Hi,
Static analysis on linux-next with Coverity has detected an issue with
the following commit:
529 static int ov02a10_s_stream(struct v4l2_subdev *sd, int on)
530 {
531 struct ov02a10 *ov02a10 = to_ov02a10(sd);
532 struct i2c_client *client =
v4l2_get_subdevdata(&ov02a10->subdev);
1. var_decl: Declaring variable ret without initializer.
533 int ret;
534
535 mutex_lock(&ov02a10->mutex);
536
2. Condition ov02a10->streaming == on, taking true branch.
537 if (ov02a10->streaming == on)
3. Jumping to label unlock_and_return.
538 goto unlock_and_return;
539
540 if (on) {
541 ret = pm_runtime_get_sync(&client->dev);
542 if (ret < 0) {
543 pm_runtime_put_noidle(&client->dev);
544 goto unlock_and_return;
545 }
546
547 ret = __ov02a10_start_stream(ov02a10);
548 if (ret) {
549 __ov02a10_stop_stream(ov02a10);
550 ov02a10->streaming = !on;
551 goto err_rpm_put;
552 }
553 } else {
554 __ov02a10_stop_stream(ov02a10);
555 pm_runtime_put(&client->dev);
556 }
557
558 ov02a10->streaming = on;
559 mutex_unlock(&ov02a10->mutex);
560
561 return 0;
562
563 err_rpm_put:
564 pm_runtime_put(&client->dev);
565 unlock_and_return:
566 mutex_unlock(&ov02a10->mutex);
567
Uninitialized scalar variable (UNINIT)
4. uninit_use: Using uninitialized value ret.
568 return ret;
569 }
Variable ret has not been initialized, so the error return value is a
garbage value. It should be initialized with some appropriate negative
error code, or ret could be removed and the return should return a
literal value of a error code.
I was unsure what value is appropriate to fix this, so instead I'm
reporting this issue.
Colin
On Thu, Dec 3, 2020 at 8:03 PM Colin Ian King <[email protected]> wrote:
> Static analysis on linux-next with Coverity has detected an issue with
> the following commit:
If you want to fix it properly, see my comments below...
> 529 static int ov02a10_s_stream(struct v4l2_subdev *sd, int on)
> 530 {
> 531 struct ov02a10 *ov02a10 = to_ov02a10(sd);
> 532 struct i2c_client *client =
> v4l2_get_subdevdata(&ov02a10->subdev);
>
> 1. var_decl: Declaring variable ret without initializer.
>
> 533 int ret;
> 534
> 535 mutex_lock(&ov02a10->mutex);
> 536
>
> 2. Condition ov02a10->streaming == on, taking true branch.
>
> 537 if (ov02a10->streaming == on)
>
> 3. Jumping to label unlock_and_return.
>
> 538 goto unlock_and_return;
> 539
> 540 if (on) {
> 541 ret = pm_runtime_get_sync(&client->dev);
> 542 if (ret < 0) {
> 543 pm_runtime_put_noidle(&client->dev);
> 544 goto unlock_and_return;
Instead of two above:
goto err_rpm_put;
> 545 }
> 546
> 547 ret = __ov02a10_start_stream(ov02a10);
> 548 if (ret) {
> 549 __ov02a10_stop_stream(ov02a10);
> 550 ov02a10->streaming = !on;
> 551 goto err_rpm_put;
> 552 }
> 553 } else {
> 554 __ov02a10_stop_stream(ov02a10);
> 555 pm_runtime_put(&client->dev);
> 556 }
> 557
> 558 ov02a10->streaming = on;
(1)
> 559 mutex_unlock(&ov02a10->mutex);
> 560
> 561 return 0;
> 562
> 563 err_rpm_put:
> 564 pm_runtime_put(&client->dev);
> 565 unlock_and_return:
Should be moved to (1).
> 566 mutex_unlock(&ov02a10->mutex);
> 567
>
> Uninitialized scalar variable (UNINIT)
> 4. uninit_use: Using uninitialized value ret.
>
> 568 return ret;
> 569 }
>
> Variable ret has not been initialized, so the error return value is a
> garbage value. It should be initialized with some appropriate negative
> error code, or ret could be removed and the return should return a
> literal value of a error code.
>
> I was unsure what value is appropriate to fix this, so instead I'm
> reporting this issue.
