From b17ed48040d9e8b6ae35bc492015bf0fe1c8bae4 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 16 Oct 2010 10:19:54 +0200 Subject: x86: Move efi to platform Signed-off-by: Thomas Gleixner Cc: Huang Ying --- arch/x86/platform/efi/efi.c | 613 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 613 insertions(+) create mode 100644 arch/x86/platform/efi/efi.c (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c new file mode 100644 index 000000000000..0fe27d7c6258 --- /dev/null +++ b/arch/x86/platform/efi/efi.c @@ -0,0 +1,613 @@ +/* + * Common EFI (Extensible Firmware Interface) support functions + * Based on Extensible Firmware Interface Specification version 1.0 + * + * Copyright (C) 1999 VA Linux Systems + * Copyright (C) 1999 Walt Drummond + * Copyright (C) 1999-2002 Hewlett-Packard Co. + * David Mosberger-Tang + * Stephane Eranian + * Copyright (C) 2005-2008 Intel Co. + * Fenghua Yu + * Bibo Mao + * Chandramouli Narayanan + * Huang Ying + * + * Copied from efi_32.c to eliminate the duplicated code between EFI + * 32/64 support code. --ying 2007-10-26 + * + * All EFI Runtime Services are not implemented yet as EFI only + * supports physical mode addressing on SoftSDV. This is to be fixed + * in a future version. --drummond 1999-07-20 + * + * Implemented EFI runtime services and virtual mode calls. --davidm + * + * Goutham Rao: + * Skip non-WB memory and ignore empty memory ranges. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#define EFI_DEBUG 1 +#define PFX "EFI: " + +int efi_enabled; +EXPORT_SYMBOL(efi_enabled); + +struct efi efi; +EXPORT_SYMBOL(efi); + +struct efi_memory_map memmap; + +static struct efi efi_phys __initdata; +static efi_system_table_t efi_systab __initdata; + +static int __init setup_noefi(char *arg) +{ + efi_enabled = 0; + return 0; +} +early_param("noefi", setup_noefi); + +int add_efi_memmap; +EXPORT_SYMBOL(add_efi_memmap); + +static int __init setup_add_efi_memmap(char *arg) +{ + add_efi_memmap = 1; + return 0; +} +early_param("add_efi_memmap", setup_add_efi_memmap); + + +static efi_status_t virt_efi_get_time(efi_time_t *tm, efi_time_cap_t *tc) +{ + return efi_call_virt2(get_time, tm, tc); +} + +static efi_status_t virt_efi_set_time(efi_time_t *tm) +{ + return efi_call_virt1(set_time, tm); +} + +static efi_status_t virt_efi_get_wakeup_time(efi_bool_t *enabled, + efi_bool_t *pending, + efi_time_t *tm) +{ + return efi_call_virt3(get_wakeup_time, + enabled, pending, tm); +} + +static efi_status_t virt_efi_set_wakeup_time(efi_bool_t enabled, efi_time_t *tm) +{ + return efi_call_virt2(set_wakeup_time, + enabled, tm); +} + +static efi_status_t virt_efi_get_variable(efi_char16_t *name, + efi_guid_t *vendor, + u32 *attr, + unsigned long *data_size, + void *data) +{ + return efi_call_virt5(get_variable, + name, vendor, attr, + data_size, data); +} + +static efi_status_t virt_efi_get_next_variable(unsigned long *name_size, + efi_char16_t *name, + efi_guid_t *vendor) +{ + return efi_call_virt3(get_next_variable, + name_size, name, vendor); +} + +static efi_status_t virt_efi_set_variable(efi_char16_t *name, + efi_guid_t *vendor, + unsigned long attr, + unsigned long data_size, + void *data) +{ + return efi_call_virt5(set_variable, + name, vendor, attr, + data_size, data); +} + +static efi_status_t virt_efi_get_next_high_mono_count(u32 *count) +{ + return efi_call_virt1(get_next_high_mono_count, count); +} + +static void virt_efi_reset_system(int reset_type, + efi_status_t status, + unsigned long data_size, + efi_char16_t *data) +{ + efi_call_virt4(reset_system, reset_type, status, + data_size, data); +} + +static efi_status_t virt_efi_set_virtual_address_map( + unsigned long memory_map_size, + unsigned long descriptor_size, + u32 descriptor_version, + efi_memory_desc_t *virtual_map) +{ + return efi_call_virt4(set_virtual_address_map, + memory_map_size, descriptor_size, + descriptor_version, virtual_map); +} + +static efi_status_t __init phys_efi_set_virtual_address_map( + unsigned long memory_map_size, + unsigned long descriptor_size, + u32 