From mboxrd@z Thu Jan 1 00:00:00 1970 From: Aaron Lu Subject: Re: [PATCH] ACPI / driver core: Store a device pointer in struct acpi_dev_node Date: Wed, 13 Nov 2013 14:57:16 +0800 Message-ID: <528322CC.2000508@intel.com> References: <3268437.YsusHvklcv@vostro.rjw.lan> <52803124.7000304@intel.com> <1683079.zjqzsOAjXB@vostro.rjw.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <1683079.zjqzsOAjXB@vostro.rjw.lan> Sender: linux-pci-owner@vger.kernel.org To: "Rafael J. Wysocki" , Lan Tianyu Cc: ACPI Devel Maling List , LKML , Linux PCI , Greg Kroah-Hartman , Bjorn Helgaas , Jarkko Nikula , Mika Westerberg , "Luck, Tony" List-Id: linux-acpi@vger.kernel.org On 11/11/2013 09:45 PM, Rafael J. Wysocki wrote: > On Monday, November 11, 2013 09:21:40 AM Lan Tianyu wrote: >> On 2013=E5=B9=B411=E6=9C=8810=E6=97=A5 08:58, Rafael J. Wysocki wrot= e: >>> From: Rafael J. Wysocki >>> >>> Modify struct acpi_dev_node to contain a pointer to struct device >>> ambedded in the struct acpi_device associated with the given device >>> object (that is, its ACPI companion device) instead of an ACPI hand= le >>> corresponding to that struct acpi_device. Introduce two new macros >>> for manipulating that pointer in a CONFIG_ACPI-safe way, >>> ACPI_COMPANION() and ACPI_COMPANION_SET(), and rework the >>> ACPI_HANDLE() macro to take the above changes into account. >>> Drop the ACPI_HANDLE_SET() macro entirely and rework its users to >>> use ACPI_COMPANION_SET() instead. For some of them who used to >>> pass the result of acpi_get_child() directly to ACPI_HANDLE_SET() >>> introduce a helper routine acpi_preset_companion() doing an >>> equivalent thing. >>> >>> The rationale for using a struct device pointer instead of a >>> struct acpi_device one as the member of struct acpi_dev_node is >>> that it allows device.h to avoid including linux/acpi.h which would >>> introduce quite a bit of compilation overhead for stuff that doesn'= t >>> care about ACPI. >>> In turn, moving the macros to linux/acpi.h forces >>> the stuff that does care about ACPI to include that file as >>> appropriate anyway. >> >> How about declaring "struct acpi_device" in the device.h? This can h= elp >> to use struct acpi_device without including linux/acpi.h. >> >> struct iommu_ops and struct iommu_group have been used by the same w= ay >> in the device.h. >=20 > Yes, they are. Well, that appears to work too. >=20 > Updated patch is appended. It also contains some fixes for problems = reported > by the auto build system and it's been tested on x86-64 now, so it sh= ould be > reasonably close to final. >=20 > Thanks, > Rafael >=20 >=20 > --- > From: Rafael J. Wysocki > Subject: ACPI / driver core: Store an ACPI device pointer in struct a= cpi_dev_node >=20 > Modify struct acpi_dev_node to contain a pointer to struct acpi_devic= e > associated with the given device object (that is, its ACPI companion > device) instead of an ACPI handle corresponding to it. Introduce two > new macros for manipulating that pointer in a CONFIG_ACPI-safe way, > ACPI_COMPANION() and ACPI_COMPANION_SET(), and rework the > ACPI_HANDLE() macro to take the above changes into account. > Drop the ACPI_HANDLE_SET() macro entirely and rework its users to > use ACPI_COMPANION_SET() instead. For some of them who used to > pass the result of acpi_get_child() directly to ACPI_HANDLE_SET() > introduce a helper routine acpi_preset_companion() doing an > equivalent thing. >=20 > The main motivation for doing this is that there are things > represented by struct acpi_device objects that don't have valid > ACPI handles (so called fixed ACPI hardware features, such as > power and sleep buttons) and we would like to create platform > device objects for them and "glue" them to their ACPI companions > in the usual way (which currently is impossible due to the > lack of valid ACPI handles). However, there are more reasons > why it may be useful. >=20 > First, struct acpi_device pointers allow of much better type checking > than void pointers which are ACPI handles, so it should be more > difficult to write buggy code using modified struct acpi_dev_node > and the new macros. Second, the change should help to reduce (over > time) the number of places in which the result of ACPI_HANDLE() is > passed to acpi_bus_get_device() in order to obtain a pointer to the > struct acpi_device associated with the given "physical" device, > because now that pointer is returned by ACPI_COMPANION() directly. > Finally, the change should make it easier to write generic code that > will build both for CONFIG_ACPI set and unset without adding explicit > compiler directives to it. >=20 > Signed-off-by: Rafael J. Wysocki Reviewed-by: Aaron Lu for ATA and SDIO part. Thanks, Aaron