From ffd642e748c867a7339b57225b8bf8b9a0dcd9c5 Mon Sep 17 00:00:00 2001 From: David Shaohua Li Date: Wed, 8 Feb 2006 17:35:00 -0500 Subject: [ACPI] enable SMP C-states on x86_64 http://bugzilla.kernel.org/show_bug.cgi?id=5653 Signed-off-by: David Shaohua Li Signed-off-by: Len Brown --- arch/x86_64/kernel/acpi/Makefile | 1 + arch/x86_64/kernel/acpi/processor.c | 72 ------------------------------------- 2 files changed, 1 insertion(+), 72 deletions(-) delete mode 100644 arch/x86_64/kernel/acpi/processor.c (limited to 'arch/x86_64') diff --git a/arch/x86_64/kernel/acpi/Makefile b/arch/x86_64/kernel/acpi/Makefile index 4fe97071f29..080b9963f1b 100644 --- a/arch/x86_64/kernel/acpi/Makefile +++ b/arch/x86_64/kernel/acpi/Makefile @@ -4,5 +4,6 @@ obj-$(CONFIG_ACPI_SLEEP) += sleep.o wakeup.o ifneq ($(CONFIG_ACPI_PROCESSOR),) obj-y += processor.o +processor-y := ../../../i386/kernel/acpi/processor.o ../../../i386/kernel/acpi/cstate.o endif diff --git a/arch/x86_64/kernel/acpi/processor.c b/arch/x86_64/kernel/acpi/processor.c deleted file mode 100644 index 3bdc2baa5bb..00000000000 --- a/arch/x86_64/kernel/acpi/processor.c +++ /dev/null @@ -1,72 +0,0 @@ -/* - * arch/x86_64/kernel/acpi/processor.c - * - * Copyright (C) 2005 Intel Corporation - * Venkatesh Pallipadi - * - Added _PDC for platforms with Intel CPUs - */ - -#include -#include -#include -#include - -#include -#include - -static void init_intel_pdc(struct acpi_processor *pr, struct cpuinfo_x86 *c) -{ - struct acpi_object_list *obj_list; - union acpi_object *obj; - u32 *buf; - - /* allocate and initialize pdc. It will be used later. */ - obj_list = kmalloc(sizeof(struct acpi_object_list), GFP_KERNEL); - if (!obj_list) { - printk(KERN_ERR "Memory allocation error\n"); - return; - } - - obj = kmalloc(sizeof(union acpi_object), GFP_KERNEL); - if (!obj) { - printk(KERN_ERR "Memory allocation error\n"); - kfree(obj_list); - return; - } - - buf = kmalloc(12, GFP_KERNEL); - if (!buf) { - printk(KERN_ERR "Memory allocation error\n"); - kfree(obj); - kfree(obj_list); - return; - } - - buf[0] = ACPI_PDC_REVISION_ID; - buf[1] = 1; - buf[2] = ACPI_PDC_EST_CAPABILITY_SMP; - - obj->type = ACPI_TYPE_BUFFER; - obj->buffer.length = 12; - obj->buffer.pointer = (u8 *) buf; - obj_list->count = 1; - obj_list->pointer = obj; - pr->pdc = obj_list; - - return; -} - -/* Initialize _PDC data based on the CPU vendor */ -void arch_acpi_processor_init_pdc(struct acpi_processor *pr) -{ - unsigned int cpu = pr->id; - struct cpuinfo_x86 *c = cpu_data + cpu; - - pr->pdc = NULL; - if (c->x86_vendor == X86_VENDOR_INTEL && cpu_has(c, X86_FEATURE_EST)) - init_intel_pdc(pr, c); - - return; -} - -EXPORT_SYMBOL(arch_acpi_processor_init_pdc); -- cgit v1.2.3 From 1300124f69cafc54331bc06e968a8dd67863f989 Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Tue, 28 Mar 2006 17:04:00 -0500 Subject: ACPI: Kconfig: ACPI should depend on, not select PCI Otherwise, illegal configurations like X86_VOYAGER=y, PCI=y are possible. This patch also fixes the options select'ing ACPI to also select PCI. Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Len Brown --- arch/x86_64/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86_64') diff --git a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig index 4310b4a311a..299d8740177 100644 --- a/arch/x86_64/Kconfig +++ b/arch/x86_64/Kconfig @@ -289,6 +289,7 @@ config X86_64_ACPI_NUMA bool "ACPI NUMA detection" depends on NUMA select ACPI + select PCI select ACPI_NUMA default y help -- cgit v1.2.3 From ead2bfeb7f739d2ad6e09dc1343f0da51feb7f51 Mon Sep 17 00:00:00 2001 From: Chuck Ebbert <76306.1226@compuserve.com> Date: Thu, 15 Jun 2006 04:41:52 -0400 Subject: [PATCH] PCI: fix issues with extended conf space when MMCONFIG disabled because of e820 On 15 Jun 2006 03:45:10 +0200, Andi Kleen wrote: > Anyways I would say that if the BIOS can't get MCFG right then > it's likely not been validated on that board and shouldn't be used. According to Petr Vandrovec: ... "What is important (and checked) is address of MMCONFIG reported by MCFG table... Unfortunately code does not bother with printing that address :-( "Another problem is that code has hardcoded that MMCONFIG area is 256MB large. Unfortunately for the code PCI specification allows any power of two between 2MB and 256MB if vendor knows that such amount of busses (from 2 to 128) will be sufficient for system. With notebook it is quite possible that not full 8 bits are implemented for MMCONFIG bus number." So here is a patch. Unfortunately my system still fails the test because it doesn't reserve any part of the MMCONFIG area, but this may fix others. Booted on x86_64, only compiled on i386. x86_64 still remaps the max area (256MB) even though only 2MB is checked... but 2.6.16 had no check at all so it is still better. PCI: reduce size of x86 MMCONFIG reserved area check 1. Print the address of the MMCONFIG area when the test for that area being reserved fails. 2. Only check if the first 2MB is reserved, as that is the minimum. Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com> Acked-by: Arjan van de Ven Signed-off-by: Greg Kroah-Hartman --- arch/x86_64/pci/mmconfig.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'arch/x86_64') diff --git a/arch/x86_64/pci/mmconfig.c b/arch/x86_64/pci/mmconfig.c index a2060e4d5de..3c55c76c6fd 100644 --- a/arch/x86_64/pci/mmconfig.c +++ b/arch/x86_64/pci/mmconfig.c @@ -13,7 +13,10 @@ #include "pci.h" -#define MMCONFIG_APER_SIZE (256*1024*1024) +/* aperture is up to 256MB but BIOS may reserve less */ +#define MMCONFIG_APER_MIN (2 * 1024*1024) +#define MMCONFIG_APER_MAX (256 * 1024*1024) + /* Verify the first 16 busses. We assume that systems with more busses get MCFG right. */ #define MAX_CHECK_BUS 16 @@ -175,9 +178,10 @@ void __init pci_mmcfg_init(void) return; if (!e820_all_mapped(pci_mmcfg_config[0].base_address, - pci_mmcfg_config[0].base_address + MMCONFIG_APER_SIZE, + pci_mmcfg_config[0].base_address + MMCONFIG_APER_MIN, E820_RESERVED)) { - printk(KERN_ERR "PCI: BIOS Bug: MCFG area is not E820-reserved\n"); + printk(KERN_ERR "PCI: BIOS Bug: MCFG area at %x is not E820-reserved\n", + pci_mmcfg_config[0].base_address); printk(KERN_ERR "PCI: Not using MMCONFIG.\n"); return; } @@ -190,7 +194,8 @@ void __init pci_mmcfg_init(void) } for (i = 0; i < pci_mmcfg_config_num; ++i) { pci_mmcfg_virt[i].cfg = &pci_mmcfg_config[i]; - pci_mmcfg_virt[i].virt = ioremap_nocache(pci_mmcfg_config[i].base_address, MMCONFIG_APER_SIZE); + pci_mmcfg_virt[i].virt = ioremap_nocache(pci_mmcfg_config[i].base_address, + MMCONFIG_APER_MAX); if (!pci_mmcfg_virt[i].virt) { printk("PCI: Cannot map mmconfig aperture for segment %d\n", pci_mmcfg_config[i].pci_segment_group_number); -- cgit v1.2.3 From dcc1a66a09420ccc5a22671bddc5a842f92d67e5 Mon Sep 17 00:00:00 2001 From: Roman Zippel Date: Thu, 22 Jun 2006 14:47:35 -0700 Subject: [PATCH] x86_64: use select for GART_IOMMU to enable AGP The AGP default doesn't work well with other selects, so use a select for GART_IOMMU as well. Remove a redundant default for SWIOTLB as