descriptor_version, + efi_memory_desc_t *virtual_map) +{ + efi_status_t status; + + efi_call_phys_prelog(); + status = efi_call_phys4(efi_phys.set_virtual_address_map, + memory_map_size, descriptor_size, + descriptor_version, virtual_map); + efi_call_phys_epilog(); + return status; +} + +static efi_status_t __init phys_efi_get_time(efi_time_t *tm, + efi_time_cap_t *tc) +{ + efi_status_t status; + + efi_call_phys_prelog(); + status = efi_call_phys2(efi_phys.get_time, tm, tc); + efi_call_phys_epilog(); + return status; +} + +int efi_set_rtc_mmss(unsigned long nowtime) +{ + int real_seconds, real_minutes; + efi_status_t status; + efi_time_t eft; + efi_time_cap_t cap; + + status = efi.get_time(&eft, &cap); + if (status != EFI_SUCCESS) { + printk(KERN_ERR "Oops: efitime: can't read time!\n"); + return -1; + } + + real_seconds = nowtime % 60; + real_minutes = nowtime / 60; + if (((abs(real_minutes - eft.minute) + 15)/30) & 1) + real_minutes += 30; + real_minutes %= 60; + eft.minute = real_minutes; + eft.second = real_seconds; + + status = efi.set_time(&eft); + if (status != EFI_SUCCESS) { + printk(KERN_ERR "Oops: efitime: can't write time!\n"); + return -1; + } + return 0; +} + +unsigned long efi_get_time(void) +{ + efi_status_t status; + efi_time_t eft; + efi_time_cap_t cap; + + status = efi.get_time(&eft, &cap); + if (status != EFI_SUCCESS) + printk(KERN_ERR "Oops: efitime: can't read time!\n"); + + return mktime(eft.year, eft.month, eft.day, eft.hour, + eft.minute, eft.second); +} + +/* + * Tell the kernel about the EFI memory map. This might include + * more than the max 128 entries that can fit in the e820 legacy + * (zeropage) memory map. + */ + +static void __init do_add_efi_memmap(void) +{ + void *p; + + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + efi_memory_desc_t *md = p; + unsigned long long start = md->phys_addr; + unsigned long long size = md->num_pages << EFI_PAGE_SHIFT; + int e820_type; + + switch (md->type) { + case EFI_LOADER_CODE: + case EFI_LOADER_DATA: + case EFI_BOOT_SERVICES_CODE: + case EFI_BOOT_SERVICES_DATA: + case EFI_CONVENTIONAL_MEMORY: + if (md->attribute & EFI_MEMORY_WB) + e820_type = E820_RAM; + else + e820_type = E820_RESERVED; + break; + case EFI_ACPI_RECLAIM_MEMORY: + e820_type = E820_ACPI; + break; + case EFI_ACPI_MEMORY_NVS: + e820_type = E820_NVS; + break; + case EFI_UNUSABLE_MEMORY: + e820_type = E820_UNUSABLE; + break; + default: + /* + * EFI_RESERVED_TYPE EFI_RUNTIME_SERVICES_CODE + * EFI_RUNTIME_SERVICES_DATA EFI_MEMORY_MAPPED_IO + * EFI_MEMORY_MAPPED_IO_PORT_SPACE EFI_PAL_CODE + */ + e820_type = E820_RESERVED; + break; + } + e820_add_region(start, size, e820_type); + } + sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); +} + +void __init efi_memblock_x86_reserve_range(void) +{ + unsigned long pmap; + +#ifdef CONFIG_X86_32 + pmap = boot_params.efi_info.efi_memmap; +#else + pmap = (boot_params.efi_info.efi_memmap | + ((__u64)boot_params.efi_info.efi_memmap_hi<<32)); +#endif + memmap.phys_map = (void *)pmap; + memmap.nr_map = boot_params.efi_info.efi_memmap_size / + boot_params.efi_info.efi_memdesc_size; + memmap.desc_version = boot_params.efi_info.efi_memdesc_version; + memmap.desc_size = boot_params.efi_info.efi_memdesc_size; + memblock_x86_reserve_range(pmap, pmap + memmap.nr_map * memmap.desc_size, + "EFI memmap"); +} + +#if EFI_DEBUG +static void __init print_efi_memmap(void) +{ + efi_memory_desc_t *md; + void *p; + int i; + + for (p = memmap.map, i = 0; + p < memmap.map_end; + p += memmap.desc_size, i++) { + md = p; + printk(KERN_INFO PFX "mem%02u: type=%u, attr=0x%llx, " + "range=[0x%016llx-0x%016llx) (%lluMB)\n", + i, md->type, md->attribute, md->phys_addr, + md->phys_addr + (md->num_pages << EFI_PAGE_SHIFT), + (md->num_pages >> (20 - EFI_PAGE_SHIFT))); + } +} +#endif /* EFI_DEBUG */ + +void __init efi_init(void) +{ + efi_config_table_t *config_tables; + efi_runtime_services_t *runtime; + efi_char16_t *c16; + char vendor[100] = "unknown"; + int i = 0; + void *tmp; + +#ifdef CONFIG_X86_32 + efi_phys.systab = (efi_system_table_t *)boot_params.efi_info.efi_systab; +#else + efi_phys.systab = (efi_system_table_t *) + (boot_params.efi_info.efi_systab | + ((__u64)boot_params.efi_info.efi_systab_hi<<32)); +#endif + + efi.systab = early_ioremap((unsigned long)efi_phys.systab, + sizeof(efi_system_table_t)); + if (efi.systab == NULL) + printk(KERN_ERR "Couldn't map the EFI system table!