2019-12-20 17:15:17

by Georgi Djakov

Subject: [PATCH] interconnect: Check for valid path in icc_set_bw()

Use IS_ERR() to ensure that the path passed to icc_set_bw() is valid.

Signed-off-by: Georgi Djakov <[email protected]>
---
drivers/interconnect/core.c | 5 +++++
1 file changed, 5 insertions(+)

diff --git a/drivers/interconnect/core.c b/drivers/interconnect/core.c
index 63c164264b73..14a6f7ade44a 100644
--- a/drivers/interconnect/core.c
+++ b/drivers/interconnect/core.c
@@ -498,6 +498,11 @@ int icc_set_bw(struct icc_path *path, u32 avg_bw, u32 peak_bw)
if (!path || !path->num_nodes)
return 0;

+ if (IS_ERR(path)) {
+ pr_err("%s: invalid path=%ld\n", __func__, PTR_ERR(path));
+ return -EINVAL;
+ }
+
mutex_lock(&icc_lock);

old_avg = path->reqs[0].avg_bw;

2019-12-20 19:05:42

by Bjorn Andersson

Subject: Re: [PATCH] interconnect: Check for valid path in icc_set_bw()

On Fri 20 Dec 09:13 PST 2019, Georgi Djakov wrote:

> Use IS_ERR() to ensure that the path passed to icc_set_bw() is valid.
>
> Signed-off-by: Georgi Djakov <[email protected]>
> ---
> drivers/interconnect/core.c | 5 +++++
> 1 file changed, 5 insertions(+)
>
> diff --git a/drivers/interconnect/core.c b/drivers/interconnect/core.c
> index 63c164264b73..14a6f7ade44a 100644
> --- a/drivers/interconnect/core.c
> +++ b/drivers/interconnect/core.c
> @@ -498,6 +498,11 @@ int icc_set_bw(struct icc_path *path, u32 avg_bw, u32 peak_bw)
> if (!path || !path->num_nodes)
> return 0;
>
> + if (IS_ERR(path)) {

This is a sign of a logical error, and the print is likely to be
ignored/lost in the noise. So I think the response should aid to help
the developer hitting this to resolve the issue.

So I think this is more visible and more useful as:

if (WARN_ON(IS_ERR(path)))
return -EINVAL;


PS. Doesn't path->num_nodes == 0 fall in this category as well? When
would you have a path object with no nodes passed to this function?

Regards,
Bjorn

> + pr_err("%s: invalid path=%ld\n", __func__, PTR_ERR(path));
> + return -EINVAL;
> + }
> +
> mutex_lock(&icc_lock);
>
> old_avg = path->reqs[0].avg_bw;

2019-12-21 02:44:01

by Georgi Djakov

Subject: Re: [PATCH] interconnect: Check for valid path in icc_set_bw()

Hi Bjorn,

Thanks for the comments!

On 20.12.19 21:04, Bjorn Andersson wrote:
> On Fri 20 Dec 09:13 PST 2019, Georgi Djakov wrote:
>
>> Use IS_ERR() to ensure that the path passed to icc_set_bw() is valid.
>>
>> Signed-off-by: Georgi Djakov <[email protected]>
>> ---
>> drivers/interconnect/core.c | 5 +++++
>> 1 file changed, 5 insertions(+)
>>
>> diff --git a/drivers/interconnect/core.c b/drivers/interconnect/core.c
>> index 63c164264b73..14a6f7ade44a 100644
>> --- a/drivers/interconnect/core.c
>> +++ b/drivers/interconnect/core.c
>> @@ -498,6 +498,11 @@ int icc_set_bw(struct icc_path *path, u32 avg_bw, u32 peak_bw)
>> if (!path || !path->num_nodes)
>> return 0;
>>
>> + if (IS_ERR(path)) {
>
> This is a sign of a logical error, and the print is likely to be
> ignored/lost in the noise. So I think the response should aid to help
> the developer hitting this to resolve the issue.
>
> So I think this is more visible and more useful as:
>
> if (WARN_ON(IS_ERR(path)))
> return -EINVAL;

That's actually what i had in mind initially, but then started
wondering whether this isn't a bit too noisy. But oh well, let's
scream loud if something is done incorrectly.

>
> PS. Doesn't path->num_nodes == 0 fall in this category as well? When
> would you have a path object with no nodes passed to this function?

Yes, will make the warning cover this case too.

Thanks,
Georgi

>
> Regards,
> Bjorn
>
>> + pr_err("%s: invalid path=%ld\n", __func__, PTR_ERR(path));
>> + return -EINVAL;
>> + }
>> +
>> mutex_lock(&icc_lock);
>>
>> old_avg = path->reqs[0].avg_bw;