well. Signed-off-by: Roman Zippel Signed-off-by: Andi Kleen Cc: Andi Kleen Cc: Dave Jones Cc: Dave Airlie Cc: Jeff Garzik Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/Kconfig | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'arch/x86_64') diff --git a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig index 408d44a5975..7d3bc5ac5db 100644 --- a/arch/x86_64/Kconfig +++ b/arch/x86_64/Kconfig @@ -389,6 +389,7 @@ config GART_IOMMU bool "K8 GART IOMMU support" default y select SWIOTLB + select AGP depends on PCI help Support for hardware IOMMU in AMD's Opteron/Athlon64 Processors @@ -401,11 +402,9 @@ config GART_IOMMU northbridge and a software emulation used on other systems without hardware IOMMU. If unsure, say Y. -# need this always enabled with GART_IOMMU for the VIA workaround +# need this always selected by GART_IOMMU for the VIA workaround config SWIOTLB bool - default y - depends on GART_IOMMU config X86_MCE bool "Machine check support" if EMBEDDED -- cgit v1.2.3 From 762834e8bf46bf41ce9034d062a7c1f8563175f3 Mon Sep 17 00:00:00 2001 From: Yasunori Goto Date: Fri, 23 Jun 2006 02:03:19 -0700 Subject: [PATCH] Unify pxm_to_node() and node_to_pxm() Consolidate the various arch-specific implementations of pxm_to_node() and node_to_pxm() into a single generic version. Signed-off-by: Yasunori Goto Cc: "Luck, Tony" Cc: Andi Kleen Cc: Dave Hansen Cc: "Brown, Len" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/mm/srat.c | 33 +-------------------------------- 1 file changed, 1 insertion(+), 32 deletions(-) (limited to 'arch/x86_64') diff --git a/arch/x86_64/mm/srat.c b/arch/x86_64/mm/srat.c index 474df22c6ed..502fce65e96 100644 --- a/arch/x86_64/mm/srat.c +++ b/arch/x86_64/mm/srat.c @@ -30,7 +30,6 @@ static struct acpi_table_slit *acpi_slit; static nodemask_t nodes_parsed __initdata; -static nodemask_t nodes_found __initdata; static struct bootnode nodes[MAX_NUMNODES] __initdata; static struct bootnode nodes_add[MAX_NUMNODES] __initdata; static int found_add_area __initdata; @@ -38,33 +37,14 @@ int hotadd_percent __initdata = 0; #ifndef RESERVE_HOTADD #define hotadd_percent 0 /* Ignore all settings */ #endif -static u8 pxm2node[256] = { [0 ... 255] = 0xff }; /* Too small nodes confuse the VM badly. Usually they result from BIOS bugs. */ #define NODE_MIN_SIZE (4*1024*1024) -static int node_to_pxm(int n); - -int pxm_to_node(int pxm) -{ - if ((unsigned)pxm >= 256) - return -1; - /* Extend 0xff to (int)-1 */ - return (signed char)pxm2node[pxm]; -} - static __init int setup_node(int pxm) { - unsigned node = pxm2node[pxm]; - if (node == 0xff) { - if (nodes_weight(nodes_found) >= MAX_NUMNODES) - return -1; - node = first_unset_node(nodes_found); - node_set(node, nodes_found); - pxm2node[pxm] = node; - } - return pxm2node[pxm]; + return acpi_map_pxm_to_node(pxm); } static __init int conflicting_nodes(unsigned long start, unsigned long end) @@ -440,17 +420,6 @@ int __init acpi_scan_nodes(unsigned long start, unsigned long end) return 0; } -static int node_to_pxm(int n) -{ - int i; - if (pxm2node[n] == n) - return n; - for (i = 0; i < 256; i++) - if (pxm2node[i] == n) - return i; - return 0; -} - void __init srat_reserve_add_area(int nodeid) { if (found_add_area && nodes_add[nodeid].end) { -- cgit v1.2.3 From 1b2db9fb7adc4d67d9ce7d16ce79c41ee84730fe Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Fri, 23 Jun 2006 02:03:56 -0700 Subject: [PATCH] sys_move_pages: 32bit support (i386, x86_64) sys_move_pages() support for 32bit (i386 plus x86_64 compat layer) Add