\n"); + memcpy(&efi_systab, efi.systab, sizeof(efi_system_table_t)); + early_iounmap(efi.systab, sizeof(efi_system_table_t)); + efi.systab = &efi_systab; + + /* + * Verify the EFI Table + */ + if (efi.systab->hdr.signature != EFI_SYSTEM_TABLE_SIGNATURE) + printk(KERN_ERR "EFI system table signature incorrect!\n"); + if ((efi.systab->hdr.revision >> 16) == 0) + printk(KERN_ERR "Warning: EFI system table version " + "%d.%02d, expected 1.00 or greater!\n", + efi.systab->hdr.revision >> 16, + efi.systab->hdr.revision & 0xffff); + + /* + * Show what we know for posterity + */ + c16 = tmp = early_ioremap(efi.systab->fw_vendor, 2); + if (c16) { + for (i = 0; i < sizeof(vendor) - 1 && *c16; ++i) + vendor[i] = *c16++; + vendor[i] = '\0'; + } else + printk(KERN_ERR PFX "Could not map the firmware vendor!\n"); + early_iounmap(tmp, 2); + + printk(KERN_INFO "EFI v%u.%.02u by %s\n", + efi.systab->hdr.revision >> 16, + efi.systab->hdr.revision & 0xffff, vendor); + + /* + * Let's see what config tables the firmware passed to us. + */ + config_tables = early_ioremap( + efi.systab->tables, + efi.systab->nr_tables * sizeof(efi_config_table_t)); + if (config_tables == NULL) + printk(KERN_ERR "Could not map EFI Configuration Table!\n"); + + printk(KERN_INFO); + for (i = 0; i < efi.systab->nr_tables; i++) { + if (!efi_guidcmp(config_tables[i].guid, MPS_TABLE_GUID)) { + efi.mps = config_tables[i].table; + printk(" MPS=0x%lx ", config_tables[i].table); + } else if (!efi_guidcmp(config_tables[i].guid, + ACPI_20_TABLE_GUID)) { + efi.acpi20 = config_tables[i].table; + printk(" ACPI 2.0=0x%lx ", config_tables[i].table); + } else if (!efi_guidcmp(config_tables[i].guid, + ACPI_TABLE_GUID)) { + efi.acpi = config_tables[i].table; + printk(" ACPI=0x%lx ", config_tables[i].table); + } else if (!efi_guidcmp(config_tables[i].guid, + SMBIOS_TABLE_GUID)) { + efi.smbios = config_tables[i].table; + printk(" SMBIOS=0x%lx ", config_tables[i].table); +#ifdef CONFIG_X86_UV + } else if (!efi_guidcmp(config_tables[i].guid, + UV_SYSTEM_TABLE_GUID)) { + efi.uv_systab = config_tables[i].table; + printk(" UVsystab=0x%lx ", config_tables[i].table); +#endif + } else if (!efi_guidcmp(config_tables[i].guid, + HCDP_TABLE_GUID)) { + efi.hcdp = config_tables[i].table; + printk(" HCDP=0x%lx ", config_tables[i].table); + } else if (!efi_guidcmp(config_tables[i].guid, + UGA_IO_PROTOCOL_GUID)) { + efi.uga = config_tables[i].table; + printk(" UGA=0x%lx ", config_tables[i].table); + } + } + printk("\n"); + early_iounmap(config_tables, + efi.systab->nr_tables * sizeof(efi_config_table_t)); + + /* + * Check out the runtime services table. We need to map + * the runtime services table so that we can grab the physical + * address of several of the EFI runtime functions, needed to + * set the firmware into virtual mode. + */ + runtime = early_ioremap((unsigned long)efi.systab->runtime, + sizeof(efi_runtime_services_t)); + if (runtime != NULL) { + /* + * We will only need *early* access to the following + * two EFI runtime services before set_virtual_address_map + * is invoked. + */ + efi_phys.get_time = (efi_get_time_t *)runtime->get_time; + efi_phys.set_virtual_address_map = + (efi_set_virtual_address_map_t *) + runtime->set_virtual_address_map; + /* + * Make efi_get_time can be called before entering + * virtual mode. + */ + efi.get_time = phys_efi_get_time; + } else + printk(KERN_ERR "Could not map the EFI runtime service " + "table!\n"); + early_iounmap(runtime, sizeof(efi_runtime_services_t)); + + /* Map the EFI memory map */ + memmap.map = early_ioremap((unsigned long)memmap.phys_map, + memmap.nr_map * memmap.desc_size); + if (memmap.map == NULL) + printk(KERN_ERR "Could not map the EFI memory map!