public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Giovanni Gherdovich <ggherdovich@suse.cz>,
	Peter Zijlstra <peterz@infradead.org>,
	"Rafael J . Wysocki" <rafael.j.wysocki@intel.com>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH AUTOSEL 5.8 05/72] x86, sched: check for counters overflow in frequency invariant accounting
Date: Sat,  8 Aug 2020 19:34:34 -0400	[thread overview]
Message-ID: <20200808233542.3617339-5-sashal@kernel.org> (raw)
In-Reply-To: <20200808233542.3617339-1-sashal@kernel.org>

From: Giovanni Gherdovich <ggherdovich@suse.cz>

[ Upstream commit e2b0d619b400ae326f954a018a1d65d736c237c5 ]

The product mcnt * arch_max_freq_ratio can overflows u64.

For context, a large value for arch_max_freq_ratio would be 5000,
corresponding to a turbo_freq/base_freq ratio of 5 (normally it's more like
1500-2000). A large increment frequency for the MPERF counter would be 5GHz
(the base clock of all CPUs on the market today is less than that). With
these figures, a CPU would need to go without a scheduler tick for around 8
days for the u64 overflow to happen. It is unlikely, but the check is
warranted.

Under similar conditions, the difference acnt of two consecutive APERF
readings can overflow as well.

In these circumstances is appropriate to disable frequency invariant
accounting: the feature relies on measures of the clock frequency done at
every scheduler tick, which need to be "fresh" to be at all meaningful.

A note on i386: prior to version 5.1, the GCC compiler didn't have the
builtin function __builtin_mul_overflow. In these GCC versions the macro
check_mul_overflow needs __udivdi3() to do (u64)a/b, which the kernel
doesn't provide. For this reason this change fails to build on i386 if
GCC<5.1, and we protect the entire frequency invariant code behind
CONFIG_X86_64 (special thanks to "kbuild test robot" <lkp@intel.com>).

Fixes: 1567c3e3467c ("x86, sched: Add support for frequency invariance")
Signed-off-by: Giovanni Gherdovich <ggherdovich@suse.cz>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lkml.kernel.org/r/20200531182453.15254-2-ggherdovich@suse.cz
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 arch/x86/include/asm/topology.h |  2 +-
 arch/x86/kernel/smpboot.c       | 33 ++++++++++++++++++++++++++++-----
 2 files changed, 29 insertions(+), 6 deletions(-)

diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h
index 79d8d54963303..f4234575f3fdb 100644
--- a/arch/x86/include/asm/topology.h
+++ b/arch/x86/include/asm/topology.h
@@ -193,7 +193,7 @@ static inline void sched_clear_itmt_support(void)
 }
 #endif /* CONFIG_SCHED_MC_PRIO */
 
-#ifdef CONFIG_SMP
+#if defined(CONFIG_SMP) && defined(CONFIG_X86_64)
 #include <asm/cpufeature.h>
 
 DECLARE_STATIC_KEY_FALSE(arch_scale_freq_key);
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index ffbd9a3d78d84..18d292fc466cb 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -56,6 +56,7 @@
 #include <linux/cpuidle.h>
 #include <linux/numa.h>
 #include <linux/pgtable.h>
+#include <linux/overflow.h>
 
 #include <asm/acpi.h>
 #include <asm/desc.h>
@@ -1777,6 +1778,7 @@ void native_play_dead(void)
 
 #endif
 
+#ifdef CONFIG_X86_64
 /*
  * APERF/MPERF frequency ratio computation.
  *
@@ -2048,11 +2050,19 @@ static void init_freq_invariance(bool secondary)
 	}
 }
 
+static void disable_freq_invariance_workfn(struct work_struct *work)
+{
+	static_branch_disable(&arch_scale_freq_key);
+}
+
+static DECLARE_WORK(disable_freq_invariance_work,
+		    disable_freq_invariance_workfn);
+
 DEFINE_PER_CPU(unsigned long, arch_freq_scale) = SCHED_CAPACITY_SCALE;
 
 void arch_scale_freq_tick(void)
 {
-	u64 freq_scale;
+	u64 freq_scale = SCHED_CAPACITY_SCALE;
 	u64 aperf, mperf;
 	u64 acnt, mcnt;
 
