2021-05-27 09:06:43

by Zhen Lei

Subject: [PATCH 1/1] drm/i915/hdcp: Simplify code in intel_hdcp_auth_downstream()

If intel_hdcp_validate_v_prime() has been successful within the allowed
number of tries, we can directly call drm_dbg_kms() and "goto out" without
jumping out of the loop and repeatedly judging whether the operation is
successful. This can help us reduce an unnecessary if judgment. And it's
a little clearer to read.

No functional change.

Signed-off-by: Zhen Lei <[email protected]>
---
drivers/gpu/drm/i915/display/intel_hdcp.c | 24 ++++++++++-------------
1 file changed, 10 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c
index d8570e14fe60..c32a854eda66 100644
--- a/drivers/gpu/drm/i915/display/intel_hdcp.c
+++ b/drivers/gpu/drm/i915/display/intel_hdcp.c
@@ -663,13 +663,13 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)

ret = shim->read_ksv_fifo(dig_port, num_downstream, ksv_fifo);
if (ret)
- goto err;
+ goto out;

if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm, ksv_fifo,
num_downstream) > 0) {
drm_err(&dev_priv->drm, "Revoked Ksv(s) in ksv_fifo\n");
ret = -EPERM;
- goto err;
+ goto out;
}

/*
@@ -680,20 +680,16 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
ret = intel_hdcp_validate_v_prime(connector, shim,
ksv_fifo, num_downstream,
bstatus);
- if (!ret)
- break;
- }
-
- if (i == tries) {
- drm_dbg_kms(&dev_priv->drm,
- "V Prime validation failed.(%d)\n", ret);
- goto err;
+ if (!ret) {
+ drm_dbg_kms(&dev_priv->drm,
+ "HDCP is enabled (%d downstream devices)\n",
+ num_downstream);
+ goto out;
+ }
}

- drm_dbg_kms(&dev_priv->drm, "HDCP is enabled (%d downstream devices)\n",
- num_downstream);
- ret = 0;
-err:
+ drm_dbg_kms(&dev_priv->drm, "V Prime validation failed.(%d)\n", ret);
+out:
kfree(ksv_fifo);
return ret;
}
--
2.25.1

2021-05-27 15:21:47

by Jani Nikula

Subject: Re: [PATCH 1/1] drm/i915/hdcp: Simplify code in intel_hdcp_auth_downstream()

On Thu, 27 May 2021, Zhen Lei <[email protected]> wrote:
> If intel_hdcp_validate_v_prime() has been successful within the allowed
> number of tries, we can directly call drm_dbg_kms() and "goto out" without
> jumping out of the loop and repeatedly judging whether the operation is
> successful. This can help us reduce an unnecessary if judgment. And it's
> a little clearer to read.

Generally I think the "happy day scenario" should be at the topmost
indentation level and not buried in the ifs with a goto exit.

BR,
Jani.

>
> No functional change.
>
> Signed-off-by: Zhen Lei <[email protected]>
> ---
> drivers/gpu/drm/i915/display/intel_hdcp.c | 24 ++++++++++-------------
> 1 file changed, 10 insertions(+), 14 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c
> index d8570e14fe60..c32a854eda66 100644
> --- a/drivers/gpu/drm/i915/display/intel_hdcp.c
> +++ b/drivers/gpu/drm/i915/display/intel_hdcp.c
> @@ -663,13 +663,13 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
>
> ret = shim->read_ksv_fifo(dig_port, num_downstream, ksv_fifo);
> if (ret)
> - goto err;
> + goto out;
>
> if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm, ksv_fifo,
> num_downstream) > 0) {
> drm_err(&dev_priv->drm, "Revoked Ksv(s) in ksv_fifo\n");
> ret = -EPERM;
> - goto err;
> + goto out;
> }
>
> /*
> @@ -680,20 +680,16 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
> ret = intel_hdcp_validate_v_prime(connector, shim,
> ksv_fifo, num_downstream,
> bstatus);
> - if (!ret)
> - break;
> - }
> -
> - if (i == tries) {
> - drm_dbg_kms(&dev_priv->drm,
> - "V Prime validation failed.(%d)\n", ret);
> - goto err;
> + if (!ret) {
> + drm_dbg_kms(&dev_priv->drm,
> + "HDCP is enabled (%d downstream devices)\n",
> + num_downstream);
> + goto out;
> + }
> }
>
> - drm_dbg_kms(&dev_priv->drm, "HDCP is enabled (%d downstream devices)\n",
> - num_downstream);
> - ret = 0;
> -err:
> + drm_dbg_kms(&dev_priv->drm, "V Prime validation failed.(%d)\n", ret);
> +out:
> kfree(ksv_fifo);
> return ret;
> }

--
Jani Nikula, Intel Open Source Graphics Center

2021-05-28 09:37:50

by Zhen Lei

Subject: Re: [PATCH 1/1] drm/i915/hdcp: Simplify code in intel_hdcp_auth_downstream()

On 2021/5/27 18:04, Jani Nikula wrote:
> On Thu, 27 May 2021, Zhen Lei <[email protected]> wrote:
>> If intel_hdcp_validate_v_prime() has been successful within the allowed
>> number of tries, we can directly call drm_dbg_kms() and "goto out" without
>> jumping out of the loop and repeatedly judging whether the operation is
>> successful. This can help us reduce an unnecessary if judgment. And it's
>> a little clearer to read.
>
> Generally I think the "happy day scenario" should be at the topmost
> indentation level and not buried in the ifs with a goto exit.

for (xxx) {
if (a == b)
return found;
}

At least this way of writing is common.


