Merge tag 'arc-4.6-rc7-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc

Pull ARC fixes from Vineet Gupta:
 "Late in the cycle, but this has fixes for couple of issues: a PAE40
  boot crash and Arnd spotting lack of barriers in BE io-accessors.

  The 3rd patch for enabling highmem in low physical mem ;-) honestly is
  more than a "fix" but its been in works for some time, seems to be
  stable in testing and enables 2 of our customers to go forward with
  4.6 kernel.

   - Fix for PTE truncation in PAE40 builds
   - Fix for big endian IO accessors lacking IO barrier
   - Allow HIGHMEM to work with low physical addresses"

* tag 'arc-4.6-rc7-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
  ARC: support HIGHMEM even without PAE40
  ARC: Fix PAE40 boot failures due to PTE truncation
  ARC: Add missing io barriers to io{read,write}{16,32}be()

arch/arc/Kconfig

@@ -58,6 +58,9 @@ config GENERIC_CSUM
 config RWSEM_GENERIC_SPINLOCK
 	def_bool y
 
+config ARCH_DISCONTIGMEM_ENABLE
+	def_bool y
+
 config ARCH_FLATMEM_ENABLE
 	def_bool y
 
@@ -347,6 +350,15 @@ config ARC_HUGEPAGE_16M
 
 endchoice
 
+config NODES_SHIFT
+	int "Maximum NUMA Nodes (as a power of 2)"
+	default "1" if !DISCONTIGMEM
+	default "2" if DISCONTIGMEM
+	depends on NEED_MULTIPLE_NODES
+	---help---
+	  Accessing memory beyond 1GB (with or w/o PAE) requires 2 memory
+	  zones.
+
 if ISA_ARCOMPACT
 
 config ARC_COMPACT_IRQ_LEVELS
@@ -455,6 +467,7 @@ config LINUX_LINK_BASE
 
 config HIGHMEM
 	bool "High Memory Support"
+	select DISCONTIGMEM
 	help
 	  With ARC 2G:2G address split, only upper 2G is directly addressable by
 	  kernel. Enable this to potentially allow access to rest of 2G and PAE

arch/arc/include/asm/io.h

@@ -13,6 +13,15 @@
 #include <asm/byteorder.h>
 #include <asm/page.h>
 
+#ifdef CONFIG_ISA_ARCV2
+#include <asm/barrier.h>
+#define __iormb()		rmb()
+#define __iowmb()		wmb()
+#else
+#define __iormb()		do { } while (0)
+#define __iowmb()		do { } while (0)
+#endif
+
 extern void __iomem *ioremap(phys_addr_t paddr, unsigned long size);
 extern void __iomem *ioremap_prot(phys_addr_t paddr, unsigned long size,
 				  unsigned long flags);
@@ -31,6 +40,15 @@ extern void iounmap(const void __iomem *addr);
 #define ioremap_wc(phy, sz)		ioremap(phy, sz)
 #define ioremap_wt(phy, sz)		ioremap(phy, sz)
 
+/*
+ * io{read,write}{16,32}be() macros
+ */
+#define ioread16be(p)		({ u16 __v = be16_to_cpu((__force __be16)__raw_readw(p)); __iormb(); __v; })
+#define ioread32be(p)		({ u32 __v = be32_to_cpu((__force __be32)__raw_readl(p)); __iormb(); __v; })
+
+#define iowrite16be(v,p)	({ __iowmb(); __raw_writew((__force u16)cpu_to_be16(v), p); })
+#define iowrite32be(v,p)	({ __iowmb(); __raw_writel((__force u32)cpu_to_be32(v), p); })
+
 /* Change struct page to physical address */
 #define page_to_phys(page)		(page_to_pfn(page) << PAGE_SHIFT)
 
@@ -108,15 +126,6 @@ static inline void __raw_writel(u32 w, volatile void __iomem *addr)
 
