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;
Subject: Re: [PATCH] x86: Don't use MWAIT if explicitly requested
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     linux-kernel@vger.kernel.org, x86@kernel.org,
        tim.c.chen@linux.intel.com, Thomas Gleixner <tglx@linutronix.de>,
        Borislav Petkov <bp@alien8.de>, Ingo Molnar <mingo@redhat.com>,
        "H. Peter Anvin" <hpa@zytor.com>,
        Boris Ostrovsky <boris.ostrovsky@oracle.com>
References: <1571370354-17736-1-git-send-email-zhenzhong.duan@oracle.com>
 <20191018071206.GZ2328@hirez.programming.kicks-ass.net>
From:   Zhenzhong Duan <zhenzhong.duan@oracle.com>
Organization: Oracle Corporation
Message-ID: <56b074ab-8718-180c-b3dd-72687025c5a8@oracle.com>
Date:   Fri, 18 Oct 2019 16:47:20 +0800
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.9.0
MIME-Version: 1.0
In-Reply-To: <20191018071206.GZ2328@hirez.programming.kicks-ass.net>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Content-Language: en-US
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


On 2019/10/18 15:12, Peter Zijlstra wrote:
> On Fri, Oct 18, 2019 at 11:45:54AM +0800, Zhenzhong Duan wrote:
>> If 'idle=nomwait' is specified or process matching what's in
>> processor_idle_dmi_table, we should't use MWAIT at bootup stage before
>> cpuidle driver loaded, even if it's preferred by default on Intel.
>>
>> Add a check so that HALT instruction is used in those cases.
> The comment in idle_setup():
>
> 	/*
> 	 * If the boot option of "idle=nomwait" is added,
> 	 * it means that mwait will be disabled for CPU C2/C3
> 	 * states. In such case it won't touch the variable
> 	 * of boot_option_idle_override.
> 	 */
> 	boot_option_idle_override = IDLE_NOMWAIT;
>
> explicitly states this option is for C2+

Yea, this is confusing. Other place referencing 
boot_option_idle_override tell me

"idle=nomwait" means not using mwait for all cstates. Maybe 'C2/C3' could be

removed from above comment?

See drivers/acpi/processor_idle.c:

                         if (cx.type == ACPI_STATE_C1 &&
                             (boot_option_idle_override == IDLE_NOMWAIT)) {
                                 /*
                                  * In most cases the C1 space_id 
obtained from
                                  * _CST object is FIXED_HARDWARE access 
mode.
                                  * But when the option of idle=halt is 
added,
                                  * the entry_method type should be 
changed from
                                  * CSTATE_FFH to CSTATE_HALT.
                                  * When the option of idle=nomwait is 
added,
                                  * the C1 entry_method type should be
                                  * CSTATE_HALT.
                                  */
                                 cx.entry_method = ACPI_CSTATE_HALT;
                                 snprintf(cx.desc, ACPI_CX_DESC_LEN, 
"ACPI HLT");
                         }

and drivers/acpi/processor_pdc.c:

         if (boot_option_idle_override == IDLE_NOMWAIT) {
                 /*
                  * If mwait is disabled for CPU C-states, the C2C3_FFH 
access
                  * mode will be disabled in the parameter of _PDC object.
                  * Of course C1_FFH access mode will also be disabled.
                  */
                 union acpi_object *obj;
                 u32 *buffer = NULL;

                 obj = pdc_in->pointer;
                 buffer = (u32 *)(obj->buffer.pointer);
                 buffer[2] &= ~(ACPI_PDC_C_C2C3_FFH | ACPI_PDC_C_C1_FFH);

         }

>
>> Signed-off-by: Zhenzhong Duan <zhenzhong.duan@oracle.com>
>> Cc: Thomas Gleixner <tglx@linutronix.de>
>> Cc: Borislav Petkov <bp@alien8.de>
>> Cc: Ingo Molnar <mingo@redhat.com>
>> Cc: "H. Peter Anvin" <hpa@zytor.com>
>> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
>> ---
>>   arch/x86/kernel/process.c | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
>> index 5e94c43..37fc577 100644
>> --- a/arch/x86/kernel/process.c
>> +++ b/arch/x86/kernel/process.c
>> @@ -667,6 +667,10 @@ static void amd_e400_idle(void)
>>    */
>>   static int prefer_mwait_c1_over_halt(const struct cpuinfo_x86 *c)
>>   {
>> +	/* Don't use MWAIT-C1 if explicitly requested */
>> +	if (boot_option_idle_override == IDLE_NOMWAIT)
>> +		return 0;
> And this is very much about C1...
>
> OTOH, "idle=halt" should be forcing HLT over MWAIT, so did you want to
> write:
>
> 	if (boot_option_idle_override == IDLE_HALT)
> 		return 0;
>
> instead?

I think it's not necessory, if 'idle=halt' specified, 
select_idle_routine() returns early,

prefer_mwait_c1_over_halt() is never called.

Zhenzhong