\n"); + memmap.map_end = memmap.map + (memmap.nr_map * memmap.desc_size); + + if (memmap.desc_size != sizeof(efi_memory_desc_t)) + printk(KERN_WARNING + "Kernel-defined memdesc doesn't match the one from EFI!\n"); + + if (add_efi_memmap) + do_add_efi_memmap(); + +#ifdef CONFIG_X86_32 + x86_platform.get_wallclock = efi_get_time; + x86_platform.set_wallclock = efi_set_rtc_mmss; +#endif + + /* Setup for EFI runtime service */ + reboot_type = BOOT_EFI; + +#if EFI_DEBUG + print_efi_memmap(); +#endif +} + +static void __init runtime_code_page_mkexec(void) +{ + efi_memory_desc_t *md; + void *p; + u64 addr, npages; + + /* Make EFI runtime service code area executable */ + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + md = p; + + if (md->type != EFI_RUNTIME_SERVICES_CODE) + continue; + + addr = md->virt_addr; + npages = md->num_pages; + memrange_efi_to_native(&addr, &npages); + set_memory_x(addr, npages); + } +} + +/* + * This function will switch the EFI runtime services to virtual mode. + * Essentially, look through the EFI memmap and map every region that + * has the runtime attribute bit set in its memory descriptor and update + * that memory descriptor with the virtual address obtained from ioremap(). + * This enables the runtime services to be called without having to + * thunk back into physical mode for every invocation. + */ +void __init efi_enter_virtual_mode(void) +{ + efi_memory_desc_t *md; + efi_status_t status; + unsigned long size; + u64 end, systab, addr, npages, end_pfn; + void *p, *va; + + efi.systab = NULL; + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + md = p; + if (!(md->attribute & EFI_MEMORY_RUNTIME)) + continue; + + size = md->num_pages << EFI_PAGE_SHIFT; + end = md->phys_addr + size; + + end_pfn = PFN_UP(end); + if (end_pfn <= max_low_pfn_mapped + || (end_pfn > (1UL << (32 - PAGE_SHIFT)) + && end_pfn <= max_pfn_mapped)) + va = __va(md->phys_addr); + else + va = efi_ioremap(md->phys_addr, size, md->type); + + md->virt_addr = (u64) (unsigned long) va; + + if (!va) { + printk(KERN_ERR PFX "ioremap of 0x%llX failed!\n", + (unsigned long long)md->phys_addr); + continue; + } + + if (!(md->attribute & EFI_MEMORY_WB)) { + addr = md->virt_addr; + npages = md->num_pages; + memrange_efi_to_native(&addr, &npages); + set_memory_uc(addr, npages); + } + + systab = (u64) (unsigned long) efi_phys.systab; + if (md->phys_addr <= systab && systab < end) { + systab += md->virt_addr - md->phys_addr; + efi.systab = (efi_system_table_t *) (unsigned long) systab; + } + } + + BUG_ON(!efi.systab); + + status = phys_efi_set_virtual_address_map( + memmap.desc_size * memmap.nr_map, + memmap.desc_size, + memmap.desc_version, + memmap.phys_map); + + if (status != EFI_SUCCESS) { + printk(KERN_ALERT "Unable to switch EFI into virtual mode " + "(status=%lx)!\n", status); + panic("EFI call to SetVirtualAddressMap() failed!"); + } + + /* + * Now that EFI is in virtual mode, update the function + * pointers in the runtime service table to the new virtual addresses. + * + * Call EFI services through wrapper functions. + */ + efi.get_time = virt_efi_get_time; + efi.set_time = virt_efi_set_time; + efi.get_wakeup_time = virt_efi_get_wakeup_time; + efi.set_wakeup_time = virt_efi_set_wakeup_time; + efi.get_variable = virt_efi_get_variable; + efi.get_next_variable = virt_efi_get_next_variable; + efi.set_variable = virt_efi_set_variable; + efi.get_next_high_mono_count = virt_efi_get_next_high_mono_count; + efi.reset_system = virt_efi_reset_system; + efi.set_virtual_address_map = virt_efi_set_virtual_address_map; + if (__supported_pte_mask & _PAGE_NX) + runtime_code_page_mkexec(); + early_iounmap(memmap.map, memmap.nr_map * memmap.desc_size); + memmap.map = NULL; +} + +/* + * Convenience functions to obtain memory types and attributes + */ +u32 efi_mem_type(unsigned long phys_addr) +{ + efi_memory_desc_t *md; + void *p; + + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + md = p; + if ((md->phys_addr <= phys_addr) && + (phys_addr < (md->phys_addr + + (md->num_pages << EFI_PAGE_SHIFT)))) + return md->type; + } + return 0; +} + +u64 efi_mem_attributes(unsigned long phys_addr) +{ + efi_memory_desc_t *md; + void *p; + + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + md = p; + if ((md->phys_addr <= phys_addr) && + (phys_addr < (md->phys_addr + + (md->num_pages << EFI_PAGE_SHIFT)))) + return md->attribute; + } + return 0; +} -- cgit v1.2.2 From 2b5e8ef35bc89eee944c0627edacbb1fea5a1b84 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Thu, 5 May 2011 15:19:42 -0400 Subject: x86, efi: Remove virtual-mode SetVirtualAddressMap call The spec says that SetVirtualAddressMap doesn't work once you're in virtual mode, so there's no point in having infrastructure for calling it from there. Signed-off-by: Matthew Garrett Link: http://lkml.kernel.org/r/1304623186-18261-1-git-send-email-mjg@redhat.com Signed-off-by: H. Peter Anvin --- arch/x86/platform/efi/efi.c | 13 +------------ 1 file changed, 1 insertion(+), 12 deletions(-) (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index 0fe27d7c6258..f2e4fe9e92a1 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -145,17 +145,6 @@ static void virt_efi_reset_system(int reset_type, data_size, data); } -static efi_status_t virt_efi_set_virtual_address_map( - unsigned long memory_map_size, - unsigned long descriptor_size, - u32 descriptor_version, - efi_memory_desc_t *virtual_map) -{ - return efi_call_virt4(set_virtual_address_map, - memory_map_size, descriptor_size, - descriptor_version, virtual_map); -} - static efi_status_t __init phys_efi_set_virtual_address_map( unsigned long memory_map_size, unsigned long descriptor_size, @@ -572,7 +561,7 @@ void __init efi_enter_virtual_mode(void) efi.set_variable = virt_efi_set_variable; efi.get_next_high_mono_count = virt_efi_get_next_high_mono_count; efi.reset_system = virt_efi_reset_system; - efi.set_virtual_address_map = virt_efi_set_virtual_address_map; + efi.set_virtual_address_map = NULL; if (__supported_pte_mask & _PAGE_NX) runtime_code_page_mkexec(); early_iounmap(memmap.map, memmap.nr_map * memmap.desc_size); -- cgit v1.2.2 From 9cd2b07c197e3ff594fc04f5fb3d86efbeab6ad8 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Thu, 5 May 2011 15:19:43 -0400 Subject: x86, efi: Consolidate EFI nx control The core EFI code and 64-bit EFI code currently have independent implementations of code for setting memory regions as executable or not. Let's consolidate them. Signed-off-by: Matthew Garrett Link: http://lkml.kernel.org/r/1304623186-18261-2-git-send-email-mjg@redhat.com Signed-off-by: H. Peter Anvin --- arch/x86/platform/efi/efi.c | 21 ++++++++++++++++----- 1 file changed, 16 insertions(+), 5 deletions(-) (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index f2e4fe9e92a1..7daae27e975e 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -457,11 +457,25 @@ void __init efi_init(void) #endif } +void __init efi_set_executable(efi_memory_desc_t *md, bool executable) +{ + u64 addr, npages; + + addr = md->virt_addr; + npages = md->num_pages; + + memrange_efi_to_native(&addr, &npages); + + if (executable) + set_memory_x(addr, npages); + else + set_memory_nx(addr, npages); +} + static void __init runtime_code_page_mkexec(void) { efi_memory_desc_t *md; void *p; - u64 addr, npages; /* Make EFI runtime service code area executable */ for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { @@ -470,10 +484,7 @@ static void __init runtime_code_page_mkexec(void) if (md->type != EFI_RUNTIME_SERVICES_CODE) continue; - addr = md->virt_addr; - npages = md->num_pages; - memrange_efi_to_native(&addr, &npages); - set_memory_x(addr, npages); + efi_set_executable(md, true); } } -- cgit v1.2.2 From 202f9d0a41809e3424af5f61489b48b622824aed Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Thu, 5 May 2011 15:19:44 -0400 Subject: x86, efi: Merge contiguous memory regions of the same type and attribute Some firmware implementations assume that physically contiguous regions will be contiguous in virtual address space. This assumption