 }
 
-#ifdef CONFIG_ISA_ARCV2
-#include <asm/barrier.h>
-#define __iormb()		rmb()
-#define __iowmb()		wmb()
-#else
-#define __iormb()		do { } while (0)
-#define __iowmb()		do { } while (0)
-#endif
-
 /*
  * MMIO can also get buffered/optimized in micro-arch, so barriers needed
  * Based on ARM model for the typical use case

arch/arc/include/asm/mmzone.h

+/*
+ * Copyright (C) 2016 Synopsys, Inc. (www.synopsys.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef _ASM_ARC_MMZONE_H
+#define _ASM_ARC_MMZONE_H
+
+#ifdef CONFIG_DISCONTIGMEM
+
+extern struct pglist_data node_data[];
+#define NODE_DATA(nid) (&node_data[nid])
+
+static inline int pfn_to_nid(unsigned long pfn)
+{
+	int is_end_low = 1;
+
+	if (IS_ENABLED(CONFIG_ARC_HAS_PAE40))
+		is_end_low = pfn <= virt_to_pfn(0xFFFFFFFFUL);
+
+	/*
+	 * node 0: lowmem:             0x8000_0000   to 0xFFFF_FFFF
+	 * node 1: HIGHMEM w/o  PAE40: 0x0           to 0x7FFF_FFFF
+	 *         HIGHMEM with PAE40: 0x1_0000_0000 to ...
+	 */
+	if (pfn >= ARCH_PFN_OFFSET && is_end_low)
+		return 0;
+
+	return 1;
+}
+
+static inline int pfn_valid(unsigned long pfn)
+{
+	int nid = pfn_to_nid(pfn);
+
+	return (pfn <= node_end_pfn(nid));
+}
+#endif /* CONFIG_DISCONTIGMEM  */
+
+#endif

arch/arc/include/asm/page.h

@@ -72,11 +72,20 @@ typedef unsigned long pgprot_t;
 
 typedef pte_t * pgtable_t;
 
+/*
+ * Use virt_to_pfn with caution:
+ * If used in pte or paddr related macros, it could cause truncation
+ * in PAE40 builds
+ * As a rule of thumb, only use it in helpers starting with virt_
+ * You have been warned !
+ */
 #define virt_to_pfn(kaddr)	(__pa(kaddr) >> PAGE_SHIFT)
 
 #define ARCH_PFN_OFFSET		virt_to_pfn(CONFIG_LINUX_LINK_BASE)
 
+#ifdef CONFIG_FLATMEM
 #define pfn_valid(pfn)		(((pfn) - ARCH_PFN_OFFSET) < max_mapnr)
+#endif
 
 /*
  * __pa, __va, virt_to_page (ALERT: deprecated, don't use them)
@@ -85,12 +94,10 @@ typedef pte_t * pgtable_t;
  * virt here means link-address/program-address as embedded in object code.
  * And for ARC, link-addr = physical address
  */
-#define __pa(vaddr)  ((unsigned long)vaddr)
+#define __pa(vaddr)  ((unsigned long)(vaddr))
 #define __va(paddr)  ((void *)((unsigned long)(paddr)))
 
-#define virt_to_page(kaddr)	\
-	(mem_map + virt_to_pfn((kaddr) - CONFIG_LINUX_LINK_BASE))
-
+#define virt_to_page(kaddr)	pfn_to_page(virt_to_pfn(kaddr))
 #define virt_addr_valid(kaddr)  pfn_valid(virt_to_pfn(kaddr))
 
 /* Default Permissions for stack/heaps pages (Non Executable) */

arch/arc/include/asm/pgtable.h

@@ -278,14 +278,13 @@ static inline void pmd_set(pmd_t *pmdp, pte_t *ptep)
 #define pmd_present(x)			(pmd_val(x))
 #define pmd_clear(xp)			do { pmd_val(*(xp)) = 0; } while (0)
 
-#define pte_page(pte)	\
-	(mem_map + virt_to_pfn(pte_val(pte) - CONFIG_LINUX_LINK_BASE))
-
+#define pte_page(pte)		pfn_to_page(pte_pfn(pte))
 #define mk_pte(page, prot)	pfn_pte(page_to_pfn(page), prot)
-#define pte_pfn(pte)		virt_to_pfn(pte_val(pte))
-#define pfn_pte(pfn, prot)	(__pte(((pte_t)(pfn) << PAGE_SHIFT) | \
-				 pgprot_val(prot)))
-#define __pte_index(addr)	(virt_to_pfn(addr) & (PTRS_PER_PTE - 1))
+#define pfn_pte(pfn, prot)	(__pte(((pte_t)(pfn) << PAGE_SHIFT) | pgprot_val(prot)))
+
+/* Don't use virt_to_pfn for macros below: could cause truncations for PAE40*/
+#define pte_pfn(pte)		(pte_val(pte) >> PAGE_SHIFT)
+#define __pte_index(addr)	(((addr) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
 
 /*
  * pte_offset gets a @ptr to PMD entry (PGD in our 2-tier paging system)

arch/arc/mm/init.c

@@ -30,11 +30,16 @@ static const unsigned long low_mem_start = CONFIG_LINUX_LINK_BASE;
 static unsigned long low_mem_sz;
 
