diff options
Diffstat (limited to 'arch/riscv/mm')
-rw-r--r-- | arch/riscv/mm/init.c | 243 |
1 files changed, 112 insertions, 131 deletions
diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c index 269fc648ef3d..c0cddf0fc22d 100644 --- a/arch/riscv/mm/init.c +++ b/arch/riscv/mm/init.c @@ -19,6 +19,7 @@ #include <linux/set_memory.h> #include <linux/dma-map-ops.h> #include <linux/crash_dump.h> +#include <linux/hugetlb.h> #include <asm/fixmap.h> #include <asm/tlbflush.h> @@ -36,6 +37,9 @@ EXPORT_SYMBOL(kernel_map); #define kernel_map (*(struct kernel_mapping *)XIP_FIXUP(&kernel_map)) #endif +phys_addr_t phys_ram_base __ro_after_init; +EXPORT_SYMBOL(phys_ram_base); + #ifdef CONFIG_XIP_KERNEL extern char _xiprom[], _exiprom[]; #endif @@ -127,10 +131,17 @@ void __init mem_init(void) } /* - * The default maximal physical memory size is -PAGE_OFFSET, - * limit the memory size via mem. + * The default maximal physical memory size is -PAGE_OFFSET for 32-bit kernel, + * whereas for 64-bit kernel, the end of the virtual address space is occupied + * by the modules/BPF/kernel mappings which reduces the available size of the + * linear mapping. + * Limit the memory size via mem. */ +#ifdef CONFIG_64BIT +static phys_addr_t memory_limit = -PAGE_OFFSET - SZ_4G; +#else static phys_addr_t memory_limit = -PAGE_OFFSET; +#endif static int __init early_mem(char *p) { @@ -152,8 +163,8 @@ static void __init setup_bootmem(void) { phys_addr_t vmlinux_end = __pa_symbol(&_end); phys_addr_t vmlinux_start = __pa_symbol(&_start); - phys_addr_t max_mapped_addr = __pa(~(ulong)0); - phys_addr_t dram_end; + phys_addr_t __maybe_unused max_mapped_addr; + phys_addr_t phys_ram_end; #ifdef CONFIG_XIP_KERNEL vmlinux_start = __pa_symbol(&_sdata); @@ -174,18 +185,28 @@ static void __init setup_bootmem(void) #endif memblock_reserve(vmlinux_start, vmlinux_end - vmlinux_start); - dram_end = memblock_end_of_DRAM(); + + phys_ram_end = memblock_end_of_DRAM(); +#ifndef CONFIG_64BIT +#ifndef CONFIG_XIP_KERNEL + phys_ram_base = memblock_start_of_DRAM(); +#endif /* * memblock allocator is not aware of the fact that last 4K bytes of * the addressable memory can not be mapped because of IS_ERR_VALUE * macro. Make sure that last 4k bytes are not usable by memblock - * if end of dram is equal to maximum addressable memory. + * if end of dram is equal to maximum addressable memory. For 64-bit + * kernel, this problem can't happen here as the end of the virtual + * address space is occupied by the kernel mapping then this check must + * be done as soon as the kernel mapping base address is determined. */ - if (max_mapped_addr == (dram_end - 1)) + max_mapped_addr = __pa(~(ulong)0); + if (max_mapped_addr == (phys_ram_end - 1)) memblock_set_current_limit(max_mapped_addr - 4096); +#endif - min_low_pfn = PFN_UP(memblock_start_of_DRAM()); - max_low_pfn = max_pfn = PFN_DOWN(dram_end); + min_low_pfn = PFN_UP(phys_ram_base); + max_low_pfn = max_pfn = PFN_DOWN(phys_ram_end); dma32_phys_limit = min(4UL * SZ_1G, (unsigned long)PFN_PHYS(max_low_pfn)); set_max_mapnr(max_low_pfn - ARCH_PFN_OFFSET); @@ -202,6 +223,8 @@ static void __init setup_bootmem(void) early_init_fdt_scan_reserved_mem(); dma_contiguous_reserve(dma32_phys_limit); + if (IS_ENABLED(CONFIG_64BIT)) + hugetlb_cma_reserve(PUD_SHIFT - PAGE_SHIFT); memblock_allow_resize(); } @@ -214,14 +237,15 @@ static struct pt_alloc_ops _pt_ops __initdata; #define pt_ops _pt_ops #endif -unsigned long pfn_base __ro_after_init; -EXPORT_SYMBOL(pfn_base); +unsigned long riscv_pfn_base __ro_after_init; +EXPORT_SYMBOL(riscv_pfn_base); pgd_t swapper_pg_dir[PTRS_PER_PGD] __page_aligned_bss; pgd_t trampoline_pg_dir[PTRS_PER_PGD] __page_aligned_bss; static pte_t fixmap_pte[PTRS_PER_PTE] __page_aligned_bss; pgd_t early_pg_dir[PTRS_PER_PGD] __initdata __aligned(PAGE_SIZE); +static pmd_t __maybe_unused early_dtb_pmd[PTRS_PER_PMD] __initdata __aligned(PAGE_SIZE); #ifdef CONFIG_XIP_KERNEL #define trampoline_pg_dir ((pgd_t *)XIP_FIXUP(trampoline_pg_dir)) @@ -302,7 +326,6 @@ static void __init create_pte_mapping(pte_t *ptep, static pmd_t trampoline_pmd[PTRS_PER_PMD] __page_aligned_bss; static pmd_t fixmap_pmd[PTRS_PER_PMD] __page_aligned_bss; static pmd_t early_pmd[PTRS_PER_PMD] __initdata __aligned(PAGE_SIZE); -static pmd_t early_dtb_pmd[PTRS_PER_PMD] __initdata __aligned(PAGE_SIZE); #ifdef CONFIG_XIP_KERNEL #define trampoline_pmd ((pmd_t *)XIP_FIXUP(trampoline_pmd)) @@ -388,6 +411,7 @@ static void __init create_pmd_mapping(pmd_t *pmdp, #define create_pgd_next_mapping(__nextp, __va, __pa, __sz, __prot) \ create_pte_mapping(__nextp, __va, __pa, __sz, __prot) #define fixmap_pgd_next fixmap_pte +#define create_pmd_mapping(__pmdp, __va, __pa, __sz, __prot) #endif void __init create_pgd_mapping(pgd_t *pgdp, @@ -495,55 +519,87 @@ static __init pgprot_t pgprot_from_va(uintptr_t va) #endif #ifdef CONFIG_XIP_KERNEL -static void __init create_kernel_page_table(pgd_t *pgdir, uintptr_t map_size, +static void __init create_kernel_page_table(pgd_t *pgdir, __always_unused bool early) { uintptr_t va, end_va; /* Map the flash resident part */ end_va = kernel_map.virt_addr + kernel_map.xiprom_sz; - for (va = kernel_map.virt_addr; va < end_va; va += map_size) + for (va = kernel_map.virt_addr; va < end_va; va += PMD_SIZE) create_pgd_mapping(pgdir, va, kernel_map.xiprom + (va - kernel_map.virt_addr), - map_size, PAGE_KERNEL_EXEC); + PMD_SIZE, PAGE_KERNEL_EXEC); /* Map the data in RAM */ end_va = kernel_map.virt_addr + XIP_OFFSET + kernel_map.size; - for (va = kernel_map.virt_addr + XIP_OFFSET; va < end_va; va += map_size) + for (va = kernel_map.virt_addr + XIP_OFFSET; va < end_va; va += PMD_SIZE) create_pgd_mapping(pgdir, va, kernel_map.phys_addr + (va - (kernel_map.virt_addr + XIP_OFFSET)), - map_size, PAGE_KERNEL); + PMD_SIZE, PAGE_KERNEL); } #else -static void __init create_kernel_page_table(pgd_t *pgdir, uintptr_t map_size, - bool early) +static void __init create_kernel_page_table(pgd_t *pgdir, bool early) { uintptr_t va, end_va; end_va = kernel_map.virt_addr + kernel_map.size; - for (va = kernel_map.virt_addr; va < end_va; va += map_size) + for (va = kernel_map.virt_addr; va < end_va; va += PMD_SIZE) create_pgd_mapping(pgdir, va, kernel_map.phys_addr + (va - kernel_map.virt_addr), - map_size, + PMD_SIZE, early ? PAGE_KERNEL_EXEC : pgprot_from_va(va)); } #endif -asmlinkage void __init setup_vm(uintptr_t dtb_pa) +/* + * Setup a 4MB mapping that encompasses the device tree: for 64-bit kernel, + * this means 2 PMD entries whereas for 32-bit kernel, this is only 1 PGDIR + * entry. + */ +static void __init create_fdt_early_page_table(pgd_t *pgdir, uintptr_t dtb_pa) { - uintptr_t __maybe_unused pa; - uintptr_t map_size; -#ifndef __PAGETABLE_PMD_FOLDED - pmd_t fix_bmap_spmd, fix_bmap_epmd; +#ifndef CONFIG_BUILTIN_DTB + uintptr_t pa = dtb_pa & ~(PMD_SIZE - 1); + + create_pgd_mapping(early_pg_dir, DTB_EARLY_BASE_VA, + IS_ENABLED(CONFIG_64BIT) ? (uintptr_t)early_dtb_pmd : pa, + PGDIR_SIZE, + IS_ENABLED(CONFIG_64BIT) ? PAGE_TABLE : PAGE_KERNEL); + + if (IS_ENABLED(CONFIG_64BIT)) { + create_pmd_mapping(early_dtb_pmd, DTB_EARLY_BASE_VA, + pa, PMD_SIZE, PAGE_KERNEL); + create_pmd_mapping(early_dtb_pmd, DTB_EARLY_BASE_VA + PMD_SIZE, + pa + PMD_SIZE, PMD_SIZE, PAGE_KERNEL); + } + + dtb_early_va = (void *)DTB_EARLY_BASE_VA + (dtb_pa & (PMD_SIZE - 1)); +#else + /* + * For 64-bit kernel, __va can't be used since it would return a linear + * mapping address whereas dtb_early_va will be used before + * setup_vm_final installs the linear mapping. For 32-bit kernel, as the + * kernel is mapped in the linear mapping, that makes no difference. + */ + dtb_early_va = kernel_mapping_pa_to_va(XIP_FIXUP(dtb_pa)); #endif + dtb_early_pa = dtb_pa; +} + +asmlinkage void __init setup_vm(uintptr_t dtb_pa) +{ + pmd_t __maybe_unused fix_bmap_spmd, fix_bmap_epmd; + kernel_map.virt_addr = KERNEL_LINK_ADDR; #ifdef CONFIG_XIP_KERNEL kernel_map.xiprom = (uintptr_t)CONFIG_XIP_PHYS_ADDR; kernel_map.xiprom_sz = (uintptr_t)(&_exiprom) - (uintptr_t)(&_xiprom); + phys_ram_base = CONFIG_PHYS_RAM_BASE; kernel_map.phys_addr = (uintptr_t)CONFIG_PHYS_RAM_BASE; kernel_map.size = (uintptr_t)(&_end) - (uintptr_t)(&_sdata); @@ -552,23 +608,22 @@ asmlinkage void __init setup_vm(uintptr_t dtb_pa) kernel_map.phys_addr = (uintptr_t)(&_start); kernel_map.size = (uintptr_t)(&_end) - kernel_map.phys_addr; #endif - kernel_map.va_pa_offset = PAGE_OFFSET - kernel_map.phys_addr; -#ifdef CONFIG_64BIT kernel_map.va_kernel_pa_offset = kernel_map.virt_addr - kernel_map.phys_addr; -#endif - pfn_base = PFN_DOWN(kernel_map.phys_addr); - - /* - * Enforce boot alignment requirements of RV32 and - * RV64 by only allowing PMD or PGD mappings. - */ - map_size = PMD_SIZE; + riscv_pfn_base = PFN_DOWN(kernel_map.phys_addr); /* Sanity check alignment and size */ BUG_ON((PAGE_OFFSET % PGDIR_SIZE) != 0); - BUG_ON((kernel_map.phys_addr % map_size) != 0); + BUG_ON((kernel_map.phys_addr % PMD_SIZE) != 0); + +#ifdef CONFIG_64BIT + /* + * The last 4K bytes of the addressable memory can not be mapped because + * of IS_ERR_VALUE macro. + */ + BUG_ON((kernel_map.virt_addr + kernel_map.size) > ADDRESS_SPACE_END - SZ_4K); +#endif pt_ops.alloc_pte = alloc_pte_early; pt_ops.get_pte_virt = get_pte_virt_early; @@ -605,50 +660,10 @@ asmlinkage void __init setup_vm(uintptr_t dtb_pa) * us to reach paging_init(). We map all memory banks later * in setup_vm_final() below. */ - create_kernel_page_table(early_pg_dir, map_size, true); + create_kernel_page_table(early_pg_dir, true); -#ifndef __PAGETABLE_PMD_FOLDED - /* Setup early PMD for DTB */ - create_pgd_mapping(early_pg_dir, DTB_EARLY_BASE_VA, - (uintptr_t)early_dtb_pmd, PGDIR_SIZE, PAGE_TABLE); -#ifndef CONFIG_BUILTIN_DTB - /* Create two consecutive PMD mappings for FDT early scan */ - pa = dtb_pa & ~(PMD_SIZE - 1); - create_pmd_mapping(early_dtb_pmd, DTB_EARLY_BASE_VA, - pa, PMD_SIZE, PAGE_KERNEL); - create_pmd_mapping(early_dtb_pmd, DTB_EARLY_BASE_VA + PMD_SIZE, - pa + PMD_SIZE, PMD_SIZE, PAGE_KERNEL); - dtb_early_va = (void *)DTB_EARLY_BASE_VA + (dtb_pa & (PMD_SIZE - 1)); -#else /* CONFIG_BUILTIN_DTB */ -#ifdef CONFIG_64BIT - /* - * __va can't be used since it would return a linear mapping address - * whereas dtb_early_va will be used before setup_vm_final installs - * the linear mapping. - */ - dtb_early_va = kernel_mapping_pa_to_va(XIP_FIXUP(dtb_pa)); -#else - dtb_early_va = __va(dtb_pa); -#endif /* CONFIG_64BIT */ -#endif /* CONFIG_BUILTIN_DTB */ -#else -#ifndef CONFIG_BUILTIN_DTB - /* Create two consecutive PGD mappings for FDT early scan */ - pa = dtb_pa & ~(PGDIR_SIZE - 