qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Zhao Liu <zhao1.liu@linux.intel.com>
To: "wangyanan (Y)" <wangyanan55@huawei.com>
Cc: qemu-devel@nongnu.org, "Zhenyu Wang" <zhenyu.z.wang@intel.com>,
	"Dapeng Mi" <dapeng1.mi@intel.com>,
	"Zhuocheng Ding" <zhuocheng.ding@intel.com>,
	"Robert Hoo" <robert.hu@linux.intel.com>,
	"Sean Christopherson" <seanjc@google.com>,
	"Like Xu" <like.xu.linux@gmail.com>,
	"Zhao Liu" <zhao1.liu@intel.com>,
	"Eduardo Habkost" <eduardo@habkost.net>,
	"Marcel Apfelbaum" <marcel.apfelbaum@gmail.com>,
	"Philippe Mathieu-Daudé" <philmd@linaro.org>,
	"Michael S . Tsirkin" <mst@redhat.com>,
	"Richard Henderson" <richard.henderson@linaro.org>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"Eric Blake" <eblake@redhat.com>,
	"Markus Armbruster" <armbru@redhat.com>
Subject: Re: [RFC 08/52] machine: Add helpers to get cpu topology info from MachineState.topo
Date: Fri, 17 Feb 2023 17:07:50 +0800	[thread overview]
Message-ID: <Y+9D5oFoa9WH6zBq@liuzhao-OptiPlex-7080> (raw)
In-Reply-To: <a6cb0be8-3eff-31b6-b5a9-66f0ec6dcb8d@huawei.com>

On Fri, Feb 17, 2023 at 03:41:30PM +0800, wangyanan (Y) wrote:
> Date: Fri, 17 Feb 2023 15:41:30 +0800
> From: "wangyanan (Y)" <wangyanan55@huawei.com>
> Subject: Re: [RFC 08/52] machine: Add helpers to get cpu topology info from
>  MachineState.topo
> 
> 在 2023/2/17 11:07, Zhao Liu 写道:
> > On Thu, Feb 16, 2023 at 04:38:38PM +0800, wangyanan (Y) wrote:
> > > Date: Thu, 16 Feb 2023 16:38:38 +0800
> > > From: "wangyanan (Y)" <wangyanan55@huawei.com>
> > > Subject: Re: [RFC 08/52] machine: Add helpers to get cpu topology info from
> > >   MachineState.topo
> > > 
> > > Hi Zhao,
> > > 
> > > 在 2023/2/13 17:49, Zhao Liu 写道:
> > > > From: Zhao Liu <zhao1.liu@intel.com>
> > > > 
> > > > When MachineState.topo is introduced, the topology related structures
> > > > become complicated. In the general case (hybrid or smp topology),
> > > > accessing the topology information needs to determine whether it is
> > > > currently smp or hybrid topology, and then access the corresponding
> > > > MachineState.topo.smp or MachineState.topo.hybrid.
> > > > 
> > > > The best way to do this is to wrap the access to the topology to
> > > > avoid having to check each time it is accessed.
> > > > 
> > > > The following helpers are provided here:
> > > > 
> > > > - General interfaces - no need to worry about whether the underlying
> > > >     topology is smp or hybrid:
> > > > 
> > > > * machine_topo_get_cpus()
> > > > * machine_topo_get_max_cpus()
> > > > * machine_topo_is_smp()
> > > > * machine_topo_get_sockets()
> > > > * machine_topo_get_dies()
> > > > * machine_topo_get_clusters()
> > > > * machine_topo_get_threads();
> > > > * machine_topo_get_cores();
> > > > * machine_topo_get_threads_by_idx()
> > > > * machine_topo_get_cores_by_idx()
> > > > * machine_topo_get_cores_per_socket()
> > > > * machine_topo_get_threads_per_socket()
> > > > 
> > > > - SMP-specific interfaces - provided for the cases that are clearly
> > > > known to be smp topology:
> > > > 
> > > > * machine_topo_get_smp_cores()
> > > > * machine_topo_get_smp_threads()
> > > > 
> > > > Since for hybrid topology, each core may has different threads, if
> > > > someone wants "cpus per core", the cpu_index is need to target a
> > > > specific core (machine_topo_get_threads_by_idx()). But for smp, there is
> > > > no need to be so troublesome, so for this case, we provide smp-specific
> > > > interfaces.
