| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <1242153926.2492.49.camel@ht.satnam>
Date: Wed, 13 May 2009 00:15:26 +0530
From: Jaswinder Singh Rajput <jaswinder@...nel.org>
To: Ingo Molnar <mingo@...e.hu>
Cc: "H. Peter Anvin" <hpa@...nel.org>,
Robert Richter <robert.richter@....com>,
Dave Jones <davej@...hat.com>,
LKML <linux-kernel@...r.kernel.org>,
x86 maintainers <x86@...nel.org>
Subject: Re: [PATCH 5/10 -tip] x86: check_powernow() for K8 and later user
of Advanced Power Management features
On Tue, 2009-05-12 at 19:48 +0200, Ingo Molnar wrote:
> * Jaswinder Singh Rajput <jaswinder@...nel.org> wrote:
>
> > - eax = cpuid_eax(CPUID_GET_MAX_CAPABILITIES);
> > - if (eax < CPUID_FREQ_VOLT_CAPABILITIES) {
> > + /* Advanced Power Management capabilities */
> > + if (c->x86_capability[9]) {
> > printk(KERN_INFO PFX
> > "No frequency change capabilities detected\n");
> > goto out;
> > }
>
> How is the new check equivalent to the old one? It isnt and this is
> a bug.
>
> Also, open-coding x86_capability[9] like that is quite unclean. Were
> we ever to reorder those bits internally, this could would break.
>
> But i see what you are trying to do. A better method might be to add
> a new helper:
>
> +#define X86_FEATURE_TS (9*32+ 0) /* Temperatue sensor */
> +#define X86_FEATURE_FID (9*32+ 1) /* Frequency ID control */
> +#define X86_FEATURE_VID (9*32+ 2) /* Voltage ID control */
> +#define X86_FEATURE_TTP (9*32+ 3) /* Thermal trip */
> +#define X86_FEATURE_HTC (9*32+ 4) /* Hardware thermal control */
> +#define X86_FEATURE_STC (9*32+ 5) /* Software thermal control */
> +#define X86_FEATURE_100MHZSTEPS (9*32+ 6) /* 100 MHz multiplier control */
> +#define X86_FEATURE_HWPSTATE (9*32+ 7) /* Hardware P-State control */
> +#define X86_FEATURE_CONSTANT_TSC (9*32+ 8) /* Constant rate TSC ticks */
>
> ... to represent the 'is any of these set' property.
>
Yes, I was also thinking about adding this helper cpufeature, may be
that's why I made above mistake ;-)
> [ btw., there's a typo in the X86_FEATURE_TS comment above ]
>
Ok I will send updated patches and new pull request :-)
[PATCH-tip] x86: check_powernow() for K8 and later user of Advanced Power Management features
use X86_FEATURE_POWER_MGMT, X86_FEATURE_FID, X86_FEATURE_VID and
X86_FEATURE_HWPSTATE to determine K8 and later PowerNOW.
Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@...il.com>
---
arch/x86/kernel/cpu/cpufreq/powernow-k8.c | 16 ++++++++--------
arch/x86/kernel/cpu/cpufreq/powernow-k8.h | 4 ----
2 files changed, 8 insertions(+), 12 deletions(-)
diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
index 4709ead..c9869fd 100644
--- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
+++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c
@@ -512,8 +512,9 @@ static int core_voltage_post_transition(struct powernow_k8_data *data,
static int check_supported_cpu(unsigned int cpu)
{
+ struct cpuinfo_x86 *c = &cpu_data(cpu);
cpumask_t oldmask;
- u32 eax, ebx, ecx, edx;
+ u32 eax;
unsigned int rc = 0;
oldmask = current->cpus_allowed;
@@ -540,23 +541,22 @@ static int check_supported_cpu(unsigned int cpu)
goto out;
}
- eax = cpuid_eax(CPUID_GET_MAX_CAPABILITIES);
- if (eax < CPUID_FREQ_VOLT_CAPABILITIES) {
+ /* Advanced Power Management capabilities */
+ if (!cpu_has(c, X86_FEATURE_POWER_MGMT)) {
printk(KERN_INFO PFX
"No frequency change capabilities detected\n");
goto out;
}
- cpuid(CPUID_FREQ_VOLT_CAPABILITIES, &eax, &ebx, &ecx, &edx);
- if ((edx & P_STATE_TRANSITION_CAPABLE)
- != P_STATE_TRANSITION_CAPABLE) {
+ /* check for frequncy and volatage ID control support */
+ if (!cpu_has(c, X86_FEATURE_FID) &&
+ !cpu_has(c, X86_FEATURE_VID)) {
printk(KERN_INFO PFX
"Power state transitions not supported\n");
goto out;
}
} else { /* must be a HW Pstate capable processor */
- cpuid(CPUID_FREQ_VOLT_CAPABILITIES, &eax, &ebx, &ecx, &edx);
- if ((edx & USE_HW_PSTATE) == USE_HW_PSTATE)
+ if (cpu_has(c, X86_FEATURE_HWPSTATE))
cpu_family = CPU_HW_PSTATE;
else
goto out;
diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h
index 6c6698f..4dfe414 100644
--- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h
+++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h
@@ -64,9 +64,6 @@ struct powernow_k8_data {
#define CPUID_XMOD_REV_MASK 0x000c0000
#define CPUID_XFAM_10H 0x00100000 /* family 0x10 */
#define CPUID_USE_XFAM_XMOD 0x00000f00
-#define CPUID_GET_MAX_CAPABILITIES 0x80000000
-#define CPUID_FREQ_VOLT_CAPABILITIES 0x80000007
-#define P_STATE_TRANSITION_CAPABLE 6
/* Model Specific Registers for p-state transitions. MSRs are 64-bit. For */
/* writes (wrmsr - opcode 0f 30), the register number is placed in ecx, and */
@@ -101,7 +98,6 @@ struct powernow_k8_data {
/* Hardware Pstate _PSS and MSR definitions */
-#define USE_HW_PSTATE 0x00000080
#define HW_PSTATE_MASK 0x00000007
#define HW_PSTATE_VALID_MASK 0x80000000
#define HW_PSTATE_MAX_MASK 0x000000f0
--
1.6.0.6
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists