Update hash MMU kernel mapping to be in sync with radix
Subsystems (5)
v3 cover Update hash MMU kernel mapping to be in sync with radix
This patch series map all the kernel regions (vmalloc, IO and vmemmap) using 0xc top nibble address. This brings hash translation kernel mapping in sync with radix. Each of these regions can now map 512TB. We use one context to map these regions and hence the 512TB limit. We also update radix to use the same limit even though we don't have context related restrictions there.
For 4K page size, Michael Ellerman requested to keep the VMALLOC_START the same as what we have with hash 64K. I did try to implement that but found that code gets complicated with no real benefit. To assist in debugging I am adding patch "powerpc/mm: Print kernel map details to dmesg" which should show different mapping region for the booted kernel.
Also to note here is we now have the same map for both hash and radix on 4K page size. This limits 4K radix to 16TB linear mapping range. This was done to make sure we have a similar mapping between hash and radix. If we think this is unnecessarily limiting radix translation mode, I can drop this.
Aneesh Kumar K.V (8):
powerpc/mm/hash64: Add a variable to track the end of IO mapping powerpc/mm/hash64: Map all the kernel regions in the same 0xc range powerpc/mm: Validate address values against different region limits powerpc/mm: Drop the unnecessary region check powerpc/mm/hash: Simplify the region id calculation. powerpc/mm: Print kernel map details to dmesg powerpc/mm: Consolidate radix and hash address map details powerpc/mm/hash: Rename KERNEL_REGION_ID to LINEAR_MAP_REGION_ID
arch/powerpc/include/asm/book3s/64/hash-4k.h | 6 -- arch/powerpc/include/asm/book3s/64/hash-64k.h | 6 -- arch/powerpc/include/asm/book3s/64/hash.h | 89 +++++++++++-------- arch/powerpc/include/asm/book3s/64/map.h | 80 +++++++++++++++++ arch/powerpc/include/asm/book3s/64/mmu-hash.h | 33 ++++--- arch/powerpc/include/asm/book3s/64/pgtable.h | 34 +------ arch/powerpc/include/asm/book3s/64/radix.h | 33 ++----- arch/powerpc/include/asm/page.h | 11 --- arch/powerpc/kernel/setup-common.c | 3 + arch/powerpc/kvm/book3s_hv_rm_xics.c | 2 +- arch/powerpc/mm/copro_fault.c | 18 ++-- arch/powerpc/mm/hash_utils_64.c | 48 ++++++---- arch/powerpc/mm/pgtable-hash64.c | 13 ++- arch/powerpc/mm/pgtable-radix.c | 23 +++-- arch/powerpc/mm/pgtable_64.c | 15 ++-- arch/powerpc/mm/ptdump/hashpagetable.c | 4 - arch/powerpc/mm/ptdump/ptdump.c | 4 - arch/powerpc/mm/slb.c | 26 +++--- arch/powerpc/platforms/cell/spu_base.c | 4 +- drivers/misc/cxl/fault.c | 2 +- drivers/misc/ocxl/link.c | 2 +- 21 files changed, 255 insertions(+), 201 deletions(-) create mode 100644 arch/powerpc/include/asm/book3s/64/map.h
Signature · 1 line
2.20.1
PATCH v3 1/8 powerpc/mm/hash64: Add a variable to track the end of IO mapping
This makes it easy to update the region mapping in the later patch
--- arch/powerpc/include/asm/book3s/64/hash.h | 3 ++- arch/powerpc/include/asm/book3s/64/pgtable.h | 8 +++++--- arch/powerpc/include/asm/book3s/64/radix.h | 1 + arch/powerpc/mm/hash_utils_64.c | 1 + arch/powerpc/mm/pgtable-radix.c | 1 + arch/powerpc/mm/pgtable_64.c | 2 ++ 6 files changed, 12 insertions(+), 4 deletions(-)
arch/powerpc/include/asm/book3s/64/hash.h +2 -1
@@ -69,7 +69,8 @@#define H_VMALLOC_SIZE (H_KERN_VIRT_SIZE - H_KERN_IO_SIZE)#define H_VMALLOC_END (H_VMALLOC_START + H_VMALLOC_SIZE) -#define H_KERN_IO_START H_VMALLOC_END +#define H_KERN_IO_START H_VMALLOC_END +#define H_KERN_IO_END (H_KERN_VIRT_START + H_KERN_VIRT_SIZE)/* * Region IDs
arch/powerpc/include/asm/book3s/64/pgtable.h +5 -3
@@ -277,9 +277,12 @@ extern unsigned long __vmalloc_end;extern unsigned long __kernel_virt_start;extern unsigned long __kernel_virt_size;extern unsigned long __kernel_io_start; +extern unsigned long __kernel_io_end;#define KERN_VIRT_START __kernel_virt_start#define KERN_VIRT_SIZE __kernel_virt_size#define KERN_IO_START __kernel_io_start +#define KERN_IO_END __kernel_io_end +extern struct page *vmemmap;extern unsigned long ioremap_bot;extern unsigned long pci_io_base;
@@ -296,8 +299,7 @@ extern unsigned long pci_io_base;#include <asm/barrier.h> /* -* The second half of the kernel virtual space is used for IO mappings, -* it's itself carved into the PIO region (ISA and PHB IO space) and +* IO space itself carved into the PIO region (ISA and PHB IO space) and * the ioremap space * * ISA_IO_BASE = KERN_IO_START, 64K reserved area
@@ -310,7 +312,7 @@ extern unsigned long pci_io_base;#define PHB_IO_BASE (ISA_IO_END)#define PHB_IO_END (KERN_IO_START + FULL_IO_SIZE)#define IOREMAP_BASE (PHB_IO_END) -#define IOREMAP_END (KERN_VIRT_START + KERN_VIRT_SIZE) +#define IOREMAP_END (KERN_IO_END)/* Advertise special mapping type for AGP */ #define HAVE_PAGE_AGP
arch/powerpc/include/asm/book3s/64/radix.h +1 -0
@@ -111,6 +111,7 @@#define RADIX_VMEMMAP_BASE (RADIX_VMALLOC_END)#define RADIX_KERN_IO_START (RADIX_KERN_VIRT_START + (RADIX_KERN_VIRT_SIZE >> 1)) +#define RADIX_KERN_IO_END (RADIX_KERN_VIRT_START + RADIX_KERN_VIRT_SIZE)#ifndef __ASSEMBLY__#define RADIX_PTE_TABLE_SIZE (sizeof(pte_t) << RADIX_PTE_INDEX_SIZE)
arch/powerpc/mm/hash_utils_64.c +1 -0
@@ -1017,6 +1017,7 @@ void __init hash__early_init_mmu(void) __vmalloc_start =H_VMALLOC_START ; __vmalloc_end =H_VMALLOC_END ; __kernel_io_start =H_KERN_IO_START ; + __kernel_io_end =H_KERN_IO_END ; vmemmap = (struct page *)H_VMEMMAP_BASE ; ioremap_bot =IOREMAP_BASE ;
arch/powerpc/mm/pgtable-radix.c +1 -0
@@ -578,6 +578,7 @@ void __init radix__early_init_mmu(void) __vmalloc_start =RADIX_VMALLOC_START ; __vmalloc_end =RADIX_VMALLOC_END ; __kernel_io_start =RADIX_KERN_IO_START ; + __kernel_io_end =RADIX_KERN_IO_END ; vmemmap = (struct page *)RADIX_VMEMMAP_BASE ; ioremap_bot =IOREMAP_BASE ;
arch/powerpc/mm/pgtable_64.c +2 -0
@@ -98,6 +98,8 @@ unsigned long __vmalloc_end;EXPORT_SYMBOL (__vmalloc_end);unsigned long __kernel_io_start;EXPORT_SYMBOL (__kernel_io_start); +unsigned long __kernel_io_end; +EXPORT_SYMBOL (__kernel_io_end);struct page *vmemmap;EXPORT_SYMBOL (vmemmap);unsigned long __pte_frag_nr;
Signature · 1 line
2.20.1
PATCH v3 2/8 powerpc/mm/hash64: Map all the kernel regions in the same 0xc range
In reply to Aneesh Kumar K.VThis patch maps vmap, IO and vmemap regions in the 0xc address range instead of the current 0xd and 0xf range. This brings the mapping closer to radix translation mode.
With hash 64K page size each of this region is 512TB whereas with 4K config we are limited by the max page table range of 64TB and hence there regions are of 16TB size.
