LinuxLists.cc - [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

2020-02-20 17:39:21

Subject: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.

The name changing as devfreq(X) breaks some user space applications,
such as Android HAL from Unisoc and Hikey [1].
The device name will be changed unexpectly after every boot depending
on module init sequence. It will make trouble to setup some system
configuration like selinux for Android.

So we'd like to revert it back to old naming rule before any better
way being found.

[1] https://lkml.org/lkml/2018/5/8/1042

Cc: John Stultz <[email protected]>
Cc: Greg Kroah-Hartman <[email protected]>
Cc: [email protected]
Signed-off-by: Orson Zhai <[email protected]>

---
drivers/devfreq/devfreq.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
index cceee8b..7dcf209 100644
--- a/drivers/devfreq/devfreq.c
+++ b/drivers/devfreq/devfreq.c
@@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
{
struct devfreq *devfreq;
struct devfreq_governor *governor;
- static atomic_t devfreq_no = ATOMIC_INIT(-1);
int err = 0;

if (!dev || !profile || !governor_name) {
@@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
atomic_set(&devfreq->suspend_count, 0);

- dev_set_name(&devfreq->dev, "devfreq%d",
- atomic_inc_return(&devfreq_no));
+ dev_set_name(&devfreq->dev, "%s", dev_name(dev));
err = device_register(&devfreq->dev);
if (err) {
mutex_unlock(&devfreq->lock);
--
2.7.4

2020-02-20 19:15:59

by Greg Kroah-Hartman

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
> This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
>
> The name changing as devfreq(X) breaks some user space applications,
> such as Android HAL from Unisoc and Hikey [1].
> The device name will be changed unexpectly after every boot depending
> on module init sequence. It will make trouble to setup some system
> configuration like selinux for Android.
>
> So we'd like to revert it back to old naming rule before any better
> way being found.
>
> [1] https://lkml.org/lkml/2018/5/8/1042
>
> Cc: John Stultz <[email protected]>
> Cc: Greg Kroah-Hartman <[email protected]>
> Cc: [email protected]
> Signed-off-by: Orson Zhai <[email protected]>
>
> ---
> drivers/devfreq/devfreq.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> index cceee8b..7dcf209 100644
> --- a/drivers/devfreq/devfreq.c
> +++ b/drivers/devfreq/devfreq.c
> @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
> {
> struct devfreq *devfreq;
> struct devfreq_governor *governor;
> - static atomic_t devfreq_no = ATOMIC_INIT(-1);
> int err = 0;
>
> if (!dev || !profile || !governor_name) {
> @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
> devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
> atomic_set(&devfreq->suspend_count, 0);
>
> - dev_set_name(&devfreq->dev, "devfreq%d",
> - atomic_inc_return(&devfreq_no));
> + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
> err = device_register(&devfreq->dev);
> if (err) {
> mutex_unlock(&devfreq->lock);
> --
> 2.7.4
>

Thanks for this, I agree, this needs to get back to the way things were
as it seems to break too many existing systems as-is.

I'll queue this up in my tree now, thanks.

greg k-h

2020-02-20 19:48:34

by John Stultz

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On Thu, Feb 20, 2020 at 11:15 AM Greg Kroah-Hartman
<[email protected]> wrote:
>
> On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
> > This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
> >
> > The name changing as devfreq(X) breaks some user space applications,
> > such as Android HAL from Unisoc and Hikey [1].
> > The device name will be changed unexpectly after every boot depending
> > on module init sequence. It will make trouble to setup some system
> > configuration like selinux for Android.
> >
> > So we'd like to revert it back to old naming rule before any better
> > way being found.
> >
> > [1] https://lkml.org/lkml/2018/5/8/1042
> >
> > Cc: John Stultz <[email protected]>
> > Cc: Greg Kroah-Hartman <[email protected]>
> > Cc: [email protected]
> > Signed-off-by: Orson Zhai <[email protected]>
> >
> > ---
> > drivers/devfreq/devfreq.c | 4 +---
> > 1 file changed, 1 insertion(+), 3 deletions(-)
> >
> > diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> > index cceee8b..7dcf209 100644
> > --- a/drivers/devfreq/devfreq.c
> > +++ b/drivers/devfreq/devfreq.c
> > @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > {
> > struct devfreq *devfreq;
> > struct devfreq_governor *governor;
> > - static atomic_t devfreq_no = ATOMIC_INIT(-1);
> > int err = 0;
> >
> > if (!dev || !profile || !governor_name) {
> > @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
> > atomic_set(&devfreq->suspend_count, 0);
> >
> > - dev_set_name(&devfreq->dev, "devfreq%d",
> > - atomic_inc_return(&devfreq_no));
> > + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
> > err = device_register(&devfreq->dev);
> > if (err) {
> > mutex_unlock(&devfreq->lock);
> > --
> > 2.7.4
> >
>
> Thanks for this, I agree, this needs to get back to the way things were
> as it seems to break too many existing systems as-is.
>
> I'll queue this up in my tree now, thanks.

Oof this old thing. I unfortunately didn't get back to look at the
devfreq name node issue or the compatibility links, since the impact
of the regression (breaking the powerHAL's interactions with the gpu)
wasn't as big as other problems we had. While the regression was
frustrating, my only hesitancy at this point is that its been this way
since 4.10, so reverting the problematic patch is likely to break any
new users since then.

thanks
-john

2020-02-20 20:04:06

by Greg Kroah-Hartman

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On Thu, Feb 20, 2020 at 11:47:41AM -0800, John Stultz wrote:
> On Thu, Feb 20, 2020 at 11:15 AM Greg Kroah-Hartman
> <[email protected]> wrote:
> >
> > On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
> > > This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
> > >
> > > The name changing as devfreq(X) breaks some user space applications,
> > > such as Android HAL from Unisoc and Hikey [1].
> > > The device name will be changed unexpectly after every boot depending
> > > on module init sequence. It will make trouble to setup some system
> > > configuration like selinux for Android.
> > >
> > > So we'd like to revert it back to old naming rule before any better
> > > way being found.
> > >
> > > [1] https://lkml.org/lkml/2018/5/8/1042
> > >
> > > Cc: John Stultz <[email protected]>
> > > Cc: Greg Kroah-Hartman <[email protected]>
> > > Cc: [email protected]
> > > Signed-off-by: Orson Zhai <[email protected]>
> > >
> > > ---
> > > drivers/devfreq/devfreq.c | 4 +---
> > > 1 file changed, 1 insertion(+), 3 deletions(-)
> > >
> > > diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> > > index cceee8b..7dcf209 100644
> > > --- a/drivers/devfreq/devfreq.c
> > > +++ b/drivers/devfreq/devfreq.c
> > > @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > > {
> > > struct devfreq *devfreq;
> > > struct devfreq_governor *governor;
> > > - static atomic_t devfreq_no = ATOMIC_INIT(-1);
> > > int err = 0;
> > >
> > > if (!dev || !profile || !governor_name) {
> > > @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > > devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
> > > atomic_set(&devfreq->suspend_count, 0);
> > >
> > > - dev_set_name(&devfreq->dev, "devfreq%d",
> > > - atomic_inc_return(&devfreq_no));
> > > + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
> > > err = device_register(&devfreq->dev);
> > > if (err) {
> > > mutex_unlock(&devfreq->lock);
> > > --
> > > 2.7.4
> > >
> >
> > Thanks for this, I agree, this needs to get back to the way things were
> > as it seems to break too many existing systems as-is.
> >
> > I'll queue this up in my tree now, thanks.
>
> Oof this old thing. I unfortunately didn't get back to look at the
> devfreq name node issue or the compatibility links, since the impact
> of the regression (breaking the powerHAL's interactions with the gpu)
> wasn't as big as other problems we had. While the regression was
> frustrating, my only hesitancy at this point is that its been this way
> since 4.10, so reverting the problematic patch is likely to break any
> new users since then.

4.11 :)

2020-02-21 07:07:12

by Greg Kroah-Hartman

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

2020-02-21 08:39:18

by Chanwoo Choi

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On 2/21/20 4:06 PM, Greg Kroah-Hartman wrote:
> On Thu, Feb 20, 2020 at 11:47:41AM -0800, John Stultz wrote:
>> On Thu, Feb 20, 2020 at 11:15 AM Greg Kroah-Hartman
>> <[email protected]> wrote:
>>>
>>> On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
>>>> This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
>>>>
>>>> The name changing as devfreq(X) breaks some user space applications,
>>>> such as Android HAL from Unisoc and Hikey [1].
>>>> The device name will be changed unexpectly after every boot depending
>>>> on module init sequence. It will make trouble to setup some system
>>>> configuration like selinux for Android.
>>>>
>>>> So we'd like to revert it back to old naming rule before any better
>>>> way being found.
>>>>
>>>> [1] https://protect2.fireeye.com/url?k=00fa721e-5d2a7af6-00fbf951-000babff32e3-95e4b92259b05656&u=https://lkml.org/lkml/2018/5/8/1042
>>>>
>>>> Cc: John Stultz <[email protected]>
>>>> Cc: Greg Kroah-Hartman <[email protected]>
>>>> Cc: [email protected]
>>>> Signed-off-by: Orson Zhai <[email protected]>
>>>>
>>>> ---
>>>> drivers/devfreq/devfreq.c | 4 +---
>>>> 1 file changed, 1 insertion(+), 3 deletions(-)
>>>>
>>>> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
>>>> index cceee8b..7dcf209 100644
>>>> --- a/drivers/devfreq/devfreq.c
>>>> +++ b/drivers/devfreq/devfreq.c
>>>> @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
>>>> {
>>>> struct devfreq *devfreq;
>>>> struct devfreq_governor *governor;
>>>> - static atomic_t devfreq_no = ATOMIC_INIT(-1);
>>>> int err = 0;
>>>>
>>>> if (!dev || !profile || !governor_name) {
>>>> @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
>>>> devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
>>>> atomic_set(&devfreq->suspend_count, 0);
>>>>
>>>> - dev_set_name(&devfreq->dev, "devfreq%d",
>>>> - atomic_inc_return(&devfreq_no));
>>>> + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
>>>> err = device_register(&devfreq->dev);
>>>> if (err) {
>>>> mutex_unlock(&devfreq->lock);
>>>> --
>>>> 2.7.4
>>>>
>>>
>>> Thanks for this, I agree, this needs to get back to the way things were
>>> as it seems to break too many existing systems as-is.
>>>
>>> I'll queue this up in my tree now, thanks.
>>
>> Oof this old thing. I unfortunately didn't get back to look at the
>> devfreq name node issue or the compatibility links, since the impact
>> of the regression (breaking the powerHAL's interactions with the gpu)
>> wasn't as big as other problems we had. While the regression was
>> frustrating, my only hesitancy at this point is that its been this way
>> since 4.10, so reverting the problematic patch is likely to break any
>> new users since then.
>
> Looks like most users just revert that commit in their trees:
> https://protect2.fireeye.com/url?k=1012ad0f-4dc2a5e7-10132640-000babff32e3-35779c5ed675ef0f&u=https://source.codeaurora.org/quic/la/kernel/msm-4.14/commit/drivers/devfreq?h=msm-4.14&id=ccf273f6d89ad0fa8032e9225305ad6f62c7770c
>
> So we should be ok here.

I'm sorry about changing the devfreq node name.

OK. Do you pick this patch to your tree?
or If not, I'll apply it to devfreq-next branch for v5.7-rc1.

And do you apply it to kernel of linux-stable tree since 4.11?

--
Best Regards,
Chanwoo Choi
Samsung Electronics

2020-02-21 11:14:51

by Greg Kroah-Hartman

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On Fri, Feb 21, 2020 at 05:11:02PM +0900, Chanwoo Choi wrote:
> On 2/21/20 4:06 PM, Greg Kroah-Hartman wrote:
> > On Thu, Feb 20, 2020 at 11:47:41AM -0800, John Stultz wrote:
> >> On Thu, Feb 20, 2020 at 11:15 AM Greg Kroah-Hartman
> >> <[email protected]> wrote:
> >>>
> >>> On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
> >>>> This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
> >>>>
> >>>> The name changing as devfreq(X) breaks some user space applications,
> >>>> such as Android HAL from Unisoc and Hikey [1].
> >>>> The device name will be changed unexpectly after every boot depending
> >>>> on module init sequence. It will make trouble to setup some system
> >>>> configuration like selinux for Android.
> >>>>
> >>>> So we'd like to revert it back to old naming rule before any better
> >>>> way being found.
> >>>>
> >>>> [1] https://protect2.fireeye.com/url?k=00fa721e-5d2a7af6-00fbf951-000babff32e3-95e4b92259b05656&u=https://lkml.org/lkml/2018/5/8/1042
> >>>>
> >>>> Cc: John Stultz <[email protected]>
> >>>> Cc: Greg Kroah-Hartman <[email protected]>
> >>>> Cc: [email protected]
> >>>> Signed-off-by: Orson Zhai <[email protected]>
> >>>>
> >>>> ---
> >>>> drivers/devfreq/devfreq.c | 4 +---
> >>>> 1 file changed, 1 insertion(+), 3 deletions(-)
> >>>>
> >>>> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> >>>> index cceee8b..7dcf209 100644
> >>>> --- a/drivers/devfreq/devfreq.c
> >>>> +++ b/drivers/devfreq/devfreq.c
> >>>> @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
> >>>> {
> >>>> struct devfreq *devfreq;
> >>>> struct devfreq_governor *governor;
> >>>> - static atomic_t devfreq_no = ATOMIC_INIT(-1);
> >>>> int err = 0;
> >>>>
> >>>> if (!dev || !profile || !governor_name) {
> >>>> @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
> >>>> devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
> >>>> atomic_set(&devfreq->suspend_count, 0);
> >>>>
> >>>> - dev_set_name(&devfreq->dev, "devfreq%d",
> >>>> - atomic_inc_return(&devfreq_no));
> >>>> + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
> >>>> err = device_register(&devfreq->dev);
> >>>> if (err) {
> >>>> mutex_unlock(&devfreq->lock);
> >>>> --
> >>>> 2.7.4
> >>>>
> >>>
> >>> Thanks for this, I agree, this needs to get back to the way things were
> >>> as it seems to break too many existing systems as-is.
> >>>
> >>> I'll queue this up in my tree now, thanks.
> >>
> >> Oof this old thing. I unfortunately didn't get back to look at the
> >> devfreq name node issue or the compatibility links, since the impact
> >> of the regression (breaking the powerHAL's interactions with the gpu)
> >> wasn't as big as other problems we had. While the regression was
> >> frustrating, my only hesitancy at this point is that its been this way
> >> since 4.10, so reverting the problematic patch is likely to break any
> >> new users since then.
> >
> > Looks like most users just revert that commit in their trees:
> > https://protect2.fireeye.com/url?k=1012ad0f-4dc2a5e7-10132640-000babff32e3-35779c5ed675ef0f&u=https://source.codeaurora.org/quic/la/kernel/msm-4.14/commit/drivers/devfreq?h=msm-4.14&id=ccf273f6d89ad0fa8032e9225305ad6f62c7770c
> >
> > So we should be ok here.
>
> I'm sorry about changing the devfreq node name.
>
> OK. Do you pick this patch to your tree?

Yes, I can do that.

> or If not, I'll apply it to devfreq-next branch for v5.7-rc1.
>
> And do you apply it to kernel of linux-stable tree since 4.11?

Yeah, I'll mark it for stable.

Can I get an ack from you for this?

thanks,

greg k-h

2020-02-21 23:56:13

by Chanwoo Choi

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On Fri, Feb 21, 2020 at 8:13 PM Greg Kroah-Hartman
<[email protected]> wrote:
>
> On Fri, Feb 21, 2020 at 05:11:02PM +0900, Chanwoo Choi wrote:
> > On 2/21/20 4:06 PM, Greg Kroah-Hartman wrote:
> > > On Thu, Feb 20, 2020 at 11:47:41AM -0800, John Stultz wrote:
> > >> On Thu, Feb 20, 2020 at 11:15 AM Greg Kroah-Hartman
> > >> <[email protected]> wrote:
> > >>>
> > >>> On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
> > >>>> This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
> > >>>>
> > >>>> The name changing as devfreq(X) breaks some user space applications,
> > >>>> such as Android HAL from Unisoc and Hikey [1].
> > >>>> The device name will be changed unexpectly after every boot depending
> > >>>> on module init sequence. It will make trouble to setup some system
> > >>>> configuration like selinux for Android.
> > >>>>
> > >>>> So we'd like to revert it back to old naming rule before any better
> > >>>> way being found.
> > >>>>
> > >>>> [1] https://protect2.fireeye.com/url?k=00fa721e-5d2a7af6-00fbf951-000babff32e3-95e4b92259b05656&u=https://lkml.org/lkml/2018/5/8/1042
> > >>>>
> > >>>> Cc: John Stultz <[email protected]>
> > >>>> Cc: Greg Kroah-Hartman <[email protected]>
> > >>>> Cc: [email protected]
> > >>>> Signed-off-by: Orson Zhai <[email protected]>
> > >>>>
> > >>>> ---
> > >>>> drivers/devfreq/devfreq.c | 4 +---
> > >>>> 1 file changed, 1 insertion(+), 3 deletions(-)
> > >>>>
> > >>>> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> > >>>> index cceee8b..7dcf209 100644
> > >>>> --- a/drivers/devfreq/devfreq.c
> > >>>> +++ b/drivers/devfreq/devfreq.c
> > >>>> @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > >>>> {
> > >>>> struct devfreq *devfreq;
> > >>>> struct devfreq_governor *governor;
> > >>>> - static atomic_t devfreq_no = ATOMIC_INIT(-1);
> > >>>> int err = 0;
> > >>>>
> > >>>> if (!dev || !profile || !governor_name) {
> > >>>> @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > >>>> devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
> > >>>> atomic_set(&devfreq->suspend_count, 0);
> > >>>>
> > >>>> - dev_set_name(&devfreq->dev, "devfreq%d",
> > >>>> - atomic_inc_return(&devfreq_no));
> > >>>> + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
> > >>>> err = device_register(&devfreq->dev);
> > >>>> if (err) {
> > >>>> mutex_unlock(&devfreq->lock);
> > >>>> --
> > >>>> 2.7.4
> > >>>>
> > >>>
> > >>> Thanks for this, I agree, this needs to get back to the way things were
> > >>> as it seems to break too many existing systems as-is.
> > >>>
> > >>> I'll queue this up in my tree now, thanks.
> > >>
> > >> Oof this old thing. I unfortunately didn't get back to look at the
> > >> devfreq name node issue or the compatibility links, since the impact
> > >> of the regression (breaking the powerHAL's interactions with the gpu)
> > >> wasn't as big as other problems we had. While the regression was
> > >> frustrating, my only hesitancy at this point is that its been this way
> > >> since 4.10, so reverting the problematic patch is likely to break any
> > >> new users since then.
> > >
> > > Looks like most users just revert that commit in their trees:
> > > https://protect2.fireeye.com/url?k=1012ad0f-4dc2a5e7-10132640-000babff32e3-35779c5ed675ef0f&u=https://source.codeaurora.org/quic/la/kernel/msm-4.14/commit/drivers/devfreq?h=msm-4.14&id=ccf273f6d89ad0fa8032e9225305ad6f62c7770c
> > >
> > > So we should be ok here.
> >
> > I'm sorry about changing the devfreq node name.
> >
> > OK. Do you pick this patch to your tree?
>
> Yes, I can do that.

If you agree, how about merging it to devfreq.git
for preventing the potential merge conflict with other devfreq patches?

>
> > or If not, I'll apply it to devfreq-next branch for v5.7-rc1.
> >
> > And do you apply it to kernel of linux-stable tree since 4.11?
>
> Yeah, I'll mark it for stable.

Thanks.

>
> Can I get an ack from you for this?

OK. but, if it will be merged to devfreq.git,
can I get an ack from you?

>
> thanks,
>
> greg k-h

--
Best Regards,
Chanwoo Choi
Samsung Electronics

2020-02-23 17:54:48

by Greg Kroah-Hartman

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On Sat, Feb 22, 2020 at 08:55:22AM +0900, Chanwoo Choi wrote:
> On Fri, Feb 21, 2020 at 8:13 PM Greg Kroah-Hartman
> <[email protected]> wrote:
> >
> > On Fri, Feb 21, 2020 at 05:11:02PM +0900, Chanwoo Choi wrote:
> > > On 2/21/20 4:06 PM, Greg Kroah-Hartman wrote:
> > > > On Thu, Feb 20, 2020 at 11:47:41AM -0800, John Stultz wrote:
> > > >> On Thu, Feb 20, 2020 at 11:15 AM Greg Kroah-Hartman
> > > >> <[email protected]> wrote:
> > > >>>
> > > >>> On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
> > > >>>> This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
> > > >>>>
> > > >>>> The name changing as devfreq(X) breaks some user space applications,
> > > >>>> such as Android HAL from Unisoc and Hikey [1].
> > > >>>> The device name will be changed unexpectly after every boot depending
> > > >>>> on module init sequence. It will make trouble to setup some system
> > > >>>> configuration like selinux for Android.
> > > >>>>
> > > >>>> So we'd like to revert it back to old naming rule before any better
> > > >>>> way being found.
> > > >>>>
> > > >>>> [1] https://protect2.fireeye.com/url?k=00fa721e-5d2a7af6-00fbf951-000babff32e3-95e4b92259b05656&u=https://lkml.org/lkml/2018/5/8/1042
> > > >>>>
> > > >>>> Cc: John Stultz <[email protected]>
> > > >>>> Cc: Greg Kroah-Hartman <[email protected]>
> > > >>>> Cc: [email protected]
> > > >>>> Signed-off-by: Orson Zhai <[email protected]>
> > > >>>>
> > > >>>> ---
> > > >>>> drivers/devfreq/devfreq.c | 4 +---
> > > >>>> 1 file changed, 1 insertion(+), 3 deletions(-)
> > > >>>>
> > > >>>> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> > > >>>> index cceee8b..7dcf209 100644
> > > >>>> --- a/drivers/devfreq/devfreq.c
> > > >>>> +++ b/drivers/devfreq/devfreq.c
> > > >>>> @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > > >>>> {
> > > >>>> struct devfreq *devfreq;
> > > >>>> struct devfreq_governor *governor;
> > > >>>> - static atomic_t devfreq_no = ATOMIC_INIT(-1);
> > > >>>> int err = 0;
> > > >>>>
> > > >>>> if (!dev || !profile || !governor_name) {
> > > >>>> @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
> > > >>>> devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
> > > >>>> atomic_set(&devfreq->suspend_count, 0);
> > > >>>>
> > > >>>> - dev_set_name(&devfreq->dev, "devfreq%d",
> > > >>>> - atomic_inc_return(&devfreq_no));
> > > >>>> + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
> > > >>>> err = device_register(&devfreq->dev);
> > > >>>> if (err) {
> > > >>>> mutex_unlock(&devfreq->lock);
> > > >>>> --
> > > >>>> 2.7.4
> > > >>>>
> > > >>>
> > > >>> Thanks for this, I agree, this needs to get back to the way things were
> > > >>> as it seems to break too many existing systems as-is.
> > > >>>
> > > >>> I'll queue this up in my tree now, thanks.
> > > >>
> > > >> Oof this old thing. I unfortunately didn't get back to look at the
> > > >> devfreq name node issue or the compatibility links, since the impact
> > > >> of the regression (breaking the powerHAL's interactions with the gpu)
> > > >> wasn't as big as other problems we had. While the regression was
> > > >> frustrating, my only hesitancy at this point is that its been this way
> > > >> since 4.10, so reverting the problematic patch is likely to break any
> > > >> new users since then.
> > > >
> > > > Looks like most users just revert that commit in their trees:
> > > > https://protect2.fireeye.com/url?k=1012ad0f-4dc2a5e7-10132640-000babff32e3-35779c5ed675ef0f&u=https://source.codeaurora.org/quic/la/kernel/msm-4.14/commit/drivers/devfreq?h=msm-4.14&id=ccf273f6d89ad0fa8032e9225305ad6f62c7770c
> > > >
> > > > So we should be ok here.
> > >
> > > I'm sorry about changing the devfreq node name.
> > >
> > > OK. Do you pick this patch to your tree?
> >
> > Yes, I can do that.
>
> If you agree, how about merging it to devfreq.git
> for preventing the potential merge conflict with other devfreq patches?

Sure, but it should go for 5.6-final, right?

> > > or If not, I'll apply it to devfreq-next branch for v5.7-rc1.
> > >
> > > And do you apply it to kernel of linux-stable tree since 4.11?
> >
> > Yeah, I'll mark it for stable.
>
> Thanks.
>
> >
> > Can I get an ack from you for this?
>
> OK. but, if it will be merged to devfreq.git,
> can I get an ack from you?

Acked-by: Greg Kroah-Hartman <[email protected]>

2020-02-24 02:09:29

by Chanwoo Choi

[permalink] [raw]

Subject: Re: [PATCH] Revert "PM / devfreq: Modify the device name as devfreq(X) for sysfs"

On 2/24/20 2:54 AM, Greg Kroah-Hartman wrote:
> On Sat, Feb 22, 2020 at 08:55:22AM +0900, Chanwoo Choi wrote:
>> On Fri, Feb 21, 2020 at 8:13 PM Greg Kroah-Hartman
>> <[email protected]> wrote:
>>>
>>> On Fri, Feb 21, 2020 at 05:11:02PM +0900, Chanwoo Choi wrote:
>>>> On 2/21/20 4:06 PM, Greg Kroah-Hartman wrote:
>>>>> On Thu, Feb 20, 2020 at 11:47:41AM -0800, John Stultz wrote:
>>>>>> On Thu, Feb 20, 2020 at 11:15 AM Greg Kroah-Hartman
>>>>>> <[email protected]> wrote:
>>>>>>>
>>>>>>> On Fri, Feb 21, 2020 at 01:37:04AM +0800, Orson Zhai wrote:
>>>>>>>> This reverts commit 4585fbcb5331fc910b7e553ad3efd0dd7b320d14.
>>>>>>>>
>>>>>>>> The name changing as devfreq(X) breaks some user space applications,
>>>>>>>> such as Android HAL from Unisoc and Hikey [1].
>>>>>>>> The device name will be changed unexpectly after every boot depending
>>>>>>>> on module init sequence. It will make trouble to setup some system
>>>>>>>> configuration like selinux for Android.
>>>>>>>>
>>>>>>>> So we'd like to revert it back to old naming rule before any better
>>>>>>>> way being found.
>>>>>>>>
>>>>>>>> [1] https://protect2.fireeye.com/url?k=00fa721e-5d2a7af6-00fbf951-000babff32e3-95e4b92259b05656&u=https://lkml.org/lkml/2018/5/8/1042
>>>>>>>>
>>>>>>>> Cc: John Stultz <[email protected]>
>>>>>>>> Cc: Greg Kroah-Hartman <[email protected]>
>>>>>>>> Cc: [email protected]
>>>>>>>> Signed-off-by: Orson Zhai <[email protected]>
>>>>>>>>
>>>>>>>> ---
>>>>>>>> drivers/devfreq/devfreq.c | 4 +---
>>>>>>>> 1 file changed, 1 insertion(+), 3 deletions(-)
>>>>>>>>
>>>>>>>> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
>>>>>>>> index cceee8b..7dcf209 100644
>>>>>>>> --- a/drivers/devfreq/devfreq.c
>>>>>>>> +++ b/drivers/devfreq/devfreq.c
>>>>>>>> @@ -738,7 +738,6 @@ struct devfreq *devfreq_add_device(struct device *dev,
>>>>>>>> {
>>>>>>>> struct devfreq *devfreq;
>>>>>>>> struct devfreq_governor *governor;
>>>>>>>> - static atomic_t devfreq_no = ATOMIC_INIT(-1);
