Damien Wyart <damien.wyart@free.fr>
Signed-off-by: Vik Heyndrickx <vik.heyndrickx@veribox.net>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Doug Smythies <dsmythies@telus.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: 0f004f5a696a ("sched: Cure more NO_HZ load average woes")
Link: http://lkml.kernel.org/r/e8d32bff-d544-7748-72b5-3c86cc71f09f@veribox.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 kernel/sched/loadavg.c |   11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

--- a/kernel/sched/loadavg.c
+++ b/kernel/sched/loadavg.c
@@ -99,10 +99,13 @@ long calc_load_fold_active(struct rq *th
 static unsigned long
 calc_load(unsigned long load, unsigned long exp, unsigned long active)
 {
-	load *= exp;
-	load += active * (FIXED_1 - exp);
-	load += 1UL << (FSHIFT - 1);
-	return load >> FSHIFT;
+	unsigned long newload;
+
+	newload = load * exp + active * (FIXED_1 - exp);
+	if (active >= load)
+		newload += FIXED_1-1;
+
+	return newload / FIXED_1;
 }
 
 #ifdef CONFIG_NO_HZ_COMMON

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 4.6 000/100] 4.6.1-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 074/100] serial: 8250_mid: recognize interrupt source in handler Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 070/100] tty/serial: atmel: fix hardware handshake selection Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 077/100] clk: bcm2835: add locking to pll*_on/off methods Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 069/100] TTY: n_gsm, fix false positive WARN_ON Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 055/100] usb: misc: usbtest: fix pattern tests for scatterlists. Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 073/100] serial: 8250_mid: use proper bar for DNV platform Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 079/100] watchdog: sp5100_tco: properly check for new register layouts Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 071/100] Fix OpenSSH pty regression on close Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 056/100] USB: leave LPM alone if possible when binding/unbinding interface drivers Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 022/100] fscrypto/f2fs: allow fs-specific key prefix for fs encryption Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 076/100] locking,qspinlock: Fix spin_is_locked() and spin_unlock_wait() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 007/100] kvm: arm64: Fix EC field in inject_abt64 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 068/100] tty: vt, return error when con_startup fails Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 067/100] x86/cpufeature, x86/mm/pkeys: Fix broken compile-time disabling of pkeys Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 042/100] mei: amthif: discard not read messages Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 008/100] perf/x86/intel/uncore: Remove WARN_ON_ONCE in uncore_pci_probe Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 066/100] xen/x86: actually allocate legacy interrupts on PV guests Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 043/100] mei: bus: call mei_cl_read_start under device lock Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 081/100] Fixing oops in callback path Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 059/100] usb: host: xhci-rcar: Avoid long wait in xhci_reset() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 044/100] USB: serial: cp210x: fix hardware flow-control disable Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 060/100] staging: comedi: das1800: fix possible NULL dereference Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:30 +0200
  [PATCH 4.6 039/100] Bluetooth: vhci: purge unhandled skbs Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:40 +0200
  [PATCH 4.6 038/100] Bluetooth: vhci: fix open_timeout vs. hdev race Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:40 +0200
  [PATCH 4.6 016/100] ring-buffer: Prevent overflow of size in ring_buffer_resize() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:40 +0200
  [PATCH 4.6 033/100] mmc: sdhci-acpi: Remove MMC_CAP_BUS_WIDTH_TEST for Intel controllers Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 037/100] mmc: sdhci-pci: Remove MMC_CAP_BUS_WIDTH_TEST for Intel controllers Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 031/100] ACPI / PM: Export acpi_device_fix_up_power() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 027/100] irqchip/gic-v3: Configure all interrupts as non-secure Group-1 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 036/100] mmc: longer timeout for long read time quirk Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 020/100] clk: qcom: msm8916: Fix crypto clock flags Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 024/100] mfd: omap-usb-tll: Fix scheduling while atomic BUG Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 018/100] crypto: talitos - fix ahash algorithms registration Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 002/100] arm64: Ensure pmd_present() returns false after pmd_mknotpresent() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 025/100] Input: pwm-beeper - fix - scheduling while atomic Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 026/100] irqchip/gic: Ensure ordering between read of INTACK and shared data Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 030/100] mmc: mmc: Fix partition switch timeout for some eMMCs Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 003/100] arm64: Implement ptep_set_access_flags() for hardware AF/DBM Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 019/100] crypto: sun4i-ss - Replace spinlock_bh by spin_lock_irq{save|restore} Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 023/100] sched/loadavg: Fix loadavg artifacts on fully idle and on fully loaded systems Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 017/100] crypto: caam - fix caam_jr_alloc() ret code Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  [PATCH 4.6 034/100] ACPI / osi: Fix an issue that acpi_osi=!* cannot disable ACPICA internal strings Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-05-30 23:50 +0200
  Re: [PATCH 4.6 000/100] 4.6.1-stable review Guenter Roeck <linux@roeck-us.net> - 2016-06-01 07:40 +0200
    Re: [PATCH 4.6 000/100] 4.6.1-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-06-01 18:20 +0200
      Re: [PATCH 4.6 000/100] 4.6.1-stable review "Xuetao Guan" <gxt@mprc.pku.edu.cn> - 2016-06-02 06:40 +0200
  Re: [PATCH 4.6 000/100] 4.6.1-stable review Shuah Khan <shuahkh@osg.samsung.com> - 2016-06-01 16:30 +0200

csiph-web