Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Vitaly Kuznetsov <vkuznets@redhat.com>
To:     linux-hyperv@vger.kernel.org
Cc:     linux-kernel@vger.kernel.org, x86@kernel.org,
        "K. Y. Srinivasan" <kys@microsoft.com>,
        Haiyang Zhang <haiyangz@microsoft.com>,
        Stephen Hemminger <sthemmin@microsoft.com>,
        Sasha Levin <sashal@kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        "H. Peter Anvin" <hpa@zytor.com>,
        Roman Kagan <rkagan@virtuozzo.com>,
        Michael Kelley <mikelley@microsoft.com>,
        Joe Perches <joe@perches.com>
Subject: [PATCH v2] x86/hyper-v: micro-optimize send_ipi_one case
Date:   Fri, 25 Oct 2019 15:15:46 +0200
Message-Id: <20191025131546.18794-1-vkuznets@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=WINDOWS-1252
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

When sending an IPI to a single CPU there is no need to deal with cpumasks.
With 2 CPU guest on WS2019 I'm seeing a minor (like 3%, 8043 -> 7761 CPU
cycles) improvement with smp_call_function_single() loop benchmark. The
optimization, however, is tiny and straitforward. Also, send_ipi_one() is
important for PV spinlock kick.

I was also wondering if it would make sense to switch to using regular
APIC IPI send for CPU > 64 case but no, it is twice as expesive (12650 CPU
cycles for __send_ipi_mask_ex() call, 26000 for orig_apic.send_IPI(cpu,
vector)).

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
---
Changes since v1:
 - Style changes [Roman, Joe]
---
 arch/x86/hyperv/hv_apic.c           | 13 ++++++++++---
 arch/x86/include/asm/trace/hyperv.h | 15 +++++++++++++++
 2 files changed, 25 insertions(+), 3 deletions(-)

diff --git a/arch/x86/hyperv/hv_apic.c b/arch/x86/hyperv/hv_apic.c
index e01078e93dd3..fd17c6341737 100644
--- a/arch/x86/hyperv/hv_apic.c
+++ b/arch/x86/hyperv/hv_apic.c
@@ -194,10 +194,17 @@ static bool __send_ipi_mask(const struct cpumask *mas=
k, int vector)
=20
 static bool __send_ipi_one(int cpu, int vector)
 {
-=09struct cpumask mask =3D CPU_MASK_NONE;
+=09trace_hyperv_send_ipi_one(cpu, vector);
=20
-=09cpumask_set_cpu(cpu, &mask);
-=09return __send_ipi_mask(&mask, vector);
+=09if (!hv_hypercall_pg || (vector < HV_IPI_LOW_VECTOR) ||
+=09    (vector > HV_IPI_HIGH_VECTOR))
+=09=09return false;
+
+=09if (cpu >=3D 64)
+=09=09return __send_ipi_mask_ex(cpumask_of(cpu), vector);
+
+=09return !hv_do_fast_hypercall16(HVCALL_SEND_IPI, vector,
+=09=09=09       BIT_ULL(hv_cpu_number_to_vp_number(cpu)));
 }
=20
 static void hv_send_ipi(int cpu, int vector)
diff --git a/arch/x86/include/asm/trace/hyperv.h b/arch/x86/include/asm/tra=
ce/hyperv.h
index ace464f09681..4d705cb4d63b 100644
--- a/arch/x86/include/asm/trace/hyperv.h
+++ b/arch/x86/include/asm/trace/hyperv.h
@@ -71,6 +71,21 @@ TRACE_EVENT(hyperv_send_ipi_mask,
 =09=09      __entry->ncpus, __entry->vector)
 =09);
=20
+TRACE_EVENT(hyperv_send_ipi_one,
+=09    TP_PROTO(int cpu,
+=09=09     int vector),
+=09    TP_ARGS(cpu, vector),
+=09    TP_STRUCT__entry(
+=09=09    __field(int, cpu)
+=09=09    __field(int, vector)
+=09=09    ),
+=09    TP_fast_assign(__entry->cpu =3D cpu;
+=09=09=09   __entry->vector =3D vector;
+=09=09    ),
+=09    TP_printk("cpu %d vector %x",
+=09=09      __entry->cpu, __entry->vector)
+=09);
+
 #endif /* CONFIG_HYPERV */
=20
 #undef TRACE_INCLUDE_PATH
--=20
2.20.1