diff options
Diffstat (limited to 'arch/s390/boot/startup.c')
| -rw-r--r-- | arch/s390/boot/startup.c | 234 |
1 files changed, 179 insertions, 55 deletions
diff --git a/arch/s390/boot/startup.c b/arch/s390/boot/startup.c index 3b3a11f95269..47ca3264c023 100644 --- a/arch/s390/boot/startup.c +++ b/arch/s390/boot/startup.c @@ -1,55 +1,37 @@ // SPDX-License-Identifier: GPL-2.0 #include <linux/string.h> #include <linux/elf.h> +#include <asm/boot_data.h> #include <asm/sections.h> +#include <asm/cpu_mf.h> #include <asm/setup.h> +#include <asm/kasan.h> #include <asm/kexec.h> #include <asm/sclp.h> #include <asm/diag.h> #include <asm/uv.h> -#include "compressed/decompressor.h" +#include <asm/abs_lowcore.h> +#include "decompressor.h" #include "boot.h" +#include "uv.h" -extern char __boot_data_start[], __boot_data_end[]; -extern char __boot_data_preserved_start[], __boot_data_preserved_end[]; unsigned long __bootdata_preserved(__kaslr_offset); +unsigned long __bootdata_preserved(__abs_lowcore); +unsigned long __bootdata_preserved(__memcpy_real_area); +unsigned long __bootdata(__amode31_base); +unsigned long __bootdata_preserved(VMALLOC_START); +unsigned long __bootdata_preserved(VMALLOC_END); +struct page *__bootdata_preserved(vmemmap); +unsigned long __bootdata_preserved(vmemmap_size); +unsigned long __bootdata_preserved(MODULES_VADDR); +unsigned long __bootdata_preserved(MODULES_END); +unsigned long __bootdata(ident_map_size); +int __bootdata(is_full_image) = 1; +struct initrd_data __bootdata(initrd_data); -/* - * Some code and data needs to stay below 2 GB, even when the kernel would be - * relocated above 2 GB, because it has to use 31 bit addresses. - * Such code and data is part of the .dma section, and its location is passed - * over to the decompressed / relocated kernel via the .boot.preserved.data - * section. - */ -extern char _sdma[], _edma[]; -extern char _stext_dma[], _etext_dma[]; -extern struct exception_table_entry _start_dma_ex_table[]; -extern struct exception_table_entry _stop_dma_ex_table[]; -unsigned long __bootdata_preserved(__sdma) = __pa(&_sdma); -unsigned long __bootdata_preserved(__edma) = __pa(&_edma); -unsigned long __bootdata_preserved(__stext_dma) = __pa(&_stext_dma); -unsigned long __bootdata_preserved(__etext_dma) = __pa(&_etext_dma); -struct exception_table_entry * - __bootdata_preserved(__start_dma_ex_table) = _start_dma_ex_table; -struct exception_table_entry * - __bootdata_preserved(__stop_dma_ex_table) = _stop_dma_ex_table; - -int _diag210_dma(struct diag210 *addr); -int _diag26c_dma(void *req, void *resp, enum diag26c_sc subcode); -int _diag14_dma(unsigned long rx, unsigned long ry1, unsigned long subcode); -void _diag0c_dma(struct hypfs_diag0c_entry *entry); -void _diag308_reset_dma(void); -struct diag_ops __bootdata_preserved(diag_dma_ops) = { - .diag210 = _diag210_dma, - .diag26c = _diag26c_dma, - .diag14 = _diag14_dma, - .diag0c = _diag0c_dma, - .diag308_reset = _diag308_reset_dma -}; -static struct diag210 _diag210_tmp_dma __section(.dma.data); -struct diag210 *__bootdata_preserved(__diag210_tmp_dma) = &_diag210_tmp_dma; -void _swsusp_reset_dma(void); -unsigned long __bootdata_preserved(__swsusp_reset_dma) = __pa(_swsusp_reset_dma); +u64 __bootdata_preserved(stfle_fac_list[16]); +u64 __bootdata_preserved(alt_stfle_fac_list[16]); +struct oldmem_data __bootdata_preserved(oldmem_data); void error(char *x) { @@ -60,6 +42,14 @@ void error(char *x) disabled_wait(); } +static void setup_lpp(void) +{ + S390_lowcore.current_pid = 0; + S390_lowcore.lpp = LPP_MAGIC; + if (test_facility(40)) + lpp(&S390_lowcore.lpp); +} + #ifdef