Merge branch 'cleanup/io-pci' into next/cleanups

From Rob Herring <robherring2@gmail.com>:

This is the 2nd part of mach/io.h removals. This series removes io.h on
platforms with PCI by creating a fixed virtual I/O mapping and a common
__io() macro.

This version has changed a bit to accommodate Tegra converting its PCIe
host to a platform driver. Now the virtual space is only reserved during
early boot before .map_io() is called. The mapping is not created until
calling pci_ioremap_io which can be done at any point after vmalloc is
initialized.

I've gone back to fixed 64K windows for each PCI bus. This allows
removing all the i/o resource setup from the individually platforms and
placing it within the common ARM PCI code.

I've only tested versatilepb under qemu (with the model hacked up to
actually enable i/o space), so any testing is appreciated. iop3xx and
mv78xx0 have some risk of breaking as the PCI bus addresses are moved
to 0 from matching the cpu host bus addesss.

* cleanup/io-pci:
  ARM: iop3xx: use fixed PCI i/o mapping
  ARM: mv78xx0: use fixed pci i/o mapping
  ARM: iop13xx: use fixed PCI i/o mapping
  iop13xx: use more regular PCI I/O space handling
  ARM: orion5x: use fixed PCI i/o mapping
  ARM: kirkwood: use fixed PCI i/o mapping
  ARM: dove: use fixed PCI i/o mapping
  ARM: footbridge: use fixed PCI i/o mapping
  ARM: shark: use fixed PCI i/o mapping
  ARM: integrator: remove trailing whitespace on pci_v3.c
  ARM: integrator: use fixed PCI i/o mapping
  ARM: tegra: use fixed PCI i/o mapping
  ARM: versatile: use fixed PCI i/o mapping
  ARM: move PCI i/o resource setup into common code
  ARM: Add fixed PCI i/o mapping
  i2c: iop3xx: use standard gpiolib functions
  i2c: iop3xx: clean-up trailing whitespace

Signed-off-by: Arnd Bergmann <arnd@arndb.de>

2 files changed, 57 insertions, 11 deletions

diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c
index 566750fa57d4..9d869f93a3da 100644
--- a/arch/arm/mm/ioremap.c
+++ b/arch/arm/mm/ioremap.c
@@ -36,6 +36,7 @@
 #include <asm/system_info.h>
 
 #include <asm/mach/map.h>
+#include <asm/mach/pci.h>
 #include "mm.h"
 
 int ioremap_page(unsigned long virt, unsigned long phys,
@@ -383,3 +384,16 @@ void __arm_iounmap(volatile void __iomem *io_addr)
 	arch_iounmap(io_addr);
 }
 EXPORT_SYMBOL(__arm_iounmap);
+
+#ifdef CONFIG_PCI
+int pci_ioremap_io(unsigned int offset, phys_addr_t phys_addr)
+{
+	BUG_ON(offset + SZ_64K > IO_SPACE_LIMIT);
+
+	return ioremap_page_range(PCI_IO_VIRT_BASE + offset,
+				  PCI_IO_VIRT_BASE + offset + SZ_64K,
+				  phys_addr,
+				  __pgprot(get_mem_type(MT_DEVICE)->prot_pte));
+}
+EXPORT_SYMBOL_GPL(pci_ioremap_io);
+#endif
diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c
index 4c2d0451e84a..512b2c042ce1 100644
--- a/arch/arm/mm/mmu.c
+++ b/arch/arm/mm/mmu.c
@@ -31,6 +31,7 @@
 
 #include <asm/mach/arch.h>
 #include <asm/mach/map.h>
+#include <asm/mach/pci.h>
 
 #include "mm.h"
 
