diff options
author | 2023-11-03 10:17:22 -1000 | |
---|---|---|
committer | 2023-11-03 10:17:22 -1000 | |
commit | e392ea4d4d00880bf94550151b1ace4f88a4b17a (patch) | |
tree | 89ecc1045aea4f49961d5abc2a625ffd6126ca2a /arch | |
parent | Merge tag 'powerpc-6.7-1' of git://git.kernel.org/pub/scm/linux/kernel/git/po... (diff) | |
parent | s390/cio: replace deprecated strncpy with strscpy (diff) | |
download | linux-e392ea4d4d00880bf94550151b1ace4f88a4b17a.tar.gz linux-e392ea4d4d00880bf94550151b1ace4f88a4b17a.tar.bz2 linux-e392ea4d4d00880bf94550151b1ace4f88a4b17a.zip |
Merge tag 's390-6.7-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Pull s390 updates from Vasily Gorbik:
- Get rid of private VM_FAULT flags
- Add word-at-a-time implementation
- Add DCACHE_WORD_ACCESS support
- Cleanup control register handling
- Disallow CPU hotplug of CPU 0 to simplify its handling complexity,
following a similar restriction in x86
- Optimize pai crypto map allocation
- Update the list of crypto express EP11 coprocessor operation modes
- Fixes and improvements for secure guests AP pass-through
- Several fixes to address incorrect page marking for address
translation with the "cmma no-dat" feature, preventing potential
incorrect guest TLB flushes
- Fix early IPI handling
- Several virtual vs physical address confusion fixes
- Various small fixes and improvements all over the code
* tag 's390-6.7-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (74 commits)
s390/cio: replace deprecated strncpy with strscpy
s390/sclp: replace deprecated strncpy with strtomem
s390/cio: fix virtual vs physical address confusion
s390/cio: export CMG value as decimal
s390: delete the unused store_prefix() function
s390/cmma: fix handling of swapper_pg_dir and invalid_pg_dir
s390/cmma: fix detection of DAT pages
s390/sclp: handle default case in sclp memory notifier
s390/pai_crypto: remove per-cpu variable assignement in event initialization
s390/pai: initialize event count once at initialization
s390/pai_crypto: use PERF_ATTACH_TASK define for per task detection
s390/mm: add missing arch_set_page_dat() call to gmap allocations
s390/mm: add missing arch_set_page_dat() call to vmem_crst_alloc()
s390/cmma: fix initial kernel address space page table walk
s390/diag: add missing virt_to_phys() translation to diag224()
s390/mm,fault: move VM_FAULT_ERROR handling to do_exception()
s390/mm,fault: remove VM_FAULT_BADMAP and VM_FAULT_BADACCESS
s390/mm,fault: remove VM_FAULT_SIGNAL
s390/mm,fault: remove VM_FAULT_BADCONTEXT
s390/mm,fault: simplify kfence fault handling
...
Diffstat (limited to 'arch')
49 files changed, 993 insertions, 801 deletions
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index ae29e4392664..b0d67ac8695f 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -131,6 +131,7 @@ config S390 select ARCH_WANT_OPTIMIZE_HUGETLB_VMEMMAP select BUILDTIME_TABLE_SORT select CLONE_BACKWARDS2 + select DCACHE_WORD_ACCESS if !KMSAN select DMA_OPS if PCI select DYNAMIC_FTRACE if FUNCTION_TRACER select FUNCTION_ALIGNMENT_8B if CC_IS_GCC diff --git a/arch/s390/boot/startup.c b/arch/s390/boot/startup.c index d3e48bd9c394..8826c4f18645 100644 --- a/arch/s390/boot/startup.c +++ b/arch/s390/boot/startup.c @@ -49,7 +49,7 @@ static void detect_facilities(void) { if (test_facility(8)) { machine.has_edat1 = 1; - __ctl_set_bit(0, 23); + local_ctl_set_bit(0, CR0_EDAT_BIT); } if (test_facility(78)) machine.has_edat2 = 1; diff --git a/arch/s390/boot/vmem.c b/arch/s390/boot/vmem.c index 442a74f113cb..3075d65e112c 100644 --- a/arch/s390/boot/vmem.c +++ b/arch/s390/boot/vmem.c @@ -5,13 +5,14 @@ #include <asm/pgalloc.h> #include <asm/facility.h> #include <asm/sections.h> +#include <asm/ctlreg.h> #include <asm/physmem_info.h> #include <asm/maccess.h> #include <asm/abs_lowcore.h> #include "decompressor.h" #include "boot.h" -unsigned long __bootdata_preserved(s390_invalid_asce); +struct ctlreg __bootdata_preserved(s390_invalid_asce); #ifdef CONFIG_PROC_FS atomic_long_t __bootdata_preserved(direct_pages_count[PG_DIRECT_MAP_MAX]); @@ -166,8 +167,6 @@ static bool kasan_pmd_populate_zero_shadow(pmd_t *pmd, unsigned long addr, static bool kasan_pte_populate_zero_shadow(pte_t *pte, enum populate_mode mode) { - pte_t entry; - if (mode == POPULATE_KASAN_ZERO_SHADOW) { set_pte(pte, pte_z); return true; @@ -426,7 +425,7 @@ void setup_vmem(unsigned long asce_limit) asce_type = _REGION3_ENTRY_EMPTY; asce_bits = _ASCE_TYPE_REGION3 | _ASCE_TABLE_LENGTH; } - s390_invalid_asce = invalid_pg_dir | _ASCE_TYPE_REGION3 | _ASCE_TABLE_LENGTH; + s390_invalid_asce.val = invalid_pg_dir | _ASCE_TYPE_REGION3 | _ASCE_TABLE_LENGTH; crst_table_init((unsigned long *)swapper_pg_dir, asce_type); crst_table_init((unsigned long *)invalid_pg_dir, _REGION3_ENTRY_EMPTY); @@ -447,12 +446,12 @@ void setup_vmem(unsigned long asce_limit) kasan_populate_shadow(); - S390_lowcore.kernel_asce = swapper_pg_dir | asce_bits; + S390_lowcore.kernel_asce.val = swapper_pg_dir | asce_bits; S390_lowcore.user_asce = s390_invalid_asce; - __ctl_load(S390_lowcore.kernel_asce, 1, 1); - __ctl_load(S390_lowcore.user_asce, 7, 7); - __ctl_load(S390_lowcore.kernel_asce, 13, 13); + local_ctl_load(1, &S390_lowcore.kernel_asce); + local_ctl_load(7, &S390_lowcore.user_asce); + local_ctl_load(13, &S390_lowcore.kernel_asce); - init_mm.context.asce = S390_lowcore.kernel_asce; + init_mm.context.asce = S390_lowcore.kernel_asce.val; } diff --git a/arch/s390/include/asm/asm-extable.h b/arch/s390/include/asm/asm-extable.h index e6532477f126..4a6b0a8b6412 100644 --- a/arch/s390/include/asm/asm-extable.h +++ b/arch/s390/include/asm/asm-extable.h @@ -13,6 +13,7 @@ #define EX_TYPE_UA_LOAD_MEM 4 #define EX_TYPE_UA_LOAD_REG 5 #define EX_TYPE_UA_LOAD_REGPAIR 6 +#define EX_TYPE_ZEROPAD 7 #define EX_DATA_REG_ERR_SHIFT 0 #define EX_DATA_REG_ERR GENMASK(3, 0) @@ -23,16 +24,7 @@ #define EX_DATA_LEN_SHIFT 8 #define EX_DATA_LEN GENMASK(11, 8) -#define __EX_TABLE(_section, _fault, _target, _type) \ - stringify_in_c(.section _section,"a";) \ - stringify_in_c(.balign 4;) \ - stringify_in_c(.long (_fault) - .;) \ - stringify_in_c(.long (_target) - .;) \ - stringify_in_c(.short (_type);) \ - stringify_in_c(.short 0;) \ - stringify_in_c(.previous) - -#define __EX_TABLE_UA(_section, _fault, _target, _type, _regerr, _regaddr, _len)\ +#define __EX_TABLE(_section, _fault, _target, _type, _regerr, _regaddr, _len) \ stringify_in_c(.section _section,"a";) \ stringify_in_c(.balign 4;) \ stringify_in_c(.long (_fault) - .;) \ @@ -72,21 +64,24 @@ stringify_in_c(.previous) #define EX_TABLE(_fault, _target) \ - __EX_TABLE(__ex_table, _fault, _target, EX_TYPE_FIXUP) + __EX_TABLE(__ex_table, _fault, _target, EX_TYPE_FIXUP, __stringify(%%r0), __stringify(%%r0), 0) #define EX_TABLE_AMODE31(_fault, _target) \ - __EX_TABLE(.amode31.ex_table, _fault, _target, EX_TYPE_FIXUP) + __EX_TABLE(.amode31.ex_table, _fault, _target, EX_TYPE_FIXUP, __stringify(%%r0), __stringify(%%r0), 0) #define EX_TABLE_UA_STORE(_fault, _target, _regerr) \ - __EX_TABLE_UA(__ex_table, _fault, _target, EX_TYPE_UA_STORE, _regerr, _regerr, 0) + __EX_TABLE(__ex_table, _fault, _target, EX_TYPE_UA_STORE, _regerr, _regerr, 0) #define EX_TABLE_UA_LOAD_MEM(_fault, _target, _regerr, _regmem, _len) \ - __EX_TABLE_UA(__ex_table, _fault, _target, EX_TYPE_UA_LOAD_MEM, _regerr, _regmem, _len) + __EX_TABLE(__ex_table, _fault, _target, EX_TYPE_UA_LOAD_MEM, _regerr, _regmem, _len) #define EX_TABLE_UA_LOAD_REG(_fault, _target, _regerr, _regzero) \ - __EX_TABLE_UA(__ex_table, _fault, _target, EX_TYPE_UA_LOAD_REG, _regerr, _regzero, 0) + __EX_TABLE(__ex_table, _fault, _target, EX_TYPE_UA_LOAD_REG, _regerr, _regzero, 0) #define EX_TABLE_UA_LOAD_REGPAIR(_fault, _target, _regerr, _regzero) \ - __EX_TABLE_UA(__ex_table, _fault, _target, EX_TYPE_UA_LOAD_REGPAIR, _regerr, _regzero, 0) + __EX_TABLE(__ex_table, _fault, _target, EX_TYPE_UA_LOAD_REGPAIR, _regerr, _regzero, 0) + +#define EX_TABLE_ZEROPAD(_fault, _target, _regdata, _regaddr) \ + __EX_TABLE(__ex_table, _fault, _target, EX_TYPE_ZEROPAD, _regdata, _regaddr, 0) #endif /* __ASM_EXTABLE_H */ diff --git a/arch/s390/include/asm/ctl_reg.h b/arch/s390/include/asm/ctl_reg.h deleted file mode 100644 index adf7d8cdac7e..000000000000 --- a/arch/s390/include/asm/ctl_reg.h +++ /dev/null @@ -1,146 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -/* - * Copyright IBM Corp. 1999, 2009 - * - * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> - */ - -#ifndef __ASM_CTL_REG_H -#define __ASM_CTL_REG_H - -#include <linux/bits.h> - -#define CR0_CLOCK_COMPARATOR_SIGN BIT(63 - 10) -#define CR0_LOW_ADDRESS_PROTECTION BIT(63 - 35) -#define CR0_FETCH_PROTECTION_OVERRIDE BIT(63 - 38) -#define CR0_STORAGE_PROTECTION_OVERRIDE BIT(63 - 39) -#define CR0_EMERGENCY_SIGNAL_SUBMASK BIT(63 - 49) -#define CR0_EXTERNAL_CALL_SUBMASK BIT(63 - 50) -#define CR0_CLOCK_COMPARATOR_SUBMASK BIT(63 - 52) -#define CR0_CPU_TIMER_SUBMASK BIT(63 - 53) -#define CR0_SERVICE_SIGNAL_SUBMASK BIT(63 - 54) -#define CR0_UNUSED_56 BIT(63 - 56) -#define CR0_INTERRUPT_KEY_SUBMASK BIT(63 - 57) -#define CR0_MEASUREMENT_ALERT_SUBMASK BIT(63 - 58) - -#define CR14_UNUSED_32 BIT(63 - 32) -#define CR14_UNUSED_33 BIT(63 - 33) -#define CR14_CHANNEL_REPORT_SUBMASK BIT(63 - 35) -#define CR14_RECOVERY_SUBMASK BIT(63 - 36) -#define CR14_DEGRADATION_SUBMASK BIT(63 - 37) -#define CR14_EXTERNAL_DAMAGE_SUBMASK BIT(63 - 38) -#define CR14_WARNING_SUBMASK BIT(63 - 39) - -#ifndef __ASSEMBLY__ - -#include <linux/bug.h> - -#define __ctl_load(array, low, high) do { \ - typedef struct { char _[sizeof(array)]; } addrtype; \ - \ - BUILD_BUG_ON(sizeof(addrtype) != (high - low + 1) * sizeof(long));\ - asm volatile( \ - " lctlg %1,%2,%0\n" \ - : \ - : "Q" (*(addrtype *)(&array)), "i" (low), "i" (high) \ - : "memory"); \ -} while (0) - -#define __ctl_store(array, low, high) do { \ - typedef struct { char _[sizeof(array)]; } addrtype; \ - \ - BUILD_BUG_ON(sizeof(addrtype) != (high - low + 1) * sizeof(long));\ - asm volatile( \ - " stctg %1,%2,%0\n" \ - : "=Q" (*(addrtype *)(&array)) \ - : "i" (low), "i" (high)); \ -} while (0) - -static __always_inline void __ctl_set_bit(unsigned int cr, unsigned int bit) -{ - unsigned long reg; - - __ctl_store(reg, cr, cr); - reg |= 1UL << bit; - __ctl_load(reg, cr, cr); -} - -static __always_inline void __ctl_clear_bit(unsigned int cr, unsigned int bit) -{ - unsigned long reg; - - __ctl_store(reg, cr, cr); - reg &= ~(1UL << bit); - __ctl_load(reg, cr, cr); -} - -void smp_ctl_set_clear_bit(int cr, int bit, bool set); - -static inline void ctl_set_bit(int cr, int bit) -{ - smp_ctl_set_clear_bit(cr, bit, true); -} - -static inline void ctl_clear_bit(int cr, int bit) -{ - smp_ctl_set_clear_bit(cr, bit, false); -} - -union ctlreg0 { - unsigned long val; - struct { - unsigned long : 8; - unsigned long tcx : 1; /* Transactional-Execution control */ - unsigned long pifo : 1; /* Transactional-Execution Program- - Interruption-Filtering Override */ - unsigned long : 3; - unsigned long ccc : 1; /* Cryptography counter control */ - unsigned long pec : 1; /* PAI extension control */ - unsigned long : 17; - unsigned long : 3; - unsigned long lap : 1; /* Low-address-protection control */ - unsigned long : 4; - unsigned long edat : 1; /* Enhanced-DAT-enablement control */ - unsigned long : 2; - unsigned long iep : 1; /* Instruction-Execution-Protection */ - unsigned long : 1; - unsigned long afp : 1; /* AFP-register control */ - unsigned long vx : 1; /* Vector enablement control */ - unsigned long : 7; - unsigned long sssm : 1; /* Service signal subclass mask */ - unsigned long : 9; - }; -}; - -union ctlreg2 { - unsigned long val; - struct { - unsigned long : 33; - unsigned long ducto : 25; - unsigned long : 1; - unsigned long gse : 1; - unsigned long : 1; - unsigned long tds : 1; - unsigned long tdc : 2; - }; -}; - -union ctlreg5 { - unsigned long val; - struct { - unsigned long : 33; - unsigned long pasteo: 25; - unsigned long : 6; - }; -}; - -union ctlreg15 { - unsigned long val; - struct { - unsigned long lsea : 61; - unsigned long : 3; - }; -}; - -#endif /* __ASSEMBLY__ */ -#endif /* __ASM_CTL_REG_H */ diff --git a/arch/s390/include/asm/ctlreg.h b/arch/s390/include/asm/ctlreg.h new file mode 100644 index 000000000000..6d4b85f2b541 --- /dev/null +++ b/arch/s390/include/asm/ctlreg.h @@ -0,0 +1,251 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Copyright IBM Corp. 1999, 2009 + * + * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> + */ + +#ifndef __ASM_S390_CTLREG_H +#define __ASM_S390_CTLREG_H + +#include <linux/bits.h> + +#define CR0_TRANSACTIONAL_EXECUTION_BIT (63 - 8) +#define CR0_CLOCK_COMPARATOR_SIGN_BIT (63 - 10) +#define CR0_CRYPTOGRAPHY_COUNTER_BIT (63 - 13) +#define CR0_PAI_EXTENSION_BIT (63 - 14) +#define CR0_CPUMF_EXTRACTION_AUTH_BIT (63 - 15) +#define CR0_WARNING_TRACK_BIT (63 - 30) +#define CR0_LOW_ADDRESS_PROTECTION_BIT (63 - 35) +#define CR0_FETCH_PROTECTION_OVERRIDE_BIT (63 - 38) +#define CR0_STORAGE_PROTECTION_OVERRIDE_BIT (63 - 39) +#define CR0_EDAT_BIT (63 - 40) +#define CR0_INSTRUCTION_EXEC_PROTECTION_BIT (63 - 43) +#define CR0_VECTOR_BIT (63 - 46) +#define CR0_MALFUNCTION_ALERT_SUBMASK_BIT (63 - 48) +#define CR0_EMERGENCY_SIGNAL_SUBMASK_BIT (63 - 49) +#define CR0_EXTERNAL_CALL_SUBMASK_BIT (63 - 50) +#define CR0_CLOCK_COMPARATOR_SUBMASK_BIT (63 - 52) +#define CR0_CPU_TIMER_SUBMASK_BIT (63 - 53) +#define CR0_SERVICE_SIGNAL_SUBMASK_BIT (63 - 54) +#define CR0_UNUSED_56_BIT (63 - 56) +#define CR0_INTERRUPT_KEY_SUBMASK_BIT (63 - 57) +#define CR0_MEASUREMENT_ALERT_SUBMASK_BIT (63 - 58) +#define CR0_ETR_SUBMASK_BIT (63 - 59) +#define CR0_IUCV_BIT (63 - 62) + +#define CR0_TRANSACTIONAL_EXECUTION BIT(CR0_TRANSACTIONAL_EXECUTION_BIT) +#define CR0_CLOCK_COMPARATOR_SIGN BIT(CR0_CLOCK_COMPARATOR_SIGN_BIT) +#define CR0_CRYPTOGRAPHY_COUNTER BIT(CR0_CRYPTOGRAPHY_COUNTER_BIT) +#define CR0_PAI_EXTENSION BIT(CR0_PAI_EXTENSION_BIT) +#define CR0_CPUMF_EXTRACTION_AUTH BIT(CR0_CPUMF_EXTRACTION_AUTH_BIT) +#define CR0_WARNING_TRACK BIT(CR0_WARNING_TRACK_BIT) +#define CR0_LOW_ADDRESS_PROTECTION BIT(CR0_LOW_ADDRESS_PROTECTION_BIT) +#define CR0_FETCH_PROTECTION_OVERRIDE BIT(CR0_FETCH_PROTECTION_OVERRIDE_BIT) +#define CR0_STORAGE_PROTECTION_OVERRIDE BIT(CR0_STORAGE_PROTECTION_OVERRIDE_BIT) +#define CR0_EDAT BIT(CR0_EDAT_BIT) +#define CR0_INSTRUCTION_EXEC_PROTECTION BIT(CR0_INSTRUCTION_EXEC_PROTECTION_BIT) +#define CR0_VECTOR BIT(CR0_VECTOR_BIT) +#define CR0_MALFUNCTION_ALERT_SUBMASK BIT(CR0_MALFUNCTION_ALERT_SUBMASK_BIT) +#define CR0_EMERGENCY_SIGNAL_SUBMASK BIT(CR0_EMERGENCY_SIGNAL_SUBMASK_BIT) +#define CR0_EXTERNAL_CALL_SUBMASK BIT(CR0_EXTERNAL_CALL_SUBMASK_BIT) +#define CR0_CLOCK_COMPARATOR_SUBMASK BIT(CR0_CLOCK_COMPARATOR_SUBMASK_BIT) +#define CR0_CPU_TIMER_SUBMASK BIT(CR0_CPU_TIMER_SUBMASK_BIT) +#define CR0_SERVICE_SIGNAL_SUBMASK BIT(CR0_SERVICE_SIGNAL_SUBMASK_BIT) +#define CR0_UNUSED_56 BIT(CR0_UNUSED_56_BIT) +#define CR0_INTERRUPT_KEY_SUBMASK BIT(CR0_INTERRUPT_KEY_SUBMASK_BIT) +#define CR0_MEASUREMENT_ALERT_SUBMASK BIT(CR0_MEASUREMENT_ALERT_SUBMASK_BIT) +#define CR0_ETR_SUBMASK BIT(CR0_ETR_SUBMASK_BIT) +#define CR0_IUCV BIT(CR0_IUCV_BIT) + +#define CR2_MIO_ADDRESSING_BIT (63 - 58) +#define CR2_GUARDED_STORAGE_BIT (63 - 59) + +#define CR2_MIO_ADDRESSING BIT(CR2_MIO_ADDRESSING_BIT) +#define CR2_GUARDED_STORAGE BIT(CR2_GUARDED_STORAGE_BIT) + +#define CR14_UNUSED_32_BIT (63 - 32) +#define CR14_UNUSED_33_BIT (63 - 33) +#define CR14_CHANNEL_REPORT_SUBMASK_BIT (63 - 35) +#define CR14_RECOVERY_SUBMASK_BIT (63 - 36) +#define CR14_DEGRADATION_SUBMASK_BIT (63 - 37) +#define CR14_EXTERNAL_DAMAGE_SUBMASK_BIT (63 - 38) +#define CR14_WARNING_SUBMASK_BIT (63 - 39) + +#define CR14_UNUSED_32 BIT(CR14_UNUSED_32_BIT) +#define CR14_UNUSED_33 BIT(CR14_UNUSED_33_BIT) +#define CR14_CHANNEL_REPORT_SUBMASK BIT(CR14_CHANNEL_REPORT_SUBMASK_BIT) +#define CR14_RECOVERY_SUBMASK BIT(CR14_RECOVERY_SUBMASK_BIT) +#define CR14_DEGRADATION_SUBMASK BIT(CR14_DEGRADATION_SUBMASK_BIT) +#define CR14_EXTERNAL_DAMAGE_SUBMASK BIT(CR14_EXTERNAL_DAMAGE_SUBMASK_BIT) +#define CR14_WARNING_SUBMASK BIT(CR14_WARNING_SUBMASK_BIT) + +#ifndef __ASSEMBLY__ + +#include <linux/bug.h> + +struct ctlreg { + unsigned long val; +}; + +#define __local_ctl_load(low, high, array) do { \ + struct addrtype { \ + char _[sizeof(array)]; \ + }; \ + int _high = high; \ + int _low = low; \ + int _esize; \ + \ + _esize = (_high - _low + 1) * sizeof(struct ctlreg); \ + BUILD_BUG_ON(sizeof(struct addrtype) != _esize); \ + typecheck(struct ctlreg, array[0]); \ + asm volatile( \ + " lctlg %[_low],%[_high],%[_arr]\n" \ + : \ + : [_arr] "Q" (*(struct addrtype *)(&array)), \ + [_low] "i" (low), [_high] "i" (high) \ + : "memory"); \ +} while (0) + +#define __local_ctl_store(low, high, array) do { \ + struct addrtype { \ + char _[sizeof(array)]; \ + }; \ + int _high = high; \ + int _low = low; \ + int _esize; \ + \ + _esize = (_high - _low + 1) * sizeof(struct ctlreg); \ + BUILD_BUG_ON(sizeof(struct addrtype) != _esize); \ + typecheck(struct ctlreg, array[0]); \ + asm volatile( \ + " stctg %[_low],%[_high],%[_arr]\n" \ + : [_arr] "=Q" (*(struct addrtype *)(&array)) \ + : [_low] "i" (low), [_high] "i" (high)); \ +} while (0) + +static __always_inline void local_ctl_load(unsigned int cr, struct ctlreg *reg) +{ + asm volatile( + " lctlg %[cr],%[cr],%[reg]\n" + : + : [reg] "Q" (*reg), [cr] "i" (cr) + : "memory"); +} + +static __always_inline void local_ctl_store(unsigned int cr, struct ctlreg *reg) +{ + asm volatile( + " stctg %[cr],%[cr],%[reg]\n" + : [reg] "=Q" (*reg) + : [cr] "i" (cr)); +} + +static __always_inline void local_ctl_set_bit(unsigned int cr, unsigned int bit) +{ + struct ctlreg reg; + + local_ctl_store(cr, ®); + reg.val |= 1UL << bit; + local_ctl_load(cr, ®); +} + +static __always_inline void local_ctl_clear_bit(unsigned int cr, unsigned int bit) +{ + struct ctlreg reg; + + local_ctl_store(cr, ®); + reg.val &= ~(1UL << bit); + local_ctl_load(cr, ®); +} + +struct lowcore; + +void system_ctlreg_lock(void); +void system_ctlreg_unlock(void); +void system_ctlreg_init_save_area(struct lowcore *lc); +void system_ctlreg_modify(unsigned int cr, unsigned long data, int request); + +enum { + CTLREG_SET_BIT, + CTLREG_CLEAR_BIT, + CTLREG_LOAD, +}; + +static inline void system_ctl_set_bit(unsigned int cr, unsigned int bit) +{ + system_ctlreg_modify(cr, bit, CTLREG_SET_BIT); +} + +static inline void system_ctl_clear_bit(unsigned int cr, unsigned int bit) +{ + system_ctlreg_modify(cr, bit, CTLREG_CLEAR_BIT); +} + +static inline void system_ctl_load(unsigned int cr, struct ctlreg *reg) +{ + system_ctlreg_modify(cr, reg->val, CTLREG_LOAD); +} + +union ctlreg0 { + unsigned long val; + struct ctlreg reg; + struct { + unsigned long : 8; + unsigned long tcx : 1; /* Transactional-Execution control */ + unsigned long pifo : 1; /* Transactional-Execution Program- + Interruption-Filtering Override */ + unsigned long : 3; + unsigned long ccc : 1; /* Cryptography counter control */ + unsigned long pec : 1; /* PAI extension control */ + unsigned long : 17; + unsigned long : 3; + unsigned long lap : 1; /* Low-address-protection control */ + unsigned long : 4; + unsigned long edat : 1; /* Enhanced-DAT-enablement control */ + unsigned long : 2; + unsigned long iep : 1; /* Instruction-Execution-Protection */ + unsigned long : 1; + unsigned long afp : 1; /* AFP-register control */ + unsigned long vx : 1; /* Vector enablement control */ + unsigned long : 7; + unsigned long sssm : 1; /* Service signal subclass mask */ + unsigned long : 9; + }; +}; + +union ctlreg2 { + unsigned long val; + struct ctlreg reg; + struct { + unsigned long : 33; + unsigned long ducto : 25; + unsigned long : 1; + unsigned long gse : 1; + unsigned long : 1; + unsigned long tds : 1; + unsigned long tdc : 2; + }; +}; + +union ctlreg5 { + unsigned long val; + struct ctlreg reg; + struct { + unsigned long : 33; + unsigned long pasteo: 25; + unsigned long : 6; + }; +}; + +union ctlreg15 { + unsigned long val; + struct ctlreg reg; + struct { + unsigned long lsea : 61; + unsigned long : 3; + }; +}; + +#endif /* __ASSEMBLY__ */ +#endif /* __ASM_S390_CTLREG_H */ diff --git a/arch/s390/include/asm/fault.h b/arch/s390/include/asm/fault.h new file mode 100644 index 000000000000..d326f56603d6 --- /dev/null +++ b/arch/s390/include/asm/fault.h @@ -0,0 +1,28 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Copyright IBM Corp. 1999, 2023 + */ +#ifndef _ASM_S390_FAULT_H +#define _ASM_S390_FAULT_H + +union teid { + unsigned long val; + struct { + unsigned long addr : 52; /* Translation-exception Address */ + unsigned long fsi : 2; /* Access Exception Fetch/Store Indication */ + unsigned long : 2; + unsigned long b56 : 1; + unsigned long : 3; + unsigned long b60 : 1; + unsigned long b61 : 1; + unsigned long as : 2; /* ASCE Identifier */ + }; +}; + +enum { + TEID_FSI_UNKNOWN = 0, /* Unknown whether fetch or store */ + TEID_FSI_STORE = 1, /* Exception was due to store operation */ + TEID_FSI_FETCH = 2 /* Exception was due to fetch operation */ +}; + +#endif /* _ASM_S390_FAULT_H */ diff --git a/arch/s390/include/asm/fpu/internal.h b/arch/s390/include/asm/fpu/internal.h index bbdadb1c9efc..8634581b9011 100644 --- a/arch/s390/include/asm/fpu/internal.h +++ b/arch/s390/include/asm/fpu/internal.h @@ -10,7 +10,6 @@ #define _ASM_S390_FPU_INTERNAL_H #include <linux/string.h> -#include <asm/ctl_reg.h> #include <asm/fpu/types.h> static inline void save_vx_regs(__vector128 *vxrs) diff --git a/arch/s390/include/asm/irq.h b/arch/s390/include/asm/irq.h index 89902f754740..54b42817f70a 100644 --- a/arch/s390/include/asm/irq.h +++ b/arch/s390/include/asm/irq.h @@ -31,6 +31,7 @@ #include <linux/percpu.h> #include <linux/cache.h> #include <linux/types.h> +#include <asm/ctlreg.h> enum interruption_class { IRQEXT_CLK, @@ -101,17 +102,17 @@ enum irq_subclass { }; #define CR0_IRQ_SUBCLASS_MASK \ - ((1UL << (63 - 30)) /* Warning Track */ | \ - (1UL << (63 - 48)) /* Malfunction Alert */ | \ - (1UL << (63 - 49)) /* Emergency Signal */ | \ - (1UL << (63 - 50)) /* External Call */ | \ - (1UL << (63 - 52)) /* Clock Comparator */ | \ - (1UL << (63 - 53)) /* CPU Timer */ | \ - (1UL << (63 - 54)) /* Service Signal */ | \ - (1UL << (63 - 57)) /* Interrupt Key */ | \ - (1UL << (63 - 58)) /* Measurement Alert */ | \ - (1UL << (63 - 59)) /* Timing Alert */ | \ - (1UL << (63 - 62))) /* IUCV */ + (CR0_WARNING_TRACK | \ + CR0_MALFUNCTION_ALERT_SUBMASK | \ + CR0_EMERGENCY_SIGNAL_SUBMASK | \ + CR0_EXTERNAL_CALL_SUBMASK | \ + CR0_CLOCK_COMPARATOR_SUBMASK | \ + CR0_CPU_TIMER_SUBMASK | \ + CR0_SERVICE_SIGNAL_SUBMASK | \ + CR0_INTERRUPT_KEY_SUBMASK | \ + CR0_MEASUREMENT_ALERT_SUBMASK | \ + CR0_ETR_SUBMASK | \ + CR0_IUCV) void irq_subclass_register(enum irq_subclass subclass); void irq_subclass_unregister(enum irq_subclass subclass); diff --git a/arch/s390/include/asm/kprobes.h b/arch/s390/include/asm/kprobes.h index 83f732ca3af4..21b9e5290c04 100644 --- a/arch/s390/include/asm/kprobes.h +++ b/arch/s390/include/asm/kprobes.h @@ -15,6 +15,7 @@ * <grundym@us.ibm.com> */ #include <linux/types.h> +#include <asm/ctlreg.h> #include <asm-generic/kprobes.h> #define BREAKPOINT_INSTRUCTION 0x0002 @@ -65,7 +66,7 @@ struct prev_kprobe { struct kprobe_ctlblk { unsigned long kprobe_status; unsigned long kprobe_saved_imask; - unsigned long kprobe_saved_ctl[3]; + struct ctlreg kprobe_saved_ctl[3]; struct prev_kprobe prev_kprobe; }; diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h index 69ccc464a430..5dc1b6345006 100644 --- a/arch/s390/include/asm/lowcore.h +++ b/arch/s390/include/asm/lowcore.h @@ -11,6 +11,7 @@ #include <linux/types.h> #include <asm/ptrace.h> +#include <asm/ctlreg.h> #include <asm/cpu.h> #include <asm/types.h> @@ -139,8 +140,8 @@ struct lowcore { __u32 restart_flags; /* 0x0384 */ /* Address space pointer. */ - __u64 kernel_asce; /* 0x0388 */ - __u64 user_asce; /* 0x0390 */ + struct ctlreg kernel_asce; /* 0x0388 */ + struct ctlreg user_asce; /* 0x0390 */ /* * The lpp and current_pid fields form a @@ -199,7 +200,7 @@ struct lowcore { __u32 clock_comp_save_area[2]; /* 0x1330 */ __u64 last_break_save_area; /* 0x1338 */ __u32 access_regs_save_area[16]; /* 0x1340 */ - __u64 cregs_save_area[16]; /* 0x1380 */ + struct ctlreg cregs_save_area[16]; /* 0x1380 */ __u8 pad_0x1400[0x1500-0x1400]; /* 0x1400 */ /* Cryptography-counter designation */ __u64 ccd; /* 0x1500 */ @@ -221,12 +222,4 @@ static inline void set_prefix(__u32 address) asm volatile("spx %0" : : "Q" (address) : "memory"); } -static inline __u32 store_prefix(void) -{ - __u32 address; - - asm volatile("stpx %0" : "=Q" (address)); - return address; -} - #endif /* _ASM_S390_LOWCORE_H */ diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h index 2a38af5a00c2..757fe6f0d802 100644 --- a/arch/s390/include/asm/mmu_context.h +++ b/arch/s390/include/asm/mmu_context.h @@ -12,7 +12,7 @@ #include <linux/uaccess.h> #include <linux/mm_types.h> #include <asm/tlbflush.h> -#include <asm/ctl_reg.h> +#include <asm/ctlreg.h> #include <asm-generic/mm_hooks.h> #define init_new_context init_new_context @@ -78,10 +78,10 @@ static inline void switch_mm_irqs_off(struct mm_struct *prev, struct mm_struct * if (next == &init_mm) S390_lowcore.user_asce = s390_invalid_asce; else - S390_lowcore.user_asce = next->context.asce; + S390_lowcore.user_asce.val = next->context.asce; cpumask_set_cpu(cpu, &next->context.cpu_attach_mask); /* Clear previous user-ASCE from CR7 */ - __ctl_load(s390_invalid_asce, 7, 7); + local_ctl_load(7, &s390_invalid_asce); if (prev != next) cpumask_clear_cpu(cpu, &prev->context.cpu_attach_mask); } @@ -111,7 +111,7 @@ static inline void finish_arch_post_lock_switch(void) __tlb_flush_mm_lazy(mm); preempt_enable(); } - __ctl_load(S390_lowcore.user_asce, 7, 7); + local_ctl_load(7, &S390_lowcore.user_asce); } #define activate_mm activate_mm @@ -120,7 +120,7 @@ static inline void activate_mm(struct mm_struct *prev, { switch_mm(prev, next, current); cpumask_set_cpu(smp_processor_id(), mm_cpumask(next)); - __ctl_load(S390_lowcore.user_asce, 7, 7); + local_ctl_load(7, &S390_lowcore.user_asce); } #include <asm-generic/mmu_context.h> diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index fb3ee7758b76..601e87fa8a9a 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -18,6 +18,7 @@ #include <linux/radix-tree.h> #include <linux/atomic.h> #include <asm/sections.h> +#include <asm/ctlreg.h> #include <asm/bug.h> #include <asm/page.h> #include <asm/uv.h> @@ -25,7 +26,7 @@ extern pgd_t swapper_pg_dir[]; extern pgd_t invalid_pg_dir[]; extern void paging_init(void); -extern unsigned long s390_invalid_asce; +extern struct ctlreg s390_invalid_asce; enum { PG_DIRECT_MAP_4K = 0, diff --git a/arch/s390/include/asm/smp.h b/arch/s390/include/asm/smp.h index 73ed2781073b..6e5b1b4b19a9 100644 --- a/arch/s390/include/asm/smp.h +++ b/arch/s390/include/asm/smp.h @@ -59,7 +59,6 @@ static inline void smp_cpus_done(unsigned int max_cpus) { } -extern int smp_reinit_ipl_cpu(void); extern int smp_rescan_cpus(void); extern void __noreturn cpu_die(void); extern void __cpu_die(unsigned int cpu); diff --git a/arch/s390/include/asm/uaccess.h b/arch/s390/include/asm/uaccess.h index 8a8c64a678c4..81ae8a98e7ec 100644 --- a/arch/s390/include/asm/uaccess.h +++ b/arch/s390/include/asm/uaccess.h @@ -15,7 +15,6 @@ */ #include <asm/asm-extable.h> #include <asm/processor.h> -#include <asm/ctl_reg.h> #include <asm/extable.h> #include <asm/facility.h> #include <asm-generic/access_ok.h> diff --git a/arch/s390/include/asm/word-at-a-time.h b/arch/s390/include/asm/word-at-a-time.h new file mode 100644 index 000000000000..2579f1694b82 --- /dev/null +++ b/arch/s390/include/asm/word-at-a-time.h @@ -0,0 +1,64 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_WORD_AT_A_TIME_H +#define _ASM_WORD_AT_A_TIME_H + +#include <linux/kernel.h> +#include <asm/asm-extable.h> +#include <asm/bitsperlong.h> + +struct word_at_a_time { + const unsigned long bits; +}; + +#define WORD_AT_A_TIME_CONSTANTS { REPEAT_BYTE(0x7f) } + +static inline unsigned long prep_zero_mask(unsigned long val, unsigned long data, const struct word_at_a_time *c) +{ + return data; +} + +static inline unsigned long create_zero_mask(unsigned long data) +{ + return __fls(data); +} + +static inline unsigned long find_zero(unsigned long data) +{ + return (data ^ (BITS_PER_LONG - 1)) >> 3; +} + +static inline unsigned long has_zero(unsigned long val, unsigned long *data, const struct word_at_a_time *c) +{ + unsigned long mask = (val & c->bits) + c->bits; + + *data = ~(mask | val | c->bits); + return *data; +} + +static inline unsigned long zero_bytemask(unsigned long data) +{ + return ~1UL << data; +} + +/* + * Load an unaligned word from kernel space. + * + * In the (very unlikely) case of the word being a page-crosser + * and the next page not being mapped, take the exception and + * return zeroes in the non-existing part. + */ +static inline unsigned long load_unaligned_zeropad(const void *addr) +{ + unsigned long data; + + asm volatile( + "0: lg %[data],0(%[addr])\n" + "1: nopr %%r7\n" + EX_TABLE_ZEROPAD(0b, 1b, %[data], %[addr]) + EX_TABLE_ZEROPAD(1b, 1b, %[data], %[addr]) + : [data] "=d" (data) + : [addr] "a" (addr), "m" (*(unsigned long *)addr)); + return data; +} + +#endif /* _ASM_WORD_AT_A_TIME_H */ diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index 0df2b88cc0da..353def93973b 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -37,7 +37,7 @@ CFLAGS_unwind_bc.o += -fno-optimize-sibling-calls obj-y := head64.o traps.o time.o process.o earlypgm.o early.o setup.o idle.o vtime.o obj-y += processor.o syscall.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o obj-y += debug.o irq.o ipl.o dis.o diag.o vdso.o cpufeature.o -obj-y += sysinfo.o lgr.o os_info.o +obj-y += sysinfo.o lgr.o os_info.o ctlreg.o obj-y += runtime_instr.o cache.o fpu.o dumpstack.o guarded_storage.o sthyi.o obj-y += entry.o reipl.o kdebugfs.o alternative.o obj-y += nospec-branch.o ipl_vmparm.o machine_kexec_reloc.o unwind_bc.o diff --git a/arch/s390/kernel/crash_dump.c b/arch/s390/kernel/crash_dump.c index 7af69948b290..514feadd4c58 100644 --- a/arch/s390/kernel/crash_dump.c +++ b/arch/s390/kernel/crash_dump.c @@ -498,7 +498,7 @@ static int get_mem_chunk_cnt(void) /* * Initialize ELF loads (new kernel) */ -static void loads_init(Elf64_Phdr *phdr, u64 loads_offset) +static void loads_init(Elf64_Phdr *phdr) { phys_addr_t start, end; u64 idx; @@ -507,7 +507,7 @@ static void loads_init(Elf64_Phdr *phdr, u64 loads_offset) phdr->p_filesz = end - start; phdr->p_type = PT_LOAD; phdr->p_offset = start; - phdr->p_vaddr = start; + phdr->p_vaddr = (unsigned long)__va(start); phdr->p_paddr = start; phdr->p_memsz = end - start; phdr->p_flags = PF_R | PF_W | PF_X; @@ -612,7 +612,7 @@ int elfcorehdr_alloc(unsigned long long *addr, unsigned long long *size) ptr = notes_init(phdr_notes, ptr, ((unsigned long) hdr) + hdr_off); /* Init loads */ hdr_off = PTR_DIFF(ptr, hdr); - loads_init(phdr_loads, hdr_off); + loads_init(phdr_loads); *addr = (unsigned long long) hdr; *size = (unsigned long long) hdr_off; BUG_ON(elfcorehdr_size > alloc_size); diff --git a/arch/s390/kernel/ctlreg.c b/arch/s390/kernel/ctlreg.c new file mode 100644 index 000000000000..8cc26cf2c64a --- /dev/null +++ b/arch/s390/kernel/ctlreg.c @@ -0,0 +1,121 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright IBM Corp. 1999, 2023 + */ + +#include <linux/irqflags.h> +#include <linux/spinlock.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/smp.h> +#include <linux/cache.h> +#include <asm/abs_lowcore.h> +#include <asm/ctlreg.h> + +/* + * ctl_lock guards access to global control register contents which + * are kept in the control register save area within absolute lowcore + * at physical address zero. + */ +static DEFINE_SPINLOCK(system_ctl_lock); + +void system_ctlreg_lock(void) + __acquires(&system_ctl_lock) +{ + spin_lock(&system_ctl_lock); +} + +void system_ctlreg_unlock(void) + __releases(&system_ctl_lock) +{ + spin_unlock(&system_ctl_lock); +} + +static bool system_ctlreg_area_init __ro_after_init; + +void __init system_ctlreg_init_save_area(struct lowcore *lc) +{ + struct lowcore *abs_lc; + + abs_lc = get_abs_lowcore(); + __local_ctl_store(0, 15, lc->cregs_save_area); + __local_ctl_store(0, 15, abs_lc->cregs_save_area); + put_abs_lowcore(abs_lc); + system_ctlreg_area_init = true; +} + +struct ctlreg_parms { + unsigned long andval; + unsigned long orval; + unsigned long val; + int request; + int cr; +}; + +static void ctlreg_callback(void *info) +{ + struct ctlreg_parms *pp = info; + struct ctlreg regs[16]; + + __local_ctl_store(0, 15, regs); + if (pp->request == CTLREG_LOAD) { + regs[pp->cr].val = pp->val; + } else { + regs[pp->cr].val &= pp->andval; + regs[pp->cr].val |= pp->orval; + } + __local_ctl_load(0, 15, regs); +} + +static void system_ctlreg_update(void *info) +{ + unsigned long flags; + + if (system_state == SYSTEM_BOOTING) { + /* + * For very early calls do not call on_each_cpu() + * since not everything might be setup. + */ + local_irq_save(flags); + ctlreg_callback(info); + local_irq_restore(flags); + } else { + on_each_cpu(ctlreg_callback, info, 1); + } +} + +void system_ctlreg_modify(unsigned int cr, unsigned long data, int request) +{ + struct ctlreg_parms pp = { .cr = cr, .request = request, }; + struct lowcore *abs_lc; + + switch (request) { + case CTLREG_SET_BIT: + pp.orval = 1UL << data; + pp.andval = -1UL; + break; + case CTLREG_CLEAR_BIT: + pp.orval = 0; + pp.andval = ~(1UL << data); + break; + case CTLREG_LOAD: + pp.val = data; + break; + } + if (system_ctlreg_area_init) { + system_ctlreg_lock(); + abs_lc = get_abs_lowcore(); + if (request == CTLREG_LOAD) { + abs_lc->cregs_save_area[cr].val = pp.val; + } else { + abs_lc->cregs_save_area[cr].val &= pp.andval; + abs_lc->cregs_save_area[cr].val |= pp.orval; + } + put_abs_lowcore(abs_lc); + system_ctlreg_update(&pp); + system_ctlreg_unlock(); + } else { + system_ctlreg_update(&pp); + } +} +EXPORT_SYMBOL(system_ctlreg_modify); diff --git a/arch/s390/kernel/diag.c b/arch/s390/kernel/diag.c index f9f06cd8fcee..92fdc35f028c 100644 --- a/arch/s390/kernel/diag.c +++ b/arch/s390/kernel/diag.c @@ -245,6 +245,7 @@ EXPORT_SYMBOL(diag8c); int diag224(void *ptr) { + unsigned long addr = __pa(ptr); int rc = -EOPNOTSUPP; diag_stat_inc(DIAG_STAT_X224); @@ -253,7 +254,7 @@ int diag224(void *ptr) "0: lhi %0,0x0\n" "1:\n" EX_TABLE(0b,1b) - : "+d" (rc) :"d" (0), "d" (ptr) : "memory"); + : "+d" (rc) :"d" (0), "d" (addr) : "memory"); return rc; } EXPORT_SYMBOL(diag224); diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 442ce0489e1a..ff1f02b54771 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -216,7 +216,7 @@ static __init void detect_machine_facilities(void) { if (test_facility(8)) { S390_lowcore.machine_flags |= MACHINE_FLAG_EDAT1; - __ctl_set_bit(0, 23); + system_ctl_set_bit(0, CR0_EDAT_BIT); } if (test_facility(78)) S390_lowcore.machine_flags |= MACHINE_FLAG_EDAT2; @@ -224,13 +224,13 @@ static __init void detect_machine_facilities(void) S390_lowcore.machine_flags |= MACHINE_FLAG_IDTE; if (test_facility(50) && test_facility(73)) { S390_lowcore.machine_flags |= MACHINE_FLAG_TE; - __ctl_set_bit(0, 55); + system_ctl_set_bit(0, CR0_TRANSACTIONAL_EXECUTION_BIT); } if (test_facility(51)) S390_lowcore.machine_flags |= MACHINE_FLAG_TLB_LC; if (test_facility(129)) { S390_lowcore.machine_flags |= MACHINE_FLAG_VX; - __ctl_set_bit(0, 17); + system_ctl_set_bit(0, CR0_VECTOR_BIT); } if (test_facility(130)) S390_lowcore.machine_flags |= MACHINE_FLAG_NX; @@ -240,7 +240,7 @@ static __init void detect_machine_facilities(void) /* Enabled signed clock comparator comparisons */ S390_lowcore.machine_flags |= MACHINE_FLAG_SCC; clock_comparator_max = -1ULL >> 1; - __ctl_set_bit(0, 53); + system_ctl_set_bit(0, CR0_CLOCK_COMPARATOR_SIGN_BIT); } if (IS_ENABLED(CONFIG_PCI) && test_facility(153)) { S390_lowcore.machine_flags |= MACHINE_FLAG_PCI_MIO; @@ -258,15 +258,9 @@ static inline void save_vector_registers(void) #endif } -static inline void setup_control_registers(void) +static inline void setup_low_address_protection(void) { - unsigned long reg; - - __ctl_store(reg, 0, 0); - reg |= CR0_LOW_ADDRESS_PROTECTION; - reg |= CR0_EMERGENCY_SIGNAL_SUBMASK; - reg |= CR0_EXTERNAL_CALL_SUBMASK; - __ctl_load(reg, 0, 0); + system_ctl_set_bit(0, CR0_LOW_ADDRESS_PROTECTION_BIT); } static inline void setup_access_registers(void) @@ -279,7 +273,7 @@ static inline void setup_access_registers(void) static int __init disable_vector_extension(char *str) { S390_lowcore.machine_flags &= ~MACHINE_FLAG_VX; - __ctl_clear_bit(0, 17); + system_ctl_clear_bit(0, CR0_VECTOR_BIT); return 0; } early_param("novx", disable_vector_extension); @@ -314,7 +308,7 @@ void __init startup_init(void) save_vector_registers(); setup_topology(); sclp_early_detect(); - setup_control_registers(); + setup_low_address_protection(); setup_access_registers(); lockdep_on(); } diff --git a/arch/s390/kernel/guarded_storage.c b/arch/s390/kernel/guarded_storage.c index d14dd1c2e524..0b68168d9566 100644 --- a/arch/s390/kernel/guarded_storage.c +++ b/arch/s390/kernel/guarded_storage.c @@ -28,7 +28,7 @@ static int gs_enable(void) return -ENOMEM; gs_cb->gsd = 25; preempt_disable(); - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); load_gs_cb(gs_cb); current->thread.gs_cb = gs_cb; preempt_enable(); @@ -42,7 +42,7 @@ static int gs_disable(void) preempt_disable(); kfree(current->thread.gs_cb); current->thread.gs_cb = NULL; - __ctl_clear_bit(2, 4); + local_ctl_clear_bit(2, CR2_GUARDED_STORAGE_BIT); preempt_enable(); } return 0; @@ -84,7 +84,7 @@ void gs_load_bc_cb(struct pt_regs *regs) if (gs_cb) { kfree(current->thread.gs_cb); current->thread.gs_bc_cb = NULL; - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); load_gs_cb(gs_cb); current->thread.gs_cb = gs_cb; } diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index 05e51666db03..cc364fce6aa9 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -2381,7 +2381,7 @@ void s390_reset_system(void) set_prefix(0); /* Disable lowcore protection */ - __ctl_clear_bit(0, 28); + local_ctl_clear_bit(0, CR0_LOW_ADDRESS_PROTECTION_BIT); diag_amode31_ops.diag308_reset(); } diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c index b020ff17d206..6f71b0ce1068 100644 --- a/arch/s390/kernel/irq.c +++ b/arch/s390/kernel/irq.c @@ -385,7 +385,7 @@ void irq_subclass_register(enum irq_subclass subclass) { spin_lock(&irq_subclass_lock); if (!irq_subclass_refcount[subclass]) - ctl_set_bit(0, subclass); + system_ctl_set_bit(0, subclass); irq_subclass_refcount[subclass]++; spin_unlock(&irq_subclass_lock); } @@ -396,7 +396,7 @@ void irq_subclass_unregister(enum irq_subclass subclass) spin_lock(&irq_subclass_lock); irq_subclass_refcount[subclass]--; if (!irq_subclass_refcount[subclass]) - ctl_clear_bit(0, subclass); + system_ctl_clear_bit(0, subclass); spin_unlock(&irq_subclass_lock); } EXPORT_SYMBOL(irq_subclass_unregister); diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c index d4b863ed0aa7..f0cf20d4b3c5 100644 --- a/arch/s390/kernel/kprobes.c +++ b/arch/s390/kernel/kprobes.c @@ -224,20 +224,27 @@ static void enable_singlestep(struct kprobe_ctlblk *kcb, struct pt_regs *regs, unsigned long ip) { - struct per_regs per_kprobe; + union { + struct ctlreg regs[3]; + struct { + struct ctlreg control; + struct ctlreg start; + struct ctlreg end; + }; + } per_kprobe; /* Set up the PER control registers %cr9-%cr11 */ - per_kprobe.control = PER_EVENT_IFETCH; - per_kprobe.start = ip; - per_kprobe.end = ip; + per_kprobe.control.val = PER_EVENT_IFETCH; + per_kprobe.start.val = ip; + per_kprobe.end.val = ip; /* Save control regs and psw mask */ - __ctl_store(kcb->kprobe_saved_ctl, 9, 11); + __local_ctl_store(9, 11, kcb->kprobe_saved_ctl); kcb->kprobe_saved_imask = regs->psw.mask & (PSW_MASK_PER | PSW_MASK_IO | PSW_MASK_EXT); /* Set PER control regs, turns on single step for the given address */ - __ctl_load(per_kprobe, 9, 11); + __local_ctl_load(9, 11, per_kprobe.regs); regs->psw.mask |= PSW_MASK_PER; regs->psw.mask &= ~(PSW_MASK_IO | PSW_MASK_EXT); regs->psw.addr = ip; @@ -249,7 +256,7 @@ static void disable_singlestep(struct kprobe_ctlblk *kcb, unsigned long ip) { /* Restore control regs and psw mask, set new psw address */ - __ctl_load(kcb->kprobe_saved_ctl, 9, 11); + __local_ctl_load(9, 11, kcb->kprobe_saved_ctl); regs->psw.mask &= ~PSW_MASK_PER; regs->psw.mask |= kcb->kprobe_saved_imask; regs->psw.addr = ip; diff --git a/arch/s390/kernel/machine_kexec.c b/arch/s390/kernel/machine_kexec.c index ce65fc01671f..bb0d4d68fcbe 100644 --- a/arch/s390/kernel/machine_kexec.c +++ b/arch/s390/kernel/machine_kexec.c @@ -94,12 +94,12 @@ static noinline void __machine_kdump(void *image) if (MACHINE_HAS_VX) save_vx_regs((__vector128 *) mcesa->vector_save_area); if (MACHINE_HAS_GS) { - __ctl_store(cr2_old.val, 2, 2); + local_ctl_store(2, &cr2_old.reg); cr2_new = cr2_old; cr2_new.gse = 1; - __ctl_load(cr2_new.val, 2, 2); + local_ctl_load(2, &cr2_new.reg); save_gs_cb((struct gs_cb *) mcesa->guarded_storage_save_area); - __ctl_load(cr2_old.val, 2, 2); + local_ctl_load(2, &cr2_old.reg); } /* * To create a good backchain for this CPU in the dump store_status diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c index 38ec0487521c..0daf0f1cdfc9 100644 --- a/arch/s390/kernel/nmi.c +++ b/arch/s390/kernel/nmi.c @@ -22,13 +22,13 @@ #include <linux/kvm_host.h> #include <linux/export.h> #include <asm/lowcore.h> +#include <asm/ctlreg.h> #include <asm/smp.h> #include <asm/stp.h> #include <asm/cputime.h> #include <asm/nmi.h> #include <asm/crw.h> #include <asm/switch_to.h> -#include <asm/ctl_reg.h> #include <asm/asm-offsets.h> #include <asm/pai.h> #include <asm/vx-insn.h> @@ -131,10 +131,10 @@ static notrace void s390_handle_damage(void) * Disable low address protection and make machine check new PSW a * disabled wait PSW. Any additional machine check cannot be handled. */ - __ctl_store(cr0.val, 0, 0); + local_ctl_store(0, &cr0.reg); cr0_new = cr0; cr0_new.lap = 0; - __ctl_load(cr0_new.val, 0, 0); + local_ctl_load(0, &cr0_new.reg); psw_save = S390_lowcore.mcck_new_psw; psw_bits(S390_lowcore.mcck_new_psw).io = 0; psw_bits(S390_lowcore.mcck_new_psw).ext = 0; @@ -146,7 +146,7 @@ static notrace void s390_handle_damage(void) * values. This makes possible system dump analysis easier. */ S390_lowcore.mcck_new_psw = psw_save; - __ctl_load(cr0.val, 0, 0); + local_ctl_load(0, &cr0.reg); disabled_wait(); while (1); } @@ -185,7 +185,7 @@ void s390_handle_mcck(void) static int mchchk_wng_posted = 0; /* Use single cpu clear, as we cannot handle smp here. */ - __ctl_clear_bit(14, 24); /* Disable WARNING MCH */ + local_ctl_clear_bit(14, CR14_WARNING_SUBMASK_BIT); if (xchg(&mchchk_wng_posted, 1) == 0) kill_cad_pid(SIGPWR, 1); } @@ -269,9 +269,9 @@ static int notrace s390_validate_registers(union mci mci) */ if (!mci.vr && !test_cpu_flag(CIF_MCCK_GUEST)) kill_task = 1; - cr0.val = S390_lowcore.cregs_save_area[0]; + cr0.reg = S390_lowcore.cregs_save_area[0]; cr0.afp = cr0.vx = 1; - __ctl_load(cr0.val, 0, 0); + local_ctl_load(0, &cr0.reg); asm volatile( " la 1,%0\n" " VLM 0,15,0,1\n" @@ -279,7 +279,7 @@ static int notrace s390_validate_registers(union mci mci) : : "Q" (*(struct vx_array *)mcesa->vector_save_area) : "1"); - __ctl_load(S390_lowcore.cregs_save_area[0], 0, 0); + local_ctl_load(0, &S390_lowcore.cregs_save_area[0]); } /* Validate access registers */ asm volatile( @@ -290,7 +290,7 @@ static int notrace s390_validate_registers(union mci mci) if (!mci.ar) kill_task = 1; /* Validate guarded storage registers */ - cr2.val = S390_lowcore.cregs_save_area[2]; + cr2.reg = S390_lowcore.cregs_save_area[2]; if (cr2.gse) { if (!mci.gs) { /* @@ -505,9 +505,9 @@ NOKPROBE_SYMBOL(s390_do_machine_check); static int __init machine_check_init(void) { - ctl_set_bit(14, 25); /* enable external damage MCH */ - ctl_set_bit(14, 27); /* enable system recovery MCH */ - ctl_set_bit(14, 24); /* enable warning MCH */ + system_ctl_set_bit(14, CR14_EXTERNAL_DAMAGE_SUBMASK_BIT); + system_ctl_set_bit(14, CR14_RECOVERY_SUBMASK_BIT); + system_ctl_set_bit(14, CR14_WARNING_SUBMASK_BIT); return 0; } early_initcall(machine_check_init); diff --git a/arch/s390/kernel/perf_cpum_cf.c b/arch/s390/kernel/perf_cpum_cf.c index 850c11ea631a..41ed6e0f0a2a 100644 --- a/arch/s390/kernel/perf_cpum_cf.c +++ b/arch/s390/kernel/perf_cpum_cf.c @@ -1193,7 +1193,7 @@ static int __init cpumf_pmu_init(void) * Clear bit 15 of cr0 to unauthorize problem-state to * extract measurement counters */ - ctl_clear_bit(0, 48); + system_ctl_clear_bit(0, CR0_CPUMF_EXTRACTION_AUTH_BIT); /* register handler for measurement-alert interruptions */ rc = register_external_irq(EXT_IRQ_MEASURE_ALERT, diff --git a/arch/s390/kernel/perf_pai_crypto.c b/arch/s390/kernel/perf_pai_crypto.c index fe7d1774ded1..77fd24e6cbb6 100644 --- a/arch/s390/kernel/perf_pai_crypto.c +++ b/arch/s390/kernel/perf_pai_crypto.c @@ -16,8 +16,7 @@ #include <linux/export.h> #include <linux/io.h> #include <linux/perf_event.h> - -#include <asm/ctl_reg.h> +#include <asm/ctlreg.h> #include <asm/pai.h> #include <asm/debug.h> @@ -41,7 +40,43 @@ struct paicrypt_map { struct perf_event *event; /* Perf event for sampling */ }; -static DEFINE_PER_CPU(struct paicrypt_map, paicrypt_map); +struct paicrypt_mapptr { + struct paicrypt_map *mapptr; +}; + +static struct paicrypt_root { /* Anchor to per CPU data */ + refcount_t refcnt; /* Overall active events */ + struct paicrypt_mapptr __percpu *mapptr; +} paicrypt_root; + +/* Free per CPU data when the last event is removed. */ +static void paicrypt_root_free(void) +{ + if (refcount_dec_and_test(&paicrypt_root.refcnt)) { + free_percpu(paicrypt_root.mapptr); + paicrypt_root.mapptr = NULL; + } + debug_sprintf_event(cfm_dbg, 5, "%s root.refcount %d\n", __func__, + refcount_read(&paicrypt_root.refcnt)); +} + +/* + * On initialization of first event also allocate per CPU data dynamically. + * Start with an array of pointers, the array size is the maximum number of + * CPUs possible, which might be larger than the number of CPUs currently + * online. + */ +static int paicrypt_root_alloc(void) +{ + if (!refcount_inc_not_zero(&paicrypt_root.refcnt)) { + /* The memory is already zeroed. */ + paicrypt_root.mapptr = alloc_percpu(struct paicrypt_mapptr); + if (!paicrypt_root.mapptr) + return -ENOMEM; + refcount_set(&paicrypt_root.refcnt, 1); + } + return 0; +} /* Release the PMU if event is the last perf event */ static DEFINE_MUTEX(pai_reserve_mutex); @@ -51,7 +86,9 @@ static DEFINE_MUTEX(pai_reserve_mutex); */ static void paicrypt_event_destroy(struct perf_event *event) { - struct paicrypt_map *cpump = per_cpu_ptr(&paicrypt_map, event->cpu); + struct paicrypt_mapptr *mp = per_cpu_ptr(paicrypt_root.mapptr, + event->cpu); + struct paicrypt_map *cpump = mp->mapptr; cpump->event = NULL; static_branch_dec(&pai_key); @@ -66,11 +103,11 @@ static void paicrypt_event_destroy(struct perf_event *event) __func__, (unsigned long)cpump->page, cpump->save); free_page((unsigned long)cpump->page); - cpump->page = NULL; kvfree(cpump->save); - cpump->save = NULL; - cpump->mode = PAI_MODE_NONE; + kfree(cpump); + mp->mapptr = NULL; } + paicrypt_root_free(); mutex_unlock(&pai_reserve_mutex); } @@ -86,7 +123,8 @@ static u64 paicrypt_getctr(struct paicrypt_map *cpump, int nr, bool kernel) */ static u64 paicrypt_getdata(struct perf_event *event, bool kernel) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; u64 sum = 0; int i; @@ -132,11 +170,31 @@ static u64 paicrypt_getall(struct perf_event *event) * * Allocate the memory for the event. */ -static int paicrypt_busy(struct perf_event_attr *a, struct paicrypt_map *cpump) +static struct paicrypt_map *paicrypt_busy(struct perf_event *event) { - int rc = 0; + struct perf_event_attr *a = &event->attr; + struct paicrypt_map *cpump = NULL; + struct paicrypt_mapptr *mp; + int rc; mutex_lock(&pai_reserve_mutex); + + /* Allocate root node */ + rc = paicrypt_root_alloc(); + if (rc) + goto unlock; + + /* Allocate node for this event */ + mp = per_cpu_ptr(paicrypt_root.mapptr, event->cpu); + cpump = mp->mapptr; + if (!cpump) { /* Paicrypt_map allocated? */ + cpump = kzalloc(sizeof(*cpump), GFP_KERNEL); + if (!cpump) { + rc = -ENOMEM; + goto free_root; + } + } + if (a->sample_period) { /* Sampling requested */ if (cpump->mode != PAI_MODE_NONE) rc = -EBUSY; /* ... sampling/counting active */ @@ -144,8 +202,15 @@ static int paicrypt_busy(struct perf_event_attr *a, struct paicrypt_map *cpump) if (cpump->mode == PAI_MODE_SAMPLING) rc = -EBUSY; /* ... and sampling active */ } + /* + * This error case triggers when there is a conflict: + * Either sampling requested and counting already active, or visa + * versa. Therefore the struct paicrypto_map for this CPU is + * needed or the error could not have occurred. Only adjust root + * node refcount. + */ if (rc) - goto unlock; + goto free_root; /* Allocate memory for counter page and counter extraction. * Only the first counting event has to allocate a page. @@ -158,30 +223,36 @@ static int paicrypt_busy(struct perf_event_attr *a, struct paicrypt_map *cpump) rc = -ENOMEM; cpump->page = (unsigned long *)get_zeroed_page(GFP_KERNEL); if (!cpump->page) - goto unlock; + goto free_paicrypt_map; cpump->save = kvmalloc_array(paicrypt_cnt + 1, sizeof(struct pai_userdata), GFP_KERNEL); if (!cpump->save) { free_page((unsigned long)cpump->page); cpump->page = NULL; - goto unlock; + goto free_paicrypt_map; } + + /* Set mode and reference count */ rc = 0; refcount_set(&cpump->refcnt, 1); - -unlock: - /* If rc is non-zero, do not set mode and reference count */ - if (!rc) { - cpump->mode = a->sample_period ? PAI_MODE_SAMPLING - : PAI_MODE_COUNTING; - } + cpump->mode = a->sample_period ? PAI_MODE_SAMPLING : PAI_MODE_COUNTING; + mp->mapptr = cpump; debug_sprintf_event(cfm_dbg, 5, "%s sample_period %#llx users %d" " mode %d refcnt %u page %#lx save %p rc %d\n", __func__, a->sample_period, cpump->active_events, cpump->mode, refcount_read(&cpump->refcnt), (unsigned long)cpump->page, cpump->save, rc); + goto unlock; + +free_paicrypt_map: + kfree(cpump); + mp->mapptr = NULL; +free_root: + paicrypt_root_free(); + +unlock: mutex_unlock(&pai_reserve_mutex); - return rc; + return rc ? ERR_PTR(rc) : cpump; } /* Might be called on different CPU than the one the event is intended for. */ @@ -189,7 +260,6 @@ static int paicrypt_event_init(struct perf_event *event) { struct perf_event_attr *a = &event->attr; struct paicrypt_map *cpump; - int rc; /* PAI crypto PMU registered as PERF_TYPE_RAW, check event type */ if (a->type != PERF_TYPE_RAW && event->pmu->type != a->type) @@ -199,16 +269,15 @@ static int paicrypt_event_init(struct perf_event *event) a->config > PAI_CRYPTO_BASE + paicrypt_cnt) return -EINVAL; /* Allow only CPU wide operation, no process context for now. */ - if (event->hw.target || event->cpu == -1) + if ((event->attach_state & PERF_ATTACH_TASK) || event->cpu == -1) return -ENOENT; /* Allow only CRYPTO_ALL for sampling. */ if (a->sample_period && a->config != PAI_CRYPTO_BASE) return -EINVAL; - cpump = per_cpu_ptr(&paicrypt_map, event->cpu); - rc = paicrypt_busy(a, cpump); - if (rc) - return rc; + cpump = paicrypt_busy(event); + if (IS_ERR(cpump)) + return PTR_ERR(cpump); /* Event initialization sets last_tag to 0. When later on the events * are deleted and re-added, do not reset the event count value to zero. @@ -216,7 +285,6 @@ static int paicrypt_event_init(struct perf_event *event) * are active at the same time. */ event->hw.last_tag = 0; - cpump->event = event; event->destroy = paicrypt_event_destroy; if (a->sample_period) { @@ -253,20 +321,20 @@ static void paicrypt_start(struct perf_event *event, int flags) if (!event->hw.last_tag) { event->hw.last_tag = 1; sum = paicrypt_getall(event); /* Get current value */ - local64_set(&event->count, 0); local64_set(&event->hw.prev_count, sum); } } static int paicrypt_add(struct perf_event *event, int flags) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; unsigned long ccd; if (++cpump->active_events == 1) { ccd = virt_to_phys(cpump->page) | PAI_CRYPTO_KERNEL_OFFSET; WRITE_ONCE(S390_lowcore.ccd, ccd); - __ctl_set_bit(0, 50); + local_ctl_set_bit(0, CR0_CRYPTOGRAPHY_COUNTER_BIT); } cpump->event = event; if (flags & PERF_EF_START && !event->attr.sample_period) { @@ -287,7 +355,8 @@ static void paicrypt_stop(struct perf_event *event, int flags) static void paicrypt_del(struct perf_event *event, int flags) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; if (event->attr.sample_period) perf_sched_cb_dec(event->pmu); @@ -295,7 +364,7 @@ static void paicrypt_del(struct perf_event *event, int flags) /* Only counting needs to read counter */ paicrypt_stop(event, PERF_EF_UPDATE); if (--cpump->active_events == 0) { - __ctl_clear_bit(0, 50); + local_ctl_clear_bit(0, CR0_CRYPTOGRAPHY_COUNTER_BIT); WRITE_ONCE(S390_lowcore.ccd, 0); } } @@ -329,7 +398,8 @@ static size_t paicrypt_copy(struct pai_userdata *userdata, static int paicrypt_push_sample(void) { - struct paicrypt_map *cpump = this_cpu_ptr(&paicrypt_map); + struct paicrypt_mapptr *mp = this_cpu_ptr(paicrypt_root.mapptr); + struct paicrypt_map *cpump = mp->mapptr; struct perf_event *event = cpump->event; struct perf_sample_data data; struct perf_raw_record raw; diff --git a/arch/s390/kernel/perf_pai_ext.c b/arch/s390/kernel/perf_pai_ext.c index c57c1a203256..8ba0f1a3a39d 100644 --- a/arch/s390/kernel/perf_pai_ext.c +++ b/arch/s390/kernel/perf_pai_ext.c @@ -17,8 +17,7 @@ #include <linux/export.h> #include <linux/io.h> #include <linux/perf_event.h> - -#include <asm/ctl_reg.h> +#include <asm/ctlreg.h> #include <asm/pai.h> #include <asm/debug.h> @@ -249,7 +248,7 @@ static int paiext_event_init(struct perf_event *event) if (rc) return rc; /* Allow only CPU wide operation, no process context for now. */ - if (event->hw.target || event->cpu == -1) + if ((event->attach_state & PERF_ATTACH_TASK) || event->cpu == -1) return -ENOENT; /* Allow only event NNPA_ALL for sampling. */ if (a->sample_period && a->config != PAI_NNPA_BASE) @@ -327,7 +326,6 @@ static void paiext_start(struct perf_event *event, int flags) event->hw.last_tag = 1; sum = paiext_getall(event); /* Get current value */ local64_set(&event->hw.prev_count, sum); - local64_set(&event->count, 0); } static int paiext_add(struct perf_event *event, int flags) @@ -340,7 +338,7 @@ static int paiext_add(struct perf_event *event, int flags) S390_lowcore.aicd = virt_to_phys(cpump->paiext_cb); pcb->acc = virt_to_phys(cpump->area) | 0x1; /* Enable CPU instruction lookup for PAIE1 control block */ - __ctl_set_bit(0, 49); + local_ctl_set_bit(0, CR0_PAI_EXTENSION_BIT); debug_sprintf_event(paiext_dbg, 4, "%s 1508 %llx acc %llx\n", __func__, S390_lowcore.aicd, pcb->acc); } @@ -376,7 +374,7 @@ static void paiext_del(struct perf_event *event, int flags) } if (--cpump->active_events == 0) { /* Disable CPU instruction lookup for PAIE1 control block */ - __ctl_clear_bit(0, 49); + local_ctl_clear_bit(0, CR0_PAI_EXTENSION_BIT); pcb->acc = 0; S390_lowcore.aicd = 0; debug_sprintf_event(paiext_dbg, 4, "%s 1508 %llx acc %llx\n", diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index ea244a73efad..046403471c5d 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -41,13 +41,20 @@ void update_cr_regs(struct task_struct *task) { struct pt_regs *regs = task_pt_regs(task); struct thread_struct *thread = &task->thread; - struct per_regs old, new; union ctlreg0 cr0_old, cr0_new; union ctlreg2 cr2_old, cr2_new; int cr0_changed, cr2_changed; - - __ctl_store(cr0_old.val, 0, 0); - __ctl_store(cr2_old.val, 2, 2); + union { + struct ctlreg regs[3]; + struct { + struct ctlreg control; + struct ctlreg start; + struct ctlreg end; + }; + } old, new; + + local_ctl_store(0, &cr0_old.reg); + local_ctl_store(2, &cr2_old.reg); cr0_new = cr0_old; cr2_new = cr2_old; /* Take care of the enable/disable of transactional execution. */ @@ -75,38 +82,38 @@ void update_cr_regs(struct task_struct *task) cr0_changed = cr0_new.val != cr0_old.val; cr2_changed = cr2_new.val != cr2_old.val; if (cr0_changed) - __ctl_load(cr0_new.val, 0, 0); + local_ctl_load(0, &cr0_new.reg); if (cr2_changed) - __ctl_load(cr2_new.val, 2, 2); + local_ctl_load(2, &cr2_new.reg); /* Copy user specified PER registers */ - new.control = thread->per_user.control; - new.start = thread->per_user.start; - new.end = thread->per_user.end; + new.control.val = thread->per_user.control; + new.start.val = thread->per_user.start; + new.end.val = thread->per_user.end; /* merge TIF_SINGLE_STEP into user specified PER registers. */ if (test_tsk_thread_flag(task, TIF_SINGLE_STEP) || test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP)) { if (test_tsk_thread_flag(task, TIF_BLOCK_STEP)) - new.control |= PER_EVENT_BRANCH; + new.control.val |= PER_EVENT_BRANCH; else - new.control |= PER_EVENT_IFETCH; - new.control |= PER_CONTROL_SUSPENSION; - new.control |= PER_EVENT_TRANSACTION_END; + new.control.val |= PER_EVENT_IFETCH; + new.control.val |= PER_CONTROL_SUSPENSION; + new.control.val |= PER_EVENT_TRANSACTION_END; if (test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP)) - new.control |= PER_EVENT_IFETCH; - new.start = 0; - new.end = -1UL; + new.control.val |= PER_EVENT_IFETCH; + new.start.val = 0; + new.end.val = -1UL; } /* Take care of the PER enablement bit in the PSW. */ - if (!(new.control & PER_EVENT_MASK)) { + if (!(new.control.val & PER_EVENT_MASK)) { regs->psw.mask &= ~PSW_MASK_PER; return; } regs->psw.mask |= PSW_MASK_PER; - __ctl_store(old, 9, 11); + __local_ctl_store(9, 11, old.regs); if (memcmp(&new, &old, sizeof(struct per_regs)) != 0) - __ctl_load(new, 9, 11); + __local_ctl_load(9, 11, new.regs); } void user_enable_single_step(struct task_struct *task) @@ -1107,7 +1114,7 @@ static int s390_gs_cb_set(struct task_struct *target, target->thread.gs_cb = data; *target->thread.gs_cb = gs_cb; if (target == current) { - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); restore_gs_cb(target->thread.gs_cb); } preempt_enable(); diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index e555b576d3c8..5701356f4f33 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -305,7 +305,7 @@ static void __init setup_zfcpdump(void) return; if (oldmem_data.start) return; - strcat(boot_command_line, " cio_ignore=all,!ipldev,!condev"); + strlcat(boot_command_line, " cio_ignore=all,!ipldev,!condev", COMMAND_LINE_SIZE); console_loglevel = 2; } #else @@ -381,12 +381,6 @@ void stack_free(unsigned long stack) #endif } -void __init __noreturn arch_call_rest_init(void) -{ - smp_reinit_ipl_cpu(); - rest_init(); -} - static unsigned long __init stack_alloc_early(void) { unsigned long stack; @@ -455,7 +449,6 @@ static void __init setup_lowcore(void) lc->restart_fn = (unsigned long) do_restart; lc->restart_data = 0; lc->restart_source = -1U; - __ctl_store(lc->cregs_save_area, 0, 15); lc->spinlock_lockval = arch_spin_lockval(0); lc->spinlock_index = 0; arch_spin_lock_setup(0); @@ -465,6 +458,7 @@ static void __init setup_lowcore(void) lc->kernel_asce = S390_lowcore.kernel_asce; lc->user_asce = S390_lowcore.user_asce; + system_ctlreg_init_save_area(lc); abs_lc = get_abs_lowcore(); abs_lc->restart_stack = lc->restart_stack; abs_lc->restart_fn = lc->restart_fn; @@ -472,7 +466,6 @@ static void __init setup_lowcore(void) abs_lc->restart_source = lc->restart_source; abs_lc->restart_psw = lc->restart_psw; abs_lc->restart_flags = RESTART_FLAG_CTLREGS; - memcpy(abs_lc->cregs_save_area, lc->cregs_save_area, sizeof(abs_lc->cregs_save_area)); abs_lc->program_new_psw = lc->program_new_psw; abs_lc->mcesad = lc->mcesad; put_abs_lowcore(abs_lc); @@ -797,15 +790,15 @@ static void __init setup_cr(void) __ctl_duct[4] = (unsigned long)__ctl_duald; /* Update control registers CR2, CR5 and CR15 */ - __ctl_store(cr2.val, 2, 2); - __ctl_store(cr5.val, 5, 5); - __ctl_store(cr15.val, 15, 15); + local_ctl_store(2, &cr2.reg); + local_ctl_store(5, &cr5.reg); + local_ctl_store(15, &cr15.reg); cr2.ducto = (unsigned long)__ctl_duct >> 6; cr5.pasteo = (unsigned long)__ctl_duct >> 6; cr15.lsea = (unsigned long)__ctl_linkage_stack >> 3; - __ctl_load(cr2.val, 2, 2); - __ctl_load(cr5.val, 5, 5); - __ctl_load(cr15.val, 15, 15); + system_ctl_load(2, &cr2.reg); + system_ctl_load(5, &cr5.reg); + system_ctl_load(15, &cr15.reg); } /* diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 214a1b67f80a..f7fcfff09acf 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -37,6 +37,7 @@ #include <linux/crash_dump.h> #include <linux/kprobes.h> #include <asm/asm-offsets.h> +#include <asm/ctlreg.h> #include <asm/pfault.h> #include <asm/diag.h> #include <asm/switch_to.h> @@ -567,54 +568,6 @@ void arch_irq_work_raise(void) } #endif -/* - * parameter area for the set/clear control bit callbacks - */ -struct ec_creg_mask_parms { - unsigned long orval; - unsigned long andval; - int cr; -}; - -/* - * callback for setting/clearing control bits - */ -static void smp_ctl_bit_callback(void *info) -{ - struct ec_creg_mask_parms *pp = info; - unsigned long cregs[16]; - - __ctl_store(cregs, 0, 15); - cregs[pp->cr] = (cregs[pp->cr] & pp->andval) | pp->orval; - __ctl_load(cregs, 0, 15); -} - -static DEFINE_SPINLOCK(ctl_lock); - -void smp_ctl_set_clear_bit(int cr, int bit, bool set) -{ - struct ec_creg_mask_parms parms = { .cr = cr, }; - struct lowcore *abs_lc; - u64 ctlreg; - - if (set) { - parms.orval = 1UL << bit; - parms.andval = -1UL; - } else { - parms.orval = 0; - parms.andval = ~(1UL << bit); - } - spin_lock(&ctl_lock); - abs_lc = get_abs_lowcore(); - ctlreg = abs_lc->cregs_save_area[cr]; - ctlreg = (ctlreg & parms.andval) | parms.orval; - abs_lc->cregs_save_area[cr] = ctlreg; - put_abs_lowcore(abs_lc); - on_each_cpu(smp_ctl_bit_callback, &parms, 1); - spin_unlock(&ctl_lock); -} -EXPORT_SYMBOL(smp_ctl_set_clear_bit); - #ifdef CONFIG_CRASH_DUMP int smp_store_status(int cpu) @@ -935,14 +888,14 @@ int __cpu_up(unsigned int cpu, struct task_struct *tidle) * Make sure global control register contents do not change * until new CPU has initialized control registers. */ - spin_lock(&ctl_lock); + system_ctlreg_lock(); pcpu_prepare_secondary(pcpu, cpu); pcpu_attach_task(pcpu, tidle); pcpu_start_fn(pcpu, smp_start_secondary, NULL); /* Wait until cpu puts itself in the online & active maps */ while (!cpu_online(cpu)) cpu_relax(); - spin_unlock(&ctl_lock); + system_ctlreg_unlock(); return 0; } @@ -957,7 +910,7 @@ early_param("possible_cpus", _setup_possible_cpus); int __cpu_disable(void) { - unsigned long cregs[16]; + struct ctlreg cregs[16]; int cpu; /* Handle possible pending IPIs */ @@ -969,11 +922,11 @@ int __cpu_disable(void) /* Disable pseudo page faults on this cpu. */ pfault_fini(); /* Disable interrupt sources via control register. */ - __ctl_store(cregs, 0, 15); - cregs[0] &= ~0x0000ee70UL; /* disable all external interrupts */ - cregs[6] &= ~0xff000000UL; /* disable all I/O interrupts */ - cregs[14] &= ~0x1f000000UL; /* disable most machine checks */ - __ctl_load(cregs, 0, 15); + __local_ctl_store(0, 15, cregs); + cregs[0].val &= ~0x0000ee70UL; /* disable all external interrupts */ + cregs[6].val &= ~0xff000000UL; /* disable all I/O interrupts */ + cregs[14].val &= ~0x1f000000UL; /* disable most machine checks */ + __local_ctl_load(0, 15, cregs); clear_cpu_flag(CIF_NOHZ_DELAY); return 0; } @@ -1013,12 +966,12 @@ void __init smp_fill_possible_mask(void) void __init smp_prepare_cpus(unsigned int max_cpus) { - /* request the 0x1201 emergency signal external interrupt */ if (register_external_irq(EXT_IRQ_EMERGENCY_SIG, do_ext_call_interrupt)) panic("Couldn't request external interrupt 0x1201"); - /* request the 0x1202 external call external interrupt */ + system_ctl_set_bit(0, 14); if (register_external_irq(EXT_IRQ_EXTERNAL_CALL, do_ext_call_interrupt)) panic("Couldn't request external interrupt 0x1202"); + system_ctl_set_bit(0, 13); } void __init smp_prepare_boot_cpu(void) @@ -1076,11 +1029,9 @@ static ssize_t cpu_configure_store(struct device *dev, cpus_read_lock(); mutex_lock(&smp_cpu_state_mutex); rc = -EBUSY; - /* disallow configuration changes of online cpus and cpu 0 */ + /* disallow configuration changes of online cpus */ cpu = dev->id; cpu = smp_get_base_cpu(cpu); - if (cpu == 0) - goto out; for (i = 0; i <= smp_cpu_mtid; i++) if (cpu_online(cpu + i)) goto out; @@ -1180,7 +1131,7 @@ static int smp_add_present_cpu(int cpu) return -ENOMEM; per_cpu(cpu_device, cpu) = c; s = &c->dev; - c->hotpluggable = 1; + c->hotpluggable = !!cpu; rc = register_cpu(c, cpu); if (rc) goto out; @@ -1258,60 +1209,3 @@ out: return rc; } subsys_initcall(s390_smp_init); - -static __always_inline void set_new_lowcore(struct lowcore *lc) -{ - union register_pair dst, src; - u32 pfx; - - src.even = (unsigned long) &S390_lowcore; - src.odd = sizeof(S390_lowcore); - dst.even = (unsigned long) lc; - dst.odd = sizeof(*lc); - pfx = __pa(lc); - - asm volatile( - " mvcl %[dst],%[src]\n" - " spx %[pfx]\n" - : [dst] "+&d" (dst.pair), [src] "+&d" (src.pair) - : [pfx] "Q" (pfx) - : "memory", "cc"); -} - -int __init smp_reinit_ipl_cpu(void) -{ - unsigned long async_stack, nodat_stack, mcck_stack; - struct lowcore *lc, *lc_ipl; - unsigned long flags, cr0; - u64 mcesad; - - lc_ipl = lowcore_ptr[0]; - lc = (struct lowcore *) __get_free_pages(GFP_KERNEL | GFP_DMA, LC_ORDER); - nodat_stack = __get_free_pages(GFP_KERNEL, THREAD_SIZE_ORDER); - async_stack = stack_alloc(); - mcck_stack = stack_alloc(); - if (!lc || !nodat_stack || !async_stack || !mcck_stack || nmi_alloc_mcesa(&mcesad)) - panic("Couldn't allocate memory"); - - local_irq_save(flags); - local_mcck_disable(); - set_new_lowcore(lc); - S390_lowcore.nodat_stack = nodat_stack + STACK_INIT_OFFSET; - S390_lowcore.async_stack = async_stack + STACK_INIT_OFFSET; - S390_lowcore.mcck_stack = mcck_stack + STACK_INIT_OFFSET; - __ctl_store(cr0, 0, 0); - __ctl_clear_bit(0, 28); /* disable lowcore protection */ - S390_lowcore.mcesad = mcesad; - __ctl_load(cr0, 0, 0); - if (abs_lowcore_map(0, lc, false)) - panic("Couldn't remap absolute lowcore"); - lowcore_ptr[0] = lc; - local_mcck_enable(); - local_irq_restore(flags); - - memblock_free_late(__pa(lc_ipl->mcck_stack - STACK_INIT_OFFSET), THREAD_SIZE); - memblock_free_late(__pa(lc_ipl->async_stack - STACK_INIT_OFFSET), THREAD_SIZE); - memblock_free_late(__pa(lc_ipl->nodat_stack - STACK_INIT_OFFSET), THREAD_SIZE); - memblock_free_late(__pa(lc_ipl), sizeof(*lc_ipl)); - return 0; -} diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index d34d3548c046..14abad953c02 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -173,10 +173,10 @@ void init_cpu_timer(void) clockevents_register_device(cd); /* Enable clock comparator timer interrupt. */ - __ctl_set_bit(0,11); + local_ctl_set_bit(0, CR0_CLOCK_COMPARATOR_SUBMASK_BIT); /* Always allow the timing alert external interrupt. */ - __ctl_set_bit(0, 4); + local_ctl_set_bit(0, CR0_ETR_SUBMASK_BIT); } static void clock_comparator_interrupt(struct ext_code ext_code, diff --git a/arch/s390/kvm/gaccess.c b/arch/s390/kvm/gaccess.c index ff8349d17b33..5bfcc50c1a68 100644 --- a/arch/s390/kvm/gaccess.c +++ b/arch/s390/kvm/gaccess.c @@ -11,7 +11,7 @@ #include <linux/err.h> #include <linux/pgtable.h> #include <linux/bitfield.h> - +#include <asm/fault.h> #include <asm/gmap.h> #include "kvm-s390.h" #include "gaccess.h" @@ -466,23 +466,6 @@ static int ar_translation(struct kvm_vcpu *vcpu, union asce *asce, u8 ar, return 0; } -struct trans_exc_code_bits { - unsigned long addr : 52; /* Translation-exception Address */ - unsigned long fsi : 2; /* Access Exception Fetch/Store Indication */ - unsigned long : 2; - unsigned long b56 : 1; - unsigned long : 3; - unsigned long b60 : 1; - unsigned long b61 : 1; - unsigned long as : 2; /* ASCE Identifier */ -}; - -enum { - FSI_UNKNOWN = 0, /* Unknown whether fetch or store */ - FSI_STORE = 1, /* Exception was due to store operation */ - FSI_FETCH = 2 /* Exception was due to fetch operation */ -}; - enum prot_type { PROT_TYPE_LA = 0, PROT_TYPE_KEYC = 1, @@ -497,11 +480,11 @@ static int trans_exc_ending(struct kvm_vcpu *vcpu, int code, unsigned long gva, enum gacc_mode mode, enum prot_type prot, bool terminate) { struct kvm_s390_pgm_info *pgm = &vcpu->arch.pgm; - struct trans_exc_code_bits *tec; + union teid *teid; memset(pgm, 0, sizeof(*pgm)); pgm->code = code; - tec = (struct trans_exc_code_bits *)&pgm->trans_exc_code; + teid = (union teid *)&pgm->trans_exc_code; switch (code) { case PGM_PROTECTION: @@ -511,25 +494,25 @@ static int trans_exc_ending(struct kvm_vcpu *vcpu, int code, unsigned long gva, WARN_ON_ONCE(1); break; case PROT_TYPE_IEP: - tec->b61 = 1; + teid->b61 = 1; fallthrough; case PROT_TYPE_LA: - tec->b56 = 1; + teid->b56 = 1; break; case PROT_TYPE_KEYC: - tec->b60 = 1; + teid->b60 = 1; break; case PROT_TYPE_ALC: - tec->b60 = 1; + teid->b60 = 1; fallthrough; case PROT_TYPE_DAT: - tec->b61 = 1; + teid->b61 = 1; break; } if (terminate) { - tec->b56 = 0; - tec->b60 = 0; - tec->b61 = 0; + teid->b56 = 0; + teid->b60 = 0; + teid->b61 = 0; } fallthrough; case PGM_ASCE_TYPE: @@ -543,9 +526,9 @@ static int trans_exc_ending(struct kvm_vcpu *vcpu, int code, unsigned long gva, * exc_access_id has to be set to 0 for some instructions. Both * cases have to be handled by the caller. */ - tec->addr = gva >> PAGE_SHIFT; - tec->fsi = mode == GACC_STORE ? FSI_STORE : FSI_FETCH; - tec->as = psw_bits(vcpu->arch.sie_block->gpsw).as; + teid->addr = gva >> PAGE_SHIFT; + teid->fsi = mode == GACC_STORE ? TEID_FSI_STORE : TEID_FSI_FETCH; + teid->as = psw_bits(vcpu->arch.sie_block->gpsw).as; fallthrough; case PGM_ALEN_TRANSLATION: case PGM_ALE_SEQUENCE: diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index 11676b81e6bf..7aa0e668488f 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -4927,7 +4927,7 @@ static void sync_regs_fmt2(struct kvm_vcpu *vcpu) } if (MACHINE_HAS_GS) { preempt_disable(); - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); if (current->thread.gs_cb) { vcpu->arch.host_gscb = current->thread.gs_cb; save_gs_cb(vcpu->arch.host_gscb); @@ -5004,13 +5004,13 @@ static void store_regs_fmt2(struct kvm_vcpu *vcpu) kvm_run->s.regs.diag318 = vcpu->arch.diag318_info.val; if (MACHINE_HAS_GS) { preempt_disable(); - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); if (vcpu->arch.gs_enabled) save_gs_cb(current->thread.gs_cb); current->thread.gs_cb = vcpu->arch.host_gscb; restore_gs_cb(vcpu->arch.host_gscb); if (!vcpu->arch.host_gscb) - __ctl_clear_bit(2, 4); + local_ctl_clear_bit(2, CR2_GUARDED_STORAGE_BIT); vcpu->arch.host_gscb = NULL; preempt_enable(); } diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c index dc4cfa8795c0..621a17fd1a1b 100644 --- a/arch/s390/kvm/priv.c +++ b/arch/s390/kvm/priv.c @@ -57,7 +57,7 @@ static int handle_gs(struct kvm_vcpu *vcpu) if (test_kvm_facility(vcpu->kvm, 133)) { VCPU_EVENT(vcpu, 3, "%s", "ENABLE: GS (lazy)"); preempt_disable(); - __ctl_set_bit(2, 4); + local_ctl_set_bit(2, CR2_GUARDED_STORAGE_BIT); current->thread.gs_cb = (struct gs_cb *)&vcpu->run->s.regs.gscb; restore_gs_cb(current->thread.gs_cb); preempt_enable(); diff --git a/arch/s390/lib/uaccess.c b/arch/s390/lib/uaccess.c index e4a13d7cab6e..61d8dcd95bbc 100644 --- a/arch/s390/lib/uaccess.c +++ b/arch/s390/lib/uaccess.c @@ -12,21 +12,22 @@ #include <linux/export.h> #include <linux/mm.h> #include <asm/asm-extable.h> +#include <asm/ctlreg.h> #ifdef CONFIG_DEBUG_ENTRY void debug_user_asce(int exit) { - unsigned long cr1, cr7; + struct ctlreg cr1, cr7; - __ctl_store(cr1, 1, 1); - __ctl_store(cr7, 7, 7); - if (cr1 == S390_lowcore.kernel_asce && cr7 == S390_lowcore.user_asce) + local_ctl_store(1, &cr1); + local_ctl_store(7, &cr7); + if (cr1.val == S390_lowcore.kernel_asce.val && cr7.val == S390_lowcore.user_asce.val) return; panic("incorrect ASCE on kernel %s\n" "cr1: %016lx cr7: %016lx\n" - "kernel: %016llx user: %016llx\n", - exit ? "exit" : "entry", cr1, cr7, - S390_lowcore.kernel_asce, S390_lowcore.user_asce); + "kernel: %016lx user: %016lx\n", + exit ? "exit" : "entry", cr1.val, cr7.val, + S390_lowcore.kernel_asce.val, S390_lowcore.user_asce.val); } #endif /*CONFIG_DEBUG_ENTRY */ diff --git a/arch/s390/mm/dump_pagetables.c b/arch/s390/mm/dump_pagetables.c index b51666967aa1..d37a8f607b71 100644 --- a/arch/s390/mm/dump_pagetables.c +++ b/arch/s390/mm/dump_pagetables.c @@ -287,7 +287,7 @@ static int pt_dump_init(void) * kernel ASCE. We need this to keep the page table walker functions * from accessing non-existent entries. */ - max_addr = (S390_lowcore.kernel_asce & _REGION_ENTRY_TYPE_MASK) >> 2; + max_addr = (S390_lowcore.kernel_asce.val & _REGION_ENTRY_TYPE_MASK) >> 2; max_addr = 1UL << (max_addr * 11 + 31); address_markers[IDENTITY_AFTER_END_NR].start_address = ident_map_size; address_markers[AMODE31_START_NR].start_address = (unsigned long)__samode31; diff --git a/arch/s390/mm/extable.c b/arch/s390/mm/extable.c index fe87291df95d..0a0738a473af 100644 --- a/arch/s390/mm/extable.c +++ b/arch/s390/mm/extable.c @@ -61,6 +61,22 @@ static bool ex_handler_ua_load_reg(const struct exception_table_entry *ex, return true; } +static bool ex_handler_zeropad(const struct exception_table_entry *ex, struct pt_regs *regs) +{ + unsigned int reg_addr = FIELD_GET(EX_DATA_REG_ADDR, ex->data); + unsigned int reg_data = FIELD_GET(EX_DATA_REG_ERR, ex->data); + unsigned long data, addr, offset; + + addr = regs->gprs[reg_addr]; + offset = addr & (sizeof(unsigned long) - 1); + addr &= ~(sizeof(unsigned long) - 1); + data = *(unsigned long *)addr; + data <<= BITS_PER_BYTE * offset; + regs->gprs[reg_data] = data; + regs->psw.addr = extable_fixup(ex); + return true; +} + bool fixup_exception(struct pt_regs *regs) { const struct exception_table_entry *ex; @@ -81,6 +97,8 @@ bool fixup_exception(struct pt_regs *regs) return ex_handler_ua_load_reg(ex, false, regs); case EX_TYPE_UA_LOAD_REGPAIR: return ex_handler_ua_load_reg(ex, true, regs); + case EX_TYPE_ZEROPAD: + return ex_handler_zeropad(ex, regs); } panic("invalid exception table entry"); } diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index b678295931c3..249aefcf7c4e 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -3,17 +3,19 @@ * S390 version * Copyright IBM Corp. 1999 * Author(s): Hartmut Penner (hp@de.ibm.com) - * Ulrich Weigand (uweigand@de.ibm.com) + * Ulrich Weigand (uweigand@de.ibm.com) * * Derived from "arch/i386/mm/fault.c" * Copyright (C) 1995 Linus Torvalds */ #include <linux/kernel_stat.h> +#include <linux/mmu_context.h> #include <linux/perf_event.h> #include <linux/signal.h> #include <linux/sched.h> #include <linux/sched/debug.h> +#include <linux/jump_label.h> #include <linux/kernel.h> #include <linux/errno.h> #include <linux/string.h> @@ -34,38 +36,27 @@ #include <linux/kfence.h> #include <asm/asm-extable.h> #include <asm/asm-offsets.h> +#include <asm/ptrace.h> +#include <asm/fault.h> #include <asm/diag.h> #include <asm/gmap.h> #include <asm/irq.h> -#include <asm/mmu_context.h> #include <asm/facility.h> #include <asm/uv.h> #include "../kernel/entry.h" -#define __FAIL_ADDR_MASK -4096L - -/* - * Allocate private vm_fault_reason from top. Please make sure it won't - * collide with vm_fault_reason. - */ -#define VM_FAULT_BADCONTEXT ((__force vm_fault_t)0x80000000) -#define VM_FAULT_BADMAP ((__force vm_fault_t)0x40000000) -#define VM_FAULT_BADACCESS ((__force vm_fault_t)0x20000000) -#define VM_FAULT_SIGNAL ((__force vm_fault_t)0x10000000) -#define VM_FAULT_PFAULT ((__force vm_fault_t)0x8000000) - enum fault_type { KERNEL_FAULT, USER_FAULT, GMAP_FAULT, }; -static unsigned long store_indication __read_mostly; +static DEFINE_STATIC_KEY_FALSE(have_store_indication); static int __init fault_init(void) { if (test_facility(75)) - store_indication = 0xc00; + static_branch_enable(&have_store_indication); return 0; } early_initcall(fault_init); @@ -75,11 +66,9 @@ early_initcall(fault_init); */ static enum fault_type get_fault_type(struct pt_regs *regs) { - unsigned long trans_exc_code; + union teid teid = { .val = regs->int_parm_long }; - trans_exc_code = regs->int_parm_long & 3; - if (likely(trans_exc_code == 0)) { - /* primary space exception */ + if (likely(teid.as == PSW_BITS_AS_PRIMARY)) { if (user_mode(regs)) return USER_FAULT; if (!IS_ENABLED(CONFIG_PGSTE)) @@ -88,83 +77,77 @@ static enum fault_type get_fault_type(struct pt_regs *regs) return GMAP_FAULT; return KERNEL_FAULT; } - if (trans_exc_code == 2) + if (teid.as == PSW_BITS_AS_SECONDARY) return USER_FAULT; - if (trans_exc_code == 1) { - /* access register mode, not used in the kernel */ + /* Access register mode, not used in the kernel */ + if (teid.as == PSW_BITS_AS_ACCREG) return USER_FAULT; - } - /* home space exception -> access via kernel ASCE */ + /* Home space -> access via kernel ASCE */ return KERNEL_FAULT; } static unsigned long get_fault_address(struct pt_regs *regs) { - unsigned long trans_exc_code = regs->int_parm_long; + union teid teid = { .val = regs->int_parm_long }; - return trans_exc_code & __FAIL_ADDR_MASK; + return teid.addr * PAGE_SIZE; } -static bool fault_is_write(struct pt_regs *regs) +static __always_inline bool fault_is_write(struct pt_regs *regs) { - unsigned long trans_exc_code = regs->int_parm_long; + union teid teid = { .val = regs->int_parm_long }; - return (trans_exc_code & store_indication) == 0x400; -} - -static int bad_address(void *p) -{ - unsigned long dummy; - - return get_kernel_nofault(dummy, (unsigned long *)p); + if (static_branch_likely(&have_store_indication)) + return teid.fsi == TEID_FSI_STORE; + return false; } static void dump_pagetable(unsigned long asce, unsigned long address) { - unsigned long *table = __va(asce & _ASCE_ORIGIN); + unsigned long entry, *table = __va(asce & _ASCE_ORIGIN); pr_alert("AS:%016lx ", asce); switch (asce & _ASCE_TYPE_MASK) { case _ASCE_TYPE_REGION1: table += (address & _REGION1_INDEX) >> _REGION1_SHIFT; - if (bad_address(table)) + if (get_kernel_nofault(entry, table)) goto bad; - pr_cont("R1:%016lx ", *table); - if (*table & _REGION_ENTRY_INVALID) + pr_cont("R1:%016lx ", entry); + if (entry & _REGION_ENTRY_INVALID) goto out; - table = __va(*table & _REGION_ENTRY_ORIGIN); + table = __va(entry & _REGION_ENTRY_ORIGIN); fallthrough; case _ASCE_TYPE_REGION2: table += (address & _REGION2_INDEX) >> _REGION2_SHIFT; - if (bad_address(table)) + if (get_kernel_nofault(entry, table)) goto bad; - pr_cont("R2:%016lx ", *table); - if (*table & _REGION_ENTRY_INVALID) + pr_cont("R2:%016lx ", entry); + if (entry & _REGION_ENTRY_INVALID) goto out; - table = __va(*table & _REGION_ENTRY_ORIGIN); + table = __va(entry & _REGION_ENTRY_ORIGIN); fallthrough; case _ASCE_TYPE_REGION3: table += (address & _REGION3_INDEX) >> _REGION3_SHIFT; - if (bad_address(table)) + if (get_kernel_nofault(entry, table)) goto bad; - pr_cont("R3:%016lx ", *table); - if (*table & (_REGION_ENTRY_INVALID | _REGION3_ENTRY_LARGE)) + pr_cont("R3:%016lx ", entry); + if (entry & (_REGION_ENTRY_INVALID | _REGION3_ENTRY_LARGE)) goto out; - table = __va(*table & _REGION_ENTRY_ORIGIN); + table = __va(entry & _REGION_ENTRY_ORIGIN); fallthrough; case _ASCE_TYPE_SEGMENT: table += (address & _SEGMENT_INDEX) >> _SEGMENT_SHIFT; - if (bad_address(table)) + if (get_kernel_nofault(entry, table)) goto bad; - pr_cont("S:%016lx ", *table); - if (*table & (_SEGMENT_ENTRY_INVALID | _SEGMENT_ENTRY_LARGE)) + pr_cont("S:%016lx ", entry); + if (entry & (_SEGMENT_ENTRY_INVALID | _SEGMENT_ENTRY_LARGE)) goto out; - table = __va(*table & _SEGMENT_ENTRY_ORIGIN); + table = __va(entry & _SEGMENT_ENTRY_ORIGIN); } table += (address & _PAGE_INDEX) >> _PAGE_SHIFT; - if (bad_address(table)) + if (get_kernel_nofault(entry, table)) goto bad; - pr_cont("P:%016lx ", *table); + pr_cont("P:%016lx ", entry); out: pr_cont("\n"); return; @@ -174,173 +157,113 @@ bad: static void dump_fault_info(struct pt_regs *regs) { + union teid teid = { .val = regs->int_parm_long }; unsigned long asce; pr_alert("Failing address: %016lx TEID: %016lx\n", - regs->int_parm_long & __FAIL_ADDR_MASK, regs->int_parm_long); + get_fault_address(regs), teid.val); pr_alert("Fault in "); - switch (regs->int_parm_long & 3) { - case 3: + switch (teid.as) { + case PSW_BITS_AS_HOME: pr_cont("home space "); break; - case 2: + case PSW_BITS_AS_SECONDARY: pr_cont("secondary space "); break; - case 1: + case PSW_BITS_AS_ACCREG: pr_cont("access register "); break; - case 0: + case PSW_BITS_AS_PRIMARY: pr_cont("primary space "); break; } pr_cont("mode while using "); switch (get_fault_type(regs)) { case USER_FAULT: - asce = S390_lowcore.user_asce; + asce = S390_lowcore.user_asce.val; pr_cont("user "); break; case GMAP_FAULT: - asce = ((struct gmap *) S390_lowcore.gmap)->asce; + asce = ((struct gmap *)S390_lowcore.gmap)->asce; pr_cont("gmap "); break; case KERNEL_FAULT: - asce = S390_lowcore.kernel_asce; + asce = S390_lowcore.kernel_asce.val; pr_cont("kernel "); break; default: unreachable(); } pr_cont("ASCE.\n"); - dump_pagetable(asce, regs->int_parm_long & __FAIL_ADDR_MASK); + dump_pagetable(asce, get_fault_address(regs)); } int show_unhandled_signals = 1; void report_user_fault(struct pt_regs *regs, long signr, int is_mm_fault) { + static DEFINE_RATELIMIT_STATE(rs, DEFAULT_RATELIMIT_INTERVAL, DEFAULT_RATELIMIT_BURST); + if ((task_pid_nr(current) > 1) && !show_unhandled_signals) return; if (!unhandled_signal(current, signr)) return; - if (!printk_ratelimit()) + if (!__ratelimit(&rs)) return; - printk(KERN_ALERT "User process fault: interruption code %04x ilc:%d ", - regs->int_code & 0xffff, regs->int_code >> 17); + pr_alert("User process fault: interruption code %04x ilc:%d ", + regs->int_code & 0xffff, regs->int_code >> 17); print_vma_addr(KERN_CONT "in ", regs->psw.addr); - printk(KERN_CONT "\n"); + pr_cont("\n"); if (is_mm_fault) dump_fault_info(regs); show_regs(regs); } -/* - * Send SIGSEGV to task. This is an external routine - * to keep the stack usage of do_page_fault small. - */ -static noinline void do_sigsegv(struct pt_regs *regs, int si_code) +static void do_sigsegv(struct pt_regs *regs, int si_code) { report_user_fault(regs, SIGSEGV, 1); - force_sig_fault(SIGSEGV, si_code, - (void __user *)(regs->int_parm_long & __FAIL_ADDR_MASK)); + force_sig_fault(SIGSEGV, si_code, (void __user *)get_fault_address(regs)); } -static noinline void do_no_context(struct pt_regs *regs, vm_fault_t fault) +static void handle_fault_error_nolock(struct pt_regs *regs, int si_code) { enum fault_type fault_type; unsigned long address; bool is_write; + if (user_mode(regs)) { + if (WARN_ON_ONCE(!si_code)) + si_code = SEGV_MAPERR; + return do_sigsegv(regs, si_code); + } if (fixup_exception(regs)) return; fault_type = get_fault_type(regs); - if ((fault_type == KERNEL_FAULT) && (fault == VM_FAULT_BADCONTEXT)) { + if (fault_type == KERNEL_FAULT) { address = get_fault_address(regs); is_write = fault_is_write(regs); if (kfence_handle_page_fault(address, is_write, regs)) return; } - /* - * Oops. The kernel tried to access some bad page. We'll have to - * terminate things with extreme prejudice. - */ if (fault_type == KERNEL_FAULT) - printk(KERN_ALERT "Unable to handle kernel pointer dereference" - " in virtual kernel address space\n"); + pr_alert("Unable to handle kernel pointer dereference in virtual kernel address space\n"); else - printk(KERN_ALERT "Unable to handle kernel paging request" - " in virtual user address space\n"); + pr_alert("Unable to handle kernel paging request in virtual user address space\n"); dump_fault_info(regs); die(regs, "Oops"); } -static noinline void do_low_address(struct pt_regs *regs) +static void handle_fault_error(struct pt_regs *regs, int si_code) { - /* Low-address protection hit in kernel mode means - NULL pointer write access in kernel mode. */ - if (regs->psw.mask & PSW_MASK_PSTATE) { - /* Low-address protection hit in user mode 'cannot happen'. */ - die (regs, "Low-address protection"); - } + struct mm_struct *mm = current->mm; - do_no_context(regs, VM_FAULT_BADACCESS); -} - -static noinline void do_sigbus(struct pt_regs *regs) -{ - /* - * Send a sigbus, regardless of whether we were in kernel - * or user mode. - */ - force_sig_fault(SIGBUS, BUS_ADRERR, - (void __user *)(regs->int_parm_long & __FAIL_ADDR_MASK)); + mmap_read_unlock(mm); + handle_fault_error_nolock(regs, si_code); } -static noinline void do_fault_error(struct pt_regs *regs, vm_fault_t fault) +static void do_sigbus(struct pt_regs *regs) { - int si_code; - - switch (fault) { - case VM_FAULT_BADACCESS: - case VM_FAULT_BADMAP: - /* Bad memory access. Check if it is kernel or user space. */ - if (user_mode(regs)) { - /* User mode accesses just cause a SIGSEGV */ - si_code = (fault == VM_FAULT_BADMAP) ? - SEGV_MAPERR : SEGV_ACCERR; - do_sigsegv(regs, si_code); - break; - } - fallthrough; - case VM_FAULT_BADCONTEXT: - case VM_FAULT_PFAULT: - do_no_context(regs, fault); - break; - case VM_FAULT_SIGNAL: - if (!user_mode(regs)) - do_no_context(regs, fault); - break; - default: /* fault & VM_FAULT_ERROR */ - if (fault & VM_FAULT_OOM) { - if (!user_mode(regs)) - do_no_context(regs, fault); - else - pagefault_out_of_memory(); - } else if (fault & VM_FAULT_SIGSEGV) { - /* Kernel mode? Handle exceptions or die */ - if (!user_mode(regs)) - do_no_context(regs, fault); - else - do_sigsegv(regs, SEGV_MAPERR); - } else if (fault & VM_FAULT_SIGBUS) { - /* Kernel mode? Handle exceptions or die */ - if (!user_mode(regs)) - do_no_context(regs, fault); - else - do_sigbus(regs); - } else - BUG(); - break; - } + force_sig_fault(SIGBUS, BUS_ADRERR, (void __user *)get_fault_address(regs)); } /* @@ -349,20 +272,20 @@ static noinline void do_fault_error(struct pt_regs *regs, vm_fault_t fault) * routines. * * interruption code (int_code): - * 04 Protection -> Write-Protection (suppression) - * 10 Segment translation -> Not present (nullification) - * 11 Page translation -> Not present (nullification) - * 3b Region third trans. -> Not present (nullification) + * 04 Protection -> Write-Protection (suppression) + * 10 Segment translation -> Not present (nullification) + * 11 Page translation -> Not present (nullification) + * 3b Region third trans. -> Not present (nullification) */ -static inline vm_fault_t do_exception(struct pt_regs *regs, int access) +static void do_exception(struct pt_regs *regs, int access) { - struct gmap *gmap; + struct vm_area_struct *vma; struct task_struct *tsk; + unsigned long address; struct mm_struct *mm; - struct vm_area_struct *vma; enum fault_type type; - unsigned long address; unsigned int flags; + struct gmap *gmap; vm_fault_t fault; bool is_write; @@ -372,31 +295,21 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access) * been nullified. Don't signal single step via SIGTRAP. */ clear_thread_flag(TIF_PER_TRAP); - if (kprobe_page_fault(regs, 14)) - return 0; - + return; mm = tsk->mm; address = get_fault_address(regs); is_write = fault_is_write(regs); - - /* - * Verify that the fault happened in user space, that - * we are not in an interrupt and that there is a - * user context. - */ - fault = VM_FAULT_BADCONTEXT; type = get_fault_type(regs); switch (type) { case KERNEL_FAULT: - goto out; + return handle_fault_error_nolock(regs, 0); case USER_FAULT: case GMAP_FAULT: if (faulthandler_disabled() || !mm) - goto out; + return handle_fault_error_nolock(regs, 0); break; } - perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address); flags = FAULT_FLAG_DEFAULT; if (user_mode(regs)) @@ -419,125 +332,117 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access) vma_end_read(vma); if (!(fault & VM_FAULT_RETRY)) { count_vm_vma_lock_event(VMA_LOCK_SUCCESS); - if (likely(!(fault & VM_FAULT_ERROR))) - fault = 0; - goto out; + if (unlikely(fault & VM_FAULT_ERROR)) + goto error; + return; } count_vm_vma_lock_event(VMA_LOCK_RETRY); /* Quick path to respond to signals */ if (fault_signal_pending(fault, regs)) { - fault = VM_FAULT_SIGNAL; - goto out; + if (!user_mode(regs)) + handle_fault_error_nolock(regs, 0); + return; } lock_mmap: mmap_read_lock(mm); - gmap = NULL; if (IS_ENABLED(CONFIG_PGSTE) && type == GMAP_FAULT) { - gmap = (struct gmap *) S390_lowcore.gmap; + gmap = (struct gmap *)S390_lowcore.gmap; current->thread.gmap_addr = address; current->thread.gmap_write_flag = !!(flags & FAULT_FLAG_WRITE); current->thread.gmap_int_code = regs->int_code & 0xffff; address = __gmap_translate(gmap, address); - if (address == -EFAULT) { - fault = VM_FAULT_BADMAP; - goto out_up; - } + if (address == -EFAULT) + return handle_fault_error(regs, SEGV_MAPERR); if (gmap->pfault_enabled) flags |= FAULT_FLAG_RETRY_NOWAIT; } - retry: - fault = VM_FAULT_BADMAP; vma = find_vma(mm, address); if (!vma) - goto out_up; - + return handle_fault_error(regs, SEGV_MAPERR); if (unlikely(vma->vm_start > address)) { if (!(vma->vm_flags & VM_GROWSDOWN)) - goto out_up; + return handle_fault_error(regs, SEGV_MAPERR); vma = expand_stack(mm, address); if (!vma) - goto out; + return handle_fault_error_nolock(regs, SEGV_MAPERR); } - - /* - * Ok, we have a good vm_area for this memory access, so - * we can handle it.. - */ - fault = VM_FAULT_BADACCESS; if (unlikely(!(vma->vm_flags & access))) - goto out_up; - - /* - * If for any reason at all we couldn't handle the fault, - * make sure we exit gracefully rather than endlessly redo - * the fault. - */ + return handle_fault_error(regs, SEGV_ACCERR); fault = handle_mm_fault(vma, address, flags, regs); if (fault_signal_pending(fault, regs)) { - fault = VM_FAULT_SIGNAL; if (flags & FAULT_FLAG_RETRY_NOWAIT) - goto out_up; - goto out; + mmap_read_unlock(mm); + if (!user_mode(regs)) + handle_fault_error_nolock(regs, 0); + return; } - /* The fault is fully completed (including releasing mmap lock) */ if (fault & VM_FAULT_COMPLETED) { if (gmap) { mmap_read_lock(mm); - goto out_gmap; + goto gmap; } - fault = 0; - goto out; + return; + } + if (unlikely(fault & VM_FAULT_ERROR)) { + mmap_read_unlock(mm); + goto error; } - - if (unlikely(fault & VM_FAULT_ERROR)) - goto out_up; - if (fault & VM_FAULT_RETRY) { - if (IS_ENABLED(CONFIG_PGSTE) && gmap && - (flags & FAULT_FLAG_RETRY_NOWAIT)) { + if (IS_ENABLED(CONFIG_PGSTE) && gmap && (flags & FAULT_FLAG_RETRY_NOWAIT)) { /* - * FAULT_FLAG_RETRY_NOWAIT has been set, mmap_lock has - * not been released + * FAULT_FLAG_RETRY_NOWAIT has been set, + * mmap_lock has not been released */ current->thread.gmap_pfault = 1; - fault = VM_FAULT_PFAULT; - goto out_up; + return handle_fault_error(regs, 0); } flags &= ~FAULT_FLAG_RETRY_NOWAIT; flags |= FAULT_FLAG_TRIED; mmap_read_lock(mm); goto retry; } -out_gmap: +gmap: if (IS_ENABLED(CONFIG_PGSTE) && gmap) { address = __gmap_link(gmap, current->thread.gmap_addr, address); - if (address == -EFAULT) { - fault = VM_FAULT_BADMAP; - goto out_up; - } + if (address == -EFAULT) + return handle_fault_error(regs, SEGV_MAPERR); if (address == -ENOMEM) { fault = VM_FAULT_OOM; - goto out_up; + mmap_read_unlock(mm); + goto error; } } - fault = 0; -out_up: mmap_read_unlock(mm); -out: - return fault; + return; +error: + if (fault & VM_FAULT_OOM) { + if (!user_mode(regs)) + handle_fault_error_nolock(regs, 0); + else + pagefault_out_of_memory(); + } else if (fault & VM_FAULT_SIGSEGV) { + if (!user_mode(regs)) + handle_fault_error_nolock(regs, 0); + else + do_sigsegv(regs, SEGV_MAPERR); + } else if (fault & VM_FAULT_SIGBUS) { + if (!user_mode(regs)) + handle_fault_error_nolock(regs, 0); + else + do_sigbus(regs); + } else { + BUG(); + } } void do_protection_exception(struct pt_regs *regs) { - unsigned long trans_exc_code; - int access; - vm_fault_t fault; + union teid teid = { .val = regs->int_parm_long }; - trans_exc_code = regs->int_parm_long; /* * Protection exceptions are suppressing, decrement psw address. * The exception to this rule are aborted transactions, for these @@ -550,33 +455,28 @@ void do_protection_exception(struct pt_regs *regs) * as a special case because the translation exception code * field is not guaranteed to contain valid data in this case. */ - if (unlikely(!(trans_exc_code & 4))) { - do_low_address(regs); - return; + if (unlikely(!teid.b61)) { + if (user_mode(regs)) { + /* Low-address protection in user mode: cannot happen */ + die(regs, "Low-address protection"); + } + /* + * Low-address protection in kernel mode means + * NULL pointer write access in kernel mode. + */ + return handle_fault_error_nolock(regs, 0); } - if (unlikely(MACHINE_HAS_NX && (trans_exc_code & 0x80))) { - regs->int_parm_long = (trans_exc_code & ~PAGE_MASK) | - (regs->psw.addr & PAGE_MASK); - access = VM_EXEC; - fault = VM_FAULT_BADACCESS; - } else { - access = VM_WRITE; - fault = do_exception(regs, access); + if (unlikely(MACHINE_HAS_NX && teid.b56)) { + regs->int_parm_long = (teid.addr * PAGE_SIZE) | (regs->psw.addr & PAGE_MASK); + return handle_fault_error_nolock(regs, SEGV_ACCERR); } - if (unlikely(fault)) - do_fault_error(regs, fault); + do_exception(regs, VM_WRITE); } NOKPROBE_SYMBOL(do_protection_exception); void do_dat_exception(struct pt_regs *regs) { - int access; - vm_fault_t fault; - - access = VM_ACCESS_FLAGS; - fault = do_exception(regs, access); - if (unlikely(fault)) - do_fault_error(regs, fault); + do_exception(regs, VM_ACCESS_FLAGS); } NOKPROBE_SYMBOL(do_dat_exception); @@ -584,7 +484,8 @@ NOKPROBE_SYMBOL(do_dat_exception); void do_secure_storage_access(struct pt_regs *regs) { - unsigned long addr = regs->int_parm_long & __FAIL_ADDR_MASK; + union teid teid = { .val = regs->int_parm_long }; + unsigned long addr = get_fault_address(regs); struct vm_area_struct *vma; struct mm_struct *mm; struct page *page; @@ -592,14 +493,12 @@ void do_secure_storage_access(struct pt_regs *regs) int rc; /* - * bit 61 tells us if the address is valid, if it's not we - * have a major problem and should stop the kernel or send a - * SIGSEGV to the process. Unfortunately bit 61 is not - * reliable without the misc UV feature so we need to check - * for that as well. + * Bit 61 indicates if the address is valid, if it is not the + * kernel should be stopped or SIGSEGV should be sent to the + * process. Bit 61 is not reliable without the misc UV feature, + * therefore this needs to be checked too. */ - if (uv_has_feature(BIT_UV_FEAT_MISC) && - !test_bit_inv(61, ®s->int_parm_long)) { + if (uv_has_feature(BIT_UV_FEAT_MISC) && !teid.b61) { /* * When this happens, userspace did something that it * was not supposed to do, e.g. branching into secure @@ -609,14 +508,12 @@ void do_secure_storage_access(struct pt_regs *regs) send_sig(SIGSEGV, current, 0); return; } - /* - * The kernel should never run into this case and we - * have no way out of this situation. + * The kernel should never run into this case and + * there is no way out of this situation. */ panic("Unexpected PGM 0x3d with TEID bit 61=0"); } - switch (get_fault_type(regs)) { case GMAP_FAULT: mm = current->mm; @@ -624,20 +521,15 @@ void do_secure_storage_access(struct pt_regs *regs) mmap_read_lock(mm); addr = __gmap_translate(gmap, addr); mmap_read_unlock(mm); - if (IS_ERR_VALUE(addr)) { - do_fault_error(regs, VM_FAULT_BADMAP); - break; - } + if (IS_ERR_VALUE(addr)) + return handle_fault_error_nolock(regs, SEGV_MAPERR); fallthrough; case USER_FAULT: mm = current->mm; mmap_read_lock(mm); vma = find_vma(mm, addr); - if (!vma) { - mmap_read_unlock(mm); - do_fault_error(regs, VM_FAULT_BADMAP); - break; - } + if (!vma) + return handle_fault_error(regs, SEGV_MAPERR); page = follow_page(vma, addr, FOLL_WRITE | FOLL_GET); if (IS_ERR_OR_NULL(page)) { mmap_read_unlock(mm); @@ -658,23 +550,18 @@ void do_secure_storage_access(struct pt_regs *regs) BUG(); break; default: - do_fault_error(regs, VM_FAULT_BADMAP); - WARN_ON_ONCE(1); + unreachable(); } } NOKPROBE_SYMBOL(do_secure_storage_access); void do_non_secure_storage_access(struct pt_regs *regs) { - unsigned long gaddr = regs->int_parm_long & __FAIL_ADDR_MASK; struct gmap *gmap = (struct gmap *)S390_lowcore.gmap; + unsigned long gaddr = get_fault_address(regs); - if (get_fault_type(regs) != GMAP_FAULT) { - do_fault_error(regs, VM_FAULT_BADMAP); - WARN_ON_ONCE(1); - return; - } - + if (WARN_ON_ONCE(get_fault_type(regs) != GMAP_FAULT)) + return handle_fault_error_nolock(regs, SEGV_MAPERR); if (gmap_convert_to_secure(gmap, gaddr) == -EINVAL) send_sig(SIGSEGV, current, 0); } @@ -682,8 +569,8 @@ NOKPROBE_SYMBOL(do_non_secure_storage_access); void do_secure_storage_violation(struct pt_regs *regs) { - unsigned long gaddr = regs->int_parm_long & __FAIL_ADDR_MASK; struct gmap *gmap = (struct gmap *)S390_lowcore.gmap; + unsigned long gaddr = get_fault_address(regs); /* * If the VM has been rebooted, its address space might still contain @@ -699,9 +586,8 @@ void do_secure_storage_violation(struct pt_regs *regs) * This exception is only triggered when a guest 2 is running * and can therefore never occur in kernel context. */ - printk_ratelimited(KERN_WARNING - "Secure storage violation in task: %s, pid %d\n", - current->comm, current->pid); + pr_warn_ratelimited("Secure storage violation in task: %s, pid %d\n", + current->comm, current->pid); send_sig(SIGSEGV, current, 0); } diff --git a/arch/s390/mm/gmap.c b/arch/s390/mm/gmap.c index 906a7bfc2a78..20786f6883b2 100644 --- a/arch/s390/mm/gmap.c +++ b/arch/s390/mm/gmap.c @@ -21,10 +21,22 @@ #include <asm/pgalloc.h> #include <asm/gmap.h> +#include <asm/page.h> #include <asm/tlb.h> #define GMAP_SHADOW_FAKE_TABLE 1ULL +static struct page *gmap_alloc_crst(void) +{ + struct page *page; + + page = alloc_pages(GFP_KERNEL_ACCOUNT, CRST_ALLOC_ORDER); + if (!page) + return NULL; + arch_set_page_dat(page, CRST_ALLOC_ORDER); + return page; +} + /** * gmap_alloc - allocate and initialize a guest address space * @limit: maximum address of the gmap address space @@ -67,7 +79,7 @@ static struct gmap *gmap_alloc(unsigned long limit) spin_lock_init(&gmap->guest_table_lock); spin_lock_init(&gmap->shadow_lock); refcount_set(&gmap->ref_count, 1); - page = alloc_pages(GFP_KERNEL_ACCOUNT, CRST_ALLOC_ORDER); + page = gmap_alloc_crst(); if (!page) goto out_free; page->index = 0; @@ -308,7 +320,7 @@ static int gmap_alloc_table(struct gmap *gmap, unsigned long *table, unsigned long *new; /* since we dont free the gmap table until gmap_free we can unlock */ - page = alloc_pages(GFP_KERNEL_ACCOUNT, CRST_ALLOC_ORDER); + page = gmap_alloc_crst(); if (!page) return -ENOMEM; new = page_to_virt(page); @@ -1759,7 +1771,7 @@ int gmap_shadow_r2t(struct gmap *sg, unsigned long saddr, unsigned long r2t, BUG_ON(!gmap_is_shadow(sg)); /* Allocate a shadow region second table */ - page = alloc_pages(GFP_KERNEL_ACCOUNT, CRST_ALLOC_ORDER); + page = gmap_alloc_crst(); if (!page) return -ENOMEM; page->index = r2t & _REGION_ENTRY_ORIGIN; @@ -1843,7 +1855,7 @@ int gmap_shadow_r3t(struct gmap *sg, unsigned long saddr, unsigned long r3t, BUG_ON(!gmap_is_shadow(sg)); /* Allocate a shadow region second table */ - page = alloc_pages(GFP_KERNEL_ACCOUNT, CRST_ALLOC_ORDER); + page = gmap_alloc_crst(); if (!page) return -ENOMEM; page->index = r3t & _REGION_ENTRY_ORIGIN; @@ -1927,7 +1939,7 @@ int gmap_shadow_sgt(struct gmap *sg, unsigned long saddr, unsigned long sgt, BUG_ON(!gmap_is_shadow(sg) || (sgt & _REGION3_ENTRY_LARGE)); /* Allocate a shadow segment table */ - page = alloc_pages(GFP_KERNEL_ACCOUNT, CRST_ALLOC_ORDER); + page = gmap_alloc_crst(); if (!page) return -ENOMEM; page->index = sgt & _REGION_ENTRY_ORIGIN; @@ -2855,7 +2867,7 @@ int s390_replace_asce(struct gmap *gmap) if ((gmap->asce & _ASCE_TYPE_MASK) == _ASCE_TYPE_SEGMENT) return -EINVAL; - page = alloc_pages(GFP_KERNEL_ACCOUNT, CRST_ALLOC_ORDER); + page = gmap_alloc_crst(); if (!page) return -ENOMEM; page->index = 0; diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index 8b94d2212d33..7eca10c32caa 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -35,6 +35,7 @@ #include <asm/processor.h> #include <linux/uaccess.h> #include <asm/pgalloc.h> +#include <asm/ctlreg.h> #include <asm/kfence.h> #include <asm/ptdump.h> #include <asm/dma.h> @@ -42,7 +43,6 @@ #include <asm/tlb.h> #include <asm/tlbflush.h> #include <asm/sections.h> -#include <asm/ctl_reg.h> #include <asm/sclp.h> #include <asm/set_memory.h> #include <asm/kasan.h> @@ -54,7 +54,7 @@ pgd_t swapper_pg_dir[PTRS_PER_PGD] __section(".bss..swapper_pg_dir"); pgd_t invalid_pg_dir[PTRS_PER_PGD] __section(".bss..invalid_pg_dir"); -unsigned long __bootdata_preserved(s390_invalid_asce); +struct ctlreg __bootdata_preserved(s390_invalid_asce); unsigned long empty_zero_page, zero_page_mask; EXPORT_SYMBOL(empty_zero_page); diff --git a/arch/s390/mm/maccess.c b/arch/s390/mm/maccess.c index c805b3e2592b..632c3a55feed 100644 --- a/arch/s390/mm/maccess.c +++ b/arch/s390/mm/maccess.c @@ -15,10 +15,10 @@ #include <linux/uio.h> #include <linux/io.h> #include <asm/asm-extable.h> -#include <asm/ctl_reg.h> #include <asm/abs_lowcore.h> #include <asm/stacktrace.h> #include <asm/maccess.h> +#include <asm/ctlreg.h> unsigned long __bootdata_preserved(__memcpy_real_area); pte_t *__bootdata_preserved(memcpy_real_ptep); diff --git a/arch/s390/mm/page-states.c b/arch/s390/mm/page-states.c index 1e2ea706aa22..a31acb2c4ef2 100644 --- a/arch/s390/mm/page-states.c +++ b/arch/s390/mm/page-states.c @@ -121,7 +121,7 @@ static void mark_kernel_pud(p4d_t *p4d, unsigned long addr, unsigned long end) continue; if (!pud_folded(*pud)) { page = phys_to_page(pud_val(*pud)); - for (i = 0; i < 3; i++) + for (i = 0; i < 4; i++) set_bit(PG_arch_1, &page[i].flags); } mark_kernel_pmd(pud, addr, next); @@ -142,7 +142,7 @@ static void mark_kernel_p4d(pgd_t *pgd, unsigned long addr, unsigned long end) continue; if (!p4d_folded(*p4d)) { page = phys_to_page(p4d_val(*p4d)); - for (i = 0; i < 3; i++) + for (i = 0; i < 4; i++) set_bit(PG_arch_1, &page[i].flags); } mark_kernel_pud(p4d, addr, next); @@ -151,24 +151,31 @@ static void mark_kernel_p4d(pgd_t *pgd, unsigned long addr, unsigned long end) static void mark_kernel_pgd(void) { - unsigned long addr, next; + unsigned long addr, next, max_addr; struct page *page; pgd_t *pgd; int i; addr = 0; + /* + * Figure out maximum virtual address accessible with the + * kernel ASCE. This is required to keep the page table walker + * from accessing non-existent entries. + */ + max_addr = (S390_lowcore.kernel_asce.val & _ASCE_TYPE_MASK) >> 2; + max_addr = 1UL << (max_addr * 11 + 31); pgd = pgd_offset_k(addr); do { - next = pgd_addr_end(addr, MODULES_END); + next = pgd_addr_end(addr, max_addr); if (pgd_none(*pgd)) continue; if (!pgd_folded(*pgd)) { page = phys_to_page(pgd_val(*pgd)); - for (i = 0; i < 3; i++) + for (i = 0; i < 4; i++) set_bit(PG_arch_1, &page[i].flags); } mark_kernel_p4d(pgd, addr, next); - } while (pgd++, addr = next, addr != MODULES_END); + } while (pgd++, addr = next, addr != max_addr); } void __init cmma_init_nodat(void) @@ -181,6 +188,12 @@ void __init cmma_init_nodat(void) return; /* Mark pages used in kernel page tables */ mark_kernel_pgd(); + page = virt_to_page(&swapper_pg_dir); + for (i = 0; i < 4; i++) + set_bit(PG_arch_1, &page[i].flags); + page = virt_to_page(&invalid_pg_dir); + for (i = 0; i < 4; i++) + set_bit(PG_arch_1, &page[i].flags); /* Set all kernel pages not used for page tables to stable/no-dat */ for_each_mem_pfn_range(i, MAX_NUMNODES, &start, &end, NULL) { diff --git a/arch/s390/mm/pageattr.c b/arch/s390/mm/pageattr.c index b87e96c64b61..631e3a4ee2de 100644 --- a/arch/s390/mm/pageattr.c +++ b/arch/s390/mm/pageattr.c @@ -75,7 +75,7 @@ static void pgt_set(unsigned long *old, unsigned long new, unsigned long addr, break; } table = (unsigned long *)((unsigned long)old & mask); - crdte(*old, new, table, dtt, addr, S390_lowcore.kernel_asce); + crdte(*old, new, table, dtt, addr, S390_lowcore.kernel_asce.val); } else if (MACHINE_HAS_IDTE) { cspg(old, *old, new); } else { diff --git a/arch/s390/mm/pgalloc.c b/arch/s390/mm/pgalloc.c index 75e1039f2ec5..5488ae17318e 100644 --- a/arch/s390/mm/pgalloc.c +++ b/arch/s390/mm/pgalloc.c @@ -61,8 +61,8 @@ static void __crst_table_upgrade(void *arg) /* change all active ASCEs to avoid the creation of new TLBs */ if (current->active_mm == mm) { - S390_lowcore.user_asce = mm->context.asce; - __ctl_load(S390_lowcore.user_asce, 7, 7); + S390_lowcore.user_asce.val = mm->context.asce; + local_ctl_load(7, &S390_lowcore.user_asce); } __tlb_flush_local(); } @@ -145,6 +145,7 @@ struct page *page_table_alloc_pgste(struct mm_struct *mm) ptdesc = pagetable_alloc(GFP_KERNEL, 0); if (ptdesc) { table = (u64 *)ptdesc_to_virt(ptdesc); + arch_set_page_dat(virt_to_page(table), 0); memset64(table, _PAGE_INVALID, PTRS_PER_PTE); memset64(table + PTRS_PER_PTE, 0, PTRS_PER_PTE); } @@ -487,11 +488,10 @@ static unsigned long *base_crst_alloc(unsigned long val) unsigned long *table; struct ptdesc *ptdesc; - ptdesc = pagetable_alloc(GFP_KERNEL & ~__GFP_HIGHMEM, CRST_ALLOC_ORDER); + ptdesc = pagetable_alloc(GFP_KERNEL, CRST_ALLOC_ORDER); if (!ptdesc) return NULL; table = ptdesc_address(ptdesc); - crst_table_init(table, val); return table; } diff --git a/arch/s390/mm/vmem.c b/arch/s390/mm/vmem.c index 6957d2ed97bf..2e8a1064f103 100644 --- a/arch/s390/mm/vmem.c +++ b/arch/s390/mm/vmem.c @@ -12,8 +12,10 @@ #include <linux/hugetlb.h> #include <linux/slab.h> #include <linux/sort.h> +#include <asm/page-states.h> #include <asm/cacheflush.h> #include <asm/nospec-branch.h> +#include <asm/ctlreg.h> #include <asm/pgalloc.h> #include <asm/setup.h> #include <asm/tlbflush.h> @@ -45,8 +47,11 @@ void *vmem_crst_alloc(unsigned long val) unsigned long *table; table = vmem_alloc_pages(CRST_ALLOC_ORDER); - if (table) - crst_table_init(table, val); + if (!table) + return NULL; + crst_table_init(table, val); + if (slab_is_available()) + arch_set_page_dat(virt_to_page(table), CRST_ALLOC_ORDER); return table; } @@ -498,6 +503,8 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node, return ret; } +#ifdef CONFIG_MEMORY_HOTPLUG + void vmemmap_free(unsigned long start, unsigned long end, struct vmem_altmap *altmap) { @@ -506,6 +513,8 @@ void vmemmap_free(unsigned long start, unsigned long end, mutex_unlock(&vmem_mutex); } +#endif + void vmem_remove_mapping(unsigned long start, unsigned long size) { mutex_lock(&vmem_mutex); @@ -659,7 +668,7 @@ void __init vmem_map_init(void) __set_memory_4k(__va(0), RELOC_HIDE(__va(0), ident_map_size)); } if (MACHINE_HAS_NX) - ctl_set_bit(0, 20); + system_ctl_set_bit(0, CR0_INSTRUCTION_EXEC_PROTECTION_BIT); pr_info("Write protected kernel read-only data: %luk\n", (unsigned long)(__end_rodata - _stext) >> 10); } diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c index d34d5813d006..6fab5c085565 100644 --- a/arch/s390/pci/pci.c +++ b/arch/s390/pci/pci.c @@ -1094,7 +1094,7 @@ static int __init pci_base_init(void) if (MACHINE_HAS_PCI_MIO) { static_branch_enable(&have_mio); - ctl_set_bit(2, 5); + system_ctl_set_bit(2, CR2_MIO_ADDRESSING_BIT); } rc = zpci_debug_init(); |