is, obviously, entirely unjustifiable. Said firmware implementations lack the good grace to handle their failings in a measured and reasonable manner, instead tending to shit all over address space and oopsing the kernel. In an ideal universe these firmware implementations would simultaneously catch fire and cease to be a problem, but since some of them are present in attractively thin and shiny metal devices vanity wins out and some poor developer spends an extended period of time surrounded by a growing array of empty bottles until the underlying reason becomes apparent. Said developer presents this patch, which simply merges adjacent regions if they happen to be contiguous and have the same EFI memory type and caching attributes. Signed-off-by: Matthew Garrett Link: http://lkml.kernel.org/r/1304623186-18261-3-git-send-email-mjg@redhat.com Signed-off-by: H. Peter Anvin --- arch/x86/platform/efi/efi.c | 30 +++++++++++++++++++++++++++++- 1 file changed, 29 insertions(+), 1 deletion(-) (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index 7daae27e975e..a46a73ecc5f3 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -498,13 +498,41 @@ static void __init runtime_code_page_mkexec(void) */ void __init efi_enter_virtual_mode(void) { - efi_memory_desc_t *md; + efi_memory_desc_t *md, *prev_md = NULL; efi_status_t status; unsigned long size; u64 end, systab, addr, npages, end_pfn; void *p, *va; efi.systab = NULL; + + /* Merge contiguous regions of the same type and attribute */ + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + u64 prev_size; + md = p; + + if (!prev_md) { + prev_md = md; + continue; + } + + if (prev_md->type != md->type || + prev_md->attribute != md->attribute) { + prev_md = md; + continue; + } + + prev_size = prev_md->num_pages << EFI_PAGE_SHIFT; + + if (md->phys_addr == (prev_md->phys_addr + prev_size)) { + prev_md->num_pages += md->num_pages; + md->type = EFI_RESERVED_TYPE; + md->attribute = 0; + continue; + } + prev_md = md; + } + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { md = p; if (!(md->attribute & EFI_MEMORY_RUNTIME)) -- cgit v1.2.2 From 7cb00b72876ea2451eb79d468da0e8fb9134aa8a Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Thu, 5 May 2011 15:19:45 -0400 Subject: x86, efi: Pass a minimal map to SetVirtualAddressMap() Experimentation with various EFI implementations has shown that functions outside runtime services will still update their pointers if SetVirtualAddressMap() is called with memory descriptors outside the runtime area. This is obviously insane, and therefore is unsurprising. Evidence from instrumenting another EFI implementation suggests that it only passes the set of descriptors covering runtime regions, so let's avoid any problems by doing the same. Runtime descriptors are copied to a separate memory map, and only that map is passed back to the firmware. Signed-off-by: Matthew Garrett Link: http://lkml.kernel.org/r/1304623186-18261-4-git-send-email-mjg@redhat.com Signed-off-by: H. Peter Anvin --- arch/x86/platform/efi/efi.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index a46a73ecc5f3..b30aa26a8df2 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -502,7 +502,8 @@ void __init efi_enter_virtual_mode(void) efi_status_t status; unsigned long size; u64 end, systab, addr, npages, end_pfn; - void *p, *va; + void *p, *va, *new_memmap = NULL; + int count = 0; efi.systab = NULL; @@ -569,15 +570,21 @@ void __init efi_enter_virtual_mode(void) systab += md->virt_addr - md->phys_addr; efi.systab = (efi_system_table_t *) (unsigned long) systab; } + new_memmap = krealloc(new_memmap, + (count + 1) * memmap.desc_size, + GFP_KERNEL); + memcpy(new_memmap + (count * memmap.desc_size), md, + memmap.desc_size); + count++; } BUG_ON(!efi.systab); status = phys_efi_set_virtual_address_map( - memmap.desc_size * memmap.nr_map, + memmap.desc_size * count, memmap.desc_size, memmap.desc_version, - memmap.phys_map); + (efi_memory_desc_t *)__pa(new_memmap)); if (status != EFI_SUCCESS) { printk(KERN_ALERT "Unable to switch EFI into virtual