@@ -2064,19 +2074,32 @@ void arch_scale_freq_tick(void)
 
 	acnt = aperf - this_cpu_read(arch_prev_aperf);
 	mcnt = mperf - this_cpu_read(arch_prev_mperf);
-	if (!mcnt)
-		return;
 
 	this_cpu_write(arch_prev_aperf, aperf);
 	this_cpu_write(arch_prev_mperf, mperf);
 
-	acnt <<= 2*SCHED_CAPACITY_SHIFT;
-	mcnt *= arch_max_freq_ratio;
+	if (check_shl_overflow(acnt, 2*SCHED_CAPACITY_SHIFT, &acnt))
+		goto error;
+
+	if (check_mul_overflow(mcnt, arch_max_freq_ratio, &mcnt) || !mcnt)
+		goto error;
 
 	freq_scale = div64_u64(acnt, mcnt);
+	if (!freq_scale)
+		goto error;
 
 	if (freq_scale > SCHED_CAPACITY_SCALE)
 		freq_scale = SCHED_CAPACITY_SCALE;
 
 	this_cpu_write(arch_freq_scale, freq_scale);
+	return;
+
+error:
+	pr_warn("Scheduler frequency invariance went wobbly, disabling!\n");
+	schedule_work(&disable_freq_invariance_work);
+}
+#else
+static inline void init_freq_invariance(bool secondary)
+{
 }
+#endif /* CONFIG_X86_64 */
-- 
2.25.1


  parent reply	other threads:[~2020-08-08 23:35 UTC|newest]