--
With Best Regards,
Andy Shevchenko
On 03/12/2020 18:10, Andy Shevchenko wrote:
> On Thu, Dec 3, 2020 at 8:03 PM Colin Ian King <[email protected]> wrote:
>
>> Static analysis on linux-next with Coverity has detected an issue with
>> the following commit:
>
> If you want to fix it properly, see my comments below...
>
>> 529 static int ov02a10_s_stream(struct v4l2_subdev *sd, int on)
>> 530 {
>> 531 struct ov02a10 *ov02a10 = to_ov02a10(sd);
>> 532 struct i2c_client *client =
>> v4l2_get_subdevdata(&ov02a10->subdev);
>>
>> 1. var_decl: Declaring variable ret without initializer.
>>
>> 533 int ret;
>> 534
>> 535 mutex_lock(&ov02a10->mutex);
>> 536
>>
>> 2. Condition ov02a10->streaming == on, taking true branch.
>>
>> 537 if (ov02a10->streaming == on)
>>
>> 3. Jumping to label unlock_and_return.
>>
>> 538 goto unlock_and_return;
>> 539
>> 540 if (on) {
>> 541 ret = pm_runtime_get_sync(&client->dev);
>> 542 if (ret < 0) {
>
>> 543 pm_runtime_put_noidle(&client->dev);
>> 544 goto unlock_and_return;
>
> Instead of two above:
> goto err_rpm_put;
>
>> 545 }
>> 546
>> 547 ret = __ov02a10_start_stream(ov02a10);
>> 548 if (ret) {
>> 549 __ov02a10_stop_stream(ov02a10);
>> 550 ov02a10->streaming = !on;
>> 551 goto err_rpm_put;
>> 552 }
>> 553 } else {
>> 554 __ov02a10_stop_stream(ov02a10);
>> 555 pm_runtime_put(&client->dev);
>> 556 }
>> 557
>> 558 ov02a10->streaming = on;
>
> (1)
>
>> 559 mutex_unlock(&ov02a10->mutex);
>> 560
>> 561 return 0;
>> 562
>> 563 err_rpm_put:
>> 564 pm_runtime_put(&client->dev);
>
>> 565 unlock_and_return:
>
> Should be moved to (1).
>
>> 566 mutex_unlock(&ov02a10->mutex);
>> 567
>>
>> Uninitialized scalar variable (UNINIT)
>> 4. uninit_use: Using uninitialized value ret.
>>
>> 568 return ret;
>> 569 }
>>
>> Variable ret has not been initialized, so the error return value is a
>> garbage value. It should be initialized with some appropriate negative
>> error code, or ret could be removed and the return should return a
>> literal value of a error code.
>>
>> I was unsure what value is appropriate to fix this, so instead I'm
>> reporting this issue.
>
Not sure I fully understand how that fixes it. Given that ret is
currently not initialized when we hit:
if (ov02a10->streaming == on)
goto unlock_and_return;
either we initialize ret to 0 at the start of the function, or do:
if (ov02a10->streaming == on) {
ret = 0;
goto unlock_and_return;
}
(assuming the intention is to return zero for this specific case).
Colin
On Thu, Dec 3, 2020 at 8:24 PM Colin Ian King <[email protected]> wrote:
> On 03/12/2020 18:10, Andy Shevchenko wrote:
> > On Thu, Dec 3, 2020 at 8:03 PM Colin Ian King <[email protected]> wrote:
> >
> >> Static analysis on linux-next with Coverity has detected an issue with
> >> the following commit:
> >
> > If you want to fix it properly, see my comments below...
> >
> >> 529 static int ov02a10_s_stream(struct v4l2_subdev *sd, int on)
> >> 530 {
> >> 531 struct ov02a10 *ov02a10 = to_ov02a10(sd);
> >> 532 struct i2c_client *client =
> >> v4l2_get_subdevdata(&ov02a10->subdev);
> >>
> >> 1. var_decl: Declaring variable ret without initializer.
> >>
> >> 533 int ret;
> >> 534
> >> 535 mutex_lock(&ov02a10->mutex);
> >> 536
> >>
> >> 2. Condition ov02a10->streaming == on, taking true branch.
> >>
> >> 537 if (ov02a10->streaming == on)
> >>
> >> 3. Jumping to label unlock_and_return.