@@ -216,7 +217,7 @@ static struct mem_type mem_types[] = {
 		.prot_l1	= PMD_TYPE_TABLE,
 		.prot_sect	= PROT_SECT_DEVICE | PMD_SECT_WB,
 		.domain		= DOMAIN_IO,
-	},	
+	},
 	[MT_DEVICE_WC] = {	/* ioremap_wc */
 		.prot_pte	= PROT_PTE_DEVICE | L_PTE_MT_DEV_WC,
 		.prot_l1	= PMD_TYPE_TABLE,
@@ -777,14 +778,27 @@ void __init iotable_init(struct map_desc *io_desc, int nr)
 		create_mapping(md);
 		vm->addr = (void *)(md->virtual & PAGE_MASK);
 		vm->size = PAGE_ALIGN(md->length + (md->virtual & ~PAGE_MASK));
-		vm->phys_addr = __pfn_to_phys(md->pfn); 
-		vm->flags = VM_IOREMAP | VM_ARM_STATIC_MAPPING; 
+		vm->phys_addr = __pfn_to_phys(md->pfn);
+		vm->flags = VM_IOREMAP | VM_ARM_STATIC_MAPPING;
 		vm->flags |= VM_ARM_MTYPE(md->type);
 		vm->caller = iotable_init;
 		vm_area_add_early(vm++);
 	}
 }
 
+void __init vm_reserve_area_early(unsigned long addr, unsigned long size,
+				  void *caller)
+{
+	struct vm_struct *vm;
+
+	vm = early_alloc_aligned(sizeof(*vm), __alignof__(*vm));
+	vm->addr = (void *)addr;
+	vm->size = size;
+	vm->flags = VM_IOREMAP | VM_ARM_STATIC_MAPPING;
+	vm->caller = caller;
+	vm_area_add_early(vm);
+}
+
 #ifndef CONFIG_ARM_LPAE
 
 /*
@@ -802,14 +816,7 @@ void __init iotable_init(struct map_desc *io_desc, int nr)
 
 static void __init pmd_empty_section_gap(unsigned long addr)
 {
-	struct vm_struct *vm;
-
-	vm = early_alloc_aligned(sizeof(*vm), __alignof__(*vm));
-	vm->addr = (void *)addr;
-	vm->size = SECTION_SIZE;
-	vm->flags = VM_IOREMAP | VM_ARM_STATIC_MAPPING;
-	vm->caller = pmd_empty_section_gap;
-	vm_area_add_early(vm);
+	vm_reserve_area_early(addr, SECTION_SIZE, pmd_empty_section_gap);
 }
 
 static void __init fill_pmd_gaps(void)
@@ -858,6 +865,28 @@ static void __init fill_pmd_gaps(void)
 #define fill_pmd_gaps() do { } while (0)
 #endif
 
+#if defined(CONFIG_PCI) && !defined(CONFIG_NEED_MACH_IO_H)
+static void __init pci_reserve_io(void)
+{
+	struct vm_struct *vm;
+	unsigned long addr;
+
+	/* we're still single threaded hence no lock needed here */
+	for (vm = vmlist; vm; vm = vm->next) {
+		if (!(vm->flags & VM_ARM_STATIC_MAPPING))
+			continue;
+		addr = (unsigned long)vm->addr;
+		addr &= ~(SZ_2M - 1);
+		if (addr == PCI_IO_VIRT_BASE)
+			return;
+
+	}
+	vm_reserve_area_early(PCI_IO_VIRT_BASE, SZ_2M, pci_reserve_io);
+}
+#else
+#define pci_reserve_io() do { } while (0)
+#endif
+
 static void * __initdata vmalloc_min =
 	(void *)(VMALLOC_END - (240 << 20) - VMALLOC_OFFSET);
 
@@ -1141,6 +1170,9 @@ static void __init devicemaps_init(struct machine_desc *mdesc)
 		mdesc->map_io();
 	fill_pmd_gaps();
 
+	/* Reserve fixed i/o space in VMALLOC region */
+	pci_reserve_io();
+
 	/*
 	 * Finally flush the caches and tlb to ensure that we're in a
 	 * consistent state wrt the writebuffer.  This also ensures that