CONFIG_KERNEL_UNCOMPRESSED unsigned long mem_safe_offset(void) { @@ -71,12 +61,12 @@ static void rescue_initrd(unsigned long addr) { if (!IS_ENABLED(CONFIG_BLK_DEV_INITRD)) return; - if (!INITRD_START || !INITRD_SIZE) + if (!initrd_data.start || !initrd_data.size) return; - if (addr <= INITRD_START) + if (addr <= initrd_data.start) return; - memmove((void *)addr, (void *)INITRD_START, INITRD_SIZE); - INITRD_START = addr; + memmove((void *)addr, (void *)initrd_data.start, initrd_data.size); + initrd_data.start = addr; } static void copy_bootdata(void) @@ -120,41 +110,176 @@ static void handle_relocs(unsigned long offset) } } +/* + * Merge information from several sources into a single ident_map_size value. + * "ident_map_size" represents the upper limit of physical memory we may ever + * reach. It might not be all online memory, but also include standby (offline) + * memory. "ident_map_size" could be lower then actual standby or even online + * memory present, due to limiting factors. We should never go above this limit. + * It is the size of our identity mapping. + * + * Consider the following factors: + * 1. max_physmem_end - end of physical memory online or standby. + * Always <= end of the last online memory block (get_mem_detect_end()). + * 2. CONFIG_MAX_PHYSMEM_BITS - the maximum size of physical memory the + * kernel is able to support. + * 3. "mem=" kernel command line option which limits physical memory usage. + * 4. OLDMEM_BASE which is a kdump memory limit when the kernel is executed as + * crash kernel. + * 5. "hsa" size which is a memory limit when the kernel is executed during + * zfcp/nvme dump. + */ +static void setup_ident_map_size(unsigned long max_physmem_end) +{ + unsigned long hsa_size; + + ident_map_size = max_physmem_end; + if (memory_limit) + ident_map_size = min(ident_map_size, memory_limit); + ident_map_size = min(ident_map_size, 1UL << MAX_PHYSMEM_BITS); + +#ifdef CONFIG_CRASH_DUMP + if (oldmem_data.start) { + kaslr_enabled = 0; + ident_map_size = min(ident_map_size, oldmem_data.size); + } else if (ipl_block_valid && is_ipl_block_dump()) { + kaslr_enabled = 0; + if (!sclp_early_get_hsa_size(&hsa_size) && hsa_size) + ident_map_size = min(ident_map_size, hsa_size); + } +#endif +} + +static void setup_kernel_memory_layout(void) +{ + unsigned long vmemmap_start; + unsigned long rte_size; + unsigned long pages; + unsigned long vmax; + + pages = ident_map_size / PAGE_SIZE; + /* vmemmap contains a multiple of PAGES_PER_SECTION struct pages */ + vmemmap_size = SECTION_ALIGN_UP(pages) * sizeof(struct page); + + /* choose kernel address space layout: 4 or 3 levels. */ + vmemmap_start = round_up(ident_map_size, _REGION3_SIZE); + if (IS_ENABLED(CONFIG_KASAN) || + vmalloc_size > _REGION2_SIZE || + vmemmap_start + vmemmap_size + vmalloc_size + MODULES_LEN > + _REGION2_SIZE) { + vmax = _REGION1_SIZE; + rte_size = _REGION2_SIZE; + } else { + vmax = _REGION2_SIZE; + rte_size = _REGION3_SIZE; + } + /* + * forcing modules and vmalloc area under the ultravisor + * secure storage limit, so that any vmalloc allocation + * we do could be used to back secure guest storage. + */ + vmax = adjust_to_uv_max(vmax); +#ifdef CONFIG_KASAN + /* force vmalloc and modules below kasan shadow */ + vmax = min(vmax, KASAN_SHADOW_START); +#endif + __memcpy_real_area = round_down(vmax - PAGE_SIZE, PAGE_SIZE); + __abs_lowcore = round_down(__memcpy_real_area - ABS_LOWCORE_MAP_SIZE, + sizeof(struct lowcore)); + MODULES_END = round_down(__abs_lowcore, _SEGMENT_SIZE); + MODULES_VADDR = MODULES_END - MODULES_LEN; + VMALLOC_END = MODULES_VADDR; + + /* allow