diff options
Diffstat (limited to 'arch/s390/kernel')
-rw-r--r-- | arch/s390/kernel/Makefile | 65 | ||||
-rw-r--r-- | arch/s390/kernel/asm-offsets.c | 17 | ||||
-rw-r--r-- | arch/s390/kernel/compat_signal.c | 4 | ||||
-rw-r--r-- | arch/s390/kernel/early.c | 50 | ||||
-rw-r--r-- | arch/s390/kernel/entry.S | 51 | ||||
-rw-r--r-- | arch/s390/kernel/head.S | 2 | ||||
-rw-r--r-- | arch/s390/kernel/head64.S | 7 | ||||
-rw-r--r-- | arch/s390/kernel/ipl.c | 7 | ||||
-rw-r--r-- | arch/s390/kernel/irq.c | 2 | ||||
-rw-r--r-- | arch/s390/kernel/lgr.c | 5 | ||||
-rw-r--r-- | arch/s390/kernel/perf_cpum_sf.c | 53 | ||||
-rw-r--r-- | arch/s390/kernel/process.c | 6 | ||||
-rw-r--r-- | arch/s390/kernel/ptrace.c | 14 | ||||
-rw-r--r-- | arch/s390/kernel/setup.c | 22 | ||||
-rw-r--r-- | arch/s390/kernel/signal.c | 14 | ||||
-rw-r--r-- | arch/s390/kernel/smp.c | 32 | ||||
-rw-r--r-- | arch/s390/kernel/swsusp.S | 2 | ||||
-rw-r--r-- | arch/s390/kernel/sysinfo.c | 33 | ||||
-rw-r--r-- | arch/s390/kernel/time.c | 191 | ||||
-rw-r--r-- | arch/s390/kernel/topology.c | 53 | ||||
-rw-r--r-- | arch/s390/kernel/vdso32/clock_gettime.S | 23 | ||||
-rw-r--r-- | arch/s390/kernel/vdso32/gettimeofday.S | 23 | ||||
-rw-r--r-- | arch/s390/kernel/vdso64/clock_gettime.S | 11 | ||||
-rw-r--r-- | arch/s390/kernel/vdso64/gettimeofday.S | 11 | ||||
-rw-r--r-- | arch/s390/kernel/vtime.c | 26 |
25 files changed, 435 insertions, 289 deletions
diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index 1f0fe98f6db9..36b5101c8606 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -2,20 +2,47 @@ # Makefile for the linux kernel. # -KCOV_INSTRUMENT_early.o := n -KCOV_INSTRUMENT_sclp.o := n -KCOV_INSTRUMENT_als.o := n - ifdef CONFIG_FUNCTION_TRACER -# Don't trace early setup code and tracing code -CFLAGS_REMOVE_early.o = $(CC_FLAGS_FTRACE) -CFLAGS_REMOVE_ftrace.o = $(CC_FLAGS_FTRACE) + +# Do not trace tracer code +CFLAGS_REMOVE_ftrace.o = $(CC_FLAGS_FTRACE) + +# Do not trace early setup code +CFLAGS_REMOVE_als.o = $(CC_FLAGS_FTRACE) +CFLAGS_REMOVE_early.o = $(CC_FLAGS_FTRACE) +CFLAGS_REMOVE_sclp.o = $(CC_FLAGS_FTRACE) + +endif + +GCOV_PROFILE_als.o := n +GCOV_PROFILE_early.o := n +GCOV_PROFILE_sclp.o := n + +KCOV_INSTRUMENT_als.o := n +KCOV_INSTRUMENT_early.o := n +KCOV_INSTRUMENT_sclp.o := n + +UBSAN_SANITIZE_als.o := n +UBSAN_SANITIZE_early.o := n +UBSAN_SANITIZE_sclp.o := n + +# +# Use -march=z900 for sclp.c and als.c to be able to print an error +# message if the kernel is started on a machine which is too old +# +ifneq ($(CC_FLAGS_MARCH),-march=z900) +CFLAGS_REMOVE_als.o += $(CC_FLAGS_MARCH) +CFLAGS_als.o += -march=z900 +CFLAGS_REMOVE_sclp.o += $(CC_FLAGS_MARCH) +CFLAGS_sclp.o += -march=z900 +AFLAGS_REMOVE_head.o += $(CC_FLAGS_MARCH) +AFLAGS_head.o += -march=z900 endif # # Passing null pointers is ok for smp code, since we access the lowcore here. # -CFLAGS_smp.o := -Wno-nonnull +CFLAGS_smp.o := -Wno-nonnull # # Disable tailcall optimizations for stack / callchain walking functions @@ -30,27 +57,7 @@ CFLAGS_dumpstack.o += -fno-optimize-sibling-calls # CFLAGS_ptrace.o += -DUTS_MACHINE='"$(UTS_MACHINE)"' -CFLAGS_sysinfo.o += -w - -# -# Use -march=z900 for sclp.c and als.c to be able to print an error -# message if the kernel is started on a machine which is too old -# -CFLAGS_REMOVE_sclp.o = $(CC_FLAGS_FTRACE) -CFLAGS_REMOVE_als.o = $(CC_FLAGS_FTRACE) -ifneq ($(CC_FLAGS_MARCH),-march=z900) -CFLAGS_REMOVE_sclp.o += $(CC_FLAGS_MARCH) -CFLAGS_sclp.o += -march=z900 -CFLAGS_REMOVE_als.o += $(CC_FLAGS_MARCH) -CFLAGS_als.o += -march=z900 -AFLAGS_REMOVE_head.o += $(CC_FLAGS_MARCH) -AFLAGS_head.o += -march=z900 -endif -GCOV_PROFILE_sclp.o := n -GCOV_PROFILE_als.o := n -UBSAN_SANITIZE_als.o := n -UBSAN_SANITIZE_early.o := n -UBSAN_SANITIZE_sclp.o := n +CFLAGS_sysinfo.o += -w obj-y := traps.o time.o process.o base.o early.o setup.o idle.o vtime.o obj-y += processor.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c index f3df9e0a5dec..c4b3570ded5b 100644 --- a/arch/s390/kernel/asm-offsets.c +++ b/arch/s390/kernel/asm-offsets.c @@ -25,12 +25,14 @@ int main(void) { /* task struct offsets */ - OFFSET(__TASK_thread_info, task_struct, stack); + OFFSET(__TASK_stack, task_struct, stack); OFFSET(__TASK_thread, task_struct, thread); OFFSET(__TASK_pid, task_struct, pid); BLANK(); /* thread struct offsets */ OFFSET(__THREAD_ksp, thread_struct, ksp); + OFFSET(__THREAD_sysc_table, thread_struct, sys_call_table); + OFFSET(__THREAD_last_break, thread_struct, last_break); OFFSET(__THREAD_FPU_fpc, thread_struct, fpu.fpc); OFFSET(__THREAD_FPU_regs, thread_struct, fpu.regs); OFFSET(__THREAD_per_cause, thread_struct, per_event.cause); @@ -39,14 +41,7 @@ int main(void) OFFSET(__THREAD_trap_tdb, thread_struct, trap_tdb); BLANK(); /* thread info offsets */ - OFFSET(__TI_task, thread_info, task); - OFFSET(__TI_flags, thread_info, flags); - OFFSET(__TI_sysc_table, thread_info, sys_call_table); - OFFSET(__TI_cpu, thread_info, cpu); - OFFSET(__TI_precount, thread_info, preempt_count); - OFFSET(__TI_user_timer, thread_info, user_timer); - OFFSET(__TI_system_timer, thread_info, system_timer); - OFFSET(__TI_last_break, thread_info, last_break); + OFFSET(__TI_flags, task_struct, thread_info.flags); BLANK(); /* pt_regs offsets */ OFFSET(__PT_ARGS, pt_regs, args); @@ -79,6 +74,8 @@ int main(void) OFFSET(__VDSO_ECTG_OK, vdso_data, ectg_available); OFFSET(__VDSO_TK_MULT, vdso_data, tk_mult); OFFSET(__VDSO_TK_SHIFT, vdso_data, tk_shift); + OFFSET(__VDSO_TS_DIR, vdso_data, ts_dir); + OFFSET(__VDSO_TS_END, vdso_data, ts_end); OFFSET(__VDSO_ECTG_BASE, vdso_per_cpu_data, ectg_timer_base); OFFSET(__VDSO_ECTG_USER, vdso_per_cpu_data, ectg_user_time); OFFSET(__VDSO_CPU_NR, vdso_per_cpu_data, cpu_nr); @@ -159,7 +156,6 @@ int main(void) OFFSET(__LC_INT_CLOCK, lowcore, int_clock); OFFSET(__LC_MCCK_CLOCK, lowcore, mcck_clock); OFFSET(__LC_CURRENT, lowcore, current_task); - OFFSET(__LC_THREAD_INFO, lowcore, thread_info); OFFSET(__LC_KERNEL_STACK, lowcore, kernel_stack); OFFSET(__LC_ASYNC_STACK, lowcore, async_stack); OFFSET(__LC_PANIC_STACK, lowcore, panic_stack); @@ -173,6 +169,7 @@ int