linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
To: Scott Branden <scott.branden@broadcom.com>
Cc: Ley Foon Tan <ley.foon.tan@intel.com>,
	Jan Kiszka <jan.kiszka@siemens.com>,
	linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org,
	Will Deacon <will.deacon@arm.com>,
	Bjorn Helgaas <bhelgaas@google.com>,
	Rob Herring <robh@kernel.org>,
	Russell King <linux@armlinux.org.uk>,
	Guenter Roeck <linux@roeck-us.net>
Subject: Re: [PATCH v2] PCI: controller: Move PCI_DOMAINS selection to arch Kconfig
Date: Tue, 26 Jun 2018 12:23:53 +0100	[thread overview]
Message-ID: <20180626112344.GA6850@red-moon> (raw)
In-Reply-To: <1cf91447-6e17-a7cb-c827-210835179db9@broadcom.com>

On Mon, Jun 25, 2018 at 11:48:11PM -0700, Scott Branden wrote:
> Hi Lorenzo,
> 
> 
> On 18-06-25 02:27 AM, Lorenzo Pieralisi wrote:
> >On Fri, Jun 22, 2018 at 11:59:08PM +0800, Ley Foon Tan wrote:
> >>On Thu, 2018-06-21 at 22:08 -0700, Scott Branden wrote:
> >>>
> >>>On 18-06-21 09:54 AM, Lorenzo Pieralisi wrote:
> >>>>
> >>>>On Wed, Jun 20, 2018 at 10:07:33AM +0200, Jan Kiszka wrote:
> >>>>>
> >>>>>On 2018-06-19 13:21, Lorenzo Pieralisi wrote:
> >>>>>>
> >>>>>>Commit 51bc085d6454 ("PCI: Improve host drivers compile test
> >>>>>>coverage")
> >>>>>>added configuration options to allow PCI host controller
> >>>>>>drivers to be
> >>>>>>compile tested on all architectures.
> >>>>>>
> >>>>>>Some host controller drivers (eg PCIE_ALTERA) config entries
> >>>>>>select
> >>>>>>the PCI_DOMAINS config option to enable PCI domains management
> >>>>>>in
> >>>>>>the kernel. Now that host controller drivers can be compiled on
> >>>>>>all architectures, this triggers build regressions on arches
> >>>>>>that
> >>>>>>do not implement the PCI_DOMAINS required API (ie
> >>>>>>pci_domain_nr()):
> >>>>>>
> >>>>>>drivers/ata/pata_ali.c: In function 'ali_init_chipset':
> >>>>>>drivers/ata/pata_ali.c:469:38: error:
> >>>>>>          implicit declaration of function 'pci_domain_nr';
> >>>>>>                                            did you mean
> >>>>>>'pci_iomap_wc'?
> >>>>>>
> >>>>>>Furthemore, some software configurations (ie Jailhouse) require
> >>>>>>a
> >>>>>>PCI_DOMAINS enabled kernel to configure multiple host
> >>>>>>controllers
> >>>>>>without having an explicit dependency on the ARM platform on
> >>>>>>which
> >>>>>>they run.
> >>>>>>
> >>>>>>Make PCI_DOMAINS a visible configuration option on ARM so that
> >>>>>>software
> >>>>>>configurations that need it can manually select it and move the
> >>>>>>PCI_DOMAINS selection from PCI controllers configuration file
> >>>>>>to ARM
> >>>>>>sub-arch config entries that currently require it, fixing the
> >>>>>>issue.
> >>>>>>
> >>>>>>Fixes: 51bc085d6454 ("PCI: Improve host drivers compile test
> >>>>>>coverage")
> >>>>>>Link: https://lkml.kernel.org/r/20180612170229.GA10141@roeck-us
> >>>>>>.net
> >>>>>>Reported-by: Guenter Roeck <linux@roeck-us.net>
> >>>>>>Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> >>>>>>Cc: Scott Branden <scott.branden@broadcom.com>
> >>>>>>Cc: Will Deacon <will.deacon@arm.com>
> >>>>>>Cc: Bjorn Helgaas <bhelgaas@google.com>
> >>>>>>Cc: Rob Herring <robh@kernel.org>
> >>>>>>Cc: Russell King <linux@armlinux.org.uk>
> >>>>>>Cc: Jan Kiszka <jan.kiszka@siemens.com>
> >>>>>>Cc: Guenter Roeck <linux@roeck-us.net>
> >>>>>>Cc: Ley Foon Tan <lftan@altera.com>
> >>>>>>---
> >>>>>>v1 -> v2
> >>>>>>     - Removed ARCH_VIRT PCI_DOMAINS selection
> >>>>>>     - Added PCI_DOMAINS visible config option
> >>>>>>
> >>>>>>v1: https://marc.info/?l=linux-pci&m=152932092612352&w=2
> >>>>>>
> >>>>>>   arch/arm/Kconfig               | 8 +++++++-
> >>>>>>   arch/arm/mach-bcm/Kconfig      | 1 +
> >>>>>>   arch/arm/mach-socfpga/Kconfig  | 1 +
> >>>>>>   drivers/pci/controller/Kconfig | 3 ---
> >>>>>>   4 files changed, 9 insertions(+), 4 deletions(-)
> >>>>>>
