Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
Date:   Tue,  7 Jun 2022 09:50:42 -0700
Message-Id: <20220607165105.639716-1-kaleshsingh@google.com>
Mime-Version: 1.0
Subject: [PATCH v3 0/5] KVM nVHE Hypervisor stack unwinder
From:   Kalesh Singh <kaleshsingh@google.com>
To:     mark.rutland@arm.com, broonie@kernel.org, maz@kernel.org
Cc:     will@kernel.org, qperret@google.com, tabba@google.com,
        surenb@google.com, tjmercier@google.com, kernel-team@android.com,
        Kalesh Singh <kaleshsingh@google.com>,
        James Morse <james.morse@arm.com>,
        Alexandru Elisei <alexandru.elisei@arm.com>,
        Suzuki K Poulose <suzuki.poulose@arm.com>,
        Catalin Marinas <catalin.marinas@arm.com>,
        Masami Hiramatsu <mhiramat@kernel.org>,
        Alexei Starovoitov <ast@kernel.org>,
        "Madhavan T. Venkataraman" <madvenka@linux.microsoft.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Andrew Jones <drjones@redhat.com>,
        Kefeng Wang <wangkefeng.wang@huawei.com>,
        Keir Fraser <keirf@google.com>,
        Zenghui Yu <yuzenghui@huawei.com>,
        Ard Biesheuvel <ardb@kernel.org>,
        Oliver Upton <oupton@google.com>,
        linux-arm-kernel@lists.infradead.org, kvmarm@lists.cs.columbia.edu,
        linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi all,

This v3 of the nVHE hypervisor stack unwinder. The previous versions were
posted at:
v2: https://lore.kernel.org/r/20220502191222.4192768-1-kaleshsingh@google.com/
v1: https://lore.kernel.org/r/20220427184716.1949239-1-kaleshsingh@google.com/

The version is rebased on 5.19-rc1 which resolves the previously unmerged
dependencies, and adds some reviewed tags from Mark Brown.

The cover letter are copied below for convenience.

Thanks,
Kalesh

-------

This new version of the unwinder splits the unwinding and dumping
of the stack between the hypervisor and host:
  - The hypervisor unwinds its stack and dumps the address entries
    into a page shared with the host.
  - The host then symnolizes and prints the hyp stacktrace from
    the shared page.

The new approach doesn't depend on CONFIG_NVHE_EL2_DEBUG,
and allows dumping hyp stacktraces in prodcution environments
(!CONFIG_NVHE_EL2_DEBUG).

arm64/kernel/stacktrace.c is compiled twice: stacktrace.o for the
host kernel and stacktrace.nvhe.o for the hypervisor. This allows
reusing most of the host unwinding logic in the nVHE hypervisor.

Kalesh Singh (5):
  KVM: arm64: Factor out common stack unwinding logic
  KVM: arm64: Compile stacktrace.nvhe.o
  KVM: arm64: Add hypervisor overflow stack
  KVM: arm64: Allocate shared stacktrace pages
  KVM: arm64: Unwind and dump nVHE hypervisor stacktrace

 arch/arm64/include/asm/kvm_asm.h    |   1 +
 arch/arm64/include/asm/stacktrace.h |  58 +++++++++--
 arch/arm64/kernel/stacktrace.c      | 151 +++++++++++++++++++++++-----
 arch/arm64/kvm/arm.c                |  34 +++++++
 arch/arm64/kvm/handle_exit.c        |   4 +
 arch/arm64/kvm/hyp/nvhe/Makefile    |   3 +-
 arch/arm64/kvm/hyp/nvhe/host.S      |   9 +-
 arch/arm64/kvm/hyp/nvhe/setup.c     |  11 ++
 arch/arm64/kvm/hyp/nvhe/switch.c    |   4 +
 9 files changed, 231 insertions(+), 44 deletions(-)


base-commit: f2906aa863381afb0015a9eb7fefad885d4e5a56
-- 
2.36.1.255.ge46751e96f-goog