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 13 +++ arch/powerpc/include/asm/book3s/64/hash-64k.h | 11 +++ arch/powerpc/include/asm/book3s/64/hash.h | 95 ++++++++++++------- arch/powerpc/include/asm/book3s/64/mmu-hash.h | 31 +++--- arch/powerpc/include/asm/book3s/64/pgtable.h | 1 - arch/powerpc/include/asm/book3s/64/radix.h | 41 ++++---- arch/powerpc/include/asm/page.h | 3 +- arch/powerpc/kvm/book3s_hv_rm_xics.c | 2 +- arch/powerpc/mm/copro_fault.c | 14 ++- arch/powerpc/mm/hash_utils_64.c | 26 ++--- arch/powerpc/mm/pgtable-radix.c | 3 +- arch/powerpc/mm/pgtable_64.c | 2 - arch/powerpc/mm/ptdump/hashpagetable.c | 2 +- arch/powerpc/mm/ptdump/ptdump.c | 3 +- arch/powerpc/mm/slb.c | 22 +++-- arch/powerpc/platforms/cell/spu_base.c | 4 +- drivers/misc/cxl/fault.c | 2 +- drivers/misc/ocxl/link.c | 2 +- 18 files changed, 170 insertions(+), 107 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-4k.h +13 -0
@@ -13,6 +13,19 @@*/ #define MAX_EA_BITS_PER_CONTEXT 46 +/* +* Our page table limit us to 64TB. Hence for the kernel mapping, +* each MAP area is limited to 16 TB. +* The four map areas are: linear mapping, vmap, IO and vmemmap +*/ +#define H_KERN_MAP_SIZE (ASM_CONST(1) << (MAX_EA_BITS_PER_CONTEXT - 2)) + +/* +* Define the address range of the kernel non-linear virtual area +* 16TB +*/ +#define H_KERN_VIRT_START ASM_CONST(0xc000100000000000) +#ifndef __ASSEMBLY__#define H_PTE_TABLE_SIZE (sizeof(pte_t) << H_PTE_INDEX_SIZE)#define H_PMD_TABLE_SIZE (sizeof(pmd_t) << H_PMD_INDEX_SIZE)
arch/powerpc/include/asm/book3s/64/hash-64k.h +11 -0
@@ -13,6 +13,17 @@*/ #define MAX_EA_BITS_PER_CONTEXT 49 +/* +* We use one context for each MAP area. +*/ +#define H_KERN_MAP_SIZE (1UL << MAX_EA_BITS_PER_CONTEXT) + +/* +* Define the address range of the kernel non-linear virtual area +* 2PB +*/ +#define H_KERN_VIRT_START ASM_CONST(0xc008000000000000) +/* * 64k aligned address free up few of the lower bits of RPN for us * We steal that here. For more deatils look at pte_pfn/pfn_pte()
arch/powerpc/include/asm/book3s/64/hash.h +63 -32
128 lines not on this page · raw message
arch/powerpc/include/asm/book3s/64/mmu-hash.h +15 -16
@@ -588,7 +588,8 @@ extern void slb_set_size(u16 size);#endif #define MAX_VMALLOC_CTX_CNT 1 -#define MAX_MEMMAP_CTX_CNT 1 +#define MAX_IO_CTX_CNT 1 +#define MAX_VMEMMAP_CTX_CNT 1/* * 256MB segment
@@ -601,13 +602,10 @@ extern void slb_set_size(u16 size);* would give a protovsid of 0x1fffffffff. That will result in a VSID 0 * because of the modulo operation in vsid scramble. * -* We add one extra context to MIN_USER_CONTEXT so that we can map kernel -* context easily. The +1 is to map the unused 0xe region mapping. */ #define MAX_USER_CONTEXT ((ASM_CONST(1) << CONTEXT_BITS) - 2)#define MIN_USER_CONTEXT (MAX_KERNEL_CTX_CNT + MAX_VMALLOC_CTX_CNT + \ - MAX_MEMMAP_CTX_CNT + 2) - + MAX_IO_CTX_CNT + MAX_VMEMMAP_CTX_CNT)/* * For platforms that support on 65bit VA we limit the context bits */
@@ -747,7 +745,7 @@ static inline unsigned long get_vsid(unsigned long context, unsigned long ea,/* * Bad address. We return VSID 0 for that */ -if ((ea & ~REGION_MASK ) >=H_PGTABLE_RANGE ) +if ((ea &EA_MASK ) >=H_PGTABLE_RANGE )return 0 ;if (!mmu_has_feature (MMU_FTR_68_BIT_VA ))
@@ -774,28 +772,29 @@ static inline unsigned long get_vsid(unsigned long context, unsigned long ea,* 0x00002 - [ 0xc002000000000000 - 0xc003ffffffffffff] * 0x00003 - [ 0xc004000000000000 - 0xc005ffffffffffff] * 0x00004 - [ 0xc006000000000000 - 0xc007ffffffffffff] - -* 0x00005 - [ 0xd000000000000000 - 0xd001ffffffffffff ] -* 0x00006 - Not used - Can map 0xe000000000000000 range. -* 0x00007 - [ 0xf000000000000000 - 0xf001ffffffffffff ] * -* So we can compute the context from the region (top nibble) by -* subtracting 11, or 0xc - 1. +* vmap, IO, vmemap +* +* 0x00005 - [ 0xc008000000000000 - 0xc009ffffffffffff] +* 0x00006 - [ 0xc00a000000000000 - 0xc00bffffffffffff] +* 0x00007 - [ 0xc00c000000000000 - 0xc00dffffffffffff] +* */ static inline unsigned long get_kernel_context (unsigned long ea) { -unsigned long region_id =REGION_ID (ea); +unsigned long region_id =get_region_id (ea);unsigned long ctx;/* -* For linear mapping we do support multiple context +* Depending on Kernel config, kernel region can have one context +* or more. */ if (region_id ==KERNEL_REGION_ID ) {/* * We already verified ea to be not beyond the addr limit. */ - ctx =1 + ((ea & ~REGION_MASK ) >>MAX_EA_BITS_PER_CONTEXT ); + ctx =1 + ((ea &EA_MASK ) >>MAX_EA_BITS_PER_CONTEXT ); }else - ctx = (region_id -0xc ) +MAX_KERNEL_CTX_CNT ; + ctx = region_id +MAX_KERNEL_CTX_CNT -2 ;return ctx; }
arch/powerpc/include/asm/book3s/64/pgtable.h +0 -1
@@ -279,7 +279,6 @@ extern unsigned long __kernel_virt_size;extern unsigned long __kernel_io_start;extern unsigned long __kernel_io_end;#define KERN_VIRT_START __kernel_virt_start -#define KERN_VIRT_SIZE __kernel_virt_size#define KERN_IO_START __kernel_io_start#define KERN_IO_END __kernel_io_end
arch/powerpc/include/asm/book3s/64/radix.h +19 -22
63 lines not on this page · raw message
arch/powerpc/include/asm/page.h +2 -1
@@ -139,7 +139,8 @@ static inline bool pfn_valid(unsigned long pfn)* return true for some vmalloc addresses, which is incorrect. So explicitly * check that the address is in the kernel region. */ -#define virt_addr_valid (kaddr) (REGION_ID(kaddr) == KERNEL_REGION_ID && \ +/* may be can drop get_region_id */ +#define virt_addr_valid (kaddr) (get_region_id((unsigned long)kaddr) == KERNEL_REGION_ID && \ pfn_valid(virt_to_pfn(kaddr)))#else #define virt_addr_valid (kaddr) pfn_valid(virt_to_pfn(kaddr))
arch/powerpc/kvm/book3s_hv_rm_xics.c +1 -1
@@ -822,7 +822,7 @@ static inline void this_cpu_inc_rm(unsigned int __percpu *addr) raddr =per_cpu_ptr (addr, cpu); l = (unsigned long )raddr; -if (REGION_ID (l) ==VMALLOC_REGION_ID ) { +if (get_region_id (l) ==VMALLOC_REGION_ID ) { l =vmalloc_to_phys (raddr); raddr = (unsigned int *)l; }
arch/powerpc/mm/copro_fault.c +9 -5
@@ -105,7 +105,7 @@ int copro_calculate_slb(struct mm_struct *mm, u64 ea, struct copro_slb *slb)u64 vsid, vsidkey;int psize, ssize; -switch (REGION_ID (ea)) { +switch (get_region_id (ea)) {case USER_REGION_ID :pr_devel ("%s: 0x%llx -- USER_REGION_ID\n" , __func__, ea);if (mm ==NULL )
