[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <158557963557.22376.16443742264836974381.tglx@nanos.tec.linutronix.de>
Date: Mon, 30 Mar 2020 14:47:15 -0000
From: Thomas Gleixner <tglx@...utronix.de>
To: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: linux-kernel@...r.kernel.org, x86@...nel.org
Subject: [GIT pull] x86/splitlock for v5.7
Linus,
please pull the latest x86/splitlock branch from:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git x86-splitlock-2020-03-30
up to: a6a60741035b: x86/split_lock: Avoid runtime reads of the TEST_CTRL MSR
Support for "split lock" detection:
- Atomic operations (lock prefixed instructions) which span two cache
lines have to acquire the global bus lock. This is at least 1k cycles
slower than an atomic operation within a cache line and disrupts
performance on other cores. Aside of performance disruption this is
a unpriviledged form of DoS.
Some newer CPUs have the capability to raise an #AC trap when such an
operation is attempted. The detection is by default enabled in warning
mode which will warn once when a user space application is caught. A
command line option allows to disable the detection or to select fatal
mode which will terminate offending applications with SIGBUS.
Thanks,
tglx
------------------>
Peter Zijlstra (Intel) (1):
x86/split_lock: Enable split lock detection by kernel
Xiaoyao Li (2):
x86/split_lock: Rework the initialization flow of split lock detection
x86/split_lock: Avoid runtime reads of the TEST_CTRL MSR
Documentation/admin-guide/kernel-parameters.txt | 22 +++
arch/x86/include/asm/cpu.h | 12 ++
arch/x86/include/asm/cpufeatures.h | 2 +
arch/x86/include/asm/msr-index.h | 9 ++
arch/x86/include/asm/thread_info.h | 4 +-
arch/x86/kernel/cpu/common.c | 2 +
arch/x86/kernel/cpu/intel.c | 183 ++++++++++++++++++++++++
arch/x86/kernel/process.c | 3 +
arch/x86/kernel/traps.c | 24 +++-
9 files changed, 258 insertions(+), 3 deletions(-)
Powered by blists - more mailing lists