From: Wei Wang <wei.w.wang@intel.com>
To: jbeulich@suse.com, andrew.cooper3@citrix.com, xen-devel@lists.xen.org
Cc: Wei Wang <wei.w.wang@intel.com>
Subject: [PATCH v5 2/9] x86/intel_pstate: APERF/MPERF feature detect
Date: Mon, 14 Sep 2015 10:32:03 +0800 [thread overview]
Message-ID: <1442197930-3738-3-git-send-email-wei.w.wang@intel.com> (raw)
In-Reply-To: <1442197930-3738-1-git-send-email-wei.w.wang@intel.com>
Add support to detect the APERF/MPERF feature. Also, remove the identical
code in cpufreq.c and powernow.c. This patch is independent of the
earlier patches.
Signed-off-by: Wei Wang <wei.w.wang@intel.com>
---
xen/arch/x86/acpi/cpufreq/cpufreq.c | 6 ++----
xen/arch/x86/acpi/cpufreq/powernow.c | 6 ++----
xen/arch/x86/cpu/common.c | 4 ++++
xen/include/asm-x86/cpufeature.h | 5 +++++
4 files changed, 13 insertions(+), 8 deletions(-)
changes in v5:
1) define macros for 0x1 and CPUID leaf5;
2) add a statement stating that this patch is independent of the
previous ones.
diff --git a/xen/arch/x86/acpi/cpufreq/cpufreq.c b/xen/arch/x86/acpi/cpufreq/cpufreq.c
index ef79f77..8494fa0 100644
--- a/xen/arch/x86/acpi/cpufreq/cpufreq.c
+++ b/xen/arch/x86/acpi/cpufreq/cpufreq.c
@@ -50,7 +50,6 @@ enum {
};
#define INTEL_MSR_RANGE (0xffffull)
-#define CPUID_6_ECX_APERFMPERF_CAPABILITY (0x1)
struct acpi_cpufreq_data *cpufreq_drv_data[NR_CPUS];
@@ -351,10 +350,9 @@ static unsigned int get_cur_freq_on_cpu(unsigned int cpu)
static void feature_detect(void *info)
{
struct cpufreq_policy *policy = info;
- unsigned int eax, ecx;
+ unsigned int eax;
- ecx = cpuid_ecx(6);
- if (ecx & CPUID_6_ECX_APERFMPERF_CAPABILITY) {
+ if ( cpu_has_aperfmperf ) {
policy->aperf_mperf = 1;
acpi_cpufreq_driver.getavg = get_measured_perf;
}
diff --git a/xen/arch/x86/acpi/cpufreq/powernow.c b/xen/arch/x86/acpi/cpufreq/powernow.c
index 4de6f8d..d11da1a 100644
--- a/xen/arch/x86/acpi/cpufreq/powernow.c
+++ b/xen/arch/x86/acpi/cpufreq/powernow.c
@@ -37,7 +37,6 @@
#include <acpi/acpi.h>
#include <acpi/cpufreq/cpufreq.h>
-#define CPUID_6_ECX_APERFMPERF_CAPABILITY (0x1)
#define CPUID_FREQ_VOLT_CAPABILITIES 0x80000007
#define CPB_CAPABLE 0x00000200
#define USE_HW_PSTATE 0x00000080
@@ -211,10 +210,9 @@ static int powernow_cpufreq_verify(struct cpufreq_policy *policy)
static void feature_detect(void *info)
{
struct cpufreq_policy *policy = info;
- unsigned int ecx, edx;
+ unsigned int edx;
- ecx = cpuid_ecx(6);
- if (ecx & CPUID_6_ECX_APERFMPERF_CAPABILITY) {
+ if ( cpu_has_aperfmperf ) {
policy->aperf_mperf = 1;
powernow_cpufreq_driver.getavg = get_measured_perf;
}
diff --git a/xen/arch/x86/cpu/common.c b/xen/arch/x86/cpu/common.c
index 35ef21b..5224d10 100644
--- a/xen/arch/x86/cpu/common.c
+++ b/xen/arch/x86/cpu/common.c
@@ -239,6 +239,10 @@ static void __cpuinit generic_identify(struct cpuinfo_x86 *c)
if ( cpu_has(c, X86_FEATURE_CLFLSH) )
c->x86_clflush_size = ((ebx >> 8) & 0xff) * 8;
+ if ( (c->cpuid_level > CPUID_PM_LEAF) &&
+ (cpuid_ecx(CPUID_PM_LEAF) & CPUID6_ECX_APERFMPERF_CAPABILITY) )
+ set_bit(X86_FEATURE_APERFMPERF, c->x86_capability);
+
/* AMD-defined flags: level 0x80000001 */
c->extended_cpuid_level = cpuid_eax(0x80000000);
if ( (c->extended_cpuid_level & 0xffff0000) == 0x80000000 ) {
diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h
index 9a01563..d5f532b 100644
--- a/xen/include/asm-x86/cpufeature.h
+++ b/xen/include/asm-x86/cpufeature.h
@@ -69,6 +69,7 @@