> > > > 
> > > > Signed-off-by: Zhao Liu <zhao1.liu@intel.com>
> > > > ---
> > > >    hw/core/machine-topo.c | 142 +++++++++++++++++++++++++++++++++++++++++
> > > >    include/hw/boards.h    |  35 ++++++++++
> > > >    2 files changed, 177 insertions(+)
> > > > 
> > > > diff --git a/hw/core/machine-topo.c b/hw/core/machine-topo.c
> > > > index 7223f73f99b0..b20160479629 100644
> > > > --- a/hw/core/machine-topo.c
> > > > +++ b/hw/core/machine-topo.c
> > > > @@ -21,6 +21,148 @@
> > > >    #include "hw/boards.h"
> > > >    #include "qapi/error.h"
> > > > +unsigned int machine_topo_get_sockets(const MachineState *ms)
> > > > +{
> > > > +    return machine_topo_is_smp(ms) ? ms->topo.smp.sockets :
> > > > +                                     ms->topo.hybrid.sockets;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_dies(const MachineState *ms)
> > > > +{
> > > > +    return machine_topo_is_smp(ms) ? ms->topo.smp.dies :
> > > > +                                     ms->topo.hybrid.dies;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_clusters(const MachineState *ms)
> > > > +{
> > > > +    return machine_topo_is_smp(ms) ? ms->topo.smp.clusters :
> > > > +                                     ms->topo.hybrid.clusters;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_smp_cores(const MachineState *ms)
> > > > +{
> > > > +    g_assert(machine_topo_is_smp(ms));
> > > > +    return ms->topo.smp.cores;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_smp_threads(const MachineState *ms)
> > > > +{
> > > > +    g_assert(machine_topo_is_smp(ms));
> > > > +    return ms->topo.smp.threads;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_threads(const MachineState *ms,
> > > > +                                      unsigned int cluster_id,
> > > > +                                      unsigned int core_id)
> > > > +{
> > > > +    if (machine_topo_is_smp(ms)) {
> > > > +        return ms->topo.smp.threads;
> > > > +    } else {
> > > > +        return ms->topo.hybrid.cluster_list[cluster_id]
> > > > +                   .core_list[core_id].threads;
> > > > +    }
> > > > +
> > > > +    return 0;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_cores(const MachineState *ms,
> > > > +                                    unsigned int cluster_id)
> > > > +{
> > > > +    if (machine_topo_is_smp(ms)) {
> > > > +        return ms->topo.smp.cores;
> > > > +    } else {
> > > > +        return ms->topo.hybrid.cluster_list[cluster_id].cores;
> > > > +    }
> > > > +}
> > > Is it possible to use variadic function so that those two smp specific
> > > helpers can be avoided? It's a bit wired that we have the generic
> > > machine_topo_get_threads but also need machine_topo_get_smp_threads
> > > at the same time.
> > I am not sure about this, because variadic functions unify function
> > naming, but eliminate the "smp-specific" information from the name.
> > 
> > Trying to get the cres/threads without considering the cpu index can
> > only be used in smp scenarios, and I think the caller needs to
> > understand that he knows it's smp.
> Ok, I get the point.
> When it comes to the naming, would it be more concise to remove the
> *_get_* in the fun name, such as machine_topo_get_cpus to
> machine_topo_cpus, machine_topo_get_clusters to machine_topo_clusters.

Good, thanks!

> 
> And maybe rename machine_topo_get_cores(int cluster_id, int core_id) to
> machine_topo_cores_by_ids?
> 
> Or machine_topo_get_cores() to machine_topo_cores_by_topo_ids()
> and machine_topo_get_cores_by_idx to machine_topo_cores_by_cpu_idx()

I like the latter, nice name.