> >>
> >> 538 goto unlock_and_return;
> >> 539
> >> 540 if (on) {
> >> 541 ret = pm_runtime_get_sync(&client->dev);
> >> 542 if (ret < 0) {
> >
> >> 543 pm_runtime_put_noidle(&client->dev);
> >> 544 goto unlock_and_return;
> >
> > Instead of two above:
> > goto err_rpm_put;
> >
> >> 545 }
> >> 546
> >> 547 ret = __ov02a10_start_stream(ov02a10);
> >> 548 if (ret) {
> >> 549 __ov02a10_stop_stream(ov02a10);
> >> 550 ov02a10->streaming = !on;
> >> 551 goto err_rpm_put;
> >> 552 }
> >> 553 } else {
> >> 554 __ov02a10_stop_stream(ov02a10);
> >> 555 pm_runtime_put(&client->dev);
> >> 556 }
> >> 557
> >> 558 ov02a10->streaming = on;
> >
> > (1)
> >
> >> 559 mutex_unlock(&ov02a10->mutex);
> >> 560
> >> 561 return 0;
> >> 562
> >> 563 err_rpm_put:
> >> 564 pm_runtime_put(&client->dev);
> >
> >> 565 unlock_and_return:
> >
> > Should be moved to (1).
> >
> >> 566 mutex_unlock(&ov02a10->mutex);
> >> 567
> >>
> >> Uninitialized scalar variable (UNINIT)
> >> 4. uninit_use: Using uninitialized value ret.
> >>
> >> 568 return ret;
> >> 569 }
> >>
> >> Variable ret has not been initialized, so the error return value is a
> >> garbage value. It should be initialized with some appropriate negative
> >> error code, or ret could be removed and the return should return a
> >> literal value of a error code.
> >>
> >> I was unsure what value is appropriate to fix this, so instead I'm
> >> reporting this issue.
> >
> Not sure I fully understand how that fixes it.
If you are not sure and have no means to test, then don't bother. This
is not the priority driver anyway.
> Given that ret is
> currently not initialized when we hit:
>
> if (ov02a10->streaming == on)
> goto unlock_and_return;
>
> either we initialize ret to 0 at the start of the function, or do:
>
> if (ov02a10->streaming == on) {
> ret = 0;
> goto unlock_and_return;
> }
>
> (assuming the intention is to return zero for this specific case).
No, please read my message again.
--
With Best Regards,
Andy Shevchenko
Hi Andy,
On Thu, 2020-12-03 at 20:10 +0200, Andy Shevchenko wrote:
> On Thu, Dec 3, 2020 at 8:03 PM Colin Ian King <[email protected]> wrote:
>
> > Static analysis on linux-next with Coverity has detected an issue with
> > the following commit:
>
> If you want to fix it properly, see my comments below...
>
> > 529 static int ov02a10_s_stream(struct v4l2_subdev *sd, int on)
> > 530 {
> > 531 struct ov02a10 *ov02a10 = to_ov02a10(sd);
> > 532 struct i2c_client *client =
> > v4l2_get_subdevdata(&ov02a10->subdev);
> >
> > 1. var_decl: Declaring variable ret without initializer.
> >
> > 533 int ret;
> > 534
> > 535 mutex_lock(&ov02a10->mutex);
> > 536
> >
> > 2. Condition ov02a10->streaming == on, taking true branch.
> >
> > 537 if (ov02a10->streaming == on)
> >
> > 3. Jumping to label unlock_and_return.
> >
> > 538 goto unlock_and_return;
> > 539
> > 540 if (on) {
> > 541 ret = pm_runtime_get_sync(&client->dev);
> > 542 if (ret < 0) {
>
> > 543 pm_runtime_put_noidle(&client->dev);
> > 544 goto unlock_and_return;
>
> Instead of two above:
From the document, pm_runtime_put_noidle is to decrease the runtime PM
usage counter of a device unless it is 0 already; while pm_runtime_put
would additionally run pm_request_idle to turn off the power if usage
counter is zero.
So here maybe we can really use pm_runtime_put instead of
pm_runtime_put_noidle, although it seems that 'pm_runtime_get_sync' and
'pm_runtime_put_noidle' often appear in pairs.