1); - create_pgd_mapping(early_pg_dir, DTB_EARLY_BASE_VA, - pa, PGDIR_SIZE, PAGE_KERNEL); - create_pgd_mapping(early_pg_dir, DTB_EARLY_BASE_VA + PGDIR_SIZE, - pa + PGDIR_SIZE, PGDIR_SIZE, PAGE_KERNEL); - dtb_early_va = (void *)DTB_EARLY_BASE_VA + (dtb_pa & (PGDIR_SIZE - 1)); -#else /* CONFIG_BUILTIN_DTB */ -#ifdef CONFIG_64BIT - dtb_early_va = kernel_mapping_pa_to_va(XIP_FIXUP(dtb_pa)); -#else - dtb_early_va = __va(dtb_pa); -#endif /* CONFIG_64BIT */ -#endif /* CONFIG_BUILTIN_DTB */ -#endif - dtb_early_pa = dtb_pa; + /* Setup early mapping for FDT early scan */ + create_fdt_early_page_table(early_pg_dir, dtb_pa); /* * Bootime fixmap only can handle PMD_SIZE mapping. Thus, boot-ioremap @@ -709,6 +724,8 @@ static void __init setup_vm_final(void) if (start <= __pa(PAGE_OFFSET) && __pa(PAGE_OFFSET) < end) start = __pa(PAGE_OFFSET); + if (end >= __pa(PAGE_OFFSET) + memory_limit) + end = __pa(PAGE_OFFSET) + memory_limit; map_size = best_map_size(start, end - start); for (pa = start; pa < end; pa += map_size) { @@ -721,7 +738,7 @@ static void __init setup_vm_final(void) #ifdef CONFIG_64BIT /* Map the kernel */ - create_kernel_page_table(swapper_pg_dir, PMD_SIZE, false); + create_kernel_page_table(swapper_pg_dir, false); #endif /* Clear fixmap PTE and PMD mappings */ @@ -788,38 +805,22 @@ static void __init reserve_crashkernel(void) crash_size = PAGE_ALIGN(crash_size); - if (crash_base == 0) { - /* - * Current riscv boot protocol requires 2MB alignment for - * RV64 and 4MB alignment for RV32 (hugepage size) - */ - crash_base = memblock_find_in_range(search_start, search_end, - crash_size, PMD_SIZE); - - if (crash_base == 0) { - pr_warn("crashkernel: couldn't allocate %lldKB\n", - crash_size >> 10); - return; - } - } else { - /* User specifies base address explicitly. */ - if (!memblock_is_region_memory(crash_base, crash_size)) { - pr_warn("crashkernel: requested region is not memory\n"); - return; - } - - if (memblock_is_region_reserved(crash_base, crash_size)) { - pr_warn("crashkernel: requested region is reserved\n"); - return; - } - + if (crash_base) { + search_start = crash_base; + search_end = crash_base + crash_size; + } - if (!IS_ALIGNED(crash_base, PMD_SIZE)) { - pr_warn("crashkernel: requested region is misaligned\n"); - return; - } + /* + * Current riscv boot protocol requires 2MB alignment for + * RV64 and 4MB alignment for RV32 (hugepage size) + */ + crash_base = memblock_phys_alloc_range(crash_size, PMD_SIZE, + search_start, search_end); + if (crash_base == 0) { + pr_warn("crashkernel: couldn't allocate %lldKB\n", + crash_size >> 10); + return; } - memblock_reserve(crash_base, crash_size); pr_info("crashkernel: reserved 0x%016llx - 0x%016llx (%lld MB)\n", crash_base, crash_base + crash_size, crash_size >> 20); @@ -829,26 +830,6 @@ static void __init reserve_crashkernel(void) } #endif /* CONFIG_KEXEC_CORE */ -#ifdef CONFIG_CRASH_DUMP -/* - * We keep track of the ELF core header of the crashed - * kernel with a reserved-memory region with compatible - * string "linux,elfcorehdr". Here we register a callback - * to populate elfcorehdr_addr/size when this region is - * present. Note that this region will be marked as - * reserved once we call early_init_fdt_scan_reserved_mem() - * later on. - */ -static int __init elfcore_hdr_setup(struct reserved_mem *rmem) -{ - elfcorehdr_addr = rmem->base; - elfcorehdr_size = rmem->size; - return 0; -} - -RESERVEDMEM_OF_DECLARE(elfcorehdr, "linux,elfcorehdr", elfcore_hdr_setup); -#endif - void __init paging_init(void) { setup_bootmem(); |