> > > > +
> > > > +unsigned int machine_topo_get_threads_by_idx(const MachineState *ms,
> > > > +                                             unsigned int cpu_index)
> > > > +{
> > > > +    unsigned cpus_per_die;
> > > > +    unsigned tmp_idx;
> > > > +    HybridCluster *cluster;
> > > > +    HybridCore *core;
> > > > +
> > > > +    if (machine_topo_is_smp(ms)) {
> > > > +        return ms->topo.smp.threads;
> > > > +    }
> > > > +
> > > > +    cpus_per_die = ms->topo.max_cpus / (ms->topo.hybrid.sockets *
> > > > +                            ms->topo.hybrid.dies);
> > > > +    tmp_idx = cpu_index % cpus_per_die;
> > > > +
> > > > +    for (int i = 0; i < ms->topo.hybrid.clusters; i++) {
> > > > +        cluster = &ms->topo.hybrid.cluster_list[i];
> > > > +
> > > > +        for (int j = 0; j < cluster->cores; j++) {
> > > > +            core = &cluster->core_list[j];
> > > > +
> > > > +            if (tmp_idx < core->threads) {
> > > > +                return core->threads;
> > > > +            } else {
> > > > +                tmp_idx -= core->threads;
> > > > +            }
> > > > +        }
> > > > +    }
> > > > +
> > > > +    return 0;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_cores_by_idx(const MachineState *ms,
> > > > +                                           unsigned int cpu_index)
> > > > +{
> > > > +    unsigned cpus_per_die;
> > > > +    unsigned tmp_idx;
> > > > +    HybridCluster *cluster;
> > > > +    HybridCore *core;
> > > > +
> > > > +    if (machine_topo_is_smp(ms)) {
> > > > +        return ms->topo.smp.cores;
> > > > +    }
> > > > +
> > > > +    cpus_per_die = ms->topo.max_cpus / (ms->topo.hybrid.sockets *
> > > > +                            ms->topo.hybrid.dies);
> > > > +    tmp_idx = cpu_index % cpus_per_die;
> > > > +
> > > > +    for (int i = 0; i < ms->topo.hybrid.clusters; i++) {
> > > > +        cluster = &ms->topo.hybrid.cluster_list[i];
> > > > +
> > > > +        for (int j = 0; j < cluster->cores; j++) {
> > > > +            core = &cluster->core_list[j];
> > > > +
> > > > +            if (tmp_idx < core->threads) {
> > > > +                return cluster->cores;
> > > > +            } else {
> > > > +                tmp_idx -= core->threads;
> > > > +            }
> > > > +        }
> > > > +    }
> > > > +
> > > > +    return 0;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_cores_per_socket(const MachineState *ms)
> > > > +{
> > > > +    unsigned int cores_per_die = 0;
> > > > +
> > > > +    if (machine_topo_is_smp(ms)) {
> > > > +        return ms->topo.smp.cores * ms->topo.smp.clusters * ms->topo.smp.dies;
> > > > +    }
> > > > +
> > > > +    for (int i = 0; i < ms->topo.hybrid.clusters; i++) {
> > > > +        cores_per_die += ms->topo.hybrid.cluster_list[i].cores;
> > > > +    }
> > > > +
> > > > +    return cores_per_die * ms->topo.hybrid.dies;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_threads_per_socket(const MachineState *ms)
> > > > +{
> > > > +    unsigned int sockets = machine_topo_is_smp(ms) ? ms->topo.smp.sockets :
> > > > +                           ms->topo.hybrid.sockets;
> > > > +    return ms->topo.max_cpus / sockets;
> > > > +}
> > > > +
> > > >    /*
> > > >     * Report information of a machine's supported CPU topology hierarchy.
