Hi,

On Friday 30 May 2014 07:45 PM, Karicheri, Muralidharan wrote:

quoted

-----Original Message-----
From: Murali Karicheri [mailto:m-karicheri2 at ti.com]
Sent: Thursday, May 29, 2014 12:32 PM
To: ABRAHAM, KISHON VIJAY
Cc: devicetree at vger.kernel.org; linux-doc at vger.kernel.org; linux-arm-
kernel at lists.infradead.org; linux-omap at vger.kernel.org; linux-pci at vger.kernel.org; linux-
kernel at vger.kernel.org; arnd at arndb.de; tony at atomide.com; jg1.han at samsung.com;
Jason Gunthorpe; Bjorn Helgaas; Mohit Kumar; Marek Vasut
Subject: Re: [PATCH v2 03/18] PCI: designware: Configuration space should be specified
in 'reg'

On 5/29/2014 2:38 AM, ABRAHAM, KISHON VIJAY wrote:

quoted

The configuration address space has so far been specified in *ranges*,
however it should be specified in *reg* making it a platform MEM resource.
Hence used 'platform_get_resource_*' API to get configuration address
space in the designware driver.

Cc: Jason Gunthorpe <redacted>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Mohit Kumar <redacted>
Cc: Jingoo Han <redacted>
Cc: Marek Vasut <marex@denx.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kishon Vijay Abraham I <redacted>
---
  .../devicetree/bindings/pci/designware-pcie.txt    |    1 +
  drivers/pci/host/pcie-designware.c                 |   17 +++++++++++++++--
  2 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/Documentation/devicetree/bindings/pci/designware-pcie.txt

b/Documentation/devicetree/bindings/pci/designware-pcie.txt
index d6fae13..8314360 100644

--- a/Documentation/devicetree/bindings/pci/designware-pcie.txt
+++ b/Documentation/devicetree/bindings/pci/designware-pcie.txt

@@ -6,6 +6,7 @@ Required properties:
  	as "samsung,exynos5440-pcie" or "fsl,imx6q-pcie".
  - reg: base addresses and lengths of the pcie controller,
  	the phy controller, additional register for the phy controller.
+	The configuration address space should also be specified here.

Kishon,

I am working on the Keystone PCI driver for which v1 is already posted.
Want to clarify
following.
1. Original text for reg states "base addresses and lengths of the pcie controller,
        the phy controller, additional register for the phy controller"
and you added
        "The configuration address space should also be specified here"

   and the code below added resource name "config"

Does PCI designware follow some convention? Does it mean after applying this patch
config name is mandatory or optional? Below code you are not returning error. Can you or
author of PCI designware clarify what is expected to be present as mandatory and what is
optional.

Does config refers to RC's config space or EP's config space or both?
The code below divide
the size by 2. So it appears to be RC's + EP's config space. Please clarify.

quoted

  - interrupts: interrupt values for level interrupt,
  	pulse interrupt, special interrupt.
  - clocks: from common clock binding: handle to pci clock.

diff --git a/drivers/pci/host/pcie-designware.c

b/drivers/pci/host/pcie-designware.c
index c4e3732..603b386 100644

--- a/drivers/pci/host/pcie-designware.c
+++ b/drivers/pci/host/pcie-designware.c

@@ -20,6 +20,7 @@
  #include <linux/of_pci.h>
  #include <linux/pci.h>
  #include <linux/pci_regs.h>
+#include <linux/platform_device.h>
  #include <linux/types.h>

  #include "pcie-designware.h"

@@ -392,11 +393,23 @@ static const struct irq_domain_ops msi_domain_ops = {
  int __init dw_pcie_host_init(struct pcie_port *pp)
  {
  	struct device_node *np = pp->dev->of_node;
+	struct platform_device *pdev = to_platform_device(pp->dev);
  	struct of_pci_range range;
  	struct of_pci_range_parser parser;
+	struct resource *cfg_res;
  	u32 val;
  	int i;

+	cfg_res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "config");
+	if (cfg_res) {
+		pp->config.cfg0_size = resource_size(cfg_res)/2;
+		pp->config.cfg1_size = resource_size(cfg_res)/2;
+		pp->cfg0_base = cfg_res->start;
+		pp->cfg1_base = cfg_res->start + pp->config.cfg0_size;
+	} else {
+		dev_err(pp->dev, "missing *config* reg space\n");

This should return error -EINVAL.

Just read the other thread and Grant Likely suggested the host controller
driver should be backward compatible [1]. So we can't return -EINVAL here.
So I'd assume this patch is fine as is? Arnd? Jingoo?

[1] -> https://lkml.org/lkml/2014/6/3/124

Thanks
Kishon