From d9676fa152c83b82137af950b1d4f629045d90c9 Mon Sep 17 00:00:00 2001 From: Evgeny Voevodin Date: Wed, 23 Mar 2016 12:26:52 +0300 Subject: ARCv2: Enable LOCKDEP - The asm helpers for calling into irq tracer were missing - Add calls to above helpers in low level assembly entry code for ARCv2 - irq_save() uses CLRI to disable interrupts and returns the prev interrupt state (in STATUS32) in a specific encoding (and not the raw value of STATUS32). This is usable with SETI in irq_restore(). However save_flags() reads the raw value of STATUS32 which doesn't pair with irq_save/restore() and thus needs fixing. Signed-off-by: Evgeny Voevodin [vgupta: updated changelog and also added some comments] Signed-off-by: Vineet Gupta --- arch/arc/include/asm/irqflags-arcv2.h | 36 ++++++++++++++++++++++++++++++++++- arch/arc/kernel/entry-arcv2.S | 10 +++++++++- arch/arc/kernel/entry-compact.S | 3 +++ 3 files changed, 47 insertions(+), 2 deletions(-) (limited to 'arch/arc') diff --git a/arch/arc/include/asm/irqflags-arcv2.h b/arch/arc/include/asm/irqflags-arcv2.h index 37c2f75..d1ec7f6 100644 --- a/arch/arc/include/asm/irqflags-arcv2.h +++ b/arch/arc/include/asm/irqflags-arcv2.h @@ -18,6 +18,12 @@ #define STATUS_AD_MASK (1< Date: Tue, 26 Apr 2016 19:29:33 +0300 Subject: ARC: support generic per-device coherent dma mem Signed-off-by: Alexey Brodkin Cc: linux-kernel@vger.kernel.org Signed-off-by: Vineet Gupta --- arch/arc/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/arc') diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig index 12d0284..28d47f8 100644 --- a/arch/arc/Kconfig +++ b/arch/arc/Kconfig @@ -37,6 +37,7 @@ config ARC select OF_EARLY_FLATTREE select PERF_USE_VMALLOC select HAVE_DEBUG_STACKOVERFLOW + select HAVE_GENERIC_DMA_COHERENT config MIGHT_HAVE_PCI bool -- cgit v1.1 From 1b10cb21d888c021bedbe678f7c26aee1bf04ffa Mon Sep 17 00:00:00 2001 From: Alexey Brodkin Date: Tue, 26 Apr 2016 19:29:34 +0300 Subject: ARC: add support for reserved memory defined by device tree Enable reserved memory initialization from device tree. Signed-off-by: Alexey Brodkin Cc: Grant Likely Cc: Marek Szyprowski Cc: linux-kernel@vger.kernel.org Signed-off-by: Vineet Gupta --- arch/arc/Kconfig | 1 + arch/arc/mm/init.c | 4 ++++ 2 files changed, 5 insertions(+) (limited to 'arch/arc') diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig index 28d47f8..ec4791e 100644 --- a/arch/arc/Kconfig +++ b/arch/arc/Kconfig @@ -35,6 +35,7 @@ config ARC select NO_BOOTMEM select OF select OF_EARLY_FLATTREE + select OF_RESERVED_MEM select PERF_USE_VMALLOC select HAVE_DEBUG_STACKOVERFLOW select HAVE_GENERIC_DMA_COHERENT diff --git a/arch/arc/mm/init.c b/arch/arc/mm/init.c index 7d2c4fb..5487d0b 100644 --- a/arch/arc/mm/init.c +++ b/arch/arc/mm/init.c @@ -13,6 +13,7 @@ #ifdef CONFIG_BLK_DEV_INITRD #include #endif +#include #include #include #include @@ -136,6 +137,9 @@ void __init setup_arch_memory(void) memblock_reserve(__pa(initrd_start), initrd_end - initrd_start); #endif + early_init_fdt_reserve_self(); + early_init_fdt_scan_reserved_mem(); + memblock_dump_all(); /*----------------- node/zones setup --------------------------*/ -- cgit v1.1 From e5bc0478ab6cf565619224536d75ecb2aedca43b Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Thu, 5 May 2016 13:32:34 +0530 Subject: ARC: Add missing io barriers to io{read,write}{16,32}be() While reviewing a different change to asm-generic/io.h Arnd spotted that ARC ioread32 and ioread32be both of which come from asm-generic versions are not symmetrical in terms of calling the io barriers. generic ioread32 -> ARC readl() [ has barriers] generic ioread32be -> __be32_to_cpu(__raw_readl()) [ lacks barriers] While generic ioread32be is being remediated to call readl(), that involves a swab32(), causing double swaps on ioread32be() on Big Endian systems. So provide our versions of big endian IO accessors to ensure io barrier calls while also keeping them optimal Suggested-by: Arnd Bergmann Acked-by: Arnd Bergmann Cc: stable@vger.kernel.org [4.2+] Signed-off-by: Vineet Gupta --- arch/arc/include/asm/io.h | 27 ++++++++++++++++++--------- 1 file changed, 18 insertions(+), 9 deletions(-) (limited to 'arch/arc') diff --git a/arch/arc/include/asm/io.h b/arch/arc/include/asm/io.h index 17f85c9..c22b181 100644 --- a/arch/arc/include/asm/io.h +++ b/arch/arc/include/asm/io.h @@ -13,6 +13,15 @@ #include #include +#ifdef CONFIG_ISA_ARCV2 +#include +#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 -#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 -- cgit v1.1 From 2519d753676bdf2460fbbcde276d5b6ba8d6b695 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Thu, 5 May 2016 14:53:48 +0530 Subject: ARC: Fix PAE40 boot failures due to PTE truncation So a benign looking cleanup which macro'ized PAGE_SHIFT shifts turned out to be bad (since it was done non-sensically across the board). It caused boot failures with PAE40 as forced cast to (unsigned long) from newly introduced virt_to_pfn() was causing truncatiion of the (long long) pte/paddr values. It is OK to use this in accessors dealing with kernel virtual address, pointers etc, but not for PTE values themelves. Fixes: cJ2ff5cf2735c ("ARC: mm: Use virt_to_pfn() for addr >> PAGE_SHIFT pattern) Signed-off-by: Vineet Gupta --- arch/arc/include/asm/page.h | 9 ++++++++- arch/arc/include/asm/pgtable.h | 13 ++++++------- 2 files changed, 14 insertions(+), 8 deletions(-) (limited to 'arch/arc') diff --git a/arch/arc/include/asm/page.h b/arch/arc/include/asm/page.h index 36da89e..533e5a0 100644 --- a/arch/arc/include/asm/page.h +++ b/arch/arc/include/asm/page.h @@ -72,6 +72,13 @@ 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) @@ -85,7 +92,7 @@ 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) \ diff --git a/arch/arc/include/asm/pgtable.h b/arch/arc/include/asm/pgtable.h index 7d6c93e..10d4b8b 100644 --- a/arch/arc/include/asm/pgtable.h +++ b/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) -- cgit v1.1 From 26f9d5fd82ca20fe536cb493ec7cf5628f8997e5 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Mon, 18 Apr 2016 10:49:56 +0530 Subject: ARC: support HIGHMEM even without PAE40 Initial HIGHMEM support on ARC was introduced for PAE40 where the low memory (0x8000_0000 based) and high memory (0x1_0000_0000) were physically contiguous. So CONFIG_FLATMEM sufficed (despite a peipheral hole in the middle, which wasted a bit of struct page memory, but things worked). However w/o PAE, highmem was not possible and we could only reach ~1.75GB of DDR. Now there is a use case to access ~4GB of DDR w/o PAE40 The idea is to have low memory at canonical 0x8000_0000 and highmem at 0 so enire 4GB address space is available for physical addressing This needs additional platform/interconnect mapping to convert the non contiguous physical addresses into linear bus adresses. From Linux point of view, non contiguous divide means FLATMEM no longer works and DISCONTIGMEM is needed to track the pfns in the 2 regions. This scheme would also work for PAE40, only better in that we don't waste struct page memory for the peripheral hole. The DT description will be something like memory { ... reg = <0x80000000 0x200000000 /* 512MB: lowmem */ 0x00000000 0x10000000>; /* 256MB: highmem */ } Signed-off-by: Noam Camus Signed-off-by: Vineet Gupta --- arch/arc/Kconfig | 13 +++++++++++ arch/arc/include/asm/mmzone.h | 43 ++++++++++++++++++++++++++++++++++ arch/arc/include/asm/page.h | 6 ++--- arch/arc/mm/init.c | 54 +++++++++++++++++++++++++++++++------------ 4 files changed, 98 insertions(+), 18 deletions(-) create mode 100644 arch/arc/include/asm/mmzone.h (limited to 'arch/arc') diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig index ec4791e..a876743 100644 --- a/arch/arc/Kconfig +++ b/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 diff --git a/arch/arc/include/asm/mmzone.h b/arch/arc/include/asm/mmzone.h new file mode 100644 index 0000000..8e97136 --- /dev/null +++ b/arch/arc/include/asm/mmzone.h @@ -0,0 +1,43 @@ +/* + * 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 diff --git a/arch/arc/include/asm/page.h b/arch/arc/include/asm/page.h index 533e5a0..0d53854 100644 --- a/arch/arc/include/asm/page.h +++ b/arch/arc/include/asm/page.h @@ -83,7 +83,9 @@ typedef pte_t * pgtable_t; #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) @@ -95,9 +97,7 @@ typedef pte_t * pgtable_t; #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) */ diff --git a/arch/arc/mm/init.c b/arch/arc/mm/init.c index 5487d0b..8be9303 100644 --- a/arch/arc/mm/init.c +++ b/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 -- cgit v1.1