> > > >     * Topology members will be ordered from the largest to the smallest
> > > > diff --git a/include/hw/boards.h b/include/hw/boards.h
> > > > index 0a61855499e3..34b64b012022 100644
> > > > --- a/include/hw/boards.h
> > > > +++ b/include/hw/boards.h
> > > > @@ -461,4 +461,39 @@ extern const size_t hw_compat_2_2_len;
> > > >    extern GlobalProperty hw_compat_2_1[];
> > > >    extern const size_t hw_compat_2_1_len;
> > > > +static inline
> > > > +unsigned int machine_topo_get_cpus(const MachineState *ms)
> > > > +{
> > > > +    return ms->topo.cpus;
> > > > +}
> > > > +
> > > > +static inline
> > > > +unsigned int machine_topo_get_max_cpus(const MachineState *ms)
> > > > +{
> > > > +    return ms->topo.max_cpus;
> > > > +}
> > > > +
> > > > +static inline
> > > > +bool machine_topo_is_smp(const MachineState *ms)
> > > > +{
> > > > +    return ms->topo.topo_type == CPU_TOPO_TYPE_SMP;
> > > > +}
> > > > +
> > > > +unsigned int machine_topo_get_sockets(const MachineState *ms);
> > > > +unsigned int machine_topo_get_dies(const MachineState *ms);
> > > > +unsigned int machine_topo_get_clusters(const MachineState *ms);
> > > > +unsigned int machine_topo_get_smp_cores(const MachineState *ms);
> > > > +unsigned int machine_topo_get_smp_threads(const MachineState *ms);
> > > > +unsigned int machine_topo_get_threads(const MachineState *ms,
> > > > +                                      unsigned int cluster_id,
> > > > +                                      unsigned int core_id);
> > > > +unsigned int machine_topo_get_cores(const MachineState *ms,
> > > > +                                    unsigned int cluster_id);
> > > > +unsigned int machine_topo_get_threads_by_idx(const MachineState *ms,
> > > > +                                             unsigned int cpu_index);
> > > > +unsigned int machine_topo_get_cores_by_idx(const MachineState *ms,
> > > > +                                           unsigned int cpu_index);
> > > > +unsigned int machine_topo_get_cores_per_socket(const MachineState *ms);
> > > > +unsigned int machine_topo_get_threads_per_socket(const MachineState *ms);
> > > > +
> > > >    #endif
> > > I think it's necessary to document the ablity for each helper.
> > > For example, at a flance, I cant figure out what
> > > machine_topo_get_threads_idx
> > > does. Add some something like:
> > > /*
> > >   * Get number of threads within the CPU core where a processor locates,
> > >   * according to the processor index.
> > >   *
> > >   * @param: ...
> > >   */
> > > will be friendly to future users.
> > Yeah, thanks! I will.
> > 
> > > Thanks,
> > > Yanan
> 


  reply	other threads:[~2023-02-17  9:00 UTC|newest]

Thread overview: 113+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-02-13  9:49 [RFC 00/52] Introduce hybrid CPU topology Zhao Liu
2023-02-13  9:49 ` [RFC 01/52] hw/smbios: Fix smbios_smp_sockets caculation Zhao Liu
2023-02-13  9:49 ` [RFC 02/52] hw/smbios: Fix thread count in type4 Zhao Liu
2023-02-13  9:49 ` [RFC 03/52] hw/smbios: Fix core " Zhao Liu
2023-02-13  9:49 ` [RFC 04/52] i386/WHPX: Fix error message when fail to set ProcessorCount Zhao Liu
2023-02-13 13:41   ` Daniel P. Berrangé
2023-02-15  2:29     ` Zhao Liu
2023-02-13  9:49 ` [RFC 05/52] hw/core/machine: Rename machine-smp.c to machine-topo.c Zhao Liu
2023-02-13 12:52   ` wangyanan (Y) via
2023-02-14  8:50     ` Zhao Liu
2023-02-13  9:49 ` [RFC 06/52] hw/cpu: Introduce hybrid CPU topology Zhao Liu