mode " @@ -605,6 +612,7 @@ void __init efi_enter_virtual_mode(void) runtime_code_page_mkexec(); early_iounmap(memmap.map, memmap.nr_map * memmap.desc_size); memmap.map = NULL; + kfree(new_memmap); } /* -- cgit v1.2.2 From 916f676f8dc016103f983c7ec54c18ecdbb6e349 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Wed, 25 May 2011 09:53:13 -0400 Subject: x86, efi: Retain boot service code until after switching to virtual mode UEFI stands for "Unified Extensible Firmware Interface", where "Firmware" is an ancient African word meaning "Why do something right when you can do it so wrong that children will weep and brave adults will cower before you", and "UEI" is Celtic for "We missed DOS so we burned it into your ROMs". The UEFI specification provides for runtime services (ie, another way for the operating system to be forced to depend on the firmware) and we rely on these for certain trivial tasks such as setting up the bootloader. But some hardware fails to work if we attempt to use these runtime services from physical mode, and so we have to switch into virtual mode. So far so dreadful. The specification makes it clear that the operating system is free to do whatever it wants with boot services code after ExitBootServices() has been called. SetVirtualAddressMap() can't be called until ExitBootServices() has been. So, obviously, a whole bunch of EFI implementations call into boot services code when we do that. Since we've been charmingly naive and trusted that the specification may be somehow relevant to the real world, we've already stuffed a picture of a penguin or something in that address space. And just to make things more entertaining, we've also marked it non-executable. This patch allocates the boot services regions during EFI init and makes sure that they're executable. Then, after SetVirtualAddressMap(), it discards them and everyone lives happily ever after. Except for the ones who have to work on EFI, who live sad lives haunted by the knowledge that someone's eventually going to write yet another firmware specification. [ hpa: adding this to urgent with a stable tag since it fixes currently-broken hardware. However, I do not know what the dependencies are and so I do not know which -stable versions this may be a candidate for. ] Signed-off-by: Matthew Garrett Link: http://lkml.kernel.org/r/1306331593-28715-1-git-send-email-mjg@redhat.com Signed-off-by: H. Peter Anvin Cc: Tony Luck Cc: --- arch/x86/platform/efi/efi.c | 45 ++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 44 insertions(+), 1 deletion(-) (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index b30aa26a8df2..0d3a4fa34560 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -304,6 +304,40 @@ static void __init print_efi_memmap(void) } #endif /* EFI_DEBUG */ +void __init efi_reserve_boot_services(void) +{ + void *p; + + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + efi_memory_desc_t *md = p; + unsigned long long start = md->phys_addr; + unsigned long long size = md->num_pages << EFI_PAGE_SHIFT; + + if (md->type != EFI_BOOT_SERVICES_CODE && + md->type != EFI_BOOT_SERVICES_DATA) + continue; + + memblock_x86_reserve_range(start, start + size, "EFI Boot"); + } +} + +static void __init efi_free_boot_services(void) +{ + void *p; + + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + efi_memory_desc_t *md = p; + unsigned long long start = md->phys_addr; + unsigned long long size = md->num_pages << EFI_PAGE_SHIFT; + + if (md->type != EFI_BOOT_SERVICES_CODE && + md->type != EFI_BOOT_SERVICES_DATA) + continue; + + free_bootmem_late(start, size); + } +} + void __init efi_init(void) { efi_config_table_t *config_tables; @@ -536,7 +570,9 @@ void __init efi_enter_virtual_mode(void) for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { md = p; - if (!(md->attribute & EFI_MEMORY_RUNTIME)) + if (!