 #ifdef CONFIG_HIGHMEM
-static unsigned long min_high_pfn;
+static unsigned long min_high_pfn, max_high_pfn;
 static u64 high_mem_start;
 static u64 high_mem_sz;
 #endif
 
+#ifdef CONFIG_DISCONTIGMEM
+struct pglist_data node_data[MAX_NUMNODES] __read_mostly;
+EXPORT_SYMBOL(node_data);
+#endif
+
 /* User can over-ride above with "mem=nnn[KkMm]" in cmdline */
 static int __init setup_mem_sz(char *str)
 {
@@ -109,13 +114,11 @@ void __init setup_arch_memory(void)
 	/* Last usable page of low mem */
 	max_low_pfn = max_pfn = PFN_DOWN(low_mem_start + low_mem_sz);
 
-#ifdef CONFIG_HIGHMEM
-	min_high_pfn = PFN_DOWN(high_mem_start);
-	max_pfn = PFN_DOWN(high_mem_start + high_mem_sz);
+#ifdef CONFIG_FLATMEM
+	/* pfn_valid() uses this */
+	max_mapnr = max_low_pfn - min_low_pfn;
 #endif
 
-	max_mapnr = max_pfn - min_low_pfn;
-
 	/*------------- bootmem allocator setup -----------------------*/
 
 	/*
@@ -129,7 +132,7 @@ void __init setup_arch_memory(void)
 	 * the crash
 	 */
 
-	memblock_add(low_mem_start, low_mem_sz);
+	memblock_add_node(low_mem_start, low_mem_sz, 0);
 	memblock_reserve(low_mem_start, __pa(_end) - low_mem_start);
 
 #ifdef CONFIG_BLK_DEV_INITRD
@@ -149,13 +152,6 @@ void __init setup_arch_memory(void)
 	zones_size[ZONE_NORMAL] = max_low_pfn - min_low_pfn;
 	zones_holes[ZONE_NORMAL] = 0;
 
-#ifdef CONFIG_HIGHMEM
-	zones_size[ZONE_HIGHMEM] = max_pfn - max_low_pfn;
-
-	/* This handles the peripheral address space hole */
-	zones_holes[ZONE_HIGHMEM] = min_high_pfn - max_low_pfn;
-#endif
-
 	/*
 	 * We can't use the helper free_area_init(zones[]) because it uses
 	 * PAGE_OFFSET to compute the @min_low_pfn which would be wrong
@@ -168,6 +164,34 @@ void __init setup_arch_memory(void)
 			    zones_holes);	/* holes */
 
 #ifdef CONFIG_HIGHMEM
+	/*
+	 * Populate a new node with highmem
+	 *
+	 * On ARC (w/o PAE) HIGHMEM addresses are actually smaller (0 based)
+	 * than addresses in normal ala low memory (0x8000_0000 based).
+	 * Even with PAE, the huge peripheral space hole would waste a lot of
+	 * mem with single mem_map[]. This warrants a mem_map per region design.
+	 * Thus HIGHMEM on ARC is imlemented with DISCONTIGMEM.
+	 *
+	 * DISCONTIGMEM in turns requires multiple nodes. node 0 above is
+	 * populated with normal memory zone while node 1 only has highmem
+	 */
+	node_set_online(1);
+
+	min_high_pfn = PFN_DOWN(high_mem_start);
+	max_high_pfn = PFN_DOWN(high_mem_start + high_mem_sz);
+
+	zones_size[ZONE_NORMAL] = 0;
+	zones_holes[ZONE_NORMAL] = 0;
+
+	zones_size[ZONE_HIGHMEM] = max_high_pfn - min_high_pfn;
+	zones_holes[ZONE_HIGHMEM] = 0;
+
+	free_area_init_node(1,			/* node-id */
+			    zones_size,		/* num pages per zone */
+			    min_high_pfn,	/* first pfn of node */
+			    zones_holes);	/* holes */
+
 	high_memory = (void *)(min_high_pfn << PAGE_SHIFT);
 	kmap_init();
 #endif
@@ -185,7 +209,7 @@ void __init mem_init(void)
 	unsigned long tmp;
 
 	reset_all_zones_managed_pages();
-	for (tmp = min_high_pfn; tmp < max_pfn; tmp++)
+	for (tmp = min_high_pfn; tmp < max_high_pfn; tmp++)
 		free_highmem_page(pfn_to_page(tmp));
 #endif