2023-02-13 13:10   ` Philippe Mathieu-Daudé
2023-02-14  9:30     ` Zhao Liu
2023-02-14  9:27       ` Philippe Mathieu-Daudé
2023-02-15  3:15         ` Zhao Liu
2023-02-13 13:18   ` wangyanan (Y) via
2023-02-14 10:16     ` Zhao Liu
2023-02-14 11:23       ` wangyanan (Y) via
2023-02-15  3:22         ` Zhao Liu
2023-02-13  9:49 ` [RFC 07/52] hw/core/machine: Add the new topology support in MachineState Zhao Liu
2023-02-13  9:49 ` [RFC 08/52] machine: Add helpers to get cpu topology info from MachineState.topo Zhao Liu
2023-02-14  1:12   ` Mi, Dapeng1
2023-02-15  2:31     ` Zhao Liu
2023-02-16  8:38   ` wangyanan (Y) via
2023-02-17  3:07     ` Zhao Liu
2023-02-17  7:41       ` wangyanan (Y) via
2023-02-17  9:07         ` Zhao Liu [this message]
2023-02-13  9:49 ` [RFC 09/52] hw/machine: Introduce core type for hybrid topology Zhao Liu
2023-02-13 13:13   ` Philippe Mathieu-Daudé
2023-02-14  9:41     ` Zhao Liu
2023-02-14  1:14   ` Mi, Dapeng1
2023-02-15  2:40     ` Zhao Liu
2023-02-13  9:49 ` [RFC 10/52] machine: Replace MachineState.topo.smp access with topology helpers Zhao Liu
2023-02-13  9:49 ` [RFC 11/52] accel/kvm: Add hybrid info when check cpu num Zhao Liu
2023-02-13  9:49 ` [RFC 12/52] hw/acpi: Replace MachineState.smp access with topology helpers Zhao Liu
2023-02-16  9:31   ` wangyanan (Y) via
2023-02-17  3:14     ` Zhao Liu
2023-02-17  7:54       ` wangyanan (Y) via
2023-02-13  9:49 ` [RFC 13/52] cpu/core: Use generic topology helper for "help" to set nr_threads Zhao Liu
2023-02-13  9:49 ` [RFC 14/52] hw/smbios: Use generic topology name and helper Zhao Liu
2023-02-13  9:49 ` [RFC 15/52] migration/postcopy-ram: " Zhao Liu
2023-02-13 10:07   ` Juan Quintela
2023-02-14  8:12     ` Zhao Liu
2023-02-13 10:16   ` Juan Quintela
2023-02-14  8:16     ` Zhao Liu
2023-02-13  9:49 ` [RFC 16/52] plugins: " Zhao Liu
2023-02-13  9:50 ` [RFC 17/52] softmmu/cpus: Use generic topology helper in vcpus initialization Zhao Liu
2023-02-13  9:50 ` [RFC 18/52] general: Replace MachineState.smp access with topology helpers Zhao Liu
2023-02-13  9:50 ` [RFC 19/52] i386: " Zhao Liu
2023-02-13  9:50 ` [RFC 20/52] s390x: " Zhao Liu
2023-02-16 13:38   ` Thomas Huth
2023-02-17  3:38     ` Zhao Liu
2023-02-13  9:50 ` [RFC 21/52] ppc: " Zhao Liu
2023-02-13  9:50 ` [RFC 22/52] riscv: " Zhao Liu
2023-02-14  2:17   ` Mi, Dapeng1
2023-02-15  2:57     ` Zhao Liu
2023-03-01 23:43       ` Palmer Dabbelt
2023-02-13  9:50 ` [RFC 23/52] arm: " Zhao Liu
2023-02-16 10:46   ` wangyanan (Y) via
2023-02-17  3:21     ` Zhao Liu
2023-02-13  9:50 ` [RFC 24/52] loongarch: " Zhao Liu
2023-02-13  9:50 ` [RFC 25/52] mips: " Zhao Liu
2023-02-14  3:40   ` Mi, Dapeng1
2023-02-15  3:08     ` Zhao Liu
2023-02-13  9:50 ` [RFC 26/52] hw: Replace MachineState.smp access with topology helpers for all remaining archs Zhao Liu
2023-02-13  9:50 ` [RFC 27/52] test/test-smp-parse: Check fields of MachineState.topo.smp Zhao Liu
2023-02-13  9:50 ` [RFC 28/52] hw/core/machine: Remove support of MachineState.smp Zhao Liu
2023-02-13  9:50 ` [RFC 29/52] hw/core/cpu: Introduce TopologyState in CPUState Zhao Liu
2023-02-13  9:50 ` [RFC 30/52] i386: Drop nr_dies and nr_modules CPUX86State Zhao Liu
2023-02-13 13:22   ` Philippe Mathieu-Daudé
2023-02-13  9:50 ` [RFC 31/52] i386/cpu: Use CPUState.topo to replace X86CPUTopoInfo to get topology info Zhao Liu
2023-02-13  9:50 ` [RFC 32/52] i386: Rename X86CPUTopoInfo and its members to reflect relationship with APIC ID Zhao Liu