(md->attribute & EFI_MEMORY_RUNTIME) && + md->type != EFI_BOOT_SERVICES_CODE && + md->type != EFI_BOOT_SERVICES_DATA) continue; size = md->num_pages << EFI_PAGE_SHIFT; @@ -592,6 +628,13 @@ void __init efi_enter_virtual_mode(void) panic("EFI call to SetVirtualAddressMap() failed!"); } + /* + * Thankfully, it does seem that no runtime services other than + * SetVirtualAddressMap() will touch boot services code, so we can + * get rid of it all at this point + */ + efi_free_boot_services(); + /* * Now that EFI is in virtual mode, update the function * pointers in the runtime service table to the new virtual addresses. -- cgit v1.2.2 From 7d68dc3f1003a38948c55c803c32d1989dd49198 Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Tue, 14 Jun 2011 19:53:09 +0200 Subject: x86, efi: Do not reserve boot services regions within reserved areas Commit 916f676f8dc started reserving boot service code since some systems require you to keep that code around until SetVirtualAddressMap is called. However, in some cases those areas will overlap with reserved regions. The proper medium-term fix is to fix the bootloader to prevent the conflicts from occurring by moving the kernel to a better position, but the kernel should check for this possibility, and only reserve regions which can be reserved. Signed-off-by: Maarten Lankhorst Link: http://lkml.kernel.org/r/4DF7A005.1050407@gmail.com Acked-by: Matthew Garrett Signed-off-by: H. Peter Anvin Signed-off-by: Ingo Molnar --- arch/x86/platform/efi/efi.c | 29 +++++++++++++++++++++++++---- 1 file changed, 25 insertions(+), 4 deletions(-) (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index 0d3a4fa34560..474356b98ede 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -310,14 +310,31 @@ void __init efi_reserve_boot_services(void) for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { efi_memory_desc_t *md = p; - unsigned long long start = md->phys_addr; - unsigned long long size = md->num_pages << EFI_PAGE_SHIFT; + u64 start = md->phys_addr; + u64 size = md->num_pages << EFI_PAGE_SHIFT; if (md->type != EFI_BOOT_SERVICES_CODE && md->type != EFI_BOOT_SERVICES_DATA) continue; - - memblock_x86_reserve_range(start, start + size, "EFI Boot"); + /* Only reserve where possible: + * - Not within any already allocated areas + * - Not over any memory area (really needed, if above?) + * - Not within any part of the kernel + * - Not the bios reserved area + */ + if ((start+size >= virt_to_phys(_text) + && start <= virt_to_phys(_end)) || + !e820_all_mapped(start, start+size, E820_RAM) || + memblock_x86_check_reserved_size(&start, &size, + 1<num_pages = 0; + memblock_dbg(PFX "Could not reserve boot range " + "[0x%010llx-0x%010llx]\n", + start, start+size-1); + } else + memblock_x86_reserve_range(start, start+size, + "EFI Boot"); } } @@ -334,6 +351,10 @@ static void __init efi_free_boot_services(void) md->type != EFI_BOOT_SERVICES_DATA) continue; + /* Could not reserve boot area */ + if (!size) + continue; + free_bootmem_late(start, size); } } -- cgit v1.2.2 From f70e957cda22d309c769805cbb932407a5232219 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Wed, 6 Jul 2011 16:52:37 -0400 Subject: x86: Don't use the EFI reboot method by default Testing suggests that at least some Lenovos and some Intels will fail to reboot via EFI, attempting to jump to an unmapped physical address. In the long run we could handle this by providing a page table with a 1:1 mapping of physical addresses, but for now it's probably just easier to assume that ACPI or legacy methods will be present and reboot via those. Signed-off-by: Matthew Garrett Cc: Linus Torvalds Cc: Andrew Morton Cc: Alan Cox Link: http://lkml.kernel.org/r/1309985557-15350-1-git-send-email-mjg@redhat.com Signed-off-by: Ingo Molnar --- arch/x86/platform/efi/efi.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'arch/x86/platform/efi/efi.c') diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index 474356b98ede..899e393d8e73 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -504,9 +504,6 @@ void __init efi_init(void) x86_platform.set_wallclock = efi_set_rtc_mmss; #endif - /* Setup for EFI runtime service */ - reboot_type = BOOT_EFI; - #if EFI_DEBUG print_efi_memmap(); #endif -- cgit v1.2.2