>
> BR,
> Jani.
>
>>
>> No functional change.
>>
>> Signed-off-by: Zhen Lei <[email protected]>
>> ---
>> drivers/gpu/drm/i915/display/intel_hdcp.c | 24 ++++++++++-------------
>> 1 file changed, 10 insertions(+), 14 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c
>> index d8570e14fe60..c32a854eda66 100644
>> --- a/drivers/gpu/drm/i915/display/intel_hdcp.c
>> +++ b/drivers/gpu/drm/i915/display/intel_hdcp.c
>> @@ -663,13 +663,13 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
>>
>> ret = shim->read_ksv_fifo(dig_port, num_downstream, ksv_fifo);
>> if (ret)
>> - goto err;
>> + goto out;
>>
>> if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm, ksv_fifo,
>> num_downstream) > 0) {
>> drm_err(&dev_priv->drm, "Revoked Ksv(s) in ksv_fifo\n");
>> ret = -EPERM;
>> - goto err;
>> + goto out;
>> }
>>
>> /*
>> @@ -680,20 +680,16 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
>> ret = intel_hdcp_validate_v_prime(connector, shim,
>> ksv_fifo, num_downstream,
>> bstatus);
>> - if (!ret)
>> - break;
>> - }
>> -
>> - if (i == tries) {
>> - drm_dbg_kms(&dev_priv->drm,
>> - "V Prime validation failed.(%d)\n", ret);
>> - goto err;
>> + if (!ret) {
>> + drm_dbg_kms(&dev_priv->drm,
>> + "HDCP is enabled (%d downstream devices)\n",
>> + num_downstream);
>> + goto out;
>> + }
>> }
>>
>> - drm_dbg_kms(&dev_priv->drm, "HDCP is enabled (%d downstream devices)\n",
>> - num_downstream);
>> - ret = 0;
>> -err:
>> + drm_dbg_kms(&dev_priv->drm, "V Prime validation failed.(%d)\n", ret);
>> +out:
>> kfree(ksv_fifo);
>> return ret;
>> }
>

2021-05-31 10:37:53

by Jani Nikula

Subject: Re: [PATCH 1/1] drm/i915/hdcp: Simplify code in intel_hdcp_auth_downstream()

On Fri, 28 May 2021, "Leizhen (ThunderTown)" <[email protected]> wrote:
> On 2021/5/27 18:04, Jani Nikula wrote:
>> On Thu, 27 May 2021, Zhen Lei <[email protected]> wrote:
>>> If intel_hdcp_validate_v_prime() has been successful within the allowed
>>> number of tries, we can directly call drm_dbg_kms() and "goto out" without
>>> jumping out of the loop and repeatedly judging whether the operation is
>>> successful. This can help us reduce an unnecessary if judgment. And it's
>>> a little clearer to read.
>>
>> Generally I think the "happy day scenario" should be at the topmost
>> indentation level and not buried in the ifs with a goto exit.
>
> for (xxx) {
> if (a == b)
> return found;
> }
>
> At least this way of writing is common.

Yes, if the loop is abstracted to a separate function.

BR,
Jani.

>
>
>>
>> BR,
>> Jani.
>>
>>>
>>> No functional change.
>>>
>>> Signed-off-by: Zhen Lei <[email protected]>
>>> ---
>>> drivers/gpu/drm/i915/display/intel_hdcp.c | 24 ++++++++++-------------
>>> 1 file changed, 10 insertions(+), 14 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c
>>> index d8570e14fe60..c32a854eda66 100644
>>> --- a/drivers/gpu/drm/i915/display/intel_hdcp.c
>>> +++ b/drivers/gpu/drm/i915/display/intel_hdcp.c
>>> @@ -663,13 +663,13 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
>>>
>>> ret = shim->read_ksv_fifo(dig_port, num_downstream, ksv_fifo);
>>> if (ret)
>>> - goto err;
>>> + goto out;
>>>
>>> if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm, ksv_fifo,
>>> num_downstream) > 0) {
>>> drm_err(&dev_priv->drm, "Revoked Ksv(s) in ksv_fifo\n");
>>> ret = -EPERM;
>>> - goto err;
>>> + goto out;
>>> }
>>>
>>> /*
>>> @@ -680,20 +680,16 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
>>> ret = intel_hdcp_validate_v_prime(connector, shim,
>>> ksv_fifo, num_downstream,
>>> bstatus);
>>> - if (!ret)
>>> - break;
>>> - }
>>> -
>>> - if (i == tries) {
>>> - drm_dbg_kms(&dev_priv->drm,
>>> - "V Prime validation failed.(%d)\n", ret);
>>> - goto err;
>>> + if (!ret) {
>>> + drm_dbg_kms(&dev_priv->drm,
>>> + "HDCP is enabled (%d downstream devices)\n",
>>> + num_downstream);
>>> + goto out;
>>> + }
>>> }
>>>
>>> - drm_dbg_kms(&dev_priv->drm, "HDCP is enabled (%d downstream devices)\n",
>>> - num_downstream);
>>> - ret = 0;
>>> -err:
>>> + drm_dbg_kms(&dev_priv->drm, "V Prime validation failed.(%d)\n", ret);
>>> +out:
>>> kfree(ksv_fifo);
>>> return ret;
>>> }
>>
>

--
Jani Nikula, Intel Open Source Graphics Center