vmalloc area to occupy up to about 1/2 of the rest virtual space left */ + vmalloc_size = min(vmalloc_size, round_down(VMALLOC_END / 2, _REGION3_SIZE)); + VMALLOC_START = VMALLOC_END - vmalloc_size; + + /* split remaining virtual space between 1:1 mapping & vmemmap array */ + pages = VMALLOC_START / (PAGE_SIZE + sizeof(struct page)); + pages = SECTION_ALIGN_UP(pages); + /* keep vmemmap_start aligned to a top level region table entry */ + vmemmap_start = round_down(VMALLOC_START - pages * sizeof(struct page), rte_size); + /* vmemmap_start is the future VMEM_MAX_PHYS, make sure it is within MAX_PHYSMEM */ + vmemmap_start = min(vmemmap_start, 1UL << MAX_PHYSMEM_BITS); + /* make sure identity map doesn't overlay with vmemmap */ + ident_map_size = min(ident_map_size, vmemmap_start); + vmemmap_size = SECTION_ALIGN_UP(ident_map_size / PAGE_SIZE) * sizeof(struct page); + /* make sure vmemmap doesn't overlay with vmalloc area */ + VMALLOC_START = max(vmemmap_start + vmemmap_size, VMALLOC_START); + vmemmap = (struct page *)vmemmap_start; +} + +/* + * This function clears the BSS section of the decompressed Linux kernel and NOT the decompressor's. + */ static void clear_bss_section(void) { memset((void *)vmlinux.default_lma + vmlinux.image_size, 0, vmlinux.bss_size); } +/* + * Set vmalloc area size to an 8th of (potential) physical memory + * size, unless size has been set by kernel command line parameter. + */ +static void setup_vmalloc_size(void) +{ + unsigned long size; + + if (vmalloc_size_set) + return; + size = round_up(ident_map_size / 8, _SEGMENT_SIZE); + vmalloc_size = max(size, vmalloc_size); +} + +static void offset_vmlinux_info(unsigned long offset) +{ + vmlinux.default_lma += offset; + *(unsigned long *)(&vmlinux.entry) += offset; + vmlinux.bootdata_off += offset; + vmlinux.bootdata_preserved_off += offset; + vmlinux.rela_dyn_start += offset; + vmlinux.rela_dyn_end += offset; + vmlinux.dynsym_start += offset; +} + +static unsigned long reserve_amode31(unsigned long safe_addr) +{ + __amode31_base = PAGE_ALIGN(safe_addr); + return safe_addr + vmlinux.amode31_size; +} + void startup_kernel(void) { unsigned long random_lma; unsigned long safe_addr; void *img; + initrd_data.start = parmarea.initrd_start; + initrd_data.size = parmarea.initrd_size; + oldmem_data.start = parmarea.oldmem_base; + oldmem_data.size = parmarea.oldmem_size; + + setup_lpp(); store_ipl_parmblock(); safe_addr = mem_safe_offset(); + safe_addr = reserve_amode31(safe_addr); safe_addr = read_ipl_report(safe_addr); uv_query_info(); rescue_initrd(safe_addr); sclp_early_read_info(); setup_boot_command_line(); parse_boot_command_line(); - setup_memory_end(); - detect_memory(); + sanitize_prot_virt_host(); + setup_ident_map_size(detect_memory()); + setup_vmalloc_size(); + setup_kernel_memory_layout(); - random_lma = __kaslr_offset = 0; if (IS_ENABLED(CONFIG_RANDOMIZE_BASE) && kaslr_enabled) { random_lma = get_random_base(safe_addr); if (random_lma) { __kaslr_offset = random_lma - vmlinux.default_lma; img = (void *)vmlinux.default_lma; - vmlinux.default_lma += __kaslr_offset; - vmlinux.entry += __kaslr_offset; - vmlinux.bootdata_off += __kaslr_offset; - vmlinux.bootdata_preserved_off += __kaslr_offset; - vmlinux.rela_dyn_start += __kaslr_offset; - vmlinux.rela_dyn_end += __kaslr_offset; - vmlinux.dynsym_start += __kaslr_offset; + offset_vmlinux_info(__kaslr_offset); } } @@ -166,8 +291,7 @@ void startup_kernel(void) clear_bss_section(); copy_bootdata(); - if (IS_ENABLED(CONFIG_RELOCATABLE)) - handle_relocs(__kaslr_offset); + handle_relocs(__kaslr_offset); if (__kaslr_offset) { /* |