@@ -117,10 +117,14 @@ int copro_calculate_slb(struct mm_struct *mm, u64 ea, struct copro_slb *slb)break ;case VMALLOC_REGION_ID :pr_devel ("%s: 0x%llx -- VMALLOC_REGION_ID\n" , __func__, ea); -if (ea <VMALLOC_END ) - psize = mmu_vmalloc_psize; -else - psize = mmu_io_psize; + psize = mmu_vmalloc_psize; + ssize = mmu_kernel_ssize; + vsid =get_kernel_vsid (ea, mmu_kernel_ssize); + vsidkey =SLB_VSID_KERNEL ; +break ; +case IO_REGION_ID : +pr_devel ("%s: 0x%llx -- IO_REGION_ID\n" , __func__, ea); + psize = mmu_io_psize; ssize = mmu_kernel_ssize; vsid =get_kernel_vsid (ea, mmu_kernel_ssize); vsidkey =SLB_VSID_KERNEL ;
arch/powerpc/mm/hash_utils_64.c +15 -11
@@ -1013,12 +1013,11 @@ void __init hash__early_init_mmu(void) __pgd_val_bits =HASH_PGD_VAL_BITS ; __kernel_virt_start =H_KERN_VIRT_START ; - __kernel_virt_size =H_KERN_VIRT_SIZE ; __vmalloc_start =H_VMALLOC_START ; __vmalloc_end =H_VMALLOC_END ; __kernel_io_start =H_KERN_IO_START ; __kernel_io_end =H_KERN_IO_END ; - vmemmap = (struct page *)H_VMEMMAP_BASE ; + vmemmap = (struct page *)H_VMEMMAP_START ; ioremap_bot =IOREMAP_BASE ;#ifdef CONFIG_PCI
@@ -1239,7 +1238,7 @@ int hash_page_mm(struct mm_struct *mm, unsigned long ea,trace_hash_fault (ea, access, trap);/* Get region & vsid */ -switch (REGION_ID (ea)) { +switch (get_region_id (ea)) {case USER_REGION_ID : user_region =1 ;if (! mm) {
@@ -1253,10 +1252,13 @@ int hash_page_mm(struct mm_struct *mm, unsigned long ea,break ;case VMALLOC_REGION_ID : vsid =get_kernel_vsid (ea, mmu_kernel_ssize); -if (ea <VMALLOC_END ) - psize = mmu_vmalloc_psize; -else - psize = mmu_io_psize; + psize = mmu_vmalloc_psize; + ssize = mmu_kernel_ssize; +break ; + +case IO_REGION_ID : + vsid =get_kernel_vsid (ea, mmu_kernel_ssize); + psize = mmu_io_psize; ssize = mmu_kernel_ssize;break ;default :
@@ -1422,7 +1424,8 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap,unsigned long flags =0 ;struct mm_struct *mm = current->mm; -if (REGION_ID (ea) ==VMALLOC_REGION_ID ) +if ((get_region_id (ea) ==VMALLOC_REGION_ID ) || + (get_region_id (ea) ==IO_REGION_ID )) mm = &init_mm;if (dsisr &DSISR_NOHPTE )
@@ -1438,8 +1441,9 @@ int __hash_page(unsigned long ea, unsigned long msr, unsigned long trap,unsigned long access = _PAGE_PRESENT | _PAGE_READ;unsigned long flags =0 ;struct mm_struct *mm = current->mm; +unsigned int region_id =get_region_id (ea); -if (REGION_ID (ea) ==VMALLOC_REGION_ID ) +if ((region_id ==VMALLOC_REGION_ID ) || (region_id ==IO_REGION_ID )) mm = &init_mm;if (dsisr &DSISR_NOHPTE )
@@ -1456,7 +1460,7 @@ int __hash_page(unsigned long ea, unsigned long msr, unsigned long trap,* 2) user space access kernel space. */ access |= _PAGE_PRIVILEGED; -if ((msr &MSR_PR ) || (REGION_ID (ea) ==USER_REGION_ID )) +if ((msr &MSR_PR ) || (region_id ==USER_REGION_ID )) access &= ~_PAGE_PRIVILEGED;if (trap ==0x400 )
@@ -1500,7 +1504,7 @@ void hash_preload(struct mm_struct *mm, unsigned long ea,int rc, ssize, update_flags =0 ;unsigned long access = _PAGE_PRESENT | _PAGE_READ | (is_exec ? _PAGE_EXEC :0 ); -BUG_ON (REGION_ID (ea) !=USER_REGION_ID ); +BUG_ON (get_region_id (ea) !=USER_REGION_ID );if (!should_hash_preload (mm, ea))return ;
arch/powerpc/mm/pgtable-radix.c +1 -2
@@ -574,12 +574,11 @@ void __init radix__early_init_mmu(void) __pgd_val_bits =RADIX_PGD_VAL_BITS ; __kernel_virt_start =RADIX_KERN_VIRT_START ; - __kernel_virt_size =RADIX_KERN_VIRT_SIZE ; __vmalloc_start =RADIX_VMALLOC_START ; __vmalloc_end =RADIX_VMALLOC_END ; __kernel_io_start =RADIX_KERN_IO_START ; __kernel_io_end =RADIX_KERN_IO_END ; - vmemmap = (struct page *)RADIX_VMEMMAP_BASE ; + vmemmap = (struct page *)RADIX_VMEMMAP_START ; ioremap_bot =IOREMAP_BASE ;#ifdef CONFIG_PCI
arch/powerpc/mm/pgtable_64.c +0 -2
@@ -90,8 +90,6 @@ unsigned long __pgd_val_bits;EXPORT_SYMBOL (__pgd_val_bits);unsigned long __kernel_virt_start;EXPORT_SYMBOL (__kernel_virt_start); -unsigned long __kernel_virt_size; -EXPORT_SYMBOL (__kernel_virt_size);unsigned long __vmalloc_start;EXPORT_SYMBOL (__vmalloc_start);unsigned long __vmalloc_end;
arch/powerpc/mm/ptdump/hashpagetable.c +1 -1
@@ -500,7 +500,7 @@ static void populate_markers(void) address_markers[7 ].start_address =IOREMAP_BASE ; address_markers[8 ].start_address =IOREMAP_END ;#ifdef CONFIG_PPC_BOOK3S_64 - address_markers[9 ].start_address =H_VMEMMAP_BASE ; + address_markers[9 ].start_address =H_VMEMMAP_START ;#else address_markers[9 ].start_address =VMEMMAP_BASE ;#endif
arch/powerpc/mm/ptdump/ptdump.c +2 -1
@@ -303,8 +303,9 @@ static void populate_markers(void) address_markers[i++].start_address =PHB_IO_END ; address_markers[i++].start_address =IOREMAP_BASE ; address_markers[i++].start_address =IOREMAP_END ; +/* What is the ifdef about? */ #ifdef CONFIG_PPC_BOOK3S_64 - address_markers[i++].start_address =H_VMEMMAP_BASE ; + address_markers[i++].start_address =H_VMEMMAP_START ;#else address_markers[i++].start_address =VMEMMAP_BASE ;#endif
arch/powerpc/mm/slb.c +14 -8
@@ -694,7 +694,7 @@ static long slb_allocate_kernel(unsigned long ea, unsigned long id)if (id ==KERNEL_REGION_ID ) {/* We only support upto MAX_PHYSMEM_BITS */ -if ((ea & ~REGION_MASK ) > (1UL <<MAX_PHYSMEM_BITS )) +if ((ea &EA_MASK ) > (1UL <<MAX_PHYSMEM_BITS ))return -EFAULT ; flags =SLB_VSID_KERNEL | mmu_psize_defs[mmu_linear_psize].sllp;
@@ -702,20 +702,25 @@ static long slb_allocate_kernel(unsigned long ea, unsigned long id) #ifdef CONFIG_SPARSEMEM_VMEMMAP }else if (id ==VMEMMAP_REGION_ID ) { -if ((ea & ~REGION_MASK ) >= (1ULL <<MAX_EA_BITS_PER_CONTEXT )) +if (ea >=H_VMEMMAP_END )return -EFAULT ; flags =SLB_VSID_KERNEL | mmu_psize_defs[mmu_vmemmap_psize].sllp;#endif }else if (id ==VMALLOC_REGION_ID ) { -if ((ea & ~REGION_MASK ) >= (1ULL <<MAX_EA_BITS_PER_CONTEXT )) +if (ea >=H_VMALLOC_END )return -EFAULT ; -if (ea <H_VMALLOC_END ) - flags = local_paca->vmalloc_sllp; -else - flags =SLB_VSID_KERNEL | mmu_psize_defs[mmu_io_psize].sllp; + flags = local_paca->vmalloc_sllp; + + }else if (id ==IO_REGION_ID ) { + +if (ea >=H_KERN_IO_END ) +return -EFAULT ; + + flags =SLB_VSID_KERNEL | mmu_psize_defs[mmu_io_psize].sllp; + }else {return -EFAULT ; }
@@ -725,6 +730,7 @@ static long slb_allocate_kernel(unsigned long ea, unsigned long id) ssize =MMU_SEGSIZE_256M ; context =get_kernel_context (ea); +return slb_insert_entry (ea, context, flags, ssize, true); }