> goto err_rpm_put;
>
> > 545 }
> > 546
> > 547 ret = __ov02a10_start_stream(ov02a10);
> > 548 if (ret) {
> > 549 __ov02a10_stop_stream(ov02a10);
> > 550 ov02a10->streaming = !on;
> > 551 goto err_rpm_put;
> > 552 }
> > 553 } else {
> > 554 __ov02a10_stop_stream(ov02a10);
> > 555 pm_runtime_put(&client->dev);
> > 556 }
> > 557
> > 558 ov02a10->streaming = on;
>
> (1)
>
> > 559 mutex_unlock(&ov02a10->mutex);
> > 560
> > 561 return 0;
> > 562
> > 563 err_rpm_put:
> > 564 pm_runtime_put(&client->dev);
>
> > 565 unlock_and_return:
>
> Should be moved to (1).
>
> > 566 mutex_unlock(&ov02a10->mutex);
> > 567
> >
> > Uninitialized scalar variable (UNINIT)
> > 4. uninit_use: Using uninitialized value ret.
> >
> > 568 return ret;
> > 569 }
> >
> > Variable ret has not been initialized, so the error return value is a
> > garbage value. It should be initialized with some appropriate negative
> > error code, or ret could be removed and the return should return a
> > literal value of a error code.
> >
> > I was unsure what value is appropriate to fix this, so instead I'm
> > reporting this issue.
>
On Fri, Dec 4, 2020 at 11:47 AM Dongchun Zhu <[email protected]> wrote:
>
> Hi Andy,
>
> On Thu, 2020-12-03 at 20:10 +0200, Andy Shevchenko wrote:
> > On Thu, Dec 3, 2020 at 8:03 PM Colin Ian King <[email protected]> wrote:
> >
> > > Static analysis on linux-next with Coverity has detected an issue with
> > > the following commit:
> >
> > If you want to fix it properly, see my comments below...
> >
> > > 529 static int ov02a10_s_stream(struct v4l2_subdev *sd, int on)
> > > 530 {
> > > 531 struct ov02a10 *ov02a10 = to_ov02a10(sd);
> > > 532 struct i2c_client *client =
> > > v4l2_get_subdevdata(&ov02a10->subdev);
> > >
> > > 1. var_decl: Declaring variable ret without initializer.
> > >
> > > 533 int ret;
> > > 534
> > > 535 mutex_lock(&ov02a10->mutex);
> > > 536
> > >
> > > 2. Condition ov02a10->streaming == on, taking true branch.
> > >
> > > 537 if (ov02a10->streaming == on)
> > >
> > > 3. Jumping to label unlock_and_return.
> > >
> > > 538 goto unlock_and_return;
> > > 539
> > > 540 if (on) {
> > > 541 ret = pm_runtime_get_sync(&client->dev);
> > > 542 if (ret < 0) {
> >
> > > 543 pm_runtime_put_noidle(&client->dev);
> > > 544 goto unlock_and_return;
> >
> > Instead of two above:
>
> From the document, pm_runtime_put_noidle is to decrease the runtime PM
> usage counter of a device unless it is 0 already; while pm_runtime_put
> would additionally run pm_request_idle to turn off the power if usage
> counter is zero.
>
> So here maybe we can really use pm_runtime_put instead of
> pm_runtime_put_noidle, although it seems that 'pm_runtime_get_sync' and
> 'pm_runtime_put_noidle' often appear in pairs.
>
In an error state (e.g. if pm_runtime_get_sync() fails),
pm_runtime_put() would decrement the usage counter and call rpm_idle()
which would instantly return an error code. The end result would be
the same, except that pm_runtime_put() would return a non-zero error
code, but we ignore it anyway. However strange it looks, this seems to
be an API guarantee, so Andy's suggestion is correct.
Best regards,
Tomasz
> > goto err_rpm_put;
> >
> > > 545 }
> > > 546
> > > 547 ret = __ov02a10_start_stream(ov02a10);
> > > 548 if (ret) {
> > > 549 __ov02a10_stop_stream(ov02a10);
> > > 550 ov02a10->streaming = !on;
> > > 551 goto err_rpm_put;
> > > 552 }
> > > 553 } else {
> > > 554 __ov02a10_stop_stream(ov02a10);
> > > 555 pm_runtime_put(&client->dev);
> > > 556 }
> > > 557
> > > 558 ov02a10->streaming = on;
> >
> > (1)
> >
> > > 559 mutex_unlock(&ov02a10->mutex);
> > > 560
> > > 561 return 0;
> > > 562
> > > 563 err_rpm_put:
> > > 564 pm_runtime_put(&client->dev);
> >
> > > 565 unlock_and_return:
> >
> > Should be moved to (1).
> >
> > > 566 mutex_unlock(&ov02a10->mutex);
> > > 567
> > >
> > > Uninitialized scalar variable (UNINIT)
> > > 4. uninit_use: Using uninitialized value ret.
> > >
> > > 568 return ret;
> > > 569 }
> > >
> > > Variable ret has not been initialized, so the error return value is a
> > > garbage value. It should be initialized with some appropriate negative
> > > error code, or ret could be removed and the return should return a
> > > literal value of a error code.
> > >
> > > I was unsure what value is appropriate to fix this, so instead I'm
> > > reporting this issue.
> >
>
On Thu, Dec 03, 2020 at 08:30:03PM +0200, Andy Shevchenko wrote:
> On Thu, Dec 3, 2020 at 8:24 PM Colin Ian King <[email protected]> wrote:
> > On 03/12/2020 18:10, Andy Shevchenko wrote:
> > > On Thu, Dec 3, 2020 at 8:03 PM Colin Ian King <[email protected]> wrote:
> > >
> > >> Static analysis on linux-next with Coverity has detected an issue with
> > >> the following commit:
> > >
> > > If you want to fix it properly, see my comments below...
> > >
> > >> 529 static int ov02a10_s_stream(struct v4l2_subdev *sd, int on)
> > >> 530 {
> > >> 531 struct ov02a10 *ov02a10 = to_ov02a10(sd);
> > >> 532 struct i2c_client *client =
> > >> v4l2_get_subdevdata(&ov02a10->subdev);
> > >>
> > >> 1. var_decl: Declaring variable ret without initializer.
> > >>
> > >> 533 int ret;
> > >> 534
> > >> 535 mutex_lock(&ov02a10->mutex);
> > >> 536
> > >>
> > >> 2. Condition ov02a10->streaming == on, taking true branch.
> > >>
> > >> 537 if (ov02a10->streaming == on)
> > >>
> > >> 3. Jumping to label unlock_and_return.
> > >>
> > >> 538 goto unlock_and_return;
> > >> 539
> > >> 540 if (on) {
> > >> 541 ret = pm_runtime_get_sync(&client->dev);
> > >> 542 if (ret < 0) {
> > >
> > >> 543 pm_runtime_put_noidle(&client->dev);
> > >> 544 goto unlock_and_return;
> > >
> > > Instead of two above:
> > > goto err_rpm_put;
> > >
> > >> 545 }
> > >> 546
> > >> 547 ret = __ov02a10_start_stream(ov02a10);
> > >> 548 if (ret) {
> > >> 549 __ov02a10_stop_stream(ov02a10);
> > >> 550 ov02a10->streaming = !on;
> > >> 551 goto err_rpm_put;
> > >> 552 }
> > >> 553 } else {
> > >> 554 __ov02a10_stop_stream(ov02a10);
> > >> 555 pm_runtime_put(&client->dev);
> > >> 556 }
> > >> 557
> > >> 558 ov02a10->streaming = on;
> > >
> > > (1)
> > >
> > >> 559 mutex_unlock(&ov02a10->mutex);
> > >> 560
> > >> 561 return 0;
> > >> 562
> > >> 563 err_rpm_put:
> > >> 564 pm_runtime_put(&client->dev);
> > >
> > >> 565 unlock_and_return:
> > >
> > > Should be moved to (1).
> > >
> > >> 566 mutex_unlock(&ov02a10->mutex);
> > >> 567
> > >>
> > >> Uninitialized scalar variable (UNINIT)
> > >> 4. uninit_use: Using uninitialized value ret.
> > >>
> > >> 568 return ret;
> > >> 569 }
> > >>
> > >> Variable ret has not been initialized, so the error return value is a
> > >> garbage value. It should be initialized with some appropriate negative
> > >> error code, or ret could be removed and the return should return a
> > >> literal value of a error code.
> > >>
> > >> I was unsure what value is appropriate to fix this, so instead I'm
> > >> reporting this issue.
> > >
> > Not sure I fully understand how that fixes it.
>
> If you are not sure and have no means to test, then don't bother. This
> is not the priority driver anyway.
Arnd sent a patch to address this:
<URL:https://patchwork.linuxtv.org/project/linux-media/patch/[email protected]/>
--
Sakari Ailus