[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <af6afa64-27b5-3947-387c-5162f190dc3f@huawei.com>
Date: Tue, 18 Dec 2018 15:44:23 +0800
From: Hanjun Guo <guohanjun@...wei.com>
To: Sinan Kaya <okaya@...nel.org>, <linux-acpi@...r.kernel.org>
CC: Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
Hanjun Guo <hanjun.guo@...aro.org>,
Sudeep Holla <sudeep.holla@....com>,
"Rafael J. Wysocki" <rjw@...ysocki.net>,
Len Brown <lenb@...nel.org>,
open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v11 6/7] ACPI/IORT: Stub out ACS functions when CONFIG_PCI
is not set
On 2018/12/18 10:56, Sinan Kaya wrote:
> Remove PCI dependent code out of iort.c when CONFIG_PCI is not defined.
> A quick search reveals the following functions:
> 1. pci_request_acs()
> 2. pci_domain_nr()
> 3. pci_is_root_bus()
> 4. to_pci_dev()
>
> Both pci_domain_nr() and pci_is_root_bus() are defined in linux/pci.h.
> pci_domain_nr() is a stub function when CONFIG_PCI is not set and
> pci_is_root_bus() just returns a reference to a structure member which
> is still valid without CONFIG_PCI set.
>
> to_pci_dev() is a macro that expands to container_of.
>
> pci_request_acs() is the only code that gets pulled in from drivers/pci/*.c
Actually we have pci_for_each_dma_alias() too which is from
drivers/pci/search.c without stub function in linux/pci.h, but I
didn't get the compile error at link time, I think the compiler
just do some optimization to remove the dead code because
dev_is_pci() is obvious false.
Thanks
Hanjun
Powered by blists - more mailing lists