From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:35629) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bjPBj-00026Z-3j for qemu-devel@nongnu.org; Mon, 12 Sep 2016 07:14:12 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1bjPBf-0007gu-Od for qemu-devel@nongnu.org; Mon, 12 Sep 2016 07:14:11 -0400 Received: from mail-wm0-f67.google.com ([74.125.82.67]:36193) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bjPBf-0007bf-IF for qemu-devel@nongnu.org; Mon, 12 Sep 2016 07:14:07 -0400 Received: by mail-wm0-f67.google.com with SMTP id z194so749459wmd.3 for ; Mon, 12 Sep 2016 04:13:46 -0700 (PDT) Sender: Paolo Bonzini From: Paolo Bonzini Date: Mon, 12 Sep 2016 13:12:25 +0200 Message-Id: <1473678761-8885-1-git-send-email-pbonzini@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Subject: [Qemu-devel] [PATCH v7 00/16] cpu-exec: Safe work in quiescent state List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: sergey.fedorov@linaro.org, alex.bennee@linaro.org In addition to fixing some of the issues found by Alex, safe work items need not run anymore with a mutex taken. Of course, cpu_exec_start/end and start_exclusive/end_exclusive are essentially the read and write side of a specialized rwlock, so there is still a lock in disguise looming to cause deadlocks; however, it does removes worries about recursive locking from CPU list manipulations. The new patches are 8, 12 and 13. Patch 12 of v6 has been split across patch 10 and patch 16. Paolo v6->v7: Do not separate qemu_work_item and SafeWorkItem More cleanups/optimizations of exclusive section logic Alex Bennée (1): cpus: pass CPUState to run_on_cpu helpers Paolo Bonzini (9): cpus-common: move CPU list management to common code cpus-common: fix uninitialized variable use in run_on_cpu cpus-common: move exclusive work infrastructure from linux-user docs: include formal model for TCG exclusive sections cpus-common: always defer async_run_on_cpu work items cpus-common: remove redundant call to exclusive_idle() cpus-common: simplify locking for start_exclusive/end_exclusive cpus-common: Introduce async_safe_run_on_cpu() cpus-common: lock-free fast path for cpu_exec_start/end Sergey Fedorov (6): cpus: Move common code out of {async_, }run_on_cpu() cpus: Rename flush_queued_work() linux-user: Use QemuMutex and QemuCond linux-user: Add qemu_cpu_is_self() and qemu_cpu_kick() cpus-common: move CPU work item management to common code tcg: Make tb_flush() thread safe Makefile.target | 2 +- bsd-user/main.c | 33 ++--- cpu-exec.c | 12 +- cpus-common.c | 335 +++++++++++++++++++++++++++++++++++++++++++++ cpus.c | 99 +------------- docs/tcg-exclusive.promela | 224 ++++++++++++++++++++++++++++++ exec.c | 37 +---- hw/i386/kvm/apic.c | 3 +- hw/i386/kvmvapic.c | 6 +- hw/ppc/ppce500_spin.c | 31 ++--- hw/ppc/spapr.c | 6 +- hw/ppc/spapr_hcall.c | 17 +-- include/exec/cpu-all.h | 4 + include/exec/cpu-common.h | 2 + include/exec/exec-all.h | 11 -- include/exec/tb-context.h | 2 +- include/qom/cpu.h | 99 ++++++++++++-- kvm-all.c | 21 +-- linux-user/main.c | 130 ++++++------------ target-i386/helper.c | 19 ++- target-i386/kvm.c | 6 +- target-s390x/cpu.c | 4 +- target-s390x/cpu.h | 7 +- target-s390x/kvm.c | 98 ++++++------- target-s390x/misc_helper.c | 4 +- translate-all.c | 38 +++-- vl.c | 1 + 27 files changed, 836 insertions(+), 415 deletions(-) create mode 100644 cpus-common.c create mode 100644 docs/tcg-exclusive.promela -- 2.7.4