support for move_pages() on i386 and also add the compat functions necessary to run 32 bit binaries on x86_64. Add compat_sys_move_pages to the x86_64 32bit binary layer. Note that it is not up to date so I added the missing pieces. Not sure if this is done the right way. [akpm@osdl.org: compile fix] Signed-off-by: Christoph Lameter Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/ia32/ia32entry.S | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86_64') diff --git a/arch/x86_64/ia32/ia32entry.S b/arch/x86_64/ia32/ia32entry.S index 5a92fed2d1d..4ec594ab1a9 100644 --- a/arch/x86_64/ia32/ia32entry.S +++ b/arch/x86_64/ia32/ia32entry.S @@ -696,4 +696,5 @@ ia32_sys_call_table: .quad sys_sync_file_range .quad sys_tee .quad compat_sys_vmsplice + .quad compat_sys_move_pages ia32_syscall_end: -- cgit v1.2.3 From acae9d32436122959667470b84da517d7b1b9c2d Mon Sep 17 00:00:00 2001 From: Andreas Mohr Date: Fri, 23 Jun 2006 02:04:25 -0700 Subject: [PATCH] x86: make using_apic_timer __read_mostly Signed-off-by: Andreas Mohr Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/apic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86_64') diff --git a/arch/x86_64/kernel/apic.c b/arch/x86_64/kernel/apic.c index 100a30c4004..29ef99001e0 100644 --- a/arch/x86_64/kernel/apic.c +++ b/arch/x86_64/kernel/apic.c @@ -51,7 +51,7 @@ int disable_apic_timer __initdata; static cpumask_t timer_interrupt_broadcast_ipi_mask; /* Using APIC to generate smp_local_timer_interrupt? */ -int using_apic_timer = 0; +int using_apic_timer __read_mostly = 0; static void apic_pm_activate(void); -- cgit v1.2.3 From 7b0c2d92180dbd9c7cd0c4b9bd38b06bb0f12843 Mon Sep 17 00:00:00 2001 From: Andreas Mohr Date: Fri, 23 Jun 2006 02:04:26 -0700 Subject: [PATCH] x86: make i387 mxcsr_feature_mask __read_mostly Signed-off-by: Andreas Mohr Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/i387.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86_64') diff --git a/arch/x86_64/kernel/i387.c b/arch/x86_64/kernel/i387.c index a5d7e16b928..44ddb1ec808 100644 --- a/arch/x86_64/kernel/i387.c +++ b/arch/x86_64/kernel/i387.c @@ -24,7 +24,7 @@ #include #include -unsigned int mxcsr_feature_mask = 0xffffffff; +unsigned int mxcsr_feature_mask __read_mostly = 0xffffffff; void mxcsr_feature_mask_init(void) { -- cgit v1.2.3 From 3e3318dee0878d42ed62a19c292a2ac284135db3 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Fri, 23 Jun 2006 02:04:46 -0700 Subject: [PATCH] swsusp: x86_64 mark special saveable/unsaveable pages Pages (Reserved/ACPI NVS/ACPI Data) below end_pfn will be saved/restored by S4 currently. We should mark 'Reserved' pages not saveable. Pages (Reserved/ACPI NVS/ACPI Data) above end_pfn will not be saved/restored by S4 currently. We should save the 'ACPI NVS/ACPI Data' pages. Signed-off-by: Shaohua Li Cc: Pavel Machek Cc: "Rafael J. Wysocki" Cc: Nigel Cunningham Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/setup.c | 95 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 95 insertions(+) (limited to 'arch/x86_64') diff --git a/arch/x86_64/kernel/setup.c b/arch/x86_64/kernel/setup.c index 655b9192eeb..b8d5116d737 100644 --- a/arch/x86_64/kernel/setup.c +++ b/arch/x86_64/kernel/setup.c @@ -47,6 +47,7 @@ #include #include #include +#include #include #include @@ -595,6 +596,100 @@ static void discover_ebda(void) ebda_size = 64*1024; } +#ifdef CONFIG_SOFTWARE_SUSPEND +static void __init mark_nosave_page_range(unsigned long start, unsigned long end) +{ + struct page *page; + while (start <= end) { + page = pfn_to_page(start); + SetPageNosave(page); + start++; + } +} + +static void __init e820_nosave_reserved_pages(void) +{ + int i; + unsigned long r_start = 0, r_end = 0; + + /* Assume e820 map is sorted */ + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + unsigned long start, end; + + start = round_down(ei->addr, PAGE_SIZE); + end = round_up(ei->addr + ei->size, PAGE_SIZE); + if (start >= end) + continue; + if (ei->type == E820_RESERVED) + continue; + r_end = start>>PAGE_SHIFT; + /* swsusp ignores invalid pfn, ignore these pages here */ + if (r_end > end_pfn) + r_end = end_pfn; + if (r_end > r_start) + mark_nosave_page_range(r_start, r_end-1); + if (r_end >= end_pfn) + break; + r_start = end>>PAGE_SHIFT; + } +} + +static void __init e820_save_acpi_pages(void) +{ + int i; + + /* Assume e820 map is sorted */ + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + unsigned long start, end; + + start = ei->addr, PAGE_SIZE; + end = ei->addr + ei->size; + if (start >= end) + continue; + if (ei->type != E820_ACPI && ei->type != E820_NVS) + continue; + /* + * If the region is below end_pfn, it will be + * saved/restored by swsusp follow 'RAM' type. + */ + if (start < (end_pfn << PAGE_SHIFT)) + start = end_pfn << PAGE_SHIFT; + if (end > start) + swsusp_add_arch_pages(start, end); + } +} + +extern char __start_rodata, __end_rodata; +/* + * BIOS reserved region/hole - no save/restore + * ACPI NVS - save/restore + * ACPI Data - this is a little tricky, the mem could be used by OS after OS + * reads tables from the region, but anyway save/restore the memory hasn't any + * side effect and Linux runtime module load/unload might use it. + * kernel rodata - no save/restore (kernel rodata isn't changed) + */ +static int __init mark_nosave_pages(void) +{ + unsigned long pfn_start, pfn_end; + + /* BIOS reserved regions & holes */ + e820_nosave_reserved_pages(); + + /* kernel rodata */ + pfn_start = round_up(__pa_symbol(&__start_rodata), PAGE_SIZE) >> PAGE_SHIFT; + pfn_end = round_down(__pa_symbol(&__end_rodata), PAGE_SIZE) >> PAGE_SHIFT; + mark_nosave_page_range(pfn_start, pfn_end-1); + + /* record ACPI Data/NVS as saveable */ + e820_save_acpi_pages(); + + return 0; +} +core_initcall(mark_nosave_pages); +#endif + void __init setup_arch(char **cmdline_p) { unsigned long kernel_end; -- cgit v1.2.3 From 55b2355eefc2f160246226d4d69fed431173a4d5 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Fri, 23 Jun 2006 02:04:49 -0700 Subject: [PATCH] don't use flush_tlb_all in suspend time flush_tlb_all uses on_each_cpu, which will disable/enable interrupt. In suspend/resume time, this will make interrupt wrongly enabled. Signed-off-by: Shaohua Li Cc: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/acpi/sleep.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'arch/x86_64') diff --git a/arch/x86_64/kernel/acpi/sleep.c b/arch/x86_64/kernel/acpi/sleep.c index 867a0ebee17..091bc79c888 100644 --- a/arch/x86_64/kernel/acpi/sleep.c +++ b/arch/x86_64/kernel/acpi/sleep.c @@ -35,6 +35,8 @@ #include #include #include +#include + #include #include #include @@ -66,7 +68,8 @@ static void init_low_mapping(void) pgd_t *slot0 = pgd_offset(current->mm, 0UL); low_ptr = *slot0; set_pgd(slot0, *pgd_offset(current->mm, PAGE_OFFSET)); - flush_tlb_all(); + WARN_ON(num_online_cpus() != 1); + local_flush_tlb(); } /** @@ -92,7 +95,7 @@ int acpi_save_state_mem(void) void acpi_restore_state_mem(void) { set_pgd(pgd_offset(current->mm, 0UL), low_ptr); - flush_tlb_all(); + local_flush_tlb(); } /** -- cgit v1.2.3