Drivers: hv: Use -ETIMEDOUT for HV_STATUS_TIME_OUT instead of -EIO

[PATCH] Drivers: hv: Use -ETIMEDOUT for HV_STATUS_TIME_OUT instead of -EIO

Posted by Easwar Hariharan 4 months, 1 week ago

Use the -ETIMEDOUT errno value as the correct 1:1 match for the
hypervisor timeout status

Fixes: 3817854ba89201 ("hyperv: Log hypercall status codes as strings")
Signed-off-by: Easwar Hariharan <easwar.hariharan@linux.microsoft.com>
---
 drivers/hv/hv_common.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
index 49898d10fafff..9b51b67d54cc8 100644
--- a/drivers/hv/hv_common.c
+++ b/drivers/hv/hv_common.c
@@ -781,7 +781,7 @@ static const struct hv_status_info hv_status_infos[] = {
 	_STATUS_INFO(HV_STATUS_INVALID_LP_INDEX,		-EIO),
 	_STATUS_INFO(HV_STATUS_INVALID_REGISTER_VALUE,		-EIO),
 	_STATUS_INFO(HV_STATUS_OPERATION_FAILED,		-EIO),
-	_STATUS_INFO(HV_STATUS_TIME_OUT,			-EIO),
+	_STATUS_INFO(HV_STATUS_TIME_OUT,			-ETIMEDOUT),
 	_STATUS_INFO(HV_STATUS_CALL_PENDING,			-EIO),
 	_STATUS_INFO(HV_STATUS_VTL_ALREADY_ENABLED,		-EIO),
 #undef _STATUS_INFO
-- 
2.43.0

Re: [PATCH] Drivers: hv: Use -ETIMEDOUT for HV_STATUS_TIME_OUT instead of -EIO

Posted by Stanislav Kinsburskii 4 months, 1 week ago

On Thu, Oct 02, 2025 at 10:13:46PM +0000, Easwar Hariharan wrote:
> Use the -ETIMEDOUT errno value as the correct 1:1 match for the
> hypervisor timeout status
> 

The commit message should answer the question why this change is being
made.
Is there a practical reason for this or is it only for consistensy?

Thanks,
Stanislav

> Fixes: 3817854ba89201 ("hyperv: Log hypercall status codes as strings")
> Signed-off-by: Easwar Hariharan <easwar.hariharan@linux.microsoft.com>
> ---
>  drivers/hv/hv_common.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
> index 49898d10fafff..9b51b67d54cc8 100644
> --- a/drivers/hv/hv_common.c
> +++ b/drivers/hv/hv_common.c
> @@ -781,7 +781,7 @@ static const struct hv_status_info hv_status_infos[] = {
>  	_STATUS_INFO(HV_STATUS_INVALID_LP_INDEX,		-EIO),
>  	_STATUS_INFO(HV_STATUS_INVALID_REGISTER_VALUE,		-EIO),
>  	_STATUS_INFO(HV_STATUS_OPERATION_FAILED,		-EIO),
> -	_STATUS_INFO(HV_STATUS_TIME_OUT,			-EIO),
> +	_STATUS_INFO(HV_STATUS_TIME_OUT,			-ETIMEDOUT),
>  	_STATUS_INFO(HV_STATUS_CALL_PENDING,			-EIO),
>  	_STATUS_INFO(HV_STATUS_VTL_ALREADY_ENABLED,		-EIO),
>  #undef _STATUS_INFO
> -- 
> 2.43.0

Re: [PATCH] Drivers: hv: Use -ETIMEDOUT for HV_STATUS_TIME_OUT instead of -EIO

Posted by Easwar Hariharan 4 months, 1 week ago

On 10/2/2025 3:37 PM, Stanislav Kinsburskii wrote:
> On Thu, Oct 02, 2025 at 10:13:46PM +0000, Easwar Hariharan wrote:
>> Use the -ETIMEDOUT errno value as the correct 1:1 match for the
>> hypervisor timeout status
>>
> 
> The commit message should answer the question why this change is being
> made.
> Is there a practical reason for this or is it only for consistensy?
> 
<snip>

It's for consistency and because it's a better match. See my sibling response
for other error codes that I can also fix in a v2.

Thanks,
Easwar (he/him)

Re: [PATCH] Drivers: hv: Use -ETIMEDOUT for HV_STATUS_TIME_OUT instead of -EIO

Posted by Easwar Hariharan 4 months, 1 week ago

On 10/2/2025 3:13 PM, Easwar Hariharan wrote:
> Use the -ETIMEDOUT errno value as the correct 1:1 match for the
> hypervisor timeout status
> 
> Fixes: 3817854ba89201 ("hyperv: Log hypercall status codes as strings")
> Signed-off-by: Easwar Hariharan <easwar.hariharan@linux.microsoft.com>
> ---
>  drivers/hv/hv_common.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
> index 49898d10fafff..9b51b67d54cc8 100644
> --- a/drivers/hv/hv_common.c
> +++ b/drivers/hv/hv_common.c
> @@ -781,7 +781,7 @@ static const struct hv_status_info hv_status_infos[] = {
>  	_STATUS_INFO(HV_STATUS_INVALID_LP_INDEX,		-EIO),
>  	_STATUS_INFO(HV_STATUS_INVALID_REGISTER_VALUE,		-EIO),
>  	_STATUS_INFO(HV_STATUS_OPERATION_FAILED,		-EIO),
> -	_STATUS_INFO(HV_STATUS_TIME_OUT,			-EIO),
> +	_STATUS_INFO(HV_STATUS_TIME_OUT,			-ETIMEDOUT),
>  	_STATUS_INFO(HV_STATUS_CALL_PENDING,			-EIO),
>  	_STATUS_INFO(HV_STATUS_VTL_ALREADY_ENABLED,		-EIO),
>  #undef _STATUS_INFO

Actually looking at the whole struct, it may be useful to also change at least some of
 the following codes?

        _STATUS_INFO(HV_STATUS_INVALID_ALIGNMENT,               -EIO), -> EINVAL
        _STATUS_INFO(HV_STATUS_ACCESS_DENIED,                   -EIO), -> EACCES
        _STATUS_INFO(HV_STATUS_INVALID_PARTITION_STATE,         -EIO), -> EINVAL
        _STATUS_INFO(HV_STATUS_OPERATION_DENIED,                -EIO), -> EACCES
        _STATUS_INFO(HV_STATUS_PROPERTY_VALUE_OUT_OF_RANGE,     -EIO), -> ERANGE
        _STATUS_INFO(HV_STATUS_INSUFFICIENT_BUFFERS,            -EIO), -> ENOBUFS
        _STATUS_INFO(HV_STATUS_NOT_ACKNOWLEDGED,                -EIO), -> EBUSY
        _STATUS_INFO(HV_STATUS_INVALID_VP_STATE,                -EIO), -> EINVAL
        _STATUS_INFO(HV_STATUS_INVALID_LP_INDEX,                -EIO), -> EINVAL
        _STATUS_INFO(HV_STATUS_INVALID_REGISTER_VALUE,          -EIO), -> EINVAL
        _STATUS_INFO(HV_STATUS_VTL_ALREADY_ENABLED,             -EIO), -> EBUSY

Nuno, Stas, others, what do you think?

Thanks,
Easwar (he/him)