Thread overview: 74+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-08 23:34 [PATCH AUTOSEL 5.8 01/72] ARM: dts: stm32: fix uart nodes ordering in stm32mp15-pinctrl Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 02/72] ARM: dts: stm32: fix uart7_pins_a comments " Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 03/72] x86/mce/inject: Fix a wrong assignment of i_mce.status Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 04/72] perf/x86/intel/uncore: Fix oops when counting IMC uncore events on some TGL Sasha Levin
2020-08-08 23:34 ` Sasha Levin [this message]
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 06/72] x86, sched: Bail out of frequency invariance if turbo frequency is unknown Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 07/72] x86, sched: Bail out of frequency invariance if turbo_freq/base_freq gives 0 Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 08/72] sched/fair: Fix NOHZ next idle balance Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 09/72] sched: correct SD_flags returned by tl->sd_flags() Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 10/72] arm64: dts: rockchip: fix rk3368-lion gmac reset gpio Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 11/72] arm64: dts: rockchip: fix rk3399-puma vcc5v0-host gpio Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 12/72] arm64: dts: rockchip: fix rk3399-puma gmac reset gpio Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 13/72] EDAC: Fix reference count leaks Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 14/72] crc-t10dif: Fix potential crypto notify dead-lock Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 15/72] arm64: dts: qcom: msm8916: Replace invalid bias-pull-none property Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 16/72] memory: tegra: Fix an error handling path in tegra186_emc_probe() Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 17/72] blktrace: fix debugfs use after free Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 18/72] soc: qcom: rpmh-rsc: Don't use ktime for timeout in write_tcs_reg_sync() Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 19/72] crypto: ccree - fix resource leak on error path Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 20/72] ARM: exynos: MCPM: Restore big.LITTLE cpuidle support Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 21/72] rcu/tree: Repeat the monitor if any free channel is busy Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 22/72] firmware: arm_scmi: Fix SCMI genpd domain probing Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 23/72] arm64: dts: sun50i-pinephone: dldo4 must not be >= 1.8V Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 24/72] arm64: dts: exynos: Fix silent hang after boot on Espresso Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 25/72] sched/uclamp: Fix initialization of struct uclamp_rq Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 26/72] crypto: qat - allow xts requests not multiple of block Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 27/72] clk: scmi: Fix min and max rate when registering clocks with discrete rates Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 28/72] m68k: mac: Don't send IOP message until channel is idle Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 29/72] m68k: mac: Fix IOP status/control register writes Sasha Levin
2020-08-08 23:34 ` [PATCH AUTOSEL 5.8 30/72] platform/x86: intel-hid: Fix return value check in check_acpi_dev() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 31/72] platform/x86: intel-vbtn: " Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 32/72] ARM: dts: gose: Fix ports node name for adv7180 Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 33/72] arm64: dts: renesas: Fix SD Card/eMMC interface device node names Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 34/72] ARM: dts: gose: Fix ports node name for adv7612 Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 35/72] ARM: at91: pm: add missing put_device() call in at91_pm_sram_init() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 36/72] ARM: dts: exynos: Disable frequency scaling for FSYS bus on Odroid XU3 family Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 37/72] reset: intel: fix a compile warning about REG_OFFSET redefined Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 38/72] ARM: dts: at91: sama5d3_xplained: change phy-mode Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 39/72] ARM: dts: sunxi: bananapi-m2-plus-v1.2: Add regulator supply to all CPU cores Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 40/72] ARM: dts: sunxi: bananapi-m2-plus-v1.2: Fix CPU supply voltages Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 41/72] ARM: dts: stm32: Fix spi4 pins in stm32mp15-pinctrl Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 42/72] spi: dw-dma: Fix Tx DMA channel working too fast Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 43/72] spi: lantiq: fix: Rx overflow error in full duplex mode Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 44/72] crypto: x86/crc32c - fix building with clang ias Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 45/72] tpm: Require that all digests are present in TCG_PCR_EVENT2 structures Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 46/72] recordmcount: only record relocation of type R_AARCH64_CALL26 on arm64 Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 47/72] regulator: fix memory leak on error path of regulator_register() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 48/72] io_uring: fix sq array offset calculation Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 49/72] arm64: dts: meson: misc fixups for w400 dtsi Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 50/72] arm64: dts: meson: fix mmc0 tuning error on Khadas VIM3 Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 51/72] soc: qcom: pdr: Reorder the PD state indication ack Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 52/72] spi: rockchip: Fix error in SPI slave pio read Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 53/72] ARM: socfpga: PM: add missing put_device() call in socfpga_setup_ocram_self_refresh() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 54/72] iocost: Fix check condition of iocg abs_vdebt Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 55/72] scripts/selinux/mdp: fix initial SID handling Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 56/72] irqchip/ti-sci-inta: Fix return value about devm_ioremap_resource() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 57/72] seccomp: Fix ioctl number for SECCOMP_IOCTL_NOTIF_ID_VALID Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 58/72] md: raid0/linear: fix dereference before null check on pointer mddev Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 59/72] irqchip/loongson-htvec: Fix potential resource leak Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 60/72] irqchip/loongson-htvec: Check return value of irq_domain_translate_onecell() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 61/72] irqchip/loongson-pch-pic: Check return value of irq_domain_translate_twocell() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 62/72] irqchip/loongson-liointc: Fix potential dead lock Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 63/72] kunit: tool: fix broken default args in unit tests Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 64/72] kunit: tool: fix improper treatment of file location Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 65/72] irqchip/irq-bcm7038-l1: Guard uses of cpu_logical_map Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 66/72] irqchip/gic-v4.1: Use GFP_ATOMIC flag in allocate_vpe_l1_table() Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 67/72] nvme-tcp: fix controller reset hang during traffic Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 68/72] nvme-rdma: " Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 69/72] nvme-multipath: fix logic for non-optimized paths Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 70/72] nvme-multipath: do not fall back to __nvme_find_path() " Sasha Levin
2020-08-10 15:37   ` Martin Wilck
2020-08-16 13:50     ` Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 71/72] irqchip/loongson-pch-pic: Fix the misused irq flow handler Sasha Levin
2020-08-08 23:35 ` [PATCH AUTOSEL 5.8 72/72] block: don't do revalidate zones on invalid devices Sasha Levin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200808233542.3617339-5-sashal@kernel.org \
    --to=sashal@kernel.org \
    --cc=ggherdovich@suse.cz \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peterz@infradead.org \
    --cc=rafael.j.wysocki@intel.com \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox