From 6859a8402945cf1d74af75a2e1aa4e327a506ab4 Mon Sep 17 00:00:00 2001 From: Alan Mayer Date: Wed, 26 Mar 2008 16:11:31 -0500 Subject: x86: resize NR_IRQS for large machines On machines with very large numbers of cpus, tables that are dimensioned by NR_IRQS get very large, especially the irq_desc table. They are also very sparsely used. When the cpu count is > MAX_IO_APICS, use MAX_IO_APICS to set NR_IRQS, otherwise use NR_CPUS. Signed-off-by: Alan Mayer Reviewed-by: Christoph Lameter Signed-off-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- include/linux/kernel_stat.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/kernel_stat.h b/include/linux/kernel_stat.h index e8ffce898bf9..cf9f40a91c9c 100644 --- a/include/linux/kernel_stat.h +++ b/include/linux/kernel_stat.h @@ -1,11 +1,11 @@ #ifndef _LINUX_KERNEL_STAT_H #define _LINUX_KERNEL_STAT_H -#include #include #include #include #include +#include #include /* -- cgit v1.2.3 From 988f7b5789ccf5cfed14c72e28573a49f0cb4809 Mon Sep 17 00:00:00 2001 From: "venkatesh.pallipadi@intel.com" Date: Tue, 18 Mar 2008 17:00:22 -0700 Subject: x86: PAT export resource_wc in pci sysfs For the ranges with IORESOURCE_PREFETCH, export a new resource_wc interface in pci /sysfs along with resource (which is uncached). Signed-off-by: Venkatesh Pallipadi Signed-off-by: Suresh Siddha Acked-by: Jesse Barnes Signed-off-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- Documentation/filesystems/sysfs-pci.txt | 1 + drivers/pci/pci-sysfs.c | 84 ++++++++++++++++++++++++--------- include/linux/pci.h | 1 + 3 files changed, 64 insertions(+), 22 deletions(-) (limited to 'include/linux') diff --git a/Documentation/filesystems/sysfs-pci.txt b/Documentation/filesystems/sysfs-pci.txt index 5daa2aaec2c5..68ef48839c04 100644 --- a/Documentation/filesystems/sysfs-pci.txt +++ b/Documentation/filesystems/sysfs-pci.txt @@ -36,6 +36,7 @@ files, each with their own function. local_cpus nearby CPU mask (cpumask, ro) resource PCI resource host addresses (ascii, ro) resource0..N PCI resource N, if present (binary, mmap) + resource0_wc..N_wc PCI WC map resource N, if prefetchable (binary, mmap) rom PCI ROM resource, if present (binary, ro) subsystem_device PCI subsystem device (ascii, ro) subsystem_vendor PCI subsystem vendor (ascii, ro) diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 271d41cc05ab..9ec7d3977a82 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -489,13 +489,14 @@ pci_mmap_legacy_mem(struct kobject *kobj, struct bin_attribute *attr, * @kobj: kobject for mapping * @attr: struct bin_attribute for the file being mapped * @vma: struct vm_area_struct passed into the mmap + * @write_combine: 1 for write_combine mapping * * Use the regular PCI mapping routines to map a PCI resource into userspace. * FIXME: write combining? maybe automatic for prefetchable regions? */ static int pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr, - struct vm_area_struct *vma) + struct vm_area_struct *vma, int write_combine) { struct pci_dev *pdev = to_pci_dev(container_of(kobj, struct device, kobj)); @@ -518,7 +519,21 @@ pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr, vma->vm_pgoff += start >> PAGE_SHIFT; mmap_type = res->flags & IORESOURCE_MEM ? pci_mmap_mem : pci_mmap_io; - return pci_mmap_page_range(pdev, vma, mmap_type, 0); + return pci_mmap_page_range(pdev, vma, mmap_type, write_combine); +} + +static int +pci_mmap_resource_uc(struct kobject *kobj, struct bin_attribute *attr, + struct vm_area_struct *vma) +{ + return pci_mmap_resource(kobj, attr, vma, 0); +} + +static int +pci_mmap_resource_wc(struct kobject *kobj, struct bin_attribute *attr, + struct vm_area_struct *vma) +{ + return pci_mmap_resource(kobj, attr, vma, 1); } /** @@ -541,9 +556,46 @@ pci_remove_resource_files(struct pci_dev *pdev) sysfs_remove_bin_file(&pdev->dev.kobj, res_attr); kfree(res_attr); } + + res_attr = pdev->res_attr_wc[i]; + if (res_attr) { + sysfs_remove_bin_file(&pdev->dev.kobj, res_attr); + kfree(res_attr); + } } } +static int pci_create_attr(struct pci_dev *pdev, int num, int write_combine) +{ + /* allocate attribute structure, piggyback attribute name */ + int name_len = write_combine ? 13 : 10; + struct bin_attribute *res_attr; + int retval; + + res_attr = kzalloc(sizeof(*res_attr) + name_len, GFP_ATOMIC); + if (res_attr) { + char *res_attr_name = (char *)(res_attr + 1); + + if (write_combine) { + pdev->res_attr_wc[num] = res_attr; + sprintf(res_attr_name, "resource%d_wc", num); + res_attr->mmap = pci_mmap_resource_wc; + } else { + pdev->res_attr[num] = res_attr; + sprintf(res_attr_name, "resource%d", num); + res_attr->mmap = pci_mmap_resource_uc; + } + res_attr->attr.name = res_attr_name; + res_attr->attr.mode = S_IRUSR | S_IWUSR; + res_attr->size = pci_resource_len(pdev, num); + res_attr->private = &pdev->resource[num]; + retval = sysfs_create_bin_file(&pdev->dev.kobj, res_attr); + } else + retval = -ENOMEM; + + return retval; +} + /** * pci_create_resource_files - create resource files in sysfs for @dev * @dev: dev in question @@ -557,31 +609,19 @@ static int pci_create_resource_files(struct pci_dev *pdev) /* Expose the PCI resources from this device as files */ for (i = 0; i < PCI_ROM_RESOURCE; i++) { - struct bin_attribute *res_attr; /* skip empty resources */ if (!pci_resource_len(pdev, i)) continue; - /* allocate attribute structure, piggyback attribute name */ - res_attr = kzalloc(sizeof(*res_attr) + 10, GFP_ATOMIC); - if (res_attr) { - char *res_attr_name = (char *)(res_attr + 1); - - pdev->res_attr[i] = res_attr; - sprintf(res_attr_name, "resource%d", i); - res_attr->attr.name = res_attr_name; - res_attr->attr.mode = S_IRUSR | S_IWUSR; - res_attr->size = pci_resource_len(pdev, i); - res_attr->mmap = pci_mmap_resource; - res_attr->private = &pdev->resource[i]; - retval = sysfs_create_bin_file(&pdev->dev.kobj, res_attr); - if (retval) { - pci_remove_resource_files(pdev); - return retval; - } - } else { - return -ENOMEM; + retval = pci_create_attr(pdev, i, 0); + /* for prefetchable resources, create a WC mappable file */ + if (!retval && pdev->resource[i].flags & IORESOURCE_PREFETCH) + retval = pci_create_attr(pdev, i, 1); + + if (retval) { + pci_remove_resource_files(pdev); + return retval; } } return 0; diff --git a/include/linux/pci.h b/include/linux/pci.h index 509159bcd4e7..d18b1dd49fab 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -206,6 +206,7 @@ struct pci_dev { struct bin_attribute *rom_attr; /* attribute descriptor for sysfs ROM entry */ int rom_attr_enabled; /* has display of the rom attribute been enabled? */ struct bin_attribute *res_attr[DEVICE_COUNT_RESOURCE]; /* sysfs file for resources */ + struct bin_attribute *res_attr_wc[DEVICE_COUNT_RESOURCE]; /* sysfs file for WC mapping of resources */ #ifdef CONFIG_PCI_MSI struct list_head msi_list; #endif -- cgit v1.2.3 From 63687a528c39a67c1a213cdffa09feb0e6af9dbe Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Mon, 12 May 2008 15:44:41 +0200 Subject: x86: move tracedata to RODATA .. allowing it to be write-protected just as other read-only data under CONFIG_DEBUG_RODATA. Signed-off-by: Jan Beulich Signed-off-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- arch/x86/kernel/vmlinux_32.lds.S | 7 ------- arch/x86/kernel/vmlinux_64.lds.S | 7 ------- drivers/base/power/trace.c | 2 +- include/asm-generic/vmlinux.lds.h | 14 ++++++++++++++ include/asm-x86/resume-trace.h | 2 +- include/linux/resume-trace.h | 2 +- 6 files changed, 17 insertions(+), 17 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/vmlinux_32.lds.S b/arch/x86/kernel/vmlinux_32.lds.S index ce5ed083a1e9..2674f5796275 100644 --- a/arch/x86/kernel/vmlinux_32.lds.S +++ b/arch/x86/kernel/vmlinux_32.lds.S @@ -60,13 +60,6 @@ SECTIONS BUG_TABLE :text - . = ALIGN(4); - .tracedata : AT(ADDR(.tracedata) - LOAD_OFFSET) { - __tracedata_start = .; - *(.tracedata) - __tracedata_end = .; - } - RODATA /* writeable */ diff --git a/arch/x86/kernel/vmlinux_64.lds.S b/arch/x86/kernel/vmlinux_64.lds.S index fad3674b06a5..687041bfbae5 100644 --- a/arch/x86/kernel/vmlinux_64.lds.S +++ b/arch/x86/kernel/vmlinux_64.lds.S @@ -53,13 +53,6 @@ SECTIONS RODATA - . = ALIGN(4); - .tracedata : AT(ADDR(.tracedata) - LOAD_OFFSET) { - __tracedata_start = .; - *(.tracedata) - __tracedata_end = .; - } - . = ALIGN(PAGE_SIZE); /* Align data segment to page size boundary */ /* Data */ .data : AT(ADDR(.data) - LOAD_OFFSET) { diff --git a/drivers/base/power/trace.c b/drivers/base/power/trace.c index 2b4b392dcbc1..87a7f1d02578 100644 --- a/drivers/base/power/trace.c +++ b/drivers/base/power/trace.c @@ -153,7 +153,7 @@ EXPORT_SYMBOL(set_trace_device); * it's not any guarantee, but it's a high _likelihood_ that * the match is valid). */ -void generate_resume_trace(void *tracedata, unsigned int user) +void generate_resume_trace(const void *tracedata, unsigned int user) { unsigned short lineno = *(unsigned short *)tracedata; const char *file = *(const char **)(tracedata + 2); diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index f054778e916c..f1992dc5c424 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -93,6 +93,8 @@ VMLINUX_SYMBOL(__end_rio_route_ops) = .; \ } \ \ + TRACEDATA \ + \ /* Kernel symbol table: Normal symbols */ \ __ksymtab : AT(ADDR(__ksymtab) - LOAD_OFFSET) { \ VMLINUX_SYMBOL(__start___ksymtab) = .; \ @@ -318,6 +320,18 @@ __stop___bug_table = .; \ } +#ifdef CONFIG_PM_TRACE +#define TRACEDATA \ + . = ALIGN(4); \ + .tracedata : AT(ADDR(.tracedata) - LOAD_OFFSET) { \ + __tracedata_start = .; \ + *(.tracedata) \ + __tracedata_end = .; \ + } +#else +#define TRACEDATA +#endif + #define NOTES \ .notes : AT(ADDR(.notes) - LOAD_OFFSET) { \ VMLINUX_SYMBOL(__start_notes) = .; \ diff --git a/include/asm-x86/resume-trace.h b/include/asm-x86/resume-trace.h index 2557514d7ef6..8d9f0b41ee86 100644 --- a/include/asm-x86/resume-trace.h +++ b/include/asm-x86/resume-trace.h @@ -6,7 +6,7 @@ #define TRACE_RESUME(user) \ do { \ if (pm_trace_enabled) { \ - void *tracedata; \ + const void *tracedata; \ asm volatile(_ASM_MOV_UL " $1f,%0\n" \ ".section .tracedata,\"a\"\n" \ "1:\t.word %c1\n\t" \ diff --git a/include/linux/resume-trace.h b/include/linux/resume-trace.h index f3f4f28c6960..c9ba2fdf807d 100644 --- a/include/linux/resume-trace.h +++ b/include/linux/resume-trace.h @@ -8,7 +8,7 @@ extern int pm_trace_enabled; struct device; extern void set_trace_device(struct device *); -extern void generate_resume_trace(void *tracedata, unsigned int user); +extern void generate_resume_trace(const void *tracedata, unsigned int user); #define TRACE_DEVICE(dev) do { \ if (pm_trace_enabled) \ -- cgit v1.2.3 From e9197bf0114661195bee35e7795cfc42164d9b2c Mon Sep 17 00:00:00 2001 From: Paul Jackson Date: Wed, 14 May 2008 08:15:10 -0700 Subject: x86 boot: remove some unused extern function declarations Remove three extern declarations for routines that don't exist. Fix a typo in a comment. Signed-off-by: Paul Jackson Signed-off-by: Ingo Molnar --- arch/x86/mm/numa_64.c | 2 +- include/linux/efi.h | 4 ---- 2 files changed, 1 insertion(+), 5 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index c5066d519e5d..afb07ffb931d 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -233,7 +233,7 @@ void __init setup_node_bootmem(int nodeid, unsigned long start, else bootmap_start = round_up(start, PAGE_SIZE); /* - * SMP_CAHCE_BYTES could be enough, but init_bootmem_node like + * SMP_CACHE_BYTES could be enough, but init_bootmem_node like * to use that to align to PAGE_SIZE */ bootmap = early_node_mem(nodeid, bootmap_start, end, diff --git a/include/linux/efi.h b/include/linux/efi.h index a5f359a7ad0e..807373d467f7 100644 --- a/include/linux/efi.h +++ b/include/linux/efi.h @@ -287,7 +287,6 @@ efi_guid_unparse(efi_guid_t *guid, char *out) extern void efi_init (void); extern void *efi_get_pal_addr (void); extern void efi_map_pal_code (void); -extern void efi_map_memmap(void); extern void efi_memmap_walk (efi_freemem_callback_t callback, void *arg); extern void efi_gettimeofday (struct timespec *ts); extern void efi_enter_virtual_mode (void); /* switch EFI to virtual mode, if possible */ @@ -295,14 +294,11 @@ extern u64 efi_get_iobase (void); extern u32 efi_mem_type (unsigned long phys_addr); extern u64 efi_mem_attributes (unsigned long phys_addr); extern u64 efi_mem_attribute (unsigned long phys_addr, unsigned long size); -extern int efi_mem_attribute_range (unsigned long phys_addr, unsigned long size, - u64 attr); extern int __init efi_uart_console_only (void); extern void efi_initialize_iomem_resources(struct resource *code_resource, struct resource *data_resource, struct resource *bss_resource); extern unsigned long efi_get_time(void); extern int efi_set_rtc_mmss(unsigned long nowtime); -extern int is_available_memory(efi_memory_desc_t * md); extern struct efi_memory_map memmap; /** -- cgit v1.2.3 From c801ed3860fe2f84281d4cae4c3e6ca87e81e890 Mon Sep 17 00:00:00 2001 From: Paul Jackson Date: Wed, 14 May 2008 08:15:23 -0700 Subject: x86 boot: simplify pageblock_bits enum declaration The use of #defines with '##' pre-processor concatenation is a useful way to form several symbol names with a common pattern. But when there is just a single name obtained from that #define, it's just obfuscation. Better to just write the plain symbol name, as is. The following patch is a result of my wasting ten minutes looking through the kernel to figure out what 'PB_migrate_end' meant, and forgetting what I came to do, by the time I figured out that the #define PB_range macro defined it. Signed-off-by: Paul Jackson Signed-off-by: Ingo Molnar --- include/linux/pageblock-flags.h | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/include/linux/pageblock-flags.h b/include/linux/pageblock-flags.h index e875905f7b12..e8c06122be36 100644 --- a/include/linux/pageblock-flags.h +++ b/include/linux/pageblock-flags.h @@ -25,13 +25,11 @@ #include -/* Macro to aid the definition of ranges of bits */ -#define PB_range(name, required_bits) \ - name, name ## _end = (name + required_bits) - 1 - /* Bit indices that affect a whole block of pages */ enum pageblock_bits { - PB_range(PB_migrate, 3), /* 3 bits required for migrate types */ + PB_migrate, + PB_migrate_end = PB_migrate + 3 - 1, + /* 3 bits required for migrate types */ NR_PAGEBLOCK_BITS }; -- cgit v1.2.3 From 9e124fe16ff24746d6de5a2ad685266d7bce0e08 Mon Sep 17 00:00:00 2001 From: Markus Armbruster Date: Mon, 26 May 2008 23:31:07 +0100 Subject: xen: Enable console tty by default in domU if it's not a dummy Without console= arguments on the kernel command line, the first console to register becomes enabled and the preferred console (the one behind /dev/console). This is normally tty (assuming CONFIG_VT_CONSOLE is enabled, which it commonly is). This is okay as long tty is a useful console. But unless we have the PV framebuffer, and it is enabled for this domain, tty0 in domU is merely a dummy. In that case, we want the preferred console to be the Xen console hvc0, and we want it without having to fiddle with the kernel command line. Commit b8c2d3dfbc117dff26058fbac316b8acfc2cb5f7 did that for us. Since we now have the PV framebuffer, we want to enable and prefer tty again, but only when PVFB is enabled. But even then we still want to enable the Xen console as well. Problem: when tty registers, we can't yet know whether the PVFB is enabled. By the time we can know (xenstore is up), the console setup game is over. Solution: enable console tty by default, but keep hvc as the preferred console. Change the preferred console to tty when PVFB probes successfully, unless we've been given console kernel parameters. Signed-off-by: Markus Armbruster Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Thomas Gleixner --- arch/x86/xen/enlighten.c | 4 +++- drivers/video/xen-fbfront.c | 25 +++++++++++++++++++++++++ include/linux/console.h | 2 ++ kernel/printk.c | 3 +++ 4 files changed, 33 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index 1b4b5fa498b3..6cfb708408e9 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -1256,8 +1256,10 @@ asmlinkage void __init xen_start_kernel(void) ? __pa(xen_start_info->mod_start) : 0; boot_params.hdr.ramdisk_size = xen_start_info->mod_len; - if (!is_initial_xendomain()) + if (!is_initial_xendomain()) { + add_preferred_console("tty", 0, NULL); add_preferred_console("hvc", 0, NULL); + } /* Start the world */ start_kernel(); diff --git a/drivers/video/xen-fbfront.c b/drivers/video/xen-fbfront.c index 619a6f8d65a2..4e10876e62fc 100644 --- a/drivers/video/xen-fbfront.c +++ b/drivers/video/xen-fbfront.c @@ -18,6 +18,7 @@ * frame buffer. */ +#include #include #include #include @@ -48,6 +49,7 @@ struct xenfb_info { static u32 xenfb_mem_len = XENFB_WIDTH * XENFB_HEIGHT * XENFB_DEPTH / 8; +static void xenfb_make_preferred_console(void); static int xenfb_remove(struct xenbus_device *); static void xenfb_init_shared_page(struct xenfb_info *); static int xenfb_connect_backend(struct xenbus_device *, struct xenfb_info *); @@ -348,6 +350,7 @@ static int __devinit xenfb_probe(struct xenbus_device *dev, if (ret < 0) goto error; + xenfb_make_preferred_console(); return 0; error_nomem: @@ -358,6 +361,28 @@ static int __devinit xenfb_probe(struct xenbus_device *dev, return ret; } +static __devinit void +xenfb_make_preferred_console(void) +{ + struct console *c; + + if (console_set_on_cmdline) + return; + + acquire_console_sem(); + for (c = console_drivers; c; c = c->next) { + if (!strcmp(c->name, "tty") && c->index == 0) + break; + } + release_console_sem(); + if (c) { + unregister_console(c); + c->flags |= CON_CONSDEV; + c->flags &= ~CON_PRINTBUFFER; /* don't print again */ + register_console(c); + } +} + static int xenfb_resume(struct xenbus_device *dev) { struct xenfb_info *info = dev->dev.driver_data; diff --git a/include/linux/console.h b/include/linux/console.h index a4f27fbdf549..248e6e3b9b73 100644 --- a/include/linux/console.h +++ b/include/linux/console.h @@ -108,6 +108,8 @@ struct console { struct console *next; }; +extern int console_set_on_cmdline; + extern int add_preferred_console(char *name, int idx, char *options); extern int update_console_cmdline(char *name, int idx, char *name_new, int idx_new, char *options); extern void register_console(struct console *); diff --git a/kernel/printk.c b/kernel/printk.c index 8fb01c32aa3b..028ed75d4864 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -121,6 +121,8 @@ struct console_cmdline static struct console_cmdline console_cmdline[MAX_CMDLINECONSOLES]; static int selected_console = -1; static int preferred_console = -1; +int console_set_on_cmdline; +EXPORT_SYMBOL(console_set_on_cmdline); /* Flag: console code may call schedule() */ static int console_may_schedule; @@ -890,6 +892,7 @@ static int __init console_setup(char *str) *s = 0; __add_preferred_console(buf, idx, options, brl_options); + console_set_on_cmdline = 1; return 1; } __setup("console=", console_setup); -- cgit v1.2.3 From 0e91398f2a5d4eb6b07df8115917d0d1cf3e9b58 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Mon, 26 May 2008 23:31:27 +0100 Subject: xen: implement save/restore This patch implements Xen save/restore and migration. Saving is triggered via xenbus, which is polled in drivers/xen/manage.c. When a suspend request comes in, the kernel prepares itself for saving by: 1 - Freeze all processes. This is primarily to prevent any partially-completed pagetable updates from confusing the suspend process. If CONFIG_PREEMPT isn't defined, then this isn't necessary. 2 - Suspend xenbus and other devices 3 - Stop_machine, to make sure all the other vcpus are quiescent. The Xen tools require the domain to run its save off vcpu0. 4 - Within the stop_machine state, it pins any unpinned pgds (under construction or destruction), performs canonicalizes various other pieces of state (mostly converting mfns to pfns), and finally 5 - Suspend the domain Restore reverses the steps used to save the domain, ending when all the frozen processes are thawed. Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Thomas Gleixner --- arch/x86/xen/Makefile | 2 +- arch/x86/xen/enlighten.c | 6 +-- arch/x86/xen/mmu.c | 46 +++++++++++++++++ arch/x86/xen/smp.c | 2 +- arch/x86/xen/suspend.c | 42 +++++++++++++++ arch/x86/xen/time.c | 8 +++ arch/x86/xen/xen-ops.h | 4 ++ drivers/xen/events.c | 83 +++++++++++++++++++++++++++++ drivers/xen/grant-table.c | 4 +- drivers/xen/manage.c | 126 ++++++++++++++++++++++++++++++++++++++++----- include/linux/page-flags.h | 1 + include/xen/events.h | 3 ++ include/xen/grant_table.h | 3 ++ include/xen/xen-ops.h | 9 ++++ 14 files changed, 318 insertions(+), 21 deletions(-) create mode 100644 arch/x86/xen/suspend.c (limited to 'include/linux') diff --git a/arch/x86/xen/Makefile b/arch/x86/xen/Makefile index 40b119b6b103..2ba2d1649131 100644 --- a/arch/x86/xen/Makefile +++ b/arch/x86/xen/Makefile @@ -1,4 +1,4 @@ obj-y := enlighten.o setup.o multicalls.o mmu.o \ - time.o xen-asm.o grant-table.o + time.o xen-asm.o grant-table.o suspend.o obj-$(CONFIG_SMP) += smp.o diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index ce67dc8f36af..b94f63ac228b 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -857,7 +857,7 @@ static __init void xen_pagetable_setup_start(pgd_t *base) PFN_DOWN(__pa(xen_start_info->pt_base))); } -static __init void setup_shared_info(void) +void xen_setup_shared_info(void) { if (!xen_feature(XENFEAT_auto_translated_physmap)) { unsigned long addr = fix_to_virt(FIX_PARAVIRT_BOOTMAP); @@ -894,7 +894,7 @@ static __init void xen_pagetable_setup_done(pgd_t *base) pv_mmu_ops.release_pmd = xen_release_pmd; pv_mmu_ops.set_pte = xen_set_pte; - setup_shared_info(); + xen_setup_shared_info(); /* Actually pin the pagetable down, but we can't set PG_pinned yet because the page structures don't exist yet. */ @@ -902,7 +902,7 @@ static __init void xen_pagetable_setup_done(pgd_t *base) } /* This is called once we have the cpu_possible_map */ -void __init xen_setup_vcpu_info_placement(void) +void xen_setup_vcpu_info_placement(void) { int cpu; diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index 4740cda36563..e95955968ba3 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -560,6 +560,29 @@ void xen_pgd_pin(pgd_t *pgd) xen_mc_issue(0); } +/* + * On save, we need to pin all pagetables to make sure they get their + * mfns turned into pfns. Search the list for any unpinned pgds and pin + * them (unpinned pgds are not currently in use, probably because the + * process is under construction or destruction). + */ +void xen_mm_pin_all(void) +{ + unsigned long flags; + struct page *page; + + spin_lock_irqsave(&pgd_lock, flags); + + list_for_each_entry(page, &pgd_list, lru) { + if (!PagePinned(page)) { + xen_pgd_pin((pgd_t *)page_address(page)); + SetPageSavePinned(page); + } + } + + spin_unlock_irqrestore(&pgd_lock, flags); +} + /* The init_mm pagetable is really pinned as soon as its created, but that's before we have page structures to store the bits. So do all the book-keeping now. */ @@ -617,6 +640,29 @@ static void xen_pgd_unpin(pgd_t *pgd) xen_mc_issue(0); } +/* + * On resume, undo any pinning done at save, so that the rest of the + * kernel doesn't see any unexpected pinned pagetables. + */ +void xen_mm_unpin_all(void) +{ + unsigned long flags; + struct page *page; + + spin_lock_irqsave(&pgd_lock, flags); + + list_for_each_entry(page, &pgd_list, lru) { + if (PageSavePinned(page)) { + BUG_ON(!PagePinned(page)); + printk("unpinning pinned %p\n", page_address(page)); + xen_pgd_unpin((pgd_t *)page_address(page)); + ClearPageSavePinned(page); + } + } + + spin_unlock_irqrestore(&pgd_lock, flags); +} + void xen_activate_mm(struct mm_struct *prev, struct mm_struct *next) { spin_lock(&next->page_table_lock); diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c index 74ab8968c525..d2e3c20127d7 100644 --- a/arch/x86/xen/smp.c +++ b/arch/x86/xen/smp.c @@ -35,7 +35,7 @@ #include "xen-ops.h" #include "mmu.h" -static cpumask_t xen_cpu_initialized_map; +cpumask_t xen_cpu_initialized_map; static DEFINE_PER_CPU(int, resched_irq) = -1; static DEFINE_PER_CPU(int, callfunc_irq) = -1; static DEFINE_PER_CPU(int, debug_irq) = -1; diff --git a/arch/x86/xen/suspend.c b/arch/x86/xen/suspend.c new file mode 100644 index 000000000000..7620a16fe535 --- /dev/null +++ b/arch/x86/xen/suspend.c @@ -0,0 +1,42 @@ +#include + +#include +#include +#include + +#include +#include + +#include "xen-ops.h" +#include "mmu.h" + +void xen_pre_suspend(void) +{ + xen_start_info->store_mfn = mfn_to_pfn(xen_start_info->store_mfn); + xen_start_info->console.domU.mfn = + mfn_to_pfn(xen_start_info->console.domU.mfn); + + BUG_ON(!irqs_disabled()); + + HYPERVISOR_shared_info = &xen_dummy_shared_info; + if (HYPERVISOR_update_va_mapping(fix_to_virt(FIX_PARAVIRT_BOOTMAP), + __pte_ma(0), 0)) + BUG(); +} + +void xen_post_suspend(int suspend_cancelled) +{ + if (suspend_cancelled) { + xen_start_info->store_mfn = + pfn_to_mfn(xen_start_info->store_mfn); + xen_start_info->console.domU.mfn = + pfn_to_mfn(xen_start_info->console.domU.mfn); + } else { +#ifdef CONFIG_SMP + xen_cpu_initialized_map = cpu_online_map; +#endif + } + + xen_setup_shared_info(); +} + diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c index c39e1a5aa241..0bef256e5f2d 100644 --- a/arch/x86/xen/time.c +++ b/arch/x86/xen/time.c @@ -572,6 +572,14 @@ void xen_setup_cpu_clockevents(void) clockevents_register_device(&__get_cpu_var(xen_clock_events)); } +void xen_time_suspend(void) +{ +} + +void xen_time_resume(void) +{ +} + __init void xen_time_init(void) { int cpu = smp_processor_id(); diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h index a1bc89a8f169..a0503acad664 100644 --- a/arch/x86/xen/xen-ops.h +++ b/arch/x86/xen/xen-ops.h @@ -9,6 +9,7 @@ extern const char xen_hypervisor_callback[]; extern const char xen_failsafe_callback[]; +struct trap_info; void xen_copy_trap_info(struct trap_info *traps); DECLARE_PER_CPU(unsigned long, xen_cr3); @@ -19,6 +20,7 @@ extern struct shared_info xen_dummy_shared_info; extern struct shared_info *HYPERVISOR_shared_info; void xen_setup_mfn_list_list(void); +void xen_setup_shared_info(void); char * __init xen_memory_setup(void); void __init xen_arch_setup(void); @@ -59,6 +61,8 @@ int xen_smp_call_function_single(int cpu, void (*func) (void *info), void *info, int xen_smp_call_function_mask(cpumask_t mask, void (*func)(void *), void *info, int wait); +extern cpumask_t xen_cpu_initialized_map; + /* Declare an asm function, along with symbols needed to make it inlineable */ diff --git a/drivers/xen/events.c b/drivers/xen/events.c index 70375a690761..73d78dc9b875 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -674,6 +674,89 @@ static int retrigger_dynirq(unsigned int irq) return ret; } +static void restore_cpu_virqs(unsigned int cpu) +{ + struct evtchn_bind_virq bind_virq; + int virq, irq, evtchn; + + for (virq = 0; virq < NR_VIRQS; virq++) { + if ((irq = per_cpu(virq_to_irq, cpu)[virq]) == -1) + continue; + + BUG_ON(irq_info[irq].type != IRQT_VIRQ); + BUG_ON(irq_info[irq].index != virq); + + /* Get a new binding from Xen. */ + bind_virq.virq = virq; + bind_virq.vcpu = cpu; + if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq, + &bind_virq) != 0) + BUG(); + evtchn = bind_virq.port; + + /* Record the new mapping. */ + evtchn_to_irq[evtchn] = irq; + irq_info[irq] = mk_irq_info(IRQT_VIRQ, virq, evtchn); + bind_evtchn_to_cpu(evtchn, cpu); + + /* Ready for use. */ + unmask_evtchn(evtchn); + } +} + +static void restore_cpu_ipis(unsigned int cpu) +{ + struct evtchn_bind_ipi bind_ipi; + int ipi, irq, evtchn; + + for (ipi = 0; ipi < XEN_NR_IPIS; ipi++) { + if ((irq = per_cpu(ipi_to_irq, cpu)[ipi]) == -1) + continue; + + BUG_ON(irq_info[irq].type != IRQT_IPI); + BUG_ON(irq_info[irq].index != ipi); + + /* Get a new binding from Xen. */ + bind_ipi.vcpu = cpu; + if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_ipi, + &bind_ipi) != 0) + BUG(); + evtchn = bind_ipi.port; + + /* Record the new mapping. */ + evtchn_to_irq[evtchn] = irq; + irq_info[irq] = mk_irq_info(IRQT_IPI, ipi, evtchn); + bind_evtchn_to_cpu(evtchn, cpu); + + /* Ready for use. */ + unmask_evtchn(evtchn); + + } +} + +void xen_irq_resume(void) +{ + unsigned int cpu, irq, evtchn; + + init_evtchn_cpu_bindings(); + + /* New event-channel space is not 'live' yet. */ + for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++) + mask_evtchn(evtchn); + + /* No IRQ <-> event-channel mappings. */ + for (irq = 0; irq < NR_IRQS; irq++) + irq_info[irq].evtchn = 0; /* zap event-channel binding */ + + for (evtchn = 0; evtchn < NR_EVENT_CHANNELS; evtchn++) + evtchn_to_irq[evtchn] = -1; + + for_each_possible_cpu(cpu) { + restore_cpu_virqs(cpu); + restore_cpu_ipis(cpu); + } +} + static struct irq_chip xen_dynamic_chip __read_mostly = { .name = "xen-dyn", .mask = disable_dynirq, diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c index 52b6b41b909d..e9e11168616a 100644 --- a/drivers/xen/grant-table.c +++ b/drivers/xen/grant-table.c @@ -471,14 +471,14 @@ static int gnttab_map(unsigned int start_idx, unsigned int end_idx) return 0; } -static int gnttab_resume(void) +int gnttab_resume(void) { if (max_nr_grant_frames() < nr_grant_frames) return -ENOSYS; return gnttab_map(0, nr_grant_frames - 1); } -static int gnttab_suspend(void) +int gnttab_suspend(void) { arch_gnttab_unmap_shared(shared, nr_grant_frames); return 0; diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c index aa7af9e6abc0..ba85fa2cff33 100644 --- a/drivers/xen/manage.c +++ b/drivers/xen/manage.c @@ -5,21 +5,113 @@ #include #include #include +#include +#include #include - -#define SHUTDOWN_INVALID -1 -#define SHUTDOWN_POWEROFF 0 -#define SHUTDOWN_SUSPEND 2 -/* Code 3 is SHUTDOWN_CRASH, which we don't use because the domain can only - * report a crash, not be instructed to crash! - * HALT is the same as POWEROFF, as far as we're concerned. The tools use - * the distinction when we return the reason code to them. - */ -#define SHUTDOWN_HALT 4 +#include +#include +#include +#include + +#include +#include + +enum shutdown_state { + SHUTDOWN_INVALID = -1, + SHUTDOWN_POWEROFF = 0, + SHUTDOWN_SUSPEND = 2, + /* Code 3 is SHUTDOWN_CRASH, which we don't use because the domain can only + report a crash, not be instructed to crash! + HALT is the same as POWEROFF, as far as we're concerned. The tools use + the distinction when we return the reason code to them. */ + SHUTDOWN_HALT = 4, +}; /* Ignore multiple shutdown requests. */ -static int shutting_down = SHUTDOWN_INVALID; +static enum shutdown_state shutting_down = SHUTDOWN_INVALID; + +static int xen_suspend(void *data) +{ + int *cancelled = data; + + BUG_ON(!irqs_disabled()); + + load_cr3(swapper_pg_dir); + + xen_mm_pin_all(); + gnttab_suspend(); + xen_time_suspend(); + xen_pre_suspend(); + + /* + * This hypercall returns 1 if suspend was cancelled + * or the domain was merely checkpointed, and 0 if it + * is resuming in a new domain. + */ + *cancelled = HYPERVISOR_suspend(virt_to_mfn(xen_start_info)); + + xen_post_suspend(*cancelled); + xen_time_resume(); + gnttab_resume(); + xen_mm_unpin_all(); + + if (!*cancelled) { + xen_irq_resume(); + xen_console_resume(); + } + + return 0; +} + +static void do_suspend(void) +{ + int err; + int cancelled = 1; + + shutting_down = SHUTDOWN_SUSPEND; + +#ifdef CONFIG_PREEMPT + /* If the kernel is preemptible, we need to freeze all the processes + to prevent them from being in the middle of a pagetable update + during suspend. */ + err = freeze_processes(); + if (err) { + printk(KERN_ERR "xen suspend: freeze failed %d\n", err); + return; + } +#endif + + err = device_suspend(PMSG_SUSPEND); + if (err) { + printk(KERN_ERR "xen suspend: device_suspend %d\n", err); + goto out; + } + + printk("suspending xenbus...\n"); + /* XXX use normal device tree? */ + xenbus_suspend(); + + err = stop_machine_run(xen_suspend, &cancelled, 0); + if (err) { + printk(KERN_ERR "failed to start xen_suspend: %d\n", err); + goto out; + } + + if (!cancelled) + xenbus_resume(); + else + xenbus_suspend_cancel(); + + device_resume(); + + +out: +#ifdef CONFIG_PREEMPT + thaw_processes(); +#endif + shutting_down = SHUTDOWN_INVALID; +} static void shutdown_handler(struct xenbus_watch *watch, const char **vec, unsigned int len) @@ -52,11 +144,17 @@ static void shutdown_handler(struct xenbus_watch *watch, } if (strcmp(str, "poweroff") == 0 || - strcmp(str, "halt") == 0) + strcmp(str, "halt") == 0) { + shutting_down = SHUTDOWN_POWEROFF; orderly_poweroff(false); - else if (strcmp(str, "reboot") == 0) + } else if (strcmp(str, "reboot") == 0) { + shutting_down = SHUTDOWN_POWEROFF; /* ? */ ctrl_alt_del(); - else { +#ifdef CONFIG_PM_SLEEP + } else if (strcmp(str, "suspend") == 0) { + do_suspend(); +#endif + } else { printk(KERN_INFO "Ignoring shutdown request: %s\n", str); shutting_down = SHUTDOWN_INVALID; } diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h index 590cff32415d..02955a1c3d76 100644 --- a/include/linux/page-flags.h +++ b/include/linux/page-flags.h @@ -157,6 +157,7 @@ PAGEFLAG(Active, active) __CLEARPAGEFLAG(Active, active) __PAGEFLAG(Slab, slab) PAGEFLAG(Checked, owner_priv_1) /* Used by some filesystems */ PAGEFLAG(Pinned, owner_priv_1) TESTSCFLAG(Pinned, owner_priv_1) /* Xen */ +PAGEFLAG(SavePinned, dirty); /* Xen */ PAGEFLAG(Reserved, reserved) __CLEARPAGEFLAG(Reserved, reserved) PAGEFLAG(Private, private) __CLEARPAGEFLAG(Private, private) __SETPAGEFLAG(Private, private) diff --git a/include/xen/events.h b/include/xen/events.h index a82ec0c45c38..67c4436554a9 100644 --- a/include/xen/events.h +++ b/include/xen/events.h @@ -41,4 +41,7 @@ static inline void notify_remote_via_evtchn(int port) } extern void notify_remote_via_irq(int irq); + +extern void xen_irq_resume(void); + #endif /* _XEN_EVENTS_H */ diff --git a/include/xen/grant_table.h b/include/xen/grant_table.h index 466204846121..a40f1cd91be1 100644 --- a/include/xen/grant_table.h +++ b/include/xen/grant_table.h @@ -51,6 +51,9 @@ struct gnttab_free_callback { u16 count; }; +int gnttab_suspend(void); +int gnttab_resume(void); + int gnttab_grant_foreign_access(domid_t domid, unsigned long frame, int readonly); diff --git a/include/xen/xen-ops.h b/include/xen/xen-ops.h index 10ddfe0142d0..5d7a6db54a8c 100644 --- a/include/xen/xen-ops.h +++ b/include/xen/xen-ops.h @@ -5,4 +5,13 @@ DECLARE_PER_CPU(struct vcpu_info *, xen_vcpu); +void xen_pre_suspend(void); +void xen_post_suspend(int suspend_cancelled); + +void xen_mm_pin_all(void); +void xen_mm_unpin_all(void); + +void xen_time_suspend(void); +void xen_time_resume(void); + #endif /* INCLUDE_XEN_OPS_H */ -- cgit v1.2.3 From cc1a9d86ce989083703c4bdc11b75a87e1cc404a Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sun, 8 Jun 2008 19:39:16 -0700 Subject: mm, x86: shrink_active_range() should check all Now we are using register_e820_active_regions() instead of add_active_range() directly. So end_pfn could be different between the value in early_node_map to node_end_pfn. So we need to make shrink_active_range() smarter. shrink_active_range() is a generic MM function in mm/page_alloc.c but it is only used on 32-bit x86. Should we move it back to some file in arch/x86? Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/mm/discontig_32.c | 2 +- include/linux/mm.h | 3 +-- mm/page_alloc.c | 44 ++++++++++++++++++++++++++++++++++---------- 3 files changed, 36 insertions(+), 13 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/mm/discontig_32.c b/arch/x86/mm/discontig_32.c index a89ccf3d4c14..489605bab85a 100644 --- a/arch/x86/mm/discontig_32.c +++ b/arch/x86/mm/discontig_32.c @@ -282,7 +282,7 @@ static unsigned long calculate_numa_remap_pages(void) node_end_pfn[nid] -= size; node_remap_start_pfn[nid] = node_end_pfn[nid]; - shrink_active_range(nid, old_end_pfn, node_end_pfn[nid]); + shrink_active_range(nid, node_end_pfn[nid]); } printk("Reserving total of %ld pages for numa KVA remap\n", reserve_pages); diff --git a/include/linux/mm.h b/include/linux/mm.h index c31a9cd2a30e..7cbd949f2516 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -997,8 +997,7 @@ extern void free_area_init_node(int nid, pg_data_t *pgdat, extern void free_area_init_nodes(unsigned long *max_zone_pfn); extern void add_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); -extern void shrink_active_range(unsigned int nid, unsigned long old_end_pfn, - unsigned long new_end_pfn); +extern void shrink_active_range(unsigned int nid, unsigned long new_end_pfn); extern void push_node_boundaries(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); extern void remove_all_active_ranges(void); diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 502223c3c2c6..215408684076 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3579,25 +3579,49 @@ void __init add_active_range(unsigned int nid, unsigned long start_pfn, /** * shrink_active_range - Shrink an existing registered range of PFNs * @nid: The node id the range is on that should be shrunk - * @old_end_pfn: The old end PFN of the range * @new_end_pfn: The new PFN of the range * * i386 with NUMA use alloc_remap() to store a node_mem_map on a local node. - * The map is kept at the end physical page range that has already been - * registered with add_active_range(). This function allows an arch to shrink - * an existing registered range. + * The map is kept near the end physical page range that has already been + * registered. This function allows an arch to shrink an existing registered + * range. */ -void __init shrink_active_range(unsigned int nid, unsigned long old_end_pfn, - unsigned long new_end_pfn) +void __init shrink_active_range(unsigned int nid, unsigned long new_end_pfn) { - int i; + int i, j; + int removed = 0; /* Find the old active region end and shrink */ - for_each_active_range_index_in_nid(i, nid) - if (early_node_map[i].end_pfn == old_end_pfn) { + for_each_active_range_index_in_nid(i, nid) { + if (early_node_map[i].start_pfn >= new_end_pfn) { + /* clear it */ + early_node_map[i].end_pfn = 0; + removed = 1; + continue; + } + if (early_node_map[i].end_pfn > new_end_pfn) { early_node_map[i].end_pfn = new_end_pfn; - break; + continue; } + } + + if (!removed) + return; + + /* remove the blank ones */ + for (i = nr_nodemap_entries - 1; i > 0; i--) { + if (early_node_map[i].nid != nid) + continue; + if (early_node_map[i].end_pfn) + continue; + /* we found it, get rid of it */ + for (j = i; j < nr_nodemap_entries - 1; j++) + memcpy(&early_node_map[j], &early_node_map[j+1], + sizeof(early_node_map[j])); + j = nr_nodemap_entries - 1; + memset(&early_node_map[j], 0, sizeof(early_node_map[j])); + nr_nodemap_entries--; + } } /** -- cgit v1.2.3 From 3da757daf86e498872855f0b5e101f763ba79499 Mon Sep 17 00:00:00 2001 From: Alok Kataria Date: Fri, 20 Jun 2008 15:06:33 -0700 Subject: x86: use cpu_khz for loops_per_jiffy calculation On the x86 platform we can use the value of tsc_khz computed during tsc calibration to calculate the loops_per_jiffy value. Its very important to keep the error in lpj values to minimum as any error in that may result in kernel panic in check_timer. In virtualization environment, On a highly overloaded host the guest delay calibration may sometimes result in errors beyond the ~50% that timer_irq_works can handle, resulting in the guest panicking. Does some formating changes to lpj_setup code to now have a single printk to print the bogomips value. We do this only for the boot processor because the AP's can have different base frequencies or the BIOS might boot a AP at a different frequency. Signed-off-by: Alok N Kataria Cc: Arjan van de Ven Cc: Daniel Hecht Cc: Tim Mann Cc: Zach Amsden Cc: Sahil Rihan Signed-off-by: Ingo Molnar --- arch/x86/kernel/time_64.c | 2 ++ arch/x86/kernel/tsc_32.c | 5 +++++ include/linux/delay.h | 1 + init/calibrate.c | 36 +++++++++++++++++++----------------- 4 files changed, 27 insertions(+), 17 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/time_64.c b/arch/x86/kernel/time_64.c index c737849e2ef7..12b4a71bd074 100644 --- a/arch/x86/kernel/time_64.c +++ b/arch/x86/kernel/time_64.c @@ -123,6 +123,8 @@ void __init time_init(void) (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)) cpu_khz = calculate_cpu_khz(); + lpj_tsc = ((unsigned long)tsc_khz * 1000)/HZ; + if (unsynchronized_tsc()) mark_tsc_unstable("TSCs unsynchronized"); diff --git a/arch/x86/kernel/tsc_32.c b/arch/x86/kernel/tsc_32.c index 068759db63dd..be729035b30b 100644 --- a/arch/x86/kernel/tsc_32.c +++ b/arch/x86/kernel/tsc_32.c @@ -401,6 +401,7 @@ static inline void check_geode_tsc_reliable(void) { } void __init tsc_init(void) { int cpu; + u64 lpj; if (!cpu_has_tsc || tsc_disabled) { /* Disable the TSC in case of !cpu_has_tsc */ @@ -421,6 +422,10 @@ void __init tsc_init(void) return; } + lpj = ((u64)tsc_khz * 1000); + do_div(lpj, HZ); + lpj_tsc = lpj; + printk("Detected %lu.%03lu MHz processor.\n", (unsigned long)cpu_khz / 1000, (unsigned long)cpu_khz % 1000); diff --git a/include/linux/delay.h b/include/linux/delay.h index 54552d21296e..01aec60590ab 100644 --- a/include/linux/delay.h +++ b/include/linux/delay.h @@ -41,6 +41,7 @@ static inline void ndelay(unsigned long x) #define ndelay(x) ndelay(x) #endif +extern unsigned long lpj_tsc; void calibrate_delay(void); void msleep(unsigned int msecs); unsigned long msleep_interruptible(unsigned int msecs); diff --git a/init/calibrate.c b/init/calibrate.c index ecb3822d4f70..86286974dada 100644 --- a/init/calibrate.c +++ b/init/calibrate.c @@ -8,7 +8,9 @@ #include #include #include +#include +unsigned long lpj_tsc; unsigned long preset_lpj; static int __init lpj_setup(char *str) { @@ -108,6 +110,10 @@ static unsigned long __cpuinit calibrate_delay_direct(void) {return 0;} * This is the number of bits of precision for the loops_per_jiffy. Each * bit takes on average 1.5/HZ seconds. This (like the original) is a little * better than 1% + * For the boot cpu we can skip the delay calibration and assign it a value + * calculated based on the tsc frequency. + * For the rest of the CPUs we cannot assume that the tsc frequency is same as + * the cpu frequency, hence do the calibration for those. */ #define LPS_PREC 8 @@ -118,20 +124,20 @@ void __cpuinit calibrate_delay(void) if (preset_lpj) { loops_per_jiffy = preset_lpj; - printk("Calibrating delay loop (skipped)... " - "%lu.%02lu BogoMIPS preset\n", - loops_per_jiffy/(500000/HZ), - (loops_per_jiffy/(5000/HZ)) % 100); + printk(KERN_INFO + "Calibrating delay loop (skipped) preset value.. "); + } else if ((smp_processor_id() == 0) && lpj_tsc) { + loops_per_jiffy = lpj_tsc; + printk(KERN_INFO + "Calibrating delay loop (skipped), " + "using tsc calculated value.. "); } else if ((loops_per_jiffy = calibrate_delay_direct()) != 0) { - printk("Calibrating delay using timer specific routine.. "); - printk("%lu.%02lu BogoMIPS (lpj=%lu)\n", - loops_per_jiffy/(500000/HZ), - (loops_per_jiffy/(5000/HZ)) % 100, - loops_per_jiffy); + printk(KERN_INFO + "Calibrating delay using timer specific routine.. "); } else { loops_per_jiffy = (1<<12); - printk(KERN_DEBUG "Calibrating delay loop... "); + printk(KERN_INFO "Calibrating delay loop... "); while ((loops_per_jiffy <<= 1) != 0) { /* wait for "start of" clock tick */ ticks = jiffies; @@ -161,12 +167,8 @@ void __cpuinit calibrate_delay(void) if (jiffies != ticks) /* longer than 1 tick */ loops_per_jiffy &= ~loopbit; } - - /* Round the value and print it */ - printk("%lu.%02lu BogoMIPS (lpj=%lu)\n", - loops_per_jiffy/(500000/HZ), - (loops_per_jiffy/(5000/HZ)) % 100, - loops_per_jiffy); } - + printk(KERN_INFO "%lu.%02lu BogoMIPS (lpj=%lu)\n", + loops_per_jiffy/(500000/HZ), + (loops_per_jiffy/(5000/HZ)) % 100, loops_per_jiffy); } -- cgit v1.2.3 From f3f3149f35b9195ef4b761b1353fc0766b5f53be Mon Sep 17 00:00:00 2001 From: Alok Kataria Date: Mon, 23 Jun 2008 18:21:56 -0700 Subject: x86: use cpu_khz for loops_per_jiffy calculation, cleanup As suggested by Ingo, remove all references to tsc from init/calibrate.c TSC is x86 specific, and using tsc in variable names in a generic file should be avoided. lpj_tsc is now called lpj_fine, since it is related to fine tuning of lpj value. Also tsc_rate_* is called timer_rate_* Signed-off-by: Alok N Kataria Cc: Arjan van de Ven Cc: Daniel Hecht Cc: Tim Mann Cc: Zach Amsden Cc: Sahil Rihan Signed-off-by: Ingo Molnar --- arch/x86/kernel/time_64.c | 2 +- arch/x86/kernel/tsc_32.c | 2 +- include/linux/delay.h | 2 +- init/calibrate.c | 36 +++++++++++++++++++----------------- 4 files changed, 22 insertions(+), 20 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/time_64.c b/arch/x86/kernel/time_64.c index 12b4a71bd074..39ae8511a137 100644 --- a/arch/x86/kernel/time_64.c +++ b/arch/x86/kernel/time_64.c @@ -123,7 +123,7 @@ void __init time_init(void) (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)) cpu_khz = calculate_cpu_khz(); - lpj_tsc = ((unsigned long)tsc_khz * 1000)/HZ; + lpj_fine = ((unsigned long)tsc_khz * 1000)/HZ; if (unsynchronized_tsc()) mark_tsc_unstable("TSCs unsynchronized"); diff --git a/arch/x86/kernel/tsc_32.c b/arch/x86/kernel/tsc_32.c index 0af49fb533e1..048baab77268 100644 --- a/arch/x86/kernel/tsc_32.c +++ b/arch/x86/kernel/tsc_32.c @@ -425,7 +425,7 @@ void __init tsc_init(void) lpj = ((u64)tsc_khz * 1000); do_div(lpj, HZ); - lpj_tsc = lpj; + lpj_fine = lpj; printk("Detected %lu.%03lu MHz processor.\n", (unsigned long)cpu_khz / 1000, diff --git a/include/linux/delay.h b/include/linux/delay.h index 01aec60590ab..fd832c6d419e 100644 --- a/include/linux/delay.h +++ b/include/linux/delay.h @@ -41,7 +41,7 @@ static inline void ndelay(unsigned long x) #define ndelay(x) ndelay(x) #endif -extern unsigned long lpj_tsc; +extern unsigned long lpj_fine; void calibrate_delay(void); void msleep(unsigned int msecs); unsigned long msleep_interruptible(unsigned int msecs); diff --git a/init/calibrate.c b/init/calibrate.c index 86286974dada..7963e3fc51d9 100644 --- a/init/calibrate.c +++ b/init/calibrate.c @@ -10,7 +10,7 @@ #include #include -unsigned long lpj_tsc; +unsigned long lpj_fine; unsigned long preset_lpj; static int __init lpj_setup(char *str) { @@ -35,9 +35,9 @@ static unsigned long __cpuinit calibrate_delay_direct(void) unsigned long pre_start, start, post_start; unsigned long pre_end, end, post_end; unsigned long start_jiffies; - unsigned long tsc_rate_min, tsc_rate_max; - unsigned long good_tsc_sum = 0; - unsigned long good_tsc_count = 0; + unsigned long timer_rate_min, timer_rate_max; + unsigned long good_timer_sum = 0; + unsigned long good_timer_count = 0; int i; if (read_current_timer(&pre_start) < 0 ) @@ -81,22 +81,24 @@ static unsigned long __cpuinit calibrate_delay_direct(void) } read_current_timer(&post_end); - tsc_rate_max = (post_end - pre_start) / DELAY_CALIBRATION_TICKS; - tsc_rate_min = (pre_end - post_start) / DELAY_CALIBRATION_TICKS; + timer_rate_max = (post_end - pre_start) / + DELAY_CALIBRATION_TICKS; + timer_rate_min = (pre_end - post_start) / + DELAY_CALIBRATION_TICKS; /* - * If the upper limit and lower limit of the tsc_rate is + * If the upper limit and lower limit of the timer_rate is * >= 12.5% apart, redo calibration. */ if (pre_start != 0 && pre_end != 0 && - (tsc_rate_max - tsc_rate_min) < (tsc_rate_max >> 3)) { - good_tsc_count++; - good_tsc_sum += tsc_rate_max; + (timer_rate_max - timer_rate_min) < (timer_rate_max >> 3)) { + good_timer_count++; + good_timer_sum += timer_rate_max; } } - if (good_tsc_count) - return (good_tsc_sum/good_tsc_count); + if (good_timer_count) + return (good_timer_sum/good_timer_count); printk(KERN_WARNING "calibrate_delay_direct() failed to get a good " "estimate for loops_per_jiffy.\nProbably due to long platform interrupts. Consider using \"lpj=\" boot option.\n"); @@ -111,8 +113,8 @@ static unsigned long __cpuinit calibrate_delay_direct(void) {return 0;} * bit takes on average 1.5/HZ seconds. This (like the original) is a little * better than 1% * For the boot cpu we can skip the delay calibration and assign it a value - * calculated based on the tsc frequency. - * For the rest of the CPUs we cannot assume that the tsc frequency is same as + * calculated based on the timer frequency. + * For the rest of the CPUs we cannot assume that the timer frequency is same as * the cpu frequency, hence do the calibration for those. */ #define LPS_PREC 8 @@ -126,11 +128,11 @@ void __cpuinit calibrate_delay(void) loops_per_jiffy = preset_lpj; printk(KERN_INFO "Calibrating delay loop (skipped) preset value.. "); - } else if ((smp_processor_id() == 0) && lpj_tsc) { - loops_per_jiffy = lpj_tsc; + } else if ((smp_processor_id() == 0) && lpj_fine) { + loops_per_jiffy = lpj_fine; printk(KERN_INFO "Calibrating delay loop (skipped), " - "using tsc calculated value.. "); + "value calculated using timer frequency.. "); } else if ((loops_per_jiffy = calibrate_delay_direct()) != 0) { printk(KERN_INFO "Calibrating delay using timer specific routine.. "); -- cgit v1.2.3 From d2dbf343329dc777d77488743465f7be4245971d Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Fri, 13 Jun 2008 02:00:56 -0700 Subject: x86: clean up reserve_bootmem_generic() and port it to 32-bit 1. add reserve_bootmem_generic for 32bit 2. change len to unsigned long 3. make early_res_to_bootmem to use it Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/kernel/e820.c | 2 +- arch/x86/kernel/mpparse.c | 18 ++++++------------ arch/x86/mm/init_32.c | 6 ++++++ arch/x86/mm/init_64.c | 3 ++- include/asm-x86/proto.h | 2 -- include/linux/bootmem.h | 2 ++ 6 files changed, 17 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 4f2cd5d179e2..774063f11be0 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -635,7 +635,7 @@ void __init early_res_to_bootmem(u64 start, u64 end) continue; printk(KERN_INFO " early res: %d [%llx-%llx] %s\n", i, final_start, final_end - 1, r->name); - reserve_bootmem(final_start, final_end - final_start, + reserve_bootmem_generic(final_start, final_end - final_start, BOOTMEM_DEFAULT); } } diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index 7ac1b689b70a..b62ac6ba1410 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -859,10 +859,11 @@ static int __init smp_scan_config(unsigned long base, unsigned long length, if (!reserve) return 1; -#ifdef CONFIG_X86_32 - reserve_bootmem(virt_to_phys(mpf), PAGE_SIZE, + reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE, BOOTMEM_DEFAULT); if (mpf->mpf_physptr) { + unsigned long size = PAGE_SIZE; +#ifdef CONFIG_X86_32 /* * We cannot access to MPC table to compute * table size yet, as only few megabytes from @@ -872,22 +873,15 @@ static int __init smp_scan_config(unsigned long base, unsigned long length, * PAGE_SIZE from mpg->mpf_physptr yields BUG() * in reserve_bootmem. */ - unsigned long size = PAGE_SIZE; unsigned long end = max_low_pfn * PAGE_SIZE; if (mpf->mpf_physptr + size > end) size = end - mpf->mpf_physptr; - reserve_bootmem(mpf->mpf_physptr, size, +#endif + reserve_bootmem_generic(mpf->mpf_physptr, size, BOOTMEM_DEFAULT); } -#else - reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE, - BOOTMEM_DEFAULT); - if (mpf->mpf_physptr) - reserve_bootmem_generic(mpf->mpf_physptr, - PAGE_SIZE, BOOTMEM_DEFAULT); -#endif - return 1; + return 1; } bp += 4; length -= 16; diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index 0e7bb5e81670..abadb1da70df 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -785,3 +785,9 @@ void free_initrd_mem(unsigned long start, unsigned long end) free_init_pages("initrd memory", start, end); } #endif + +int __init reserve_bootmem_generic(unsigned long phys, unsigned long len, + int flags) +{ + return reserve_bootmem(phys, len, flags); +} diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index bf7bf1de6c25..b8c2c1ef7ad5 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -799,7 +799,8 @@ void free_initrd_mem(unsigned long start, unsigned long end) } #endif -int __init reserve_bootmem_generic(unsigned long phys, unsigned len, int flags) +int __init reserve_bootmem_generic(unsigned long phys, unsigned long len, + int flags) { #ifdef CONFIG_NUMA int nid, next_nid; diff --git a/include/asm-x86/proto.h b/include/asm-x86/proto.h index a9f51472521e..3dd458c385c0 100644 --- a/include/asm-x86/proto.h +++ b/include/asm-x86/proto.h @@ -14,8 +14,6 @@ extern void ia32_syscall(void); extern void ia32_cstar_target(void); extern void ia32_sysenter_target(void); -extern int reserve_bootmem_generic(unsigned long phys, unsigned len, int flags); - extern void syscall32_cpu_init(void); extern void check_efer(void); diff --git a/include/linux/bootmem.h b/include/linux/bootmem.h index 686895bacd9d..a1d9b79078ea 100644 --- a/include/linux/bootmem.h +++ b/include/linux/bootmem.h @@ -84,6 +84,8 @@ extern int reserve_bootmem(unsigned long addr, unsigned long size, int flags); __alloc_bootmem_low(x, PAGE_SIZE, 0) #endif /* !CONFIG_HAVE_ARCH_BOOTMEM_NODE */ +extern int reserve_bootmem_generic(unsigned long addr, unsigned long size, + int flags); extern unsigned long free_all_bootmem(void); extern unsigned long free_all_bootmem_node(pg_data_t *pgdat); extern void *__alloc_bootmem_node(pg_data_t *pgdat, -- cgit v1.2.3 From cc1050bafebfb1d7935331282e948b5016318192 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Fri, 13 Jun 2008 19:08:52 -0700 Subject: x86: replace shrink_active_range() with remove_active_range() in case we have kva before ramdisk on a node, we still need to use those ranges. v2: reserve_early kva ram area, in case there are holes in highmem, to avoid those area could be treat as free high pages. Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/mm/discontig_32.c | 45 ++++++++++++++++++++++++--------------------- include/linux/mm.h | 3 ++- mm/page_alloc.c | 29 +++++++++++++++++++++++------ 3 files changed, 49 insertions(+), 28 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/mm/discontig_32.c b/arch/x86/mm/discontig_32.c index accc7c6c57fc..c3f119e99e0d 100644 --- a/arch/x86/mm/discontig_32.c +++ b/arch/x86/mm/discontig_32.c @@ -230,8 +230,8 @@ static unsigned long calculate_numa_remap_pages(void) unsigned long size, reserve_pages = 0; for_each_online_node(nid) { - u64 node_end_target; - u64 node_end_final; + u64 node_kva_target; + u64 node_kva_final; /* * The acpi/srat node info can show hot-add memroy zones @@ -254,42 +254,45 @@ static unsigned long calculate_numa_remap_pages(void) /* now the roundup is correct, convert to PAGE_SIZE pages */ size = size * PTRS_PER_PTE; - node_end_target = round_down(node_end_pfn[nid] - size, + node_kva_target = round_down(node_end_pfn[nid] - size, PTRS_PER_PTE); - node_end_target <<= PAGE_SHIFT; + node_kva_target <<= PAGE_SHIFT; do { - node_end_final = find_e820_area(node_end_target, + node_kva_final = find_e820_area(node_kva_target, ((u64)node_end_pfn[nid])<>PAGE_SHIFT) > (node_start_pfn[nid])); + node_kva_target -= LARGE_PAGE_BYTES; + } while (node_kva_final == -1ULL && + (node_kva_target>>PAGE_SHIFT) > (node_start_pfn[nid])); - if (node_end_final == -1ULL) + if (node_kva_final == -1ULL) panic("Can not get kva ram\n"); - printk("Reserving %ld pages of KVA for lmem_map of node %d\n", - size, nid); node_remap_size[nid] = size; node_remap_offset[nid] = reserve_pages; reserve_pages += size; - printk("Shrinking node %d from %ld pages to %lld pages\n", - nid, node_end_pfn[nid], node_end_final>>PAGE_SHIFT); + printk("Reserving %ld pages of KVA for lmem_map of node %d at %llx\n", + size, nid, node_kva_final>>PAGE_SHIFT); /* * prevent kva address below max_low_pfn want it on system * with less memory later. * layout will be: KVA address , KVA RAM + * + * we are supposed to only record the one less then max_low_pfn + * but we could have some hole in high memory, and it will only + * check page_is_ram(pfn) && !page_is_reserved_early(pfn) to decide + * to use it as free. + * So reserve_early here, hope we don't run out of that array */ - if ((node_end_final>>PAGE_SHIFT) < max_low_pfn) - reserve_early(node_end_final, - node_end_final+(((u64)size)<>PAGE_SHIFT; - node_remap_start_pfn[nid] = node_end_pfn[nid]; - shrink_active_range(nid, node_end_pfn[nid]); + reserve_early(node_kva_final, + node_kva_final+(((u64)size)<>PAGE_SHIFT; + remove_active_range(nid, node_remap_start_pfn[nid], + node_remap_start_pfn[nid] + size); } printk("Reserving total of %ld pages for numa KVA remap\n", reserve_pages); diff --git a/include/linux/mm.h b/include/linux/mm.h index ce8e397a61f6..034a3156d2f0 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -998,7 +998,8 @@ extern void free_area_init_node(int nid, pg_data_t *pgdat, extern void free_area_init_nodes(unsigned long *max_zone_pfn); extern void add_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); -extern void shrink_active_range(unsigned int nid, unsigned long new_end_pfn); +extern void remove_active_range(unsigned int nid, unsigned long start_pfn, + unsigned long end_pfn); extern void push_node_boundaries(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); extern void remove_all_active_ranges(void); diff --git a/mm/page_alloc.c b/mm/page_alloc.c index eee5ba7509c1..d80e1868e570 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3552,30 +3552,47 @@ void __init add_active_range(unsigned int nid, unsigned long start_pfn, } /** - * shrink_active_range - Shrink an existing registered range of PFNs + * remove_active_range - Shrink an existing registered range of PFNs * @nid: The node id the range is on that should be shrunk - * @new_end_pfn: The new PFN of the range + * @start_pfn: The new PFN of the range + * @end_pfn: The new PFN of the range * * i386 with NUMA use alloc_remap() to store a node_mem_map on a local node. * The map is kept near the end physical page range that has already been * registered. This function allows an arch to shrink an existing registered * range. */ -void __init shrink_active_range(unsigned int nid, unsigned long new_end_pfn) +void __init remove_active_range(unsigned int nid, unsigned long start_pfn, + unsigned long end_pfn) { int i, j; int removed = 0; + printk(KERN_DEBUG "remove_active_range (%d, %lu, %lu)\n", + nid, start_pfn, end_pfn); + /* Find the old active region end and shrink */ for_each_active_range_index_in_nid(i, nid) { - if (early_node_map[i].start_pfn >= new_end_pfn) { + if (early_node_map[i].start_pfn >= start_pfn && + early_node_map[i].end_pfn <= end_pfn) { /* clear it */ + early_node_map[i].start_pfn = 0; early_node_map[i].end_pfn = 0; removed = 1; continue; } - if (early_node_map[i].end_pfn > new_end_pfn) { - early_node_map[i].end_pfn = new_end_pfn; + if (early_node_map[i].start_pfn < start_pfn && + early_node_map[i].end_pfn > start_pfn) { + unsigned long temp_end_pfn = early_node_map[i].end_pfn; + early_node_map[i].end_pfn = start_pfn; + if (temp_end_pfn > end_pfn) + add_active_range(nid, end_pfn, temp_end_pfn); + continue; + } + if (early_node_map[i].start_pfn >= start_pfn && + early_node_map[i].end_pfn > end_pfn && + early_node_map[i].start_pfn < end_pfn) { + early_node_map[i].start_pfn = end_pfn; continue; } } -- cgit v1.2.3 From b5bc6c0e55000dab86b73f838f5ad02908b23755 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sat, 14 Jun 2008 18:32:52 -0700 Subject: x86, mm: use add_highpages_with_active_regions() for high pages init v2 use early_node_map to init high pages, so we can remove page_is_ram() and page_is_reserved_early() in the big loop with add_one_highpage also remove page_is_reserved_early(), it is not needed anymore. v2: fix the build of other platforms Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/kernel/e820.c | 11 -------- arch/x86/mm/discontig_32.c | 19 ++++++-------- arch/x86/mm/init_32.c | 62 ++++++++++++++++++++++++++++++++++++++-------- include/asm-x86/e820.h | 1 - include/asm-x86/highmem.h | 3 +++ include/linux/mm.h | 2 ++ mm/page_alloc.c | 8 ++++++ 7 files changed, 71 insertions(+), 35 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 5051ce744b4e..ed46b7a6bc13 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -612,17 +612,6 @@ void __init free_early(u64 start, u64 end) early_res[j - 1].end = 0; } -int __init page_is_reserved_early(unsigned long pagenr) -{ - u64 start = (u64)pagenr << PAGE_SHIFT; - int i; - struct early_res *r; - - i = find_overlapped_early(start, start + PAGE_SIZE); - r = &early_res[i]; - return (i < MAX_EARLY_RES && r->end); -} - void __init early_res_to_bootmem(u64 start, u64 end) { int i; diff --git a/arch/x86/mm/discontig_32.c b/arch/x86/mm/discontig_32.c index c3f119e99e0d..7c4d0255f8d8 100644 --- a/arch/x86/mm/discontig_32.c +++ b/arch/x86/mm/discontig_32.c @@ -100,7 +100,6 @@ unsigned long node_memmap_size_bytes(int nid, unsigned long start_pfn, #endif extern unsigned long find_max_low_pfn(void); -extern void add_one_highpage_init(struct page *, int, int); extern unsigned long highend_pfn, highstart_pfn; #define LARGE_PAGE_BYTES (PTRS_PER_PTE * PAGE_SIZE) @@ -432,10 +431,10 @@ void __init set_highmem_pages_init(int bad_ppro) { #ifdef CONFIG_HIGHMEM struct zone *zone; - struct page *page; + int nid; for_each_zone(zone) { - unsigned long node_pfn, zone_start_pfn, zone_end_pfn; + unsigned long zone_start_pfn, zone_end_pfn; if (!is_highmem(zone)) continue; @@ -443,16 +442,12 @@ void __init set_highmem_pages_init(int bad_ppro) zone_start_pfn = zone->zone_start_pfn; zone_end_pfn = zone_start_pfn + zone->spanned_pages; + nid = zone_to_nid(zone); printk("Initializing %s for node %d (%08lx:%08lx)\n", - zone->name, zone_to_nid(zone), - zone_start_pfn, zone_end_pfn); - - for (node_pfn = zone_start_pfn; node_pfn < zone_end_pfn; node_pfn++) { - if (!pfn_valid(node_pfn)) - continue; - page = pfn_to_page(node_pfn); - add_one_highpage_init(page, node_pfn, bad_ppro); - } + zone->name, nid, zone_start_pfn, zone_end_pfn); + + add_highpages_with_active_regions(nid, zone_start_pfn, + zone_end_pfn, bad_ppro); } totalram_pages += totalhigh_pages; #endif diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index abadb1da70df..ba07a489230e 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -287,10 +287,10 @@ static void __init permanent_kmaps_init(pgd_t *pgd_base) pkmap_page_table = pte; } -void __init add_one_highpage_init(struct page *page, int pfn, int bad_ppro) +static void __init +add_one_highpage_init(struct page *page, int pfn, int bad_ppro) { - if (page_is_ram(pfn) && !(bad_ppro && page_kills_ppro(pfn)) && - !page_is_reserved_early(pfn)) { + if (!(bad_ppro && page_kills_ppro(pfn))) { ClearPageReserved(page); init_page_count(page); __free_page(page); @@ -299,18 +299,58 @@ void __init add_one_highpage_init(struct page *page, int pfn, int bad_ppro) SetPageReserved(page); } +struct add_highpages_data { + unsigned long start_pfn; + unsigned long end_pfn; + int bad_ppro; +}; + +static void __init add_highpages_work_fn(unsigned long start_pfn, + unsigned long end_pfn, void *datax) +{ + int node_pfn; + struct page *page; + unsigned long final_start_pfn, final_end_pfn; + struct add_highpages_data *data; + int bad_ppro; + + data = (struct add_highpages_data *)datax; + bad_ppro = data->bad_ppro; + + final_start_pfn = max(start_pfn, data->start_pfn); + final_end_pfn = min(end_pfn, data->end_pfn); + if (final_start_pfn >= final_end_pfn) + return; + + for (node_pfn = final_start_pfn; node_pfn < final_end_pfn; + node_pfn++) { + if (!pfn_valid(node_pfn)) + continue; + page = pfn_to_page(node_pfn); + add_one_highpage_init(page, node_pfn, bad_ppro); + } + +} + +void __init add_highpages_with_active_regions(int nid, unsigned long start_pfn, + unsigned long end_pfn, + int bad_ppro) +{ + struct add_highpages_data data; + + data.start_pfn = start_pfn; + data.end_pfn = end_pfn; + data.bad_ppro = bad_ppro; + + work_with_active_regions(nid, add_highpages_work_fn, &data); +} + #ifndef CONFIG_NUMA static void __init set_highmem_pages_init(int bad_ppro) { - int pfn; + add_highpages_with_active_regions(0, highstart_pfn, highend_pfn, + bad_ppro); - for (pfn = highstart_pfn; pfn < highend_pfn; pfn++) { - /* - * Holes under sparsemem might not have no mem_map[]: - */ - if (pfn_valid(pfn)) - add_one_highpage_init(pfn_to_page(pfn), pfn, bad_ppro); - } totalram_pages += totalhigh_pages; } #endif /* !CONFIG_NUMA */ diff --git a/include/asm-x86/e820.h b/include/asm-x86/e820.h index 6b0ce745a60c..55d310596907 100644 --- a/include/asm-x86/e820.h +++ b/include/asm-x86/e820.h @@ -86,7 +86,6 @@ extern u64 find_e820_area_size(u64 start, u64 *sizep, u64 align); extern void reserve_early(u64 start, u64 end, char *name); extern void free_early(u64 start, u64 end); extern void early_res_to_bootmem(u64 start, u64 end); -extern int page_is_reserved_early(unsigned long pagenr); extern u64 early_reserve_e820(u64 startt, u64 sizet, u64 align); extern unsigned long e820_end_of_ram(void); diff --git a/include/asm-x86/highmem.h b/include/asm-x86/highmem.h index e153f3b44774..85c4fea41ff6 100644 --- a/include/asm-x86/highmem.h +++ b/include/asm-x86/highmem.h @@ -74,6 +74,9 @@ struct page *kmap_atomic_to_page(void *ptr); #define flush_cache_kmaps() do { } while (0) +extern void add_highpages_with_active_regions(int nid, unsigned long start_pfn, + unsigned long end_pfn, int bad_ppro); + #endif /* __KERNEL__ */ #endif /* _ASM_HIGHMEM_H */ diff --git a/include/linux/mm.h b/include/linux/mm.h index 034a3156d2f0..e4de460907c1 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1011,6 +1011,8 @@ extern unsigned long find_min_pfn_with_active_regions(void); extern unsigned long find_max_pfn_with_active_regions(void); extern void free_bootmem_with_active_regions(int nid, unsigned long max_low_pfn); +typedef void (*work_fn_t)(unsigned long, unsigned long, void *); +extern void work_with_active_regions(int nid, work_fn_t work_fn, void *data); extern void sparse_memory_present_with_active_regions(int nid); #ifndef CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID extern int early_pfn_to_nid(unsigned long pfn); diff --git a/mm/page_alloc.c b/mm/page_alloc.c index d80e1868e570..41c6e3aa059f 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -2929,6 +2929,14 @@ void __init free_bootmem_with_active_regions(int nid, } } +void __init work_with_active_regions(int nid, work_fn_t work_fn, void *data) +{ + int i; + + for_each_active_range_index_in_nid(i, nid) + work_fn(early_node_map[i].start_pfn, early_node_map[i].end_pfn, + data); +} /** * sparse_memory_present_with_active_regions - Call memory_present for each active range * @nid: The node to call memory_present for. If MAX_NUMNODES, all nodes will be used. -- cgit v1.2.3 From 3461b0af025251bbc6b3d56c821c6ac2de6f7209 Mon Sep 17 00:00:00 2001 From: Mike Travis Date: Mon, 12 May 2008 21:21:13 +0200 Subject: x86: remove static boot_cpu_pda array v2 * Remove the boot_cpu_pda array and pointer table from the data section. Allocate the pointer table and array during init. do_boot_cpu() will reallocate the pda in node local memory and if the cpu is being brought up before the bootmem array is released (after_bootmem = 0), then it will free the initial pda. This will happen for all cpus present at system startup. This removes 512k + 32k bytes from the data section. For inclusion into sched-devel/latest tree. Based on: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git + sched-devel/latest .../mingo/linux-2.6-sched-devel.git Signed-off-by: Mike Travis Signed-off-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- arch/x86/kernel/head64.c | 26 +++++++++++++++-- arch/x86/kernel/setup.c | 73 ++++++++++++++++++++++++++++++++++++----------- arch/x86/kernel/setup64.c | 8 ++++-- arch/x86/kernel/smpboot.c | 59 +++++++++++++++++++++++++++++--------- include/asm-x86/pda.h | 6 ++-- include/linux/mm.h | 1 + 6 files changed, 135 insertions(+), 38 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index e25c57b8aa84..0ab59edd7067 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -25,6 +25,24 @@ #include #include +/* boot cpu pda */ +static struct x8664_pda _boot_cpu_pda __read_mostly; + +#ifdef CONFIG_SMP +#ifdef CONFIG_DEBUG_PER_CPU_MAPS +/* + * We install an empty cpu_pda pointer table to trap references before + * the actual cpu_pda pointer table is created in setup_cpu_pda_map(). + */ +static struct x8664_pda *__cpu_pda[NR_CPUS] __initdata; +#else +static struct x8664_pda *__cpu_pda[1] __read_mostly; +#endif + +#else /* !CONFIG_SMP (NR_CPUS will be 1) */ +static struct x8664_pda *__cpu_pda[NR_CPUS] __read_mostly; +#endif + static void __init zap_identity_mappings(void) { pgd_t *pgd = pgd_offset_k(0UL); @@ -156,10 +174,12 @@ void __init x86_64_start_kernel(char * real_mode_data) early_printk("Kernel alive\n"); - for (i = 0; i < NR_CPUS; i++) - cpu_pda(i) = &boot_cpu_pda[i]; - + _cpu_pda = __cpu_pda; + cpu_pda(0) = &_boot_cpu_pda; pda_init(0); + + early_printk("Kernel really alive\n"); + copy_bootdata(__va(real_mode_data)); reserve_early(__pa_symbol(&_text), __pa_symbol(&_end), "TEXT DATA BSS"); diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 913af838c3c5..dd12c1c84a8f 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -101,6 +101,50 @@ static inline void setup_cpumask_of_cpu(void) { } */ unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; EXPORT_SYMBOL(__per_cpu_offset); +static inline void setup_cpu_pda_map(void) { } + +#elif !defined(CONFIG_SMP) +static inline void setup_cpu_pda_map(void) { } + +#else /* CONFIG_SMP && CONFIG_X86_64 */ + +/* + * Allocate cpu_pda pointer table and array via alloc_bootmem. + */ +static void __init setup_cpu_pda_map(void) +{ + char *pda; + struct x8664_pda **new_cpu_pda; + unsigned long size; + int cpu; + + size = roundup(sizeof(struct x8664_pda), cache_line_size()); + + /* allocate cpu_pda array and pointer table */ + { + unsigned long tsize = nr_cpu_ids * sizeof(void *); + unsigned long asize = size * (nr_cpu_ids - 1); + + tsize = roundup(tsize, cache_line_size()); + new_cpu_pda = alloc_bootmem(tsize + asize); + pda = (char *)new_cpu_pda + tsize; + } + + /* initialize pointer table to static pda's */ + for_each_possible_cpu(cpu) { + if (cpu == 0) { + /* leave boot cpu pda in place */ + new_cpu_pda[0] = cpu_pda(0); + continue; + } + new_cpu_pda[cpu] = (struct x8664_pda *)pda; + new_cpu_pda[cpu]->in_bootmem = 1; + pda += size; + } + + /* point to new pointer table */ + _cpu_pda = new_cpu_pda; +} #endif /* @@ -110,46 +154,43 @@ EXPORT_SYMBOL(__per_cpu_offset); */ void __init setup_per_cpu_areas(void) { - int i, highest_cpu = 0; - unsigned long size; + ssize_t size = PERCPU_ENOUGH_ROOM; + char *ptr; + int cpu; #ifdef CONFIG_HOTPLUG_CPU prefill_possible_map(); +#else + nr_cpu_ids = num_processors; #endif + /* Setup cpu_pda map */ + setup_cpu_pda_map(); + /* Copy section for each CPU (we discard the original) */ size = PERCPU_ENOUGH_ROOM; printk(KERN_INFO "PERCPU: Allocating %lu bytes of per cpu data\n", size); - for_each_possible_cpu(i) { - char *ptr; + for_each_possible_cpu(cpu) { #ifndef CONFIG_NEED_MULTIPLE_NODES ptr = alloc_bootmem_pages(size); #else - int node = early_cpu_to_node(i); + int node = early_cpu_to_node(cpu); if (!node_online(node) || !NODE_DATA(node)) { ptr = alloc_bootmem_pages(size); printk(KERN_INFO "cpu %d has no node %d or node-local memory\n", - i, node); + cpu, node); } else ptr = alloc_bootmem_pages_node(NODE_DATA(node), size); #endif - if (!ptr) - panic("Cannot allocate cpu data for CPU %d\n", i); -#ifdef CONFIG_X86_64 - cpu_pda(i)->data_offset = ptr - __per_cpu_start; -#else - __per_cpu_offset[i] = ptr - __per_cpu_start; -#endif + per_cpu_offset(cpu) = ptr - __per_cpu_start; memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); - highest_cpu = i; } - nr_cpu_ids = highest_cpu + 1; printk(KERN_DEBUG "NR_CPUS: %d, nr_cpu_ids: %d, nr_node_ids %d\n", NR_CPUS, nr_cpu_ids, nr_node_ids); @@ -199,7 +240,7 @@ void __cpuinit numa_set_node(int cpu, int node) { int *cpu_to_node_map = early_per_cpu_ptr(x86_cpu_to_node_map); - if (node != NUMA_NO_NODE) + if (cpu_pda(cpu) && node != NUMA_NO_NODE) cpu_pda(cpu)->nodenumber = node; if (cpu_to_node_map) diff --git a/arch/x86/kernel/setup64.c b/arch/x86/kernel/setup64.c index aee0e8200777..631ea6cc01d8 100644 --- a/arch/x86/kernel/setup64.c +++ b/arch/x86/kernel/setup64.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include #include @@ -34,9 +35,8 @@ struct boot_params boot_params; cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE; -struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly; +struct x8664_pda **_cpu_pda __read_mostly; EXPORT_SYMBOL(_cpu_pda); -struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned; struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table }; @@ -114,8 +114,10 @@ void pda_init(int cpu) __get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER); if (!pda->irqstackptr) panic("cannot allocate irqstack for cpu %d", cpu); - } + if (pda->nodenumber == 0 && cpu_to_node(cpu) != NUMA_NO_NODE) + pda->nodenumber = cpu_to_node(cpu); + } pda->irqstackptr += IRQSTACKSIZE-64; } diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 036604d3daed..bf0833487455 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -816,6 +816,43 @@ static void __cpuinit do_fork_idle(struct work_struct *work) complete(&c_idle->done); } +/* + * Allocate node local memory for the AP pda. + * + * Must be called after the _cpu_pda pointer table is initialized. + */ +static int __cpuinit get_local_pda(int cpu) +{ + struct x8664_pda *oldpda, *newpda; + unsigned long size = sizeof(struct x8664_pda); + int node = cpu_to_node(cpu); + + if (cpu_pda(cpu) && !cpu_pda(cpu)->in_bootmem) + return 0; + + oldpda = cpu_pda(cpu); + newpda = kmalloc_node(size, GFP_ATOMIC, node); + if (!newpda) { + printk(KERN_ERR "Could not allocate node local PDA " + "for CPU %d on node %d\n", cpu, node); + + if (oldpda) + return 0; /* have a usable pda */ + else + return -1; + } + + if (oldpda) { + memcpy(newpda, oldpda, size); + if (!after_bootmem) + free_bootmem((unsigned long)oldpda, size); + } + + newpda->in_bootmem = 0; + cpu_pda(cpu) = newpda; + return 0; +} + static int __cpuinit do_boot_cpu(int apicid, int cpu) /* * NOTE - on most systems this is a PHYSICAL apic ID, but on multiquad @@ -841,19 +878,11 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) } /* Allocate node local memory for AP pdas */ - if (cpu_pda(cpu) == &boot_cpu_pda[cpu]) { - struct x8664_pda *newpda, *pda; - int node = cpu_to_node(cpu); - pda = cpu_pda(cpu); - newpda = kmalloc_node(sizeof(struct x8664_pda), GFP_ATOMIC, - node); - if (newpda) { - memcpy(newpda, pda, sizeof(struct x8664_pda)); - cpu_pda(cpu) = newpda; - } else - printk(KERN_ERR - "Could not allocate node local PDA for CPU %d on node %d\n", - cpu, node); + if (cpu > 0) { + boot_error = get_local_pda(cpu); + if (boot_error) + goto restore_state; + /* if can't get pda memory, can't start cpu */ } #endif @@ -972,6 +1001,8 @@ do_rest: } } +restore_state: + if (boot_error) { /* Try to put things back the way they were before ... */ unmap_cpu_to_logical_apicid(cpu); @@ -1347,6 +1378,8 @@ __init void prefill_possible_map(void) for (i = 0; i < possible; i++) cpu_set(i, cpu_possible_map); + + nr_cpu_ids = possible; } static void __ref remove_cpu_from_maps(int cpu) diff --git a/include/asm-x86/pda.h b/include/asm-x86/pda.h index de2ad9ac35a9..b34e9a7cc80b 100644 --- a/include/asm-x86/pda.h +++ b/include/asm-x86/pda.h @@ -22,7 +22,8 @@ struct x8664_pda { offset 40!!! */ #endif char *irqstackptr; - int nodenumber; /* number of current node */ + short nodenumber; /* number of current node (32k max) */ + short in_bootmem; /* pda lives in bootmem */ unsigned int __softirq_pending; unsigned int __nmi_count; /* number of NMI on this CPUs */ short mmu_state; @@ -38,8 +39,7 @@ struct x8664_pda { unsigned irq_spurious_count; } ____cacheline_aligned_in_smp; -extern struct x8664_pda *_cpu_pda[]; -extern struct x8664_pda boot_cpu_pda[]; +extern struct x8664_pda **_cpu_pda; extern void pda_init(int); #define cpu_pda(i) (_cpu_pda[i]) diff --git a/include/linux/mm.h b/include/linux/mm.h index 586a943cab01..0ea48a5af823 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1024,6 +1024,7 @@ extern void mem_init(void); extern void show_mem(void); extern void si_meminfo(struct sysinfo * val); extern void si_meminfo_node(struct sysinfo *val, int nid); +extern int after_bootmem; #ifdef CONFIG_NUMA extern void setup_per_cpu_pageset(void); -- cgit v1.2.3 From a7bf0bd5e6af7fe69342dabf2a3b721f0163469a Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Wed, 28 May 2008 15:02:14 +0100 Subject: build: add __page_aligned_data and __page_aligned_bss Making a variable page-aligned by using __attribute__((section(".data.page_aligned"))) is fragile because if sizeof(variable) is not also a multiple of page size, it leaves variables in the remainder of the section unaligned. This patch introduces two new qualifiers, __page_aligned_data and __page_aligned_bss to set the section *and* the alignment of variables. This makes page-aligned variables more robust because the linker will make sure they're aligned properly. Unfortunately it requires *all* page-aligned data to use these macros... Signed-off-by: Ingo Molnar --- arch/x86/kernel/setup64.c | 2 +- arch/x86/mm/ioremap.c | 3 +-- include/linux/linkage.h | 4 ++++ 3 files changed, 6 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/setup64.c b/arch/x86/kernel/setup64.c index 631ea6cc01d8..fc1a56da8240 100644 --- a/arch/x86/kernel/setup64.c +++ b/arch/x86/kernel/setup64.c @@ -40,7 +40,7 @@ EXPORT_SYMBOL(_cpu_pda); struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table }; -char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned"))); +char boot_cpu_stack[IRQSTACKSIZE] __page_aligned_bss; unsigned long __supported_pte_mask __read_mostly = ~0UL; EXPORT_SYMBOL_GPL(__supported_pte_mask); diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index 416ea415f5c2..0561fde56a51 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -394,8 +394,7 @@ static int __init early_ioremap_debug_setup(char *str) early_param("early_ioremap_debug", early_ioremap_debug_setup); static __initdata int after_paging_init; -static pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] - __section(.bss.page_aligned); +static pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] __page_aligned_bss; static inline pmd_t * __init early_ioremap_pmd(unsigned long addr) { diff --git a/include/linux/linkage.h b/include/linux/linkage.h index 2119610b24f8..9fd1f859021b 100644 --- a/include/linux/linkage.h +++ b/include/linux/linkage.h @@ -1,6 +1,7 @@ #ifndef _LINUX_LINKAGE_H #define _LINUX_LINKAGE_H +#include #include #ifdef __cplusplus @@ -17,6 +18,9 @@ # define asmregparm #endif +#define __page_aligned_data __section(.data.page_aligned) __aligned(PAGE_SIZE) +#define __page_aligned_bss __section(.bss.page_aligned) __aligned(PAGE_SIZE) + /* * This is used by architectures to keep arguments on the stack * untouched by the compiler by keeping them live until the end. -- cgit v1.2.3 From d52d53b8a5b258bfaab9223a5e7284fcfdd48577 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Mon, 16 Jun 2008 20:10:55 -0700 Subject: RFC x86: try to remove arch_get_ram_range want to remove arch_get_ram_range, and use early_node_map instead. Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/mm/init_32.c | 6 ++++-- drivers/pci/intel-iommu.c | 51 ++++++++++++++++++++++++++++++++++------------- include/linux/mm.h | 2 +- mm/page_alloc.c | 10 +++++++--- 4 files changed, 49 insertions(+), 20 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index 65d55056b6e7..a0484adbf59d 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -298,7 +298,7 @@ struct add_highpages_data { unsigned long end_pfn; }; -static void __init add_highpages_work_fn(unsigned long start_pfn, +static int __init add_highpages_work_fn(unsigned long start_pfn, unsigned long end_pfn, void *datax) { int node_pfn; @@ -311,7 +311,7 @@ static void __init add_highpages_work_fn(unsigned long start_pfn, final_start_pfn = max(start_pfn, data->start_pfn); final_end_pfn = min(end_pfn, data->end_pfn); if (final_start_pfn >= final_end_pfn) - return; + return 0; for (node_pfn = final_start_pfn; node_pfn < final_end_pfn; node_pfn++) { @@ -321,6 +321,8 @@ static void __init add_highpages_work_fn(unsigned long start_pfn, add_one_highpage_init(page, node_pfn); } + return 0; + } void __init add_highpages_with_active_regions(int nid, unsigned long start_pfn, diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index 66c0fd21894b..bb0642318a95 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -1637,12 +1637,43 @@ static inline int iommu_prepare_rmrr_dev(struct dmar_rmrr_unit *rmrr, } #ifdef CONFIG_DMAR_GFX_WA -extern int arch_get_ram_range(int slot, u64 *addr, u64 *size); +struct iommu_prepare_data { + struct pci_dev *pdev; + int ret; +}; + +static int __init iommu_prepare_work_fn(unsigned long start_pfn, + unsigned long end_pfn, void *datax) +{ + struct iommu_prepare_data *data; + + data = (struct iommu_prepare_data *)datax; + + data->ret = iommu_prepare_identity_map(data->pdev, + start_pfn<ret; + +} + +static int __init iommu_prepare_with_active_regions(struct pci_dev *pdev) +{ + int nid; + struct iommu_prepare_data data; + + data.pdev = pdev; + data.ret = 0; + + for_each_online_node(nid) { + work_with_active_regions(nid, iommu_prepare_work_fn, &data); + if (data.ret) + return data.ret; + } + return data.ret; +} + static void __init iommu_prepare_gfx_mapping(void) { struct pci_dev *pdev = NULL; - u64 base, size; - int slot; int ret; for_each_pci_dev(pdev) { @@ -1651,17 +1682,9 @@ static void __init iommu_prepare_gfx_mapping(void) continue; printk(KERN_INFO "IOMMU: gfx device %s 1-1 mapping\n", pci_name(pdev)); - slot = arch_get_ram_range(0, &base, &size); - while (slot >= 0) { - ret = iommu_prepare_identity_map(pdev, - base, base + size); - if (ret) - goto error; - slot = arch_get_ram_range(slot, &base, &size); - } - continue; -error: - printk(KERN_ERR "IOMMU: mapping reserved region failed\n"); + ret = iommu_prepare_with_active_regions(pdev); + if (ret) + printk(KERN_ERR "IOMMU: mapping reserved region failed\n"); } } #endif diff --git a/include/linux/mm.h b/include/linux/mm.h index 3d647b24041f..cf1cd3a2ed78 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1011,7 +1011,7 @@ extern unsigned long find_min_pfn_with_active_regions(void); extern unsigned long find_max_pfn_with_active_regions(void); extern void free_bootmem_with_active_regions(int nid, unsigned long max_low_pfn); -typedef void (*work_fn_t)(unsigned long, unsigned long, void *); +typedef int (*work_fn_t)(unsigned long, unsigned long, void *); extern void work_with_active_regions(int nid, work_fn_t work_fn, void *data); extern void sparse_memory_present_with_active_regions(int nid); #ifndef CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 41c6e3aa059f..e25b6b24f844 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -2932,10 +2932,14 @@ void __init free_bootmem_with_active_regions(int nid, void __init work_with_active_regions(int nid, work_fn_t work_fn, void *data) { int i; + int ret; - for_each_active_range_index_in_nid(i, nid) - work_fn(early_node_map[i].start_pfn, early_node_map[i].end_pfn, - data); + for_each_active_range_index_in_nid(i, nid) { + ret = work_fn(early_node_map[i].start_pfn, + early_node_map[i].end_pfn, data); + if (ret) + break; + } } /** * sparse_memory_present_with_active_regions - Call memory_present for each active range -- cgit v1.2.3 From 3c999f142665265afd0fe9190204dd051f17e505 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Fri, 20 Jun 2008 16:11:20 -0700 Subject: x86: check command line when CONFIG_X86_MPPARSE is not set, v2 if acpi=off, acpi=noirq and pci=noacpi, we need to disable apic. Signed-off-by: Yinghai Lu Cc: Andrew Morton Cc: "Maciej W. Rozycki" Signed-off-by: Ingo Molnar --- arch/x86/kernel/acpi/boot.c | 14 ++++++++++++++ arch/x86/kernel/apic_32.c | 2 +- arch/x86/kernel/setup.c | 4 ++++ arch/x86/kernel/setup_32.c | 5 +++++ arch/x86/kernel/setup_64.c | 9 +++++++++ include/asm-x86/apic.h | 6 +++++- include/linux/acpi.h | 6 ++++++ 7 files changed, 44 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 6516359922ba..5c0107602b62 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -1787,6 +1787,20 @@ static int __init parse_pci(char *arg) } early_param("pci", parse_pci); +int __init acpi_mps_check(void) +{ +#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_X86_MPPARSE) +/* mptable code is not built-in*/ + if (acpi_disabled || acpi_noirq) { + printk(KERN_WARNING "MPS support code is not built-in.\n" + "Using acpi=off or acpi=noirq or pci=noacpi " + "may have problem\n"); + return 1; + } +#endif + return 0; +} + #ifdef CONFIG_X86_IO_APIC static int __init parse_acpi_skip_timer_override(char *arg) { diff --git a/arch/x86/kernel/apic_32.c b/arch/x86/kernel/apic_32.c index dd8de26b2786..4932d7813bcd 100644 --- a/arch/x86/kernel/apic_32.c +++ b/arch/x86/kernel/apic_32.c @@ -57,7 +57,7 @@ unsigned long mp_lapic_addr; * * -1=force-disable, +1=force-enable */ -static int enable_local_apic __initdata; +int enable_local_apic; /* Local APIC timer verification ok */ static int local_apic_timer_verify_ok; diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index c5330f601b68..56aee55cf8dc 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -161,6 +161,10 @@ void __init setup_per_cpu_areas(void) char *ptr; int cpu; + /* no processor from mptable or madt */ + if (!num_processors) + num_processors = 1; + #ifdef CONFIG_HOTPLUG_CPU prefill_possible_map(); #else diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c index 369d0fe1ff9c..cad4e893df05 100644 --- a/arch/x86/kernel/setup_32.c +++ b/arch/x86/kernel/setup_32.c @@ -677,6 +677,11 @@ void __init setup_arch(char **cmdline_p) parse_early_param(); + if (acpi_mps_check()){ + enable_local_apic = -1; + clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC); + } + finish_e820_parsing(); probe_roms(); diff --git a/arch/x86/kernel/setup_64.c b/arch/x86/kernel/setup_64.c index a93300de4da9..175c696ec536 100644 --- a/arch/x86/kernel/setup_64.c +++ b/arch/x86/kernel/setup_64.c @@ -302,6 +302,11 @@ void __init setup_arch(char **cmdline_p) parse_early_param(); + if (acpi_mps_check()) { + disable_apic = 1; + clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC); + } + #ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT if (init_ohci1394_dma_early) init_ohci1394_dma_on_all_controllers(); @@ -723,6 +728,10 @@ static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c) cpu_devs[c->x86_vendor]->c_early_init(c); validate_pat_support(c); + + /* early_param could clear that, but recall get it set again */ + if (disable_apic) + clear_cpu_cap(c, X86_FEATURE_APIC); } /* diff --git a/include/asm-x86/apic.h b/include/asm-x86/apic.h index 313bcaf4b6c3..9fe941cd843d 100644 --- a/include/asm-x86/apic.h +++ b/include/asm-x86/apic.h @@ -39,8 +39,12 @@ extern int apic_verbosity; extern int local_apic_timer_c2_ok; extern int ioapic_force; -extern int disable_apic; +#ifdef CONFIG_X86_64 +extern int disable_apic; +#else +extern int enable_local_apic; +#endif /* * Basic functions accessing APICs. */ diff --git a/include/linux/acpi.h b/include/linux/acpi.h index 41f7ce7edd7a..0601075d09a1 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -82,6 +82,7 @@ char * __acpi_map_table (unsigned long phys_addr, unsigned long size); int early_acpi_boot_init(void); int acpi_boot_init (void); int acpi_boot_table_init (void); +int acpi_mps_check (void); int acpi_numa_init (void); int acpi_table_init (void); @@ -250,6 +251,11 @@ static inline int acpi_boot_table_init(void) return 0; } +static inline int acpi_mps_check(void) +{ + return 0; +} + static inline int acpi_check_resource_conflict(struct resource *res) { return 0; -- cgit v1.2.3 From 69ac9cd629ca96e59f34eb4ccd12d00b2c8276a7 Mon Sep 17 00:00:00 2001 From: Bernhard Walle Date: Fri, 27 Jun 2008 13:12:54 +0200 Subject: sysfs: add /sys/firmware/memmap This patch adds /sys/firmware/memmap interface that represents the BIOS (or Firmware) provided memory map. The tree looks like: /sys/firmware/memmap/0/start (hex number) end (hex number) type (string) ... /1/start end type With the following shell snippet one can print the memory map in the same form the kernel prints itself when booting on x86 (the E820 map). --------- 8< -------------------------- #!/bin/sh cd /sys/firmware/memmap for dir in * ; do start=$(cat $dir/start) end=$(cat $dir/end) type=$(cat $dir/type) printf "%016x-%016x (%s)\n" $start $[ $end +1] "$type" done --------- >8 -------------------------- That patch only provides the needed interface: 1. The sysfs interface. 2. The structure and enumeration definition. 3. The function firmware_map_add() and firmware_map_add_early() that should be called from architecture code (E820/EFI, for example) to add the contents to the interface. If the kernel is compiled without CONFIG_FIRMWARE_MEMMAP, the interface does nothing without cluttering the architecture-specific code with #ifdef's. The purpose of the new interface is kexec: While /proc/iomem represents the *used* memory map (e.g. modified via kernel parameters like 'memmap' and 'mem'), the /sys/firmware/memmap tree represents the unmodified memory map provided via the firmware. So kexec can: - use the original memory map for rebooting, - use the /proc/iomem for setting up the ELF core headers for kdump case that should only represent the memory of the system. The patch has been tested on i386 and x86_64. Signed-off-by: Bernhard Walle Acked-by: Greg KH Acked-by: Vivek Goyal Cc: kexec@lists.infradead.org Cc: yhlu.kernel@gmail.com Signed-off-by: Ingo Molnar --- Documentation/ABI/testing/sysfs-firmware-memmap | 71 ++++++++ drivers/firmware/Kconfig | 10 ++ drivers/firmware/Makefile | 1 + drivers/firmware/memmap.c | 205 ++++++++++++++++++++++++ include/linux/firmware-map.h | 74 +++++++++ 5 files changed, 361 insertions(+) create mode 100644 Documentation/ABI/testing/sysfs-firmware-memmap create mode 100644 drivers/firmware/memmap.c create mode 100644 include/linux/firmware-map.h (limited to 'include/linux') diff --git a/Documentation/ABI/testing/sysfs-firmware-memmap b/Documentation/ABI/testing/sysfs-firmware-memmap new file mode 100644 index 000000000000..0d99ee6ae02e --- /dev/null +++ b/Documentation/ABI/testing/sysfs-firmware-memmap @@ -0,0 +1,71 @@ +What: /sys/firmware/memmap/ +Date: June 2008 +Contact: Bernhard Walle +Description: + On all platforms, the firmware provides a memory map which the + kernel reads. The resources from that memory map are registered + in the kernel resource tree and exposed to userspace via + /proc/iomem (together with other resources). + + However, on most architectures that firmware-provided memory + map is modified afterwards by the kernel itself, either because + the kernel merges that memory map with other information or + just because the user overwrites that memory map via command + line. + + kexec needs the raw firmware-provided memory map to setup the + parameter segment of the kernel that should be booted with + kexec. Also, the raw memory map is useful for debugging. For + that reason, /sys/firmware/memmap is an interface that provides + the raw memory map to userspace. + + The structure is as follows: Under /sys/firmware/memmap there + are subdirectories with the number of the entry as their name: + + /sys/firmware/memmap/0 + /sys/firmware/memmap/1 + /sys/firmware/memmap/2 + /sys/firmware/memmap/3 + ... + + The maximum depends on the number of memory map entries provided + by the firmware. The order is just the order that the firmware + provides. + + Each directory contains three files: + + start : The start address (as hexadecimal number with the + '0x' prefix). + end : The end address, inclusive (regardless whether the + firmware provides inclusive or exclusive ranges). + type : Type of the entry as string. See below for a list of + valid types. + + So, for example: + + /sys/firmware/memmap/0/start + /sys/firmware/memmap/0/end + /sys/firmware/memmap/0/type + /sys/firmware/memmap/1/start + ... + + Currently following types exist: + + - System RAM + - ACPI Tables + - ACPI Non-volatile Storage + - reserved + + Following shell snippet can be used to display that memory + map in a human-readable format: + + -------------------- 8< ---------------------------------------- + #!/bin/bash + cd /sys/firmware/memmap + for dir in * ; do + start=$(cat $dir/start) + end=$(cat $dir/end) + type=$(cat $dir/type) + printf "%016x-%016x (%s)\n" $start $[ $end +1] "$type" + done + -------------------- >8 ---------------------------------------- diff --git a/drivers/firmware/Kconfig b/drivers/firmware/Kconfig index dc2cec6127d1..ebb9e51deb0c 100644 --- a/drivers/firmware/Kconfig +++ b/drivers/firmware/Kconfig @@ -26,6 +26,16 @@ config EDD_OFF kernel. Say N if you want EDD enabled by default. EDD can be dynamically set using the kernel parameter 'edd={on|skipmbr|off}'. +config FIRMWARE_MEMMAP + bool "Add firmware-provided memory map to sysfs" if EMBEDDED + default (X86_64 || X86_32) + help + Add the firmware-provided (unmodified) memory map to /sys/firmware/memmap. + That memory map is used for example by kexec to set up parameter area + for the next kernel, but can also be used for debugging purposes. + + See also Documentation/ABI/testing/sysfs-firmware-memmap. + config EFI_VARS tristate "EFI Variable Support via sysfs" depends on EFI diff --git a/drivers/firmware/Makefile b/drivers/firmware/Makefile index 4c9147154df8..1c3c17343dbe 100644 --- a/drivers/firmware/Makefile +++ b/drivers/firmware/Makefile @@ -10,3 +10,4 @@ obj-$(CONFIG_DCDBAS) += dcdbas.o obj-$(CONFIG_DMIID) += dmi-id.o obj-$(CONFIG_ISCSI_IBFT_FIND) += iscsi_ibft_find.o obj-$(CONFIG_ISCSI_IBFT) += iscsi_ibft.o +obj-$(CONFIG_FIRMWARE_MEMMAP) += memmap.o diff --git a/drivers/firmware/memmap.c b/drivers/firmware/memmap.c new file mode 100644 index 000000000000..e23399c7f773 --- /dev/null +++ b/drivers/firmware/memmap.c @@ -0,0 +1,205 @@ +/* + * linux/drivers/firmware/memmap.c + * Copyright (C) 2008 SUSE LINUX Products GmbH + * by Bernhard Walle + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License v2.0 as published by + * the Free Software Foundation + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#include +#include +#include +#include +#include +#include + +/* + * Data types ------------------------------------------------------------------ + */ + +/* + * Firmware map entry. Because firmware memory maps are flat and not + * hierarchical, it's ok to organise them in a linked list. No parent + * information is necessary as for the resource tree. + */ +struct firmware_map_entry { + resource_size_t start; /* start of the memory range */ + resource_size_t end; /* end of the memory range (incl.) */ + const char *type; /* type of the memory range */ + struct list_head list; /* entry for the linked list */ + struct kobject kobj; /* kobject for each entry */ +}; + +/* + * Forward declarations -------------------------------------------------------- + */ +static ssize_t memmap_attr_show(struct kobject *kobj, + struct attribute *attr, char *buf); +static ssize_t start_show(struct firmware_map_entry *entry, char *buf); +static ssize_t end_show(struct firmware_map_entry *entry, char *buf); +static ssize_t type_show(struct firmware_map_entry *entry, char *buf); + +/* + * Static data ----------------------------------------------------------------- + */ + +struct memmap_attribute { + struct attribute attr; + ssize_t (*show)(struct firmware_map_entry *entry, char *buf); +}; + +struct memmap_attribute memmap_start_attr = __ATTR_RO(start); +struct memmap_attribute memmap_end_attr = __ATTR_RO(end); +struct memmap_attribute memmap_type_attr = __ATTR_RO(type); + +/* + * These are default attributes that are added for every memmap entry. + */ +static struct attribute *def_attrs[] = { + &memmap_start_attr.attr, + &memmap_end_attr.attr, + &memmap_type_attr.attr, + NULL +}; + +static struct sysfs_ops memmap_attr_ops = { + .show = memmap_attr_show, +}; + +static struct kobj_type memmap_ktype = { + .sysfs_ops = &memmap_attr_ops, + .default_attrs = def_attrs, +}; + +/* + * Registration functions ------------------------------------------------------ + */ + +/* + * Firmware memory map entries + */ +static LIST_HEAD(map_entries); + +/** + * Common implementation of firmware_map_add() and firmware_map_add_early() + * which expects a pre-allocated struct firmware_map_entry. + * + * @start: Start of the memory range. + * @end: End of the memory range (inclusive). + * @type: Type of the memory range. + * @entry: Pre-allocated (either kmalloc() or bootmem allocator), uninitialised + * entry. + */ +static int firmware_map_add_entry(resource_size_t start, resource_size_t end, + const char *type, + struct firmware_map_entry *entry) +{ + BUG_ON(start > end); + + entry->start = start; + entry->end = end; + entry->type = type; + INIT_LIST_HEAD(&entry->list); + kobject_init(&entry->kobj, &memmap_ktype); + + list_add_tail(&entry->list, &map_entries); + + return 0; +} + +/* + * See for documentation. + */ +int firmware_map_add(resource_size_t start, resource_size_t end, + const char *type) +{ + struct firmware_map_entry *entry; + + entry = kmalloc(sizeof(struct firmware_map_entry), GFP_ATOMIC); + WARN_ON(!entry); + if (!entry) + return -ENOMEM; + + return firmware_map_add_entry(start, end, type, entry); +} + +/* + * See for documentation. + */ +int __init firmware_map_add_early(resource_size_t start, resource_size_t end, + const char *type) +{ + struct firmware_map_entry *entry; + + entry = alloc_bootmem_low(sizeof(struct firmware_map_entry)); + WARN_ON(!entry); + if (!entry) + return -ENOMEM; + + return firmware_map_add_entry(start, end, type, entry); +} + +/* + * Sysfs functions ------------------------------------------------------------- + */ + +static ssize_t start_show(struct firmware_map_entry *entry, char *buf) +{ + return snprintf(buf, PAGE_SIZE, "0x%llx\n", entry->start); +} + +static ssize_t end_show(struct firmware_map_entry *entry, char *buf) +{ + return snprintf(buf, PAGE_SIZE, "0x%llx\n", entry->end); +} + +static ssize_t type_show(struct firmware_map_entry *entry, char *buf) +{ + return snprintf(buf, PAGE_SIZE, "%s\n", entry->type); +} + +#define to_memmap_attr(_attr) container_of(_attr, struct memmap_attribute, attr) +#define to_memmap_entry(obj) container_of(obj, struct firmware_map_entry, kobj) + +static ssize_t memmap_attr_show(struct kobject *kobj, + struct attribute *attr, char *buf) +{ + struct firmware_map_entry *entry = to_memmap_entry(kobj); + struct memmap_attribute *memmap_attr = to_memmap_attr(attr); + + return memmap_attr->show(entry, buf); +} + +/* + * Initialises stuff and adds the entries in the map_entries list to + * sysfs. Important is that firmware_map_add() and firmware_map_add_early() + * must be called before late_initcall. + */ +static int __init memmap_init(void) +{ + int i = 0; + struct firmware_map_entry *entry; + struct kset *memmap_kset; + + memmap_kset = kset_create_and_add("memmap", NULL, firmware_kobj); + WARN_ON(!memmap_kset); + if (!memmap_kset) + return -ENOMEM; + + list_for_each_entry(entry, &map_entries, list) { + entry->kobj.kset = memmap_kset; + kobject_add(&entry->kobj, NULL, "%d", i++); + } + + return 0; +} +late_initcall(memmap_init); + diff --git a/include/linux/firmware-map.h b/include/linux/firmware-map.h new file mode 100644 index 000000000000..acbdbcc16051 --- /dev/null +++ b/include/linux/firmware-map.h @@ -0,0 +1,74 @@ +/* + * include/linux/firmware-map.h: + * Copyright (C) 2008 SUSE LINUX Products GmbH + * by Bernhard Walle + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License v2.0 as published by + * the Free Software Foundation + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ +#ifndef _LINUX_FIRMWARE_MAP_H +#define _LINUX_FIRMWARE_MAP_H + +#include +#include + +/* + * provide a dummy interface if CONFIG_FIRMWARE_MEMMAP is disabled + */ +#ifdef CONFIG_FIRMWARE_MEMMAP + +/** + * Adds a firmware mapping entry. This function uses kmalloc() for memory + * allocation. Use firmware_map_add_early() if you want to use the bootmem + * allocator. + * + * That function must be called before late_initcall. + * + * @start: Start of the memory range. + * @end: End of the memory range (inclusive). + * @type: Type of the memory range. + * + * Returns 0 on success, or -ENOMEM if no memory could be allocated. + */ +int firmware_map_add(resource_size_t start, resource_size_t end, + const char *type); + +/** + * Adds a firmware mapping entry. This function uses the bootmem allocator + * for memory allocation. Use firmware_map_add() if you want to use kmalloc(). + * + * That function must be called before late_initcall. + * + * @start: Start of the memory range. + * @end: End of the memory range (inclusive). + * @type: Type of the memory range. + * + * Returns 0 on success, or -ENOMEM if no memory could be allocated. + */ +int firmware_map_add_early(resource_size_t start, resource_size_t end, + const char *type); + +#else /* CONFIG_FIRMWARE_MEMMAP */ + +static inline int firmware_map_add(resource_size_t start, resource_size_t end, + const char *type) +{ + return 0; +} + +static inline int firmware_map_add_early(resource_size_t start, + resource_size_t end, const char *type) +{ + return 0; +} + +#endif /* CONFIG_FIRMWARE_MEMMAP */ + +#endif /* _LINUX_FIRMWARE_MAP_H */ -- cgit v1.2.3