qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[PATCH v10 12/21] i386: Introduce module level cpu topology to CPUX86Sta


From: Zhao Liu
Subject: [PATCH v10 12/21] i386: Introduce module level cpu topology to CPUX86State
Date: Thu, 21 Mar 2024 22:40:39 +0800

From: Zhao Liu <zhao1.liu@intel.com>

Intel CPUs implement module level on hybrid client products (e.g.,
ADL-N, MTL, etc) and E-core server products.

A module contains a set of cores that share certain resources (in
current products, the resource usually includes L2 cache, as well as
module scoped features and MSRs).

Module level support is the prerequisite for L2 cache topology on
module level. With module level, we can implement the Guest's CPU
topology and future cache topology to be consistent with the Host's on
Intel hybrid client/E-core server platforms.

Tested-by: Yongwei Ma <yongwei.ma@intel.com>
Co-developed-by: Zhuocheng Ding <zhuocheng.ding@intel.com>
Signed-off-by: Zhuocheng Ding <zhuocheng.ding@intel.com>
Signed-off-by: Zhao Liu <zhao1.liu@intel.com>
Tested-by: Babu Moger <babu.moger@amd.com>
---
Changes since v7:
 * Mapped x86 module to smp module instead of cluster.
 * Re-wrote the commit message to explain the reason why we needs module
   level.
 * Dropped Michael/Babu's ACKed/Tested tags since the code change.
 * Re-added Yongwei's Tested tag For his re-testing.

Changes since v1:
 * The background of the introduction of the "cluster" parameter and its
   exact meaning were revised according to Yanan's explanation. (Yanan)
---
 hw/i386/x86.c     | 5 +++++
 target/i386/cpu.c | 1 +
 target/i386/cpu.h | 3 +++
 3 files changed, 9 insertions(+)

diff --git a/hw/i386/x86.c b/hw/i386/x86.c
index 0a6c59c724f1..7c94d366af03 100644
--- a/hw/i386/x86.c
+++ b/hw/i386/x86.c
@@ -313,6 +313,11 @@ void x86_cpu_pre_plug(HotplugHandler *hotplug_dev,
 
     init_topo_info(&topo_info, x86ms);
 
+    if (ms->smp.modules > 1) {
+        env->nr_modules = ms->smp.modules;
+        /* TODO: Expose module level in CPUID[0x1F]. */
+    }
+
     if (ms->smp.dies > 1) {
         env->nr_dies = ms->smp.dies;
         set_bit(CPU_TOPO_LEVEL_DIE, env->avail_cpu_topo);
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 92d85e920015..b8917c412175 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -7717,6 +7717,7 @@ static void x86_cpu_init_default_topo(X86CPU *cpu)
 {
     CPUX86State *env = &cpu->env;
 
+    env->nr_modules = 1;
     env->nr_dies = 1;
 
     /* SMT, core and package levels are set by default. */
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 2e24f457468d..095540e58f7a 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1899,6 +1899,9 @@ typedef struct CPUArchState {
     /* Number of dies within this CPU package. */
     unsigned nr_dies;
 
+    /* Number of modules within one die. */
+    unsigned nr_modules;
+
     /* Bitmap of available CPU topology levels for this CPU. */
     DECLARE_BITMAP(avail_cpu_topo, CPU_TOPO_LEVEL_MAX);
 } CPUX86State;
-- 
2.34.1




reply via email to

[Prev in Thread] Current Thread [Next in Thread]