[PATCH V2] x86/Hyper-V: don't allocate clockevent device when synthetic timer is unavailable

[PATCH V2] x86/Hyper-V: don't allocate clockevent device when synthetic timer is unavailable

[ltykernel]

From: Tianyu Lan <Tianyu.Lan@microsoft.com>

Current code initializes clock event data structure for syn timer
even when it's unavailable. Fix it.

Signed-off-by: Tianyu Lan <Tianyu.Lan@microsoft.com>
---
Change since v1:
	Update title and commit log. 

 drivers/hv/hv.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index 632d25674e7f..2e893768fc76 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -212,13 +212,16 @@ int hv_synic_alloc(void)
 		tasklet_init(&hv_cpu->msg_dpc,
 			     vmbus_on_msg_dpc, (unsigned long) hv_cpu);
 
-		hv_cpu->clk_evt = kzalloc(sizeof(struct clock_event_device),
-					  GFP_KERNEL);
-		if (hv_cpu->clk_evt == NULL) {
-			pr_err("Unable to allocate clock event device\n");
-			goto err;
+		if (ms_hyperv.features & HV_MSR_SYNTIMER_AVAILABLE) {
+			hv_cpu->clk_evt =
+				kzalloc(sizeof(struct clock_event_device),
+						  GFP_KERNEL);
+			if (hv_cpu->clk_evt == NULL) {
+				pr_err("Unable to allocate clock event device\n");
+				goto err;
+			}
+			hv_init_clockevent_device(hv_cpu->clk_evt, cpu);
 		}
-		hv_init_clockevent_device(hv_cpu->clk_evt, cpu);
 
 		hv_cpu->synic_message_page =
 			(void *)get_zeroed_page(GFP_ATOMIC);
-- 
2.14.5

Re: [PATCH V2] x86/Hyper-V: don't allocate clockevent device when synthetic timer is unavailable

[Vitaly Kuznetsov]

ltykernel@gmail.com writes:

> From: Tianyu Lan <Tianyu.Lan@microsoft.com>
>
> Current code initializes clock event data structure for syn timer
> even when it's unavailable. Fix it.
>
> Signed-off-by: Tianyu Lan <Tianyu.Lan@microsoft.com>
> ---
> Change since v1:
> 	Update title and commit log. 
>
>  drivers/hv/hv.c | 15 +++++++++------
>  1 file changed, 9 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
> index 632d25674e7f..2e893768fc76 100644
> --- a/drivers/hv/hv.c
> +++ b/drivers/hv/hv.c
> @@ -212,13 +212,16 @@ int hv_synic_alloc(void)
>  		tasklet_init(&hv_cpu->msg_dpc,
>  			     vmbus_on_msg_dpc, (unsigned long) hv_cpu);
>  
> -		hv_cpu->clk_evt = kzalloc(sizeof(struct clock_event_device),
> -					  GFP_KERNEL);
> -		if (hv_cpu->clk_evt == NULL) {
> -			pr_err("Unable to allocate clock event device\n");
> -			goto err;
> +		if (ms_hyperv.features & HV_MSR_SYNTIMER_AVAILABLE) {
> +			hv_cpu->clk_evt =
> +				kzalloc(sizeof(struct clock_event_device),
> +						  GFP_KERNEL);
> +			if (hv_cpu->clk_evt == NULL) {
> +				pr_err("Unable to allocate clock event device\n");
> +				goto err;
> +			}
> +			hv_init_clockevent_device(hv_cpu->clk_evt, cpu);
>  		}
> -		hv_init_clockevent_device(hv_cpu->clk_evt, cpu);
>  
>  		hv_cpu->synic_message_page =
>  			(void *)get_zeroed_page(GFP_ATOMIC);

Thank you for fixing the subject! I had one more question on the
previous version which still stands: which tree is this patch for?
Upstream, clockevent allocation has moved to
drivers/clocksource/hyperv_timer.c and the code there is different.

Is this intended for some stable tree?

-- 
Vitaly

Re: [PATCH V2] x86/Hyper-V: don't allocate clockevent device when synthetic timer is unavailable

[Tianyu Lan]

On 3/31/2020 7:20 PM, Vitaly Kuznetsov wrote:
> ltykernel@gmail.com writes:
> 
>> From: Tianyu Lan <Tianyu.Lan@microsoft.com>
>>
>> Current code initializes clock event data structure for syn timer
>> even when it's unavailable. Fix it.
>>
>> Signed-off-by: Tianyu Lan <Tianyu.Lan@microsoft.com>
>> ---
>> Change since v1:
>> 	Update title and commit log.
>>
>>   drivers/hv/hv.c | 15 +++++++++------
>>   1 file changed, 9 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
>> index 632d25674e7f..2e893768fc76 100644
>> --- a/drivers/hv/hv.c
>> +++ b/drivers/hv/hv.c
>> @@ -212,13 +212,16 @@ int hv_synic_alloc(void)
>>   		tasklet_init(&hv_cpu->msg_dpc,
>>   			     vmbus_on_msg_dpc, (unsigned long) hv_cpu);
>>   
>> -		hv_cpu->clk_evt = kzalloc(sizeof(struct clock_event_device),
>> -					  GFP_KERNEL);
>> -		if (hv_cpu->clk_evt == NULL) {
>> -			pr_err("Unable to allocate clock event device\n");
>> -			goto err;
>> +		if (ms_hyperv.features & HV_MSR_SYNTIMER_AVAILABLE) {
>> +			hv_cpu->clk_evt =
>> +				kzalloc(sizeof(struct clock_event_device),
>> +						  GFP_KERNEL);
>> +			if (hv_cpu->clk_evt == NULL) {
>> +				pr_err("Unable to allocate clock event device\n");
>> +				goto err;
>> +			}
>> +			hv_init_clockevent_device(hv_cpu->clk_evt, cpu);
>>   		}
>> -		hv_init_clockevent_device(hv_cpu->clk_evt, cpu);
>>   
>>   		hv_cpu->synic_message_page =
>>   			(void *)get_zeroed_page(GFP_ATOMIC);
> 
> Thank you for fixing the subject! I had one more question on the
> previous version which still stands: which tree is this patch for?
> Upstream, clockevent allocation has moved to
> drivers/clocksource/hyperv_timer.c and the code there is different.
> 
Yes, I just noticed I fixed an issue on the old code after Michael 
reminded me. Sorry for noise. Please ignore it.