2023-02-13 13:25   ` Philippe Mathieu-Daudé
2023-02-13  9:50 ` [RFC 33/52] i386: Rename init_topo_info() to init_apic_topo_info() Zhao Liu
2023-02-13 13:27   ` Philippe Mathieu-Daudé
2023-02-14 10:20     ` Zhao Liu
2023-02-13  9:50 ` [RFC 34/52] i386: Rename variable topo_info to apicid_topo Zhao Liu
2023-02-13 13:28   ` Philippe Mathieu-Daudé
2023-02-14 10:18     ` Zhao Liu
2023-02-13  9:50 ` [RFC 35/52] i386: Support APIC ID topology for hybrid CPU topology Zhao Liu
2023-02-13  9:50 ` [RFC 36/52] i386: Use init_apicid_topo_info() to initialize APIC ID topology for system emulator Zhao Liu
2023-02-13  9:50 ` [RFC 37/52] i386: Update X86CPUTopoIDs generating rule for hybrid topology Zhao Liu
2023-02-13  9:50 ` [RFC 38/52] i386: Introduce hybrid_core_type to CPUX86State Zhao Liu
2023-02-13  9:50 ` [RFC 39/52] i386/cpu: Add Intel hybrid related CPUID support Zhao Liu
2023-02-13  9:50 ` [RFC 40/52] qapi: Introduce hybrid options Zhao Liu
2023-02-13  9:50 ` [RFC 41/52] machine: Introduce core_type() hook Zhao Liu
2023-02-13 13:33   ` Philippe Mathieu-Daudé
2023-02-14 14:33     ` Zhao Liu
2023-02-13 13:35   ` Philippe Mathieu-Daudé
2023-02-14 14:51     ` Zhao Liu
2023-02-16 12:15   ` wangyanan (Y) via
2023-02-17  3:26     ` Zhao Liu
2023-02-17  7:51       ` wangyanan (Y) via
2023-02-13  9:50 ` [RFC 42/52] hw/machine: Add hybrid_supported in generic topo properties Zhao Liu
2023-02-14  1:46   ` wangyanan (Y) via
2023-02-15  2:53     ` Zhao Liu
2023-02-16 12:28       ` wangyanan (Y) via
2023-02-17  3:28         ` Zhao Liu
2023-02-13  9:50 ` [RFC 43/52] hw/machine: Rename MachineClass.smp_props to MachineClass.topo_props Zhao Liu
2023-02-13  9:50 ` [RFC 44/52] machine: Add "-hybrid" parsing rule Zhao Liu
2023-02-13  9:50 ` [RFC 45/52] hw/machine: Add hybrid cpu topology validation Zhao Liu
2023-02-13  9:50 ` [RFC 46/52] hw/machine: build core level hybrid topology form HybridCorePack Zhao Liu
2023-02-13  9:50 ` [RFC 47/52] hw/machine: Use opts_visitor to parse hybrid topo Zhao Liu
2023-02-13  9:50 ` [RFC 48/52] machine: Support "-hybrid" command Zhao Liu
2023-02-13  9:50 ` [RFC 49/52] i386/pc: Support hybrid cpu topology Zhao Liu
2023-02-13  9:50 ` [RFC 50/52] qemu-options: Add the document of hybrid command Zhao Liu
2023-02-13  9:50 ` [RFC 51/52] qapi: Expose CPU topology info in query_cpus_fast Zhao Liu
2023-02-13  9:50 ` [RFC 52/52] i386: Support cpu_index_to_core_type() for x86 Zhao Liu
2023-02-13 10:14 ` [RFC 00/52] Introduce hybrid CPU topology Alex Bennée
2023-02-14  8:48   ` Zhao Liu
2023-02-13 13:38 ` Daniel P. Berrangé
2023-02-14 17:14   ` Zhao Liu
2023-08-04 13:43   ` Zhao Liu

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=Y+9D5oFoa9WH6zBq@liuzhao-OptiPlex-7080 \
    --to=zhao1.liu@linux.intel.com \
    --cc=armbru@redhat.com \
    --cc=dapeng1.mi@intel.com \
    --cc=eblake@redhat.com \
    --cc=eduardo@habkost.net \
    --cc=like.xu.linux@gmail.com \
    --cc=marcel.apfelbaum@gmail.com \
    --cc=mst@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=philmd@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.org \
    --cc=robert.hu@linux.intel.com \
    --cc=seanjc@google.com \
    --cc=wangyanan55@huawei.com \
    --cc=zhao1.liu@intel.com \
    --cc=zhenyu.z.wang@intel.com \
    --cc=zhuocheng.ding@intel.com \
    /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;
as well as URLs for NNTP newsgroup(s).