> >>>>>>diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
> >>>>>>index 54eeb8d00bc6..843edfd000be 100644
> >>>>>>--- a/arch/arm/Kconfig
> >>>>>>+++ b/arch/arm/Kconfig
> >>>>>>@@ -1245,8 +1245,14 @@ config PCI
> >>>>>>       VESA. If you have PCI, say Y, otherwise N.
> >>>>>>
> >>>>>>   config PCI_DOMAINS
> >>>>>>-   bool
> >>>>>>+   bool "Support for multiple PCI domains"
> >>>>>>     depends on PCI
> >>>>>>+   help
> >>>>>>+     Enable PCI domains kernel management. Say Y if your
> >>>>>>machine
> >>>>>>+     has a PCI bus hierarchy that requires more than one PCI
> >>>>>>+     domain (aka segment) to be correctly managed. Say N
> >>>>>>otherwise.
> >>>>>>+
> >>>>>>+     If you don't know what to do here, say N.
> >>>>>>
> >>>>>>   config PCI_DOMAINS_GENERIC
> >>>>>>     def_bool PCI_DOMAINS
> >>>>>>diff --git a/arch/arm/mach-bcm/Kconfig b/arch/arm/mach-
> >>>>>>bcm/Kconfig
> >>>>>>index c46a728df44e..25aac6ee2ab1 100644
> >>>>>>--- a/arch/arm/mach-bcm/Kconfig
> >>>>>>+++ b/arch/arm/mach-bcm/Kconfig
> >>>>>>@@ -20,6 +20,7 @@ config ARCH_BCM_IPROC
> >>>>>>     select GPIOLIB
> >>>>>>     select ARM_AMBA
> >>>>>>     select PINCTRL
> >>>>>>+   select PCI_DOMAINS if PCI
> >>>>>>     help
> >>>>>>       This enables support for systems based on Broadcom IPROC
> >>>>>>architected SoCs.
> >>>>>>       The IPROC complex contains one or more ARM CPUs along
> >>>>>>with common
> >>>>>>diff --git a/arch/arm/mach-socfpga/Kconfig b/arch/arm/mach-
> >>>>>>socfpga/Kconfig
> >>>>>>index d0f62eacf59d..4adb901dd5eb 100644
> >>>>>>--- a/arch/arm/mach-socfpga/Kconfig
> >>>>>>+++ b/arch/arm/mach-socfpga/Kconfig
> >>>>>>@@ -10,6 +10,7 @@ menuconfig ARCH_SOCFPGA
> >>>>>>     select HAVE_ARM_SCU
> >>>>>>     select HAVE_ARM_TWD if SMP
> >>>>>>     select MFD_SYSCON
> >>>>>>+   select PCI_DOMAINS if PCI
> >>>>>>
> >>>>>>   if ARCH_SOCFPGA
> >>>>>>   config SOCFPGA_SUSPEND
> >>>>>>diff --git a/drivers/pci/controller/Kconfig
> >>>>>>b/drivers/pci/controller/Kconfig
> >>>>>>index 18fa09b3ac8f..cc9fa02d32a0 100644
> >>>>>>--- a/drivers/pci/controller/Kconfig
> >>>>>>+++ b/drivers/pci/controller/Kconfig
> >>>>>>@@ -96,7 +96,6 @@ config PCI_HOST_GENERIC
> >>>>>>     depends on OF
> >>>>>>     select PCI_HOST_COMMON
> >>>>>>     select IRQ_DOMAIN
> >>>>>>-   select PCI_DOMAINS
> >>>>>>     help
> >>>>>>       Say Y here if you want to support a simple generic PCI
> >>>>>>host
> >>>>>>       controller, such as the one emulated by kvmtool.
> >>>>>>@@ -138,7 +137,6 @@ config PCI_VERSATILE
> >>>>>>
> >>>>>>   config PCIE_IPROC
> >>>>>>     tristate
> >>>>>>-   select PCI_DOMAINS
> >>>>>>     help
> >>>>>>       This enables the iProc PCIe core controller support for
> >>>>>>Broadcom's
> >>>>>>       iProc family of SoCs. An appropriate bus interface driver
> >>>>>>needs
> >>>>>>@@ -176,7 +174,6 @@ config PCIE_IPROC_MSI
> >>>>>>   config PCIE_ALTERA
> >>>>>>     bool "Altera PCIe controller"
> >>>>>>     depends on ARM || NIOS2 || COMPILE_TEST
> >>>>>>-   select PCI_DOMAINS
> >>>>>>     help
> >>>>>>       Say Y here if you want to enable PCIe controller support
> >>>>>>on Altera
> >>>>>>       FPGA.
> >>>>>>
> >>>>>Acked-by: Jan Kiszka <jan.kiszka@siemens.com>
> >>>>Thanks a lot. Scott, Ley Foon, does it work for you ? I would like
> >>>>to ask Bjorn to send it asap since the build on sparc is still
> >>>>broken.
> >>>Change will have to be made for arm64 for ARCH_BCM_IPROC as well.
> >>Yes, need to add it to arm64 too. Others are okay.
> >On arm64 CONFIG_PCI_DOMAINS defaults to CONFIG_PCI so there is no need
> >to select it, if that's all you need changed there is nothing to
> >change and this patch is read to be merged.
> I see that now: arm64 is doing things properly.  The same should be
> done for arm: default to CONFIG_PCI so there is no need to select
> it.
> I didn't like the change to have select to begin with.