>>>>>>>> int err = 0;
>>>>>>>>
>>>>>>>> if (!dev || !profile || !governor_name) {
>>>>>>>> @@ -800,8 +799,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
>>>>>>>> devfreq->suspend_freq = dev_pm_opp_get_suspend_opp_freq(dev);
>>>>>>>> atomic_set(&devfreq->suspend_count, 0);
>>>>>>>>
>>>>>>>> - dev_set_name(&devfreq->dev, "devfreq%d",
>>>>>>>> - atomic_inc_return(&devfreq_no));
>>>>>>>> + dev_set_name(&devfreq->dev, "%s", dev_name(dev));
>>>>>>>> err = device_register(&devfreq->dev);
>>>>>>>> if (err) {
>>>>>>>> mutex_unlock(&devfreq->lock);
>>>>>>>> --
>>>>>>>> 2.7.4
>>>>>>>>
>>>>>>>
>>>>>>> Thanks for this, I agree, this needs to get back to the way things were
>>>>>>> as it seems to break too many existing systems as-is.
>>>>>>>
>>>>>>> I'll queue this up in my tree now, thanks.
>>>>>>
>>>>>> Oof this old thing. I unfortunately didn't get back to look at the
>>>>>> devfreq name node issue or the compatibility links, since the impact
>>>>>> of the regression (breaking the powerHAL's interactions with the gpu)
>>>>>> wasn't as big as other problems we had. While the regression was
>>>>>> frustrating, my only hesitancy at this point is that its been this way
>>>>>> since 4.10, so reverting the problematic patch is likely to break any
>>>>>> new users since then.
>>>>>
>>>>> Looks like most users just revert that commit in their trees:
>>>>> https://protect2.fireeye.com/url?k=1012ad0f-4dc2a5e7-10132640-000babff32e3-35779c5ed675ef0f&u=https://source.codeaurora.org/quic/la/kernel/msm-4.14/commit/drivers/devfreq?h=msm-4.14&id=ccf273f6d89ad0fa8032e9225305ad6f62c7770c
>>>>>
>>>>> So we should be ok here.
>>>>
>>>> I'm sorry about changing the devfreq node name.
>>>>
>>>> OK. Do you pick this patch to your tree?
>>>
>>> Yes, I can do that.
>>
>> If you agree, how about merging it to devfreq.git
>> for preventing the potential merge conflict with other devfreq patches?
>
> Sure, but it should go for 5.6-final, right?

OK.I'll send pull-request of this patch for v5.6-rc4.

>
>>>> or If not, I'll apply it to devfreq-next branch for v5.7-rc1.
>>>>
>>>> And do you apply it to kernel of linux-stable tree since 4.11?
>>>
>>> Yeah, I'll mark it for stable.
>>
>> Thanks.
>>
>>>
>>> Can I get an ack from you for this?
>>
>> OK. but, if it will be merged to devfreq.git,
>> can I get an ack from you?
>
> Acked-by: Greg Kroah-Hartman <[email protected]>

Thanks. I applied it to devfreq-fixes branch for v5.6-rc4.

--
Best Regards,
Chanwoo Choi
Samsung Electronics