#define X86_FEATURE_XTOPOLOGY (3*32+13) /* cpu topology enum extensions */
#define X86_FEATURE_CPUID_FAULTING (3*32+14) /* cpuid faulting */
#define X86_FEATURE_CLFLUSH_MONITOR (3*32+15) /* clflush reqd with monitor */
+#define X86_FEATURE_APERFMPERF (3*32+28) /* APERFMPERF */
/* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
#define X86_FEATURE_XMM3 (4*32+ 0) /* Streaming SIMD Extensions-3 */
@@ -165,6 +166,9 @@
#define CPUID5_ECX_EXTENSIONS_SUPPORTED 0x1
#define CPUID5_ECX_INTERRUPT_BREAK 0x2
+#define CPUID_PM_LEAF 6
+#define CPUID6_ECX_APERFMPERF_CAPABILITY 0x1
+
#define cpu_has_vme 0
#define cpu_has_de 1
#define cpu_has_pse 1
@@ -190,6 +194,7 @@
#define cpu_has_page1gb boot_cpu_has(X86_FEATURE_PAGE1GB)
#define cpu_has_efer 1
#define cpu_has_fsgsbase boot_cpu_has(X86_FEATURE_FSGSBASE)
+#define cpu_has_aperfmperf boot_cpu_has(X86_FEATURE_APERFMPERF)
#define cpu_has_smep boot_cpu_has(X86_FEATURE_SMEP)
#define cpu_has_smap boot_cpu_has(X86_FEATURE_SMAP)
--
1.9.1
next prev parent reply other threads:[~2015-09-14 2:32 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-09-14 2:32 [PATCH v5 0/9] Porting the intel_pstate driver to Xen Wei Wang
2015-09-14 2:32 ` [PATCH v5 1/9] x86/intel_pstate: add some calculation related support Wei Wang
2015-09-17 15:19 ` Andrew Cooper
2015-10-05 16:00 ` Jan Beulich
2015-09-14 2:32 ` Wei Wang [this message]
2015-09-17 15:26 ` [PATCH v5 2/9] x86/intel_pstate: APERF/MPERF feature detect Andrew Cooper
2015-10-05 16:14 ` Jan Beulich
2015-09-14 2:32 ` [PATCH v5 3/9] x86/intel_pstate: add a new driver interface, setpolicy() Wei Wang
2015-09-17 15:34 ` Andrew Cooper
2015-10-06 15:37 ` Jan Beulich
2015-09-14 2:32 ` [PATCH v5 4/9] x86/intel_pstate: relocate the driver register function Wei Wang
2015-09-17 15:38 ` Andrew Cooper
2015-09-21 13:13 ` Jan Beulich
2015-10-07 15:08 ` Jan Beulich
2015-09-14 2:32 ` [PATCH v5 5/9] x86/intel_pstate: changes in cpufreq_del_cpu for CPU offline Wei Wang
2015-09-17 15:43 ` Andrew Cooper
2015-10-07 15:28 ` Jan Beulich
2015-10-11 2:19 ` Wang, Wei W
2015-09-14 2:32 ` [PATCH v5 6/9] x86/intel_pstate: the main boby of the intel_pstate driver Wei Wang
2015-09-17 15:51 ` Andrew Cooper
2015-10-07 15:39 ` Jan Beulich
2015-09-14 2:32 ` [PATCH v5 7/9] x86/intel_pstate: add a booting param to select the driver to load Wei Wang
2015-09-17 16:08 ` Andrew Cooper
2015-09-21 13:15 ` Jan Beulich
2015-10-07 15:46 ` Jan Beulich
2015-10-23 8:09 ` Wang, Wei W
2015-10-23 8:16 ` Wang, Wei W
2015-10-23 8:18 ` Wang, Wei W
2015-10-23 8:35 ` Jan Beulich
2015-10-23 8:48 ` Wang, Wei W
2015-09-14 2:32 ` [PATCH v5 8/9] x86/intel_pstate: support the use of intel_pstate in pmstat.c Wei Wang
2015-10-07 16:10 ` Jan Beulich
2015-10-26 6:26 ` Wang, Wei W
2015-10-26 7:02 ` Jan Beulich
2015-10-26 7:59 ` Wang, Wei W
2015-10-26 9:40 ` Jan Beulich
2015-10-26 9:48 ` Wang, Wei W
2015-10-26 9:53 ` Jan Beulich
2015-10-26 10:19 ` Wang, Wei W
2015-10-26 10:35 ` Jan Beulich
2015-10-26 10:45 ` Wang, Wei W
2015-10-26 10:51 ` Jan Beulich
2015-10-26 11:03 ` Wang, Wei W
2015-09-14 2:32 ` [PATCH v5 9/9] tools: enable xenpm to control the intel_pstate driver Wei Wang
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=1442197930-3738-3-git-send-email-wei.w.wang@intel.com \
--to=wei.w.wang@intel.com \
--cc=andrew.cooper3@citrix.com \
--cc=jbeulich@suse.com \
--cc=xen-devel@lists.xen.org \
/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).