@@ -761,7 +767,7 @@ static long slb_allocate_user(struct mm_struct *mm, unsigned long ea)long do_slb_fault (struct pt_regs *regs,unsigned long ea) { -unsigned long id =REGION_ID (ea); +unsigned long id =get_region_id (ea);/* IRQs are not reconciled here, so can't check irqs_disabled */ VM_WARN_ON (mfmsr () &MSR_EE );
arch/powerpc/platforms/cell/spu_base.c +2 -2
@@ -194,7 +194,7 @@ static int __spu_trap_data_map(struct spu *spu, unsigned long ea, u64 dsisr)* faults need to be deferred to process context. */ if ((dsisr &MFC_DSISR_PTE_NOT_FOUND ) && - (REGION_ID (ea) !=USER_REGION_ID )) { + (get_region_id (ea) !=USER_REGION_ID )) {spin_unlock (&spu->register_lock); ret = hash_page(ea,
@@ -224,7 +224,7 @@ static void __spu_kernel_slb(void *addr, struct copro_slb *slb)unsigned long ea = (unsigned long )addr;u64 llp; -if (REGION_ID (ea) ==KERNEL_REGION_ID ) +if (get_region_id (ea) ==KERNEL_REGION_ID ) llp = mmu_psize_defs[mmu_linear_psize].sllp;else llp = mmu_psize_defs[mmu_virtual_psize].sllp;
drivers/misc/cxl/fault.c +1 -1
@@ -168,7 +168,7 @@ int cxl_handle_mm_fault(struct mm_struct *mm, u64 dsisr, u64 dar)if (dsisr & CXL_PSL_DSISR_An_S) access |= _PAGE_WRITE; -if (!mm && (REGION_ID (dar) !=USER_REGION_ID )) +if (!mm && (get_region_id (dar) !=USER_REGION_ID )) access |= _PAGE_PRIVILEGED;if (dsisr &DSISR_NOHPTE )
drivers/misc/ocxl/link.c +1 -1
@@ -163,7 +163,7 @@ static void xsl_fault_handler_bh(struct work_struct *fault_work)if (fault->dsisr &SPA_XSL_S ) access |= _PAGE_WRITE; -if (REGION_ID (fault->dar) !=USER_REGION_ID ) +if (get_region_id (fault->dar) !=USER_REGION_ID ) access |= _PAGE_PRIVILEGED;local_irq_save (flags);
Signature · 1 line
2.20.1
PATCH v3 3/8 powerpc/mm: Validate address values against different region limits
In reply to Aneesh Kumar K.VThis adds an explicit check in various functions.
--- arch/powerpc/mm/hash_utils_64.c | 18 +++++++++++++++--- arch/powerpc/mm/pgtable-hash64.c | 13 ++++++++++--- arch/powerpc/mm/pgtable-radix.c | 16 ++++++++++++++++ arch/powerpc/mm/pgtable_64.c | 5 +++++ 4 files changed, 46 insertions(+), 6 deletions(-)
arch/powerpc/mm/hash_utils_64.c +15 -3
@@ -786,9 +786,16 @@ void resize_hpt_for_hotplug(unsigned long new_mem_size)int hash__create_section_mapping (unsigned long start,unsigned long end,int nid) { -int rc =htab_bolt_mapping (start, end,__pa (start), -pgprot_val (PAGE_KERNEL ), mmu_linear_psize, - mmu_kernel_ssize); +int rc; + +if (end >=H_VMALLOC_START ) { +pr_warn ("Outisde the supported range\n" ); +return -1 ; + } + + rc =htab_bolt_mapping (start, end,__pa (start), +pgprot_val (PAGE_KERNEL ), mmu_linear_psize, + mmu_kernel_ssize);if (rc <0 ) { int rc2 = htab_remove_mapping(start, end, mmu_linear_psize,
@@ -929,6 +936,11 @@ static void __init htab_initialize(void)DBG ("creating mapping for region: %lx..%lx (prot: %lx)\n" , base, size, prot); +if ((base + size) >=H_VMALLOC_START ) { +pr_warn ("Outisde the supported range\n" ); +continue ; + } +BUG_ON (htab_bolt_mapping (base, base + size,__pa (base), prot, mmu_linear_psize, mmu_kernel_ssize)); }
arch/powerpc/mm/pgtable-hash64.c +10 -3
@@ -112,9 +112,16 @@ int __meminit hash__vmemmap_create_mapping(unsigned long start,unsigned long page_size, unsigned long phys) { - int rc = htab_bolt_mapping(start, start + page_size, phys, -pgprot_val (PAGE_KERNEL ), - mmu_vmemmap_psize, mmu_kernel_ssize); + int rc; + +if ((start + page_size) >=H_VMEMMAP_END ) { +pr_warn ("Outisde the supported range\n" ); +return -1 ; + } + + rc =htab_bolt_mapping (start, start + page_size, phys, +pgprot_val (PAGE_KERNEL ), + mmu_vmemmap_psize, mmu_kernel_ssize);if (rc <0 ) { int rc2 = htab_remove_mapping(start, start + page_size, mmu_vmemmap_psize,
arch/powerpc/mm/pgtable-radix.c +16 -0
@@ -334,6 +334,12 @@ void __init radix_init_pgtable(void)* page tables will be allocated within the range. No * need or a node (which we don't have yet). */ + +if ((reg->base + reg->size) >=RADIX_VMALLOC_START ) { +pr_warn ("Outisde the supported range\n" ); +continue ; + } +WARN_ON (create_physical_mapping (reg->base, reg->base + reg->size,-1 ));
@@ -866,6 +872,11 @@ static void __meminit remove_pagetable(unsigned long start, unsigned long end)int __meminit radix__create_section_mapping (unsigned long start,unsigned long end,int nid) { +if (end >=RADIX_VMALLOC_START ) { +pr_warn ("Outisde the supported range\n" ); +return -1 ; + } +return create_physical_mapping (start, end, nid); }
@@ -893,6 +904,11 @@ int __meminit radix__vmemmap_create_mapping(unsigned long start,int nid =early_pfn_to_nid (phys >>PAGE_SHIFT );int ret; +if ((start + page_size) >=RADIX_VMEMMAP_END ) { +pr_warn ("Outisde the supported range\n" ); +return -1 ; + } + ret =__map_kernel_page_nid (start, phys,__pgprot (flags), page_size, nid);BUG_ON (ret);
arch/powerpc/mm/pgtable_64.c +5 -0
@@ -121,6 +121,11 @@ void __iomem *__ioremap_at(phys_addr_t pa, void *ea, unsigned long size, pgprot_if (pgprot_val (prot) &H_PAGE_4K_PFN )return NULL ; +if ((ea + size) >= (void *)IOREMAP_END ) { +pr_warn ("Outisde the supported range\n" ); +return NULL ; + } +WARN_ON (pa & ~PAGE_MASK );WARN_ON (((unsigned long )ea) & ~PAGE_MASK );WARN_ON (size & ~PAGE_MASK );
Signature · 1 line
2.20.1
PATCH v3 4/8 powerpc/mm: Drop the unnecessary region check
In reply to Aneesh Kumar K.VAll the regions are now mapped with top nibble 0xc. Hence the region id check is not needed for virt_addr_valid()
--- arch/powerpc/include/asm/page.h | 12 ------------ 1 file changed, 12 deletions(-)
arch/powerpc/include/asm/page.h +0 -12
@@ -132,19 +132,7 @@ static inline bool pfn_valid(unsigned long pfn)#define virt_to_page (kaddr) pfn_to_page(virt_to_pfn(kaddr))#define pfn_to_kaddr (pfn) __va((pfn) << PAGE_SHIFT) -#ifdef CONFIG_PPC_BOOK3S_64 -/* -* On hash the vmalloc and other regions alias to the kernel region when passed -* through __pa(), which virt_to_pfn() uses. That means virt_addr_valid() can -* return true for some vmalloc addresses, which is incorrect. So explicitly -* check that the address is in the kernel region. -*/ -/* may be can drop get_region_id */ -#define virt_addr_valid (kaddr) (get_region_id((unsigned long)kaddr) == KERNEL_REGION_ID && \ - pfn_valid(virt_to_pfn(kaddr))) -#else #define virt_addr_valid (kaddr) pfn_valid(virt_to_pfn(kaddr)) -#endif /* * On Book-E parts we need __va to parse the device tree and we can't
Signature · 1 line
2.20.1
PATCH v3 5/8 powerpc/mm/hash: Simplify the region id calculation.
In reply to Aneesh Kumar K.VThis reduces multiple comparisons in get_region_id to a bit shift operation.
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 4 ++- arch/powerpc/include/asm/book3s/64/hash-64k.h | 1 + arch/powerpc/include/asm/book3s/64/hash.h | 31 +++++++++---------- arch/powerpc/include/asm/book3s/64/mmu-hash.h | 2 +- 4 files changed, 20 insertions(+), 18 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-4k.h +3 -1
@@ -13,12 +13,14 @@*/ #define MAX_EA_BITS_PER_CONTEXT 46 +#define REGION_SHIFT (MAX_EA_BITS_PER_CONTEXT - 2) +/* * Our page table limit us to 64TB. Hence for the kernel mapping, * each MAP area is limited to 16 TB. * The four map areas are: linear mapping, vmap, IO and vmemmap */ -#define H_KERN_MAP_SIZE (ASM_CONST(1) << (MAX_EA_BITS_PER_CONTEXT - 2)) +#define H_KERN_MAP_SIZE (ASM_CONST(1) << REGION_SHIFT)/* * Define the address range of the kernel non-linear virtual area
arch/powerpc/include/asm/book3s/64/hash-64k.h +1 -0
@@ -12,6 +12,7 @@* is handled in the hotpath. */ #define MAX_EA_BITS_PER_CONTEXT 49 +#define REGION_SHIFT MAX_EA_BITS_PER_CONTEXT/* * We use one context for each MAP area.
arch/powerpc/include/asm/book3s/64/hash.h +15 -16
@@ -87,26 +87,26 @@#define H_VMEMMAP_SIZE H_KERN_MAP_SIZE#define H_VMEMMAP_END (H_VMEMMAP_START + H_VMEMMAP_SIZE) +#define NON_LINEAR_REGION_ID (ea) ((((unsigned long)ea - H_KERN_VIRT_START) >> REGION_SHIFT) + 2) +/* * Region IDs */ -#define USER_REGION_ID 1 -#define KERNEL_REGION_ID 2 -#define VMALLOC_REGION_ID 3 -#define IO_REGION_ID 4 -#define VMEMMAP_REGION_ID 5 +#define USER_REGION_ID 0 +#define KERNEL_REGION_ID 1 +#define VMALLOC_REGION_ID NON_LINEAR_REGION_ID(H_VMALLOC_START) +#define IO_REGION_ID NON_LINEAR_REGION_ID(H_KERN_IO_START) +#define VMEMMAP_REGION_ID NON_LINEAR_REGION_ID(H_VMEMMAP_START)/* * Defines the address of the vmemap area, in its own region on * hash table CPUs. */ -#ifdef CONFIG_PPC_MM_SLICES #define HAVE_ARCH_UNMAPPED_AREA #define HAVE_ARCH_UNMAPPED_AREA_TOPDOWN #endif /* CONFIG_PPC_MM_SLICES */ -/* PTEIDX nibble */ #define _PTEIDX_SECONDARY 0x8#define _PTEIDX_GROUP_IX 0x7
@@ -117,22 +117,21 @@#ifndef __ASSEMBLY__static inline int get_region_id (unsigned long ea) { +int region_id;int id = (ea >>60UL );if (id ==0 )return USER_REGION_ID ; -VM_BUG_ON (id !=0xc ); -VM_BUG_ON (ea >=H_VMEMMAP_END ); +if (ea <H_KERN_VIRT_START ) +return KERNEL_REGION_ID ; -if (ea >=H_VMEMMAP_START ) -return VMEMMAP_REGION_ID ; -else if (ea >=H_KERN_IO_START ) -return IO_REGION_ID ; -else if (ea >=H_VMALLOC_START ) -return VMALLOC_REGION_ID ; +VM_BUG_ON (id !=0xc ); +BUILD_BUG_ON (NON_LINEAR_REGION_ID (H_VMALLOC_START ) !=2 ); -return KERNEL_REGION_ID ; + region_id =NON_LINEAR_REGION_ID (ea); +VM_BUG_ON (region_id >VMEMMAP_REGION_ID ); +return region_id; }#define hash__pmd_bad (pmd) (pmd_val(pmd) & H_PMD_BAD_BITS)
arch/powerpc/include/asm/book3s/64/mmu-hash.h +1 -1
@@ -794,7 +794,7 @@ static inline unsigned long get_kernel_context(unsigned long ea)*/ ctx =1 + ((ea &EA_MASK ) >>MAX_EA_BITS_PER_CONTEXT ); }else - ctx = region_id +MAX_KERNEL_CTX_CNT -2 ; + ctx = region_id +MAX_KERNEL_CTX_CNT -1 ;return ctx; }
Signature · 1 line
2.20.1
PATCH v3 6/8 powerpc/mm: Print kernel map details to dmesg
In reply to Aneesh Kumar K.VThis helps in debugging. We can look at the dmesg to find out different kernel mapping details.
On 4K config this shows
kernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000
On 64K config:
kernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000
--- arch/powerpc/kernel/setup-common.c | 3 +++ 1 file changed, 3 insertions(+)
arch/powerpc/kernel/setup-common.c +3 -0
@@ -831,6 +831,9 @@ static __init void print_system_info(void)pr_info ("htab_address = 0x%p\n" , htab_address);if (htab_hash_mask)pr_info ("htab_hash_mask = 0x%lx\n" , htab_hash_mask); +pr_info ("kernel vmalloc start = 0x%lx\n" ,KERN_VIRT_START ); +pr_info ("kernel IO start = 0x%lx\n" ,KERN_IO_START ); +pr_info ("kernel vmemmap start = 0x%lx\n" , (unsigned long )vmemmap);#endif #ifdef CONFIG_PPC_BOOK3S_32 if (Hash)
Signature · 1 line
2.20.1
PATCH v3 7/8 powerpc/mm: Consolidate radix and hash address map details
In reply to Aneesh Kumar K.VWe now have
4K page size config
kernel_region_map_size = 16TB
kernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000
with 64K page size config:
kernel_region_map_size = 512TB
kernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 21 ----- arch/powerpc/include/asm/book3s/64/hash-64k.h | 18 ----- arch/powerpc/include/asm/book3s/64/hash.h | 28 ++----- arch/powerpc/include/asm/book3s/64/map.h | 80 +++++++++++++++++++ arch/powerpc/include/asm/book3s/64/pgtable.h | 35 +------- arch/powerpc/include/asm/book3s/64/radix.h | 19 ----- arch/powerpc/mm/hash_utils_64.c | 11 +-- arch/powerpc/mm/pgtable-hash64.c | 2 +- arch/powerpc/mm/pgtable-radix.c | 13 +-- arch/powerpc/mm/pgtable_64.c | 10 --- arch/powerpc/mm/ptdump/hashpagetable.c | 4 - arch/powerpc/mm/ptdump/ptdump.c | 5 -- arch/powerpc/mm/slb.c | 6 +- 13 files changed, 101 insertions(+), 151 deletions(-) create mode 100644 arch/powerpc/include/asm/book3s/64/map.h
arch/powerpc/include/asm/book3s/64/hash-4k.h +0 -21
@@ -7,27 +7,6 @@#define H_PUD_INDEX_SIZE 9#define H_PGD_INDEX_SIZE 9 -/* -* Each context is 512TB. But on 4k we restrict our max TASK size to 64TB -* Hence also limit max EA bits to 64TB. -*/ -#define MAX_EA_BITS_PER_CONTEXT 46 - -#define REGION_SHIFT (MAX_EA_BITS_PER_CONTEXT - 2) - -/* -* Our page table limit us to 64TB. Hence for the kernel mapping, -* each MAP area is limited to 16 TB. -* The four map areas are: linear mapping, vmap, IO and vmemmap -*/ -#define H_KERN_MAP_SIZE (ASM_CONST(1) << REGION_SHIFT) - -/* -* Define the address range of the kernel non-linear virtual area -* 16TB -*/ -#define H_KERN_VIRT_START ASM_CONST(0xc000100000000000) -#ifndef __ASSEMBLY__#define H_PTE_TABLE_SIZE (sizeof(pte_t) << H_PTE_INDEX_SIZE)#define H_PMD_TABLE_SIZE (sizeof(pmd_t) << H_PMD_INDEX_SIZE)
arch/powerpc/include/asm/book3s/64/hash-64k.h +0 -18
@@ -7,24 +7,6 @@#define H_PUD_INDEX_SIZE 10#define H_PGD_INDEX_SIZE 8 -/* -* Each context is 512TB size. SLB miss for first context/default context -* is handled in the hotpath. -*/ -#define MAX_EA_BITS_PER_CONTEXT 49 -#define REGION_SHIFT MAX_EA_BITS_PER_CONTEXT - -/* -* We use one context for each MAP area. -*/ -#define H_KERN_MAP_SIZE (1UL << MAX_EA_BITS_PER_CONTEXT) - -/* -* Define the address range of the kernel non-linear virtual area -* 2PB -*/ -#define H_KERN_VIRT_START ASM_CONST(0xc008000000000000) -/* * 64k aligned address free up few of the lower bits of RPN for us * We steal that here. For more deatils look at pte_pfn/pfn_pte()
arch/powerpc/include/asm/book3s/64/hash.h +6 -22
@@ -45,10 +45,6 @@#define H_PUD_CACHE_INDEX (H_PUD_INDEX_SIZE)#endif -/* -* One context each will be used for vmap, IO and vmemmap -*/ -#define H_KERN_VIRT_SIZE (H_KERN_MAP_SIZE * 3)/* * +------------------------------+ * | |
@@ -75,28 +71,16 @@* +------------------------------+ Kernel linear (0xc.....) */ -#define H_VMALLOC_START H_KERN_VIRT_START -#define H_VMALLOC_SIZE H_KERN_MAP_SIZE -#define H_VMALLOC_END (H_VMALLOC_START + H_VMALLOC_SIZE) - -#define H_KERN_IO_START H_VMALLOC_END -#define H_KERN_IO_SIZE H_KERN_MAP_SIZE -#define H_KERN_IO_END (H_KERN_IO_START + H_KERN_IO_SIZE) - -#define H_VMEMMAP_START H_KERN_IO_END -#define H_VMEMMAP_SIZE H_KERN_MAP_SIZE -#define H_VMEMMAP_END (H_VMEMMAP_START + H_VMEMMAP_SIZE) - -#define NON_LINEAR_REGION_ID (ea) ((((unsigned long)ea - H_KERN_VIRT_START) >> REGION_SHIFT) + 2) +#define NON_LINEAR_REGION_ID (ea) ((((unsigned long)ea - KERN_VIRT_START) >> REGION_SHIFT) + 2)/* * Region IDs */ #define USER_REGION_ID 0#define KERNEL_REGION_ID 1 -#define VMALLOC_REGION_ID NON_LINEAR_REGION_ID(H_VMALLOC_START) -#define IO_REGION_ID NON_LINEAR_REGION_ID(H_KERN_IO_START) -#define VMEMMAP_REGION_ID NON_LINEAR_REGION_ID(H_VMEMMAP_START) +#define VMALLOC_REGION_ID NON_LINEAR_REGION_ID(VMALLOC_START) +#define IO_REGION_ID NON_LINEAR_REGION_ID(KERN_IO_START) +#define VMEMMAP_REGION_ID NON_LINEAR_REGION_ID(VMEMMAP_BASE)/* * Defines the address of the vmemap area, in its own region on
@@ -123,11 +107,11 @@ static inline int get_region_id(unsigned long ea)if (id ==0 )return USER_REGION_ID ; -if (ea <H_KERN_VIRT_START ) +if (ea <KERN_VIRT_START )return KERNEL_REGION_ID ;VM_BUG_ON (id !=0xc ); -BUILD_BUG_ON (NON_LINEAR_REGION_ID (H_VMALLOC_START ) !=2 ); +BUILD_BUG_ON (NON_LINEAR_REGION_ID (VMALLOC_START ) !=2 ); region_id =NON_LINEAR_REGION_ID (ea);VM_BUG_ON (region_id >VMEMMAP_REGION_ID );
arch/powerpc/include/asm/book3s/64/map.h new file +80 -0
80 lines not on this page · raw message
arch/powerpc/include/asm/book3s/64/pgtable.h +4 -31
@@ -269,24 +269,12 @@ enum pgtable_index { HTLB_16G_INDEX, }; -extern unsigned long __vmalloc_start; -extern unsigned long __vmalloc_end; -#define VMALLOC_START __vmalloc_start -#define VMALLOC_END __vmalloc_end - -extern unsigned long __kernel_virt_start; -extern unsigned long __kernel_virt_size; -extern unsigned long __kernel_io_start; -extern unsigned long __kernel_io_end; -#define KERN_VIRT_START __kernel_virt_start -#define KERN_IO_START __kernel_io_start -#define KERN_IO_END __kernel_io_end -extern struct page *vmemmap;extern unsigned long ioremap_bot;extern unsigned long pci_io_base;#endif /* __ASSEMBLY__ */ +#include <asm/book3s/64/map.h> #include <asm/book3s/64/hash.h> #include <asm/book3s/64/radix.h>
@@ -297,22 +285,6 @@ extern unsigned long pci_io_base;#endif #include <asm/barrier.h> -/* -* IO space itself carved into the PIO region (ISA and PHB IO space) and -* the ioremap space -* -* ISA_IO_BASE = KERN_IO_START, 64K reserved area -* PHB_IO_BASE = ISA_IO_BASE + 64K to ISA_IO_BASE + 2G, PHB IO spaces -* IOREMAP_BASE = ISA_IO_BASE + 2G to VMALLOC_START + PGTABLE_RANGE -*/ -#define FULL_IO_SIZE 0x80000000ul -#define ISA_IO_BASE (KERN_IO_START) -#define ISA_IO_END (KERN_IO_START + 0x10000ul) -#define PHB_IO_BASE (ISA_IO_END) -#define PHB_IO_END (KERN_IO_START + FULL_IO_SIZE) -#define IOREMAP_BASE (PHB_IO_END) -#define IOREMAP_END (KERN_IO_END) -/* Advertise special mapping type for AGP */ #define HAVE_PAGE_AGP
@@ -344,8 +316,9 @@ extern unsigned long pci_io_base;#endif /* __real_pte */ -static inline unsigned long pte_update (struct mm_struct *mm,unsigned long addr, -pte_t *ptep,unsigned long clr, +static inline unsigned long pte_update (struct mm_struct *mm, +unsigned long addr,pte_t *ptep, +unsigned long clr,unsigned long set,int huge) {if (radix_enabled ())
arch/powerpc/include/asm/book3s/64/radix.h +0 -19
@@ -91,25 +91,6 @@* +------------------------------+ Kernel linear (0xc.....) */ -#define RADIX_KERN_VIRT_START ASM_CONST(0xc008000000000000) -/* -* We use MAX_EA_BITS_PER_CONTEXT(hash specific) here just to make sure we pick -* the same value as hash. -*/ -#define RADIX_KERN_MAP_SIZE (1UL << MAX_EA_BITS_PER_CONTEXT) - -#define RADIX_VMALLOC_START RADIX_KERN_VIRT_START -#define RADIX_VMALLOC_SIZE RADIX_KERN_MAP_SIZE -#define RADIX_VMALLOC_END (RADIX_VMALLOC_START + RADIX_VMALLOC_SIZE) - -#define RADIX_KERN_IO_START RADIX_VMALLOC_END -#define RADIX_KERN_IO_SIZE RADIX_KERN_MAP_SIZE -#define RADIX_KERN_IO_END (RADIX_KERN_IO_START + RADIX_KERN_IO_SIZE) - -#define RADIX_VMEMMAP_START RADIX_KERN_IO_END -#define RADIX_VMEMMAP_SIZE RADIX_KERN_MAP_SIZE -#define RADIX_VMEMMAP_END (RADIX_VMEMMAP_START + RADIX_VMEMMAP_SIZE) -#ifndef __ASSEMBLY__#define RADIX_PTE_TABLE_SIZE (sizeof(pte_t) << RADIX_PTE_INDEX_SIZE)#define RADIX_PMD_TABLE_SIZE (sizeof(pmd_t) << RADIX_PMD_INDEX_SIZE)
arch/powerpc/mm/hash_utils_64.c +3 -8
@@ -788,7 +788,7 @@ int hash__create_section_mapping(unsigned long start, unsigned long end, int nid {int rc; -if (end >=H_VMALLOC_START ) { +if (end >=VMALLOC_START ) {pr_warn ("Outisde the supported range\n" );return -1 ; }
@@ -936,7 +936,7 @@ static void __init htab_initialize(void)DBG ("creating mapping for region: %lx..%lx (prot: %lx)\n" , base, size, prot); -if ((base + size) >=H_VMALLOC_START ) { +if ((base + size) >=VMALLOC_START ) {pr_warn ("Outisde the supported range\n" );continue ; }
@@ -1024,12 +1024,7 @@ void __init hash__early_init_mmu(void) __pud_val_bits =HASH_PUD_VAL_BITS ; __pgd_val_bits =HASH_PGD_VAL_BITS ; - __kernel_virt_start =H_KERN_VIRT_START ; - __vmalloc_start =H_VMALLOC_START ; - __vmalloc_end =H_VMALLOC_END ; - __kernel_io_start =H_KERN_IO_START ; - __kernel_io_end =H_KERN_IO_END ; - vmemmap = (struct page *)H_VMEMMAP_START ; + vmemmap = (struct page *)VMEMMAP_BASE ; ioremap_bot =IOREMAP_BASE ;#ifdef CONFIG_PCI
arch/powerpc/mm/pgtable-hash64.c +1 -1
@@ -114,7 +114,7 @@ int __meminit hash__vmemmap_create_mapping(unsigned long start, {int rc; -if ((start + page_size) >=H_VMEMMAP_END ) { +if ((start + page_size) >=VMEMMAP_END ) {pr_warn ("Outisde the supported range\n" );return -1 ; }
arch/powerpc/mm/pgtable-radix.c +4 -9
@@ -335,7 +335,7 @@ void __init radix_init_pgtable(void)* need or a node (which we don't have yet). */ -if ((reg->base + reg->size) >=RADIX_VMALLOC_START ) { +if ((reg->base + reg->size) >=VMALLOC_START ) {pr_warn ("Outisde the supported range\n" );continue ; }
@@ -579,12 +579,7 @@ void __init radix__early_init_mmu(void) __pud_val_bits =RADIX_PUD_VAL_BITS ; __pgd_val_bits =RADIX_PGD_VAL_BITS ; - __kernel_virt_start =RADIX_KERN_VIRT_START ; - __vmalloc_start =RADIX_VMALLOC_START ; - __vmalloc_end =RADIX_VMALLOC_END ; - __kernel_io_start =RADIX_KERN_IO_START ; - __kernel_io_end =RADIX_KERN_IO_END ; - vmemmap = (struct page *)RADIX_VMEMMAP_START ; + vmemmap = (struct page *)VMEMMAP_BASE ; ioremap_bot =IOREMAP_BASE ;#ifdef CONFIG_PCI
@@ -872,7 +867,7 @@ static void __meminit remove_pagetable(unsigned long start, unsigned long end)int __meminit radix__create_section_mapping (unsigned long start,unsigned long end,int nid) { -if (end >=RADIX_VMALLOC_START ) { +if (end >=VMALLOC_START ) {pr_warn ("Outisde the supported range\n" );return -1 ; }
@@ -904,7 +899,7 @@ int __meminit radix__vmemmap_create_mapping(unsigned long start,int nid =early_pfn_to_nid (phys >>PAGE_SHIFT );int ret; -if ((start + page_size) >=RADIX_VMEMMAP_END ) { +if ((start + page_size) >=VMEMMAP_END ) {pr_warn ("Outisde the supported range\n" );return -1 ; }
arch/powerpc/mm/pgtable_64.c +0 -10
@@ -88,16 +88,6 @@ unsigned long __pud_val_bits;EXPORT_SYMBOL (__pud_val_bits);unsigned long __pgd_val_bits;EXPORT_SYMBOL (__pgd_val_bits); -unsigned long __kernel_virt_start; -EXPORT_SYMBOL (__kernel_virt_start); -unsigned long __vmalloc_start; -EXPORT_SYMBOL (__vmalloc_start); -unsigned long __vmalloc_end; -EXPORT_SYMBOL (__vmalloc_end); -unsigned long __kernel_io_start; -EXPORT_SYMBOL (__kernel_io_start); -unsigned long __kernel_io_end; -EXPORT_SYMBOL (__kernel_io_end);struct page *vmemmap;EXPORT_SYMBOL (vmemmap);unsigned long __pte_frag_nr;
arch/powerpc/mm/ptdump/hashpagetable.c +0 -4
@@ -499,11 +499,7 @@ static void populate_markers(void) address_markers[6 ].start_address =PHB_IO_END ; address_markers[7 ].start_address =IOREMAP_BASE ; address_markers[8 ].start_address =IOREMAP_END ; -#ifdef CONFIG_PPC_BOOK3S_64 - address_markers[9 ].start_address =H_VMEMMAP_START ; -#else address_markers[9 ].start_address =VMEMMAP_BASE ; -#endif }static int ptdump_show (struct seq_file *m,void *v)
arch/powerpc/mm/ptdump/ptdump.c +0 -5
@@ -303,12 +303,7 @@ static void populate_markers(void) address_markers[i++].start_address =PHB_IO_END ; address_markers[i++].start_address =IOREMAP_BASE ; address_markers[i++].start_address =IOREMAP_END ; -/* What is the ifdef about? */ -#ifdef CONFIG_PPC_BOOK3S_64 - address_markers[i++].start_address =H_VMEMMAP_START ; -#else address_markers[i++].start_address =VMEMMAP_BASE ; -#endif #else /* !CONFIG_PPC64 */ address_markers[i++].start_address = ioremap_bot; address_markers[i++].start_address =IOREMAP_TOP ;
arch/powerpc/mm/slb.c +3 -3
@@ -702,21 +702,21 @@ static long slb_allocate_kernel(unsigned long ea, unsigned long id) #ifdef CONFIG_SPARSEMEM_VMEMMAP }else if (id ==VMEMMAP_REGION_ID ) { -if (ea >=H_VMEMMAP_END ) +if (ea >=VMEMMAP_END )return -EFAULT ; flags =SLB_VSID_KERNEL | mmu_psize_defs[mmu_vmemmap_psize].sllp;#endif }else if (id ==VMALLOC_REGION_ID ) { -if (ea >=H_VMALLOC_END ) +if (ea >=VMALLOC_END )return -EFAULT ; flags = local_paca->vmalloc_sllp; }else if (id ==IO_REGION_ID ) { -if (ea >=H_KERN_IO_END ) +if (ea >=KERN_IO_END )return -EFAULT ; flags =SLB_VSID_KERNEL | mmu_psize_defs[mmu_io_psize].sllp;
Signature · 1 line
2.20.1
PATCH v3 8/8 powerpc/mm/hash: Rename KERNEL_REGION_ID to LINEAR_MAP_REGION_ID
In reply to Aneesh Kumar K.VThe region actually point to linear map. Rename the #define to clarify thati.
--- arch/powerpc/include/asm/book3s/64/hash.h | 4 ++-- arch/powerpc/include/asm/book3s/64/mmu-hash.h | 2 +- arch/powerpc/mm/copro_fault.c | 4 ++-- arch/powerpc/mm/slb.c | 4 ++-- arch/powerpc/platforms/cell/spu_base.c | 2 +- 5 files changed, 8 insertions(+), 8 deletions(-)
arch/powerpc/include/asm/book3s/64/hash.h +2 -2
@@ -77,7 +77,7 @@* Region IDs */ #define USER_REGION_ID 0 -#define KERNEL_REGION_ID 1 +#define LINEAR_MAP_REGION_ID 1#define VMALLOC_REGION_ID NON_LINEAR_REGION_ID(VMALLOC_START)#define IO_REGION_ID NON_LINEAR_REGION_ID(KERN_IO_START)#define VMEMMAP_REGION_ID NON_LINEAR_REGION_ID(VMEMMAP_BASE)
@@ -108,7 +108,7 @@ static inline int get_region_id(unsigned long ea)return USER_REGION_ID ;if (ea <KERN_VIRT_START ) -return KERNEL_REGION_ID ; +return LINEAR_MAP_REGION_ID ;VM_BUG_ON (id !=0xc );BUILD_BUG_ON (NON_LINEAR_REGION_ID (VMALLOC_START ) !=2 );
arch/powerpc/include/asm/book3s/64/mmu-hash.h +1 -1
@@ -788,7 +788,7 @@ static inline unsigned long get_kernel_context(unsigned long ea)* Depending on Kernel config, kernel region can have one context * or more. */ -if (region_id ==KERNEL_REGION_ID ) { +if (region_id ==LINEAR_MAP_REGION_ID ) {/* * We already verified ea to be not beyond the addr limit. */
arch/powerpc/mm/copro_fault.c +2 -2
@@ -129,8 +129,8 @@ int copro_calculate_slb(struct mm_struct *mm, u64 ea, struct copro_slb *slb) vsid =get_kernel_vsid (ea, mmu_kernel_ssize); vsidkey =SLB_VSID_KERNEL ;break ; -case KERNEL_REGION_ID : -pr_devel ("%s: 0x%llx -- KERNEL_REGION_ID\n" , __func__, ea); +case LINEAR_MAP_REGION_ID : +pr_devel ("%s: 0x%llx -- LINEAR_MAP_REGION_ID\n" , __func__, ea); psize = mmu_linear_psize; ssize = mmu_kernel_ssize; vsid =get_kernel_vsid (ea, mmu_kernel_ssize);
arch/powerpc/mm/slb.c +2 -2
@@ -691,7 +691,7 @@ static long slb_allocate_kernel(unsigned long ea, unsigned long id)unsigned long flags;int ssize; -if (id ==KERNEL_REGION_ID ) { +if (id ==LINEAR_MAP_REGION_ID ) {/* We only support upto MAX_PHYSMEM_BITS */ if ((ea &EA_MASK ) > (1UL <<MAX_PHYSMEM_BITS ))
@@ -790,7 +790,7 @@ long do_slb_fault(struct pt_regs *regs, unsigned long ea)* first class kernel code. But for performance it's probably nicer * if they go via fast_exception_return too. */ -if (id >=KERNEL_REGION_ID ) { +if (id >=LINEAR_MAP_REGION_ID ) {long err;#ifdef CONFIG_DEBUG_VM /* Catch recursive kernel SLB faults. */
arch/powerpc/platforms/cell/spu_base.c +1 -1
@@ -224,7 +224,7 @@ static void __spu_kernel_slb(void *addr, struct copro_slb *slb)unsigned long ea = (unsigned long )addr;u64 llp; -if (get_region_id (ea) ==KERNEL_REGION_ID ) +if (get_region_id (ea) ==LINEAR_MAP_REGION_ID ) llp = mmu_psize_defs[mmu_linear_psize].sllp;else llp = mmu_psize_defs[mmu_virtual_psize].sllp;
Signature · 1 line
2.20.1
Nicholas Piggin
reviewerRe: [PATCH v3 7/8] powerpc/mm: Consolidate radix and hash address map details
In reply to Aneesh Kumar K.VAneesh Kumar K.V's on April 16, 2019 8:07 pm:
Quoted 9 lines
We now have
4K page size config
kernel_region_map_size= 16TBkernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000
Quoted 6 lines
with 64K page size config:
kernel_region_map_size= 512TBkernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000
Hey Aneesh,
I like the series, I like consolidating the address spaces into 0xc, and making the layouts match or similar isn't a bad thing. I don't see any real reason to force limitations on one layout or another -- you could make the argument that 4k radix should match 64k radix as much as matching 4k hash IMO.
I wouldn't like to tie them too strongly to the same base defines that force them to stay in sync.
Can we drop this patch? Or at least keep the users of the H_ and R_ defines and set them to the same thing in map.h?
Quoted 9 lines
arch/powerpc/include/asm/book3s/64/map.h new file +3
-0@@ -0,0 +1,80 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_POWERPC_BOOK3S_64_MAP_H_ +#define _ASM_POWERPC_BOOK3S_64_MAP_H_
arch/powerpc/include/asm/book3s/64/map.h line 13
Quoted 10 lines
arch/powerpc/include/asm/book3s/64/map.h new file +10
-0+ +/* +* We use MAX_EA_BITS_PER_CONTEXT (hash specific) here just to make sure we pick +* the same value for hash and radix. +*/ +#ifdef CONFIG_PPC_64K_PAGES + +/* +* Each context is 512TB size. SLB miss for first context/default context +* is handled in the hotpath.
Now everything is handled in the slowpath :P I guess that's a copy paste of the comment which my SLB miss patch should have fixed it.
Thanks,
Nick
Re: [PATCH v3 7/8] powerpc/mm: Consolidate radix and hash address map details
Quoted 25 lines · Nicholas Piggin
Aneesh Kumar K.V's on April 16, 2019 8:07 pm:
Quoted 9 lines
We now have
4K page size config
kernel_region_map_size= 16TBkernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000Quoted 6 lines
with 64K page size config:
kernel_region_map_size= 512TBkernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000Hey Aneesh,
I like the series, I like consolidating the address spaces into
0xc, and making the layouts match or similar isn't a bad thing. I don't see any real reason to force limitations on one layout or another -- you could make the argument that 4k radix should match 64k radix as much as matching 4k hash IMO.
Quoted 7 lines
I wouldn't like to tie them too strongly to the same base defines that force them to stay in sync.
Can we drop this patch? Or at least keep the users of the H_ and R_ defines and set them to the same thing in
map.h?
I did that based on the suggestion from Michael Ellerman. I guess he wanted the VMALLOC_START to match. I am not sure whether we should match the kernel_region_map_size too. I did mention that in the cover letter.
I agree with your suggestion above. I still can keep the VMALLOC_START at 16TB and keep the region_map_size as 512TB for radix 4k. I am not sure we want to do that.
I will wait for feedback from Michael to make the suggested changes.
-aneesh
Michael Ellerman
reviewerRe: [PATCH v3 7/8] powerpc/mm: Consolidate radix and hash address map details
In reply to Nicholas PigginQuoted 17 lines · Nicholas Piggin
Aneesh Kumar K.V's on April 16, 2019 8:07 pm:
Quoted 9 lines
We now have
4K page size config
kernel_region_map_size= 16TBkernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000Quoted 6 lines
with 64K page size config:
kernel_region_map_size= 512TBkernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000
Quoted 7 lines
Hey Aneesh,
I like the series, I like consolidating the address spaces into
0xc, and making the layouts match or similar isn't a bad thing. I don't see any real reason to force limitations on one layout or another -- you could make the argument that 4k radix should match 64k radix as much as matching 4k hash IMO.
I don't think I agree. The 4K and 64K layouts must be different because the page size is different and therefore the span of the page tables is different. (Unless you shrunk the 64K layouts to match 4K but that would be silly)
On the other hand there's no good reason why hash & radix need to differ, it's just that radix has a more strictly defined layout and it didn't match what hash used when we added radix. We really should have done this realignment of the hash address ranges before we merged radix.
Quoted 5 lines
I wouldn't like to tie them too strongly to the same base defines that force them to stay in sync.
Can we drop this patch? Or at least keep the users of the H_ and R_ defines and set them to the same thing in
map.h?
I don't understand why that would be a good thing?
We have all this indirection through variables at the moment, for what appear to be constants. It makes the code harder to follow and it's less efficient as well.
cheers
Nicholas Piggin
reviewerRe: [PATCH v3 7/8] powerpc/mm: Consolidate radix and hash address map details
Michael Ellerman's on April 17, 2019 10:34 pm:
Quoted 26 lines
Quoted 17 lines · Nicholas Piggin
Aneesh Kumar K.V's on April 16, 2019 8:07 pm:
Quoted 9 lines
We now have
4K page size config
kernel_region_map_size= 16TBkernel vmalloc start = 0xc000100000000000 kernel IO start = 0xc000200000000000 kernel vmemmap start = 0xc000300000000000Quoted 6 lines
with 64K page size config:
kernel_region_map_size= 512TBkernel vmalloc start = 0xc008000000000000 kernel IO start = 0xc00a000000000000 kernel vmemmap start = 0xc00c000000000000Quoted 7 lines
Hey Aneesh,
I like the series, I like consolidating the address spaces into
0xc, and making the layouts match or similar isn't a bad thing. I don't see any real reason to force limitations on one layout or another -- you could make the argument that 4k radix should match 64k radix as much as matching 4k hash IMO.
Quoted 4 lines
I don't think I agree. The 4K and 64K layouts must be different because the page size is different and therefore the span of the page tables is different. (Unless you shrunk the 64K layouts to match 4K but that would be silly)
Both 4K and 64K radix map 52 bits of EA.
Quoted 4 lines
On the other hand there's no good reason why hash & radix need to differ, it's just that radix has a more strictly defined layout and it didn't match what hash used when we added radix. We really should have done this realignment of the hash address ranges before we merged radix.
Well radix is limited by hardware, hash is more limited by software.
There is no real reason to define the addressing limit for radix any less than what the hardware supports, for example. Hash is free to match those limits if that's what we implement but it also does not need to.
Quoted 7 lines
Quoted 5 lines
I wouldn't like to tie them too strongly to the same base defines that force them to stay in sync.
Can we drop this patch? Or at least keep the users of the H_ and R_ defines and set them to the same thing in
map.h?I don't understand why that would be a good thing?
Hash could be expanded in future, radix may be expanded or get different page table layouts. I don't see why you'd tie them together artificially.
Quoted 3 lines
We have all this indirection through variables at the moment, for what appear to be constants. It makes the code harder to follow and it's less efficient as well.
What part is less efficient that matters?
Thanks,
Nick