I can't select PCI_DOMAINS by default on ARM 32-bit since this would
force it on platforms (eg non-DT legacy) that never relied on it; this
can easily trigger regressions. This patch is ready to go as-is, I do
not like it either but as I tried to explain it is the only reasonable
way forward.

Thanks,
Lorenzo

  reply	other threads:[~2018-06-26 11:22 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-06-19 11:21 [PATCH v2] PCI: controller: Move PCI_DOMAINS selection to arch Kconfig Lorenzo Pieralisi
2018-06-20  8:07 ` Jan Kiszka
2018-06-21 16:54   ` Lorenzo Pieralisi
2018-06-22  5:08     ` Scott Branden
2018-06-22 15:59       ` Ley Foon Tan
2018-06-25  9:27         ` Lorenzo Pieralisi
2018-06-26  6:48           ` Scott Branden
2018-06-26 11:23             ` Lorenzo Pieralisi [this message]
2018-06-26 15:59 ` Lorenzo Pieralisi
2018-06-26 16:50   ` Rob Herring
2018-06-26 17:10 ` Bjorn Helgaas

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180626112344.GA6850@red-moon \
    --to=lorenzo.pieralisi@arm.com \
    --cc=bhelgaas@google.com \
    --cc=jan.kiszka@siemens.com \
    --cc=ley.foon.tan@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=linux@armlinux.org.uk \
    --cc=linux@roeck-us.net \
    --cc=robh@kernel.org \
    --cc=scott.branden@broadcom.com \
    --cc=will.deacon@arm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).