linux-kernel - Re: [PATCH v1 1/3] PM: runtime: Wrapper macros for ACQUIRE()/ACQUIRE

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] [day] [month] [year] [list]
Message-ID: <aRINlvX2GpW3Ue94@lizhi-Precision-Tower-5810>
Date: Mon, 10 Nov 2025 11:06:46 -0500
From: Frank Li <Frank.li@....com>
To: "Rafael J. Wysocki" <rafael@...nel.org>
Cc: Linux PM <linux-pm@...r.kernel.org>,
	Linux ACPI <linux-acpi@...r.kernel.org>,
	Jonathan Cameron <jonathan.cameron@...wei.com>,
	Takashi Iwai <tiwai@...e.de>, LKML <linux-kernel@...r.kernel.org>,
	Zhang Qilong <zhangqilong3@...wei.com>, Dhruva Gole <d-gole@...com>,
	Dan Williams <dan.j.williams@...el.com>,
	Linux PCI <linux-pci@...r.kernel.org>,
	Bjorn Helgaas <helgaas@...nel.org>,
	Alex Williamson <alex.williamson@...hat.com>
Subject: Re: [PATCH v1 1/3] PM: runtime: Wrapper macros for
 ACQUIRE()/ACQUIRE_ERR()

On Fri, Nov 07, 2025 at 07:39:55PM +0100, Rafael J. Wysocki wrote:
> From: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
>
> Add several wrapper macros for ACQUIRE()/ACQUIRE_ERR() and runtime PM
> usage counter guards introduced recently: pm_runtime_active_try,
> pm_runtime_active_auto_try, pm_runtime_active_try_enabled, and
> pm_runtime_active_auto_try_enabled.
>
> The new macros are simpler and should be more straightforward to use.
> Moreover, they do not expose internal details that are not strictly
> related to the code using the macros.
>
> For example, they can be used for rewriting a piece of code like below:
>
>         ACQUIRE(pm_runtime_active_try, pm)(dev);
>         if ((ret = ACQUIRE_ERR(pm_runtime_active_try, &pm)))
>                 return ret;
>
> in the following way:
>
>         PM_RUNTIME_ACQUIRE(dev);
>         if ((ret = PM_RUNTIME_ACQUIRE_ERR))

Personally, I feel like PM_RUNTIME_ACQUIRE_ERR hide too much informaiton.

There are not clear connection between PM_RUNTIME_ACQUIRE and
PM_RUNTIME_ACQUIRE_ERR. but previous code, the 'pm' is good connector.

Frank
>                 return ret;
>
> If the original code does not care about the specific error code
> returned when attempting to resume the device:
>
>         ACQUIRE(pm_runtime_active_try, pm)(dev);
>         if (ACQUIRE_ERR(pm_runtime_active_try, &pm))
>                 return -ENXIO;
>
> it may be changed like this:
>
>         PM_RUNTIME_ACQUIRE(dev);
>         if (PM_RUNTIME_ACQUIRE_ERR)
>                 return -ENXIO;
>
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
> ---
>  include/linux/pm_runtime.h |   55 +++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 55 insertions(+)
>
> --- a/include/linux/pm_runtime.h
> +++ b/include/linux/pm_runtime.h
> @@ -637,6 +637,61 @@ DEFINE_GUARD_COND(pm_runtime_active_auto
>  DEFINE_GUARD_COND(pm_runtime_active_auto, _try_enabled,
>  		  pm_runtime_resume_and_get(_T), _RET == 0)
>
> +/*
> + * ACQUIRE() wrapper macros for the guards defined above.
> + *
> + * The tagged __PM_RUNTIME_ACQUIRE*() variants are for the cases in which two or
> + * more of these macros are used in the same scope and the tags are necessary to
> + * distinguish the internal guard variables from each other. Don't do that
> + * unless you have to. No, really. If they are needed, using simple tags is
> + * recommended (for example, individual digits or letters).
> + *
> + * The simpler PM_RUNTIME_ACQUIRE*() variants are wrappers around the
> + * corresponding __PM_RUNTIME_ACQUIRE*() that use the underline character
> + * as a (special) tag.  They should be suitable for the vast majority of use
> + * cases.
> + *
> + * Don't mix up PM_RUNTIME_ACQUIRE*() with __PM_RUNTIME_ACQUIRE*() even though
> + * that may work.
> + */
> +#define __PM_RUNTIME_ACQUIRE(dev, tag)	\
> +	ACQUIRE(pm_runtime_active_try, _pm_runtime_guard_var_##tag)(dev)
> +
> +#define PM_RUNTIME_ACQUIRE(dev)	\
> +	__PM_RUNTIME_ACQUIRE(dev, _)
> +
> +#define __PM_RUNTIME_ACQUIRE_AUTOSUSPEND(dev, tag)	\
> +	ACQUIRE(pm_runtime_active_auto_try, _pm_runtime_guard_var_##tag)(dev)
> +
> +#define PM_RUNTIME_ACQUIRE_AUTOSUSPEND(dev)	\
> +	__PM_RUNTIME_ACQUIRE_AUTOSUSPEND(dev, _)
> +
> +#define __PM_RUNTIME_ACQUIRE_ENABLED(dev, tag)	\
> +	ACQUIRE(pm_runtime_active_try_enabled, _pm_runtime_guard_var_##tag)(dev)
> +
> +#define PM_RUNTIME_ACQUIRE_ENABLED(dev)	\
> +	__PM_RUNTIME_ACQUIRE_ENABLED(dev, _)
> +
> +#define __PM_RUNTIME_ACQUIRE_ENABLED_AUTOSUSPEND(dev, tag)	\
> +	ACQUIRE(pm_runtime_active_auto_try_enabled, _pm_runtime_guard_var_##tag)(dev)
> +
> +#define PM_RUNTIME_ACQUIRE_ENABLED_AUTOSUSPEND(dev)	\
> +	__PM_RUNTIME_ACQUIRE_ENABLED_AUTOSUSPEND(dev, _)
> +
> +/*
> + * ACQUIRE_ERR() wrapper macros for guard pm_runtime_active.
> + *
> + * Always check __PM_RUNTIME_ACQUIRE_ERR() with a matching tag after using one
> + * of the tagged __PM_RUNTIME_ACQUIRE*() macros defined above (yes, it can be
> + * used with any of them) and avoid accessing the given device if it is nonzero.
> + * Analogously, always check PM_RUNTIME_ACQUIRE_ERR after using any of the
> + * simpler PM_RUNTIME_ACQUIRE*() macros.
> + */
> +#define __PM_RUNTIME_ACQUIRE_ERR(tag)	\
> +	ACQUIRE_ERR(pm_runtime_active, &_pm_runtime_guard_var_##tag)
> +
> +#define PM_RUNTIME_ACQUIRE_ERR	__PM_RUNTIME_ACQUIRE_ERR(_)
> +
>  /**
>   * pm_runtime_put_sync - Drop device usage counter and run "idle check" if 0.
>   * @dev: Target device.
>
>
>