netdev - 答复: [PATCH 1/2] PM: runtime: Add a general runtime get sync operation to deal with usage counter

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <d05e3d35a68e41e2ac36acfcd577ad47@huawei.com>
Date:   Mon, 9 Nov 2020 13:24:42 +0000
From:   zhangqilong <zhangqilong3@...wei.com>
To:     "Rafael J. Wysocki" <rafael@...nel.org>
CC:     "Rafael J. Wysocki" <rjw@...ysocki.net>,
        "fugang.duan@....com" <fugang.duan@....com>,
        David Miller <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Linux PM <linux-pm@...r.kernel.org>,
        netdev <netdev@...r.kernel.org>
Subject: 答复: [PATCH 1/2] PM: runtime: Add a general runtime get sync operation to deal with usage counter

Hi
> 
> On Mon, Nov 9, 2020 at 9:05 AM Zhang Qilong <zhangqilong3@...wei.com>
> wrote:
> >
> > In many case, we need to check return value of pm_runtime_get_sync,
> > but it brings a trouble to the usage counter processing. Many callers
> > forget to decrease the usage counter when it failed. It has been
> > discussed a lot[0][1]. So we add a function to deal with the usage
> > counter for better coding.
> >
> > [0]https://lkml.org/lkml/2020/6/14/88
> > [1]https://patchwork.ozlabs.org/project/linux-tegra/patch/202005200951
> > 48.10995-1-dinghao.liu@....edu.cn/
> > Signed-off-by: Zhang Qilong <zhangqilong3@...wei.com>
> > ---
> >  include/linux/pm_runtime.h | 32 ++++++++++++++++++++++++++++++++
> >  1 file changed, 32 insertions(+)
> >
> > diff --git a/include/linux/pm_runtime.h b/include/linux/pm_runtime.h
> > index 4b708f4e8eed..2b0af5b1dffd 100644
> > --- a/include/linux/pm_runtime.h
> > +++ b/include/linux/pm_runtime.h
> > @@ -386,6 +386,38 @@ static inline int pm_runtime_get_sync(struct device
> *dev)
> >         return __pm_runtime_resume(dev, RPM_GET_PUT);  }
> >
> > +/**
> > + * gene_pm_runtime_get_sync - Bump up usage counter of a device and
> resume it.
> > + * @dev: Target device.
> 
> The force argument is not documented.

(1) Good catch, I will add it in next version.

> 
> > + *
> > + * Increase runtime PM usage counter of @dev first, and carry out
> > + runtime-resume
> > + * of it synchronously. If __pm_runtime_resume return negative
> > + value(device is in
> > + * error state) or return positive value(the runtime of device is
> > + already active)
> > + * with force is true, it need decrease the usage counter of the
> > + device when
> > + * return.
> > + *
> > + * The possible return values of this function is zero or negative value.
> > + * zero:
> > + *    - it means success and the status will store the resume operation
> status
> > + *      if needed, the runtime PM usage counter of @dev remains
> incremented.
> > + * negative:
> > + *    - it means failure and the runtime PM usage counter of @dev has
> been
> > + *      decreased.
> > + * positive:
> > + *    - it means the runtime of the device is already active before that. If
> > + *      caller set force to true, we still need to decrease the usage
> counter.
> 
> Why is this needed?

(2) If caller set force, it means caller will return even the device has already been active
(__pm_runtime_resume return positive value) after calling gene_pm_runtime_get_sync,
we still need to decrease the usage count.

> 
> > + */
> > +static inline int gene_pm_runtime_get_sync(struct device *dev, bool
> > +force)
> 
> The name is not really a good one and note that pm_runtime_get() has the
> same problem as _get_sync() (ie. the usage counter is incremented regardless
> of the return value).
> 

(3) I have not thought a good name now, if you have good ideas, welcome.


Thanks, 
Zhang

> > +{
> > +       int ret = 0;
> > +
> > +       ret = __pm_runtime_resume(dev, RPM_GET_PUT);
> > +       if (ret < 0 || (ret > 0 && force))
> > +               pm_runtime_put_noidle(dev);
> > +
> > +       return ret;
> > +}
> > +
> >  /**
> >   * pm_runtime_put - Drop device usage counter and queue up "idle check"
> if 0.
> >   * @dev: Target device.
> > --
> 
> Thanks!