main(void) OFFSET(__LC_PERCPU_OFFSET, lowcore, percpu_offset); OFFSET(__LC_VDSO_PER_CPU, lowcore, vdso_per_cpu_data); OFFSET(__LC_MACHINE_FLAGS, lowcore, machine_flags); + OFFSET(__LC_PREEMPT_COUNT, lowcore, preempt_count); OFFSET(__LC_GMAP, lowcore, gmap); OFFSET(__LC_PASTE, lowcore, paste); /* software defined ABI-relevant lowcore locations 0xe00 - 0xe20 */ diff --git a/arch/s390/kernel/compat_signal.c b/arch/s390/kernel/compat_signal.c index 4af60374eba0..6f2a6ab13cb5 100644 --- a/arch/s390/kernel/compat_signal.c +++ b/arch/s390/kernel/compat_signal.c @@ -446,7 +446,7 @@ static int setup_frame32(struct ksignal *ksig, sigset_t *set, /* set extra registers only for synchronous signals */ regs->gprs[4] = regs->int_code & 127; regs->gprs[5] = regs->int_parm_long; - regs->gprs[6] = task_thread_info(current)->last_break; + regs->gprs[6] = current->thread.last_break; } return 0; @@ -523,7 +523,7 @@ static int setup_rt_frame32(struct ksignal *ksig, sigset_t *set, regs->gprs[2] = ksig->sig; regs->gprs[3] = (__force __u64) &frame->info; regs->gprs[4] = (__force __u64) &frame->uc; - regs->gprs[5] = task_thread_info(current)->last_break; + regs->gprs[5] = current->thread.last_break; return 0; } diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 2374c5b46bbc..d038c8cea6cb 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -293,6 +293,7 @@ static noinline __init void setup_lowcore_early(void) psw.addr = (unsigned long) s390_base_pgm_handler; S390_lowcore.program_new_psw = psw; s390_base_pgm_handler_fn = early_pgm_check_handler; + S390_lowcore.preempt_count = INIT_PREEMPT_COUNT; } static noinline __init void setup_facility_list(void) @@ -391,7 +392,49 @@ static int __init cad_init(void) } early_initcall(cad_init); -static __init void rescue_initrd(void) +static __init void memmove_early(void *dst, const void *src, size_t n) +{ + unsigned long addr; + long incr; + psw_t old; + + if (!n) + return; + incr = 1; + if (dst > src) { + incr = -incr; + dst += n - 1; + src += n - 1; + } + old = S390_lowcore.program_new_psw; + S390_lowcore.program_new_psw.mask = __extract_psw(); + asm volatile( + " larl %[addr],1f\n" + " stg %[addr],%[psw_pgm_addr]\n" + "0: mvc 0(1,%[dst]),0(%[src])\n" + " agr %[dst],%[incr]\n" + " agr %[src],%[incr]\n" + " brctg %[n],0b\n" + "1:\n" + : [addr] "=&d" (addr), + [psw_pgm_addr] "=&Q" (S390_lowcore.program_new_psw.addr), + [dst] "+&a" (dst), [src] "+&a" (src), [n] "+d" (n) + : [incr] "d" (incr) + : "cc", "memory"); + S390_lowcore.program_new_psw = old; +} + +static __init noinline void ipl_save_parameters(void) +{ + void *src, *dst; + + src = (void *)(unsigned long) S390_lowcore.ipl_parmblock_ptr; + dst = (void *) IPL_PARMBLOCK_ORIGIN; + memmove_early(dst, src, PAGE_SIZE); + S390_lowcore.ipl_parmblock_ptr = IPL_PARMBLOCK_ORIGIN; +} + +static __init noinline void rescue_initrd(void) { #ifdef CONFIG_BLK_DEV_INITRD unsigned long min_initrd_addr = (unsigned long) _end + (4UL << 20); @@ -405,7 +448,7 @@ static __init void rescue_initrd(void) return; if (INITRD_START >= min_initrd_addr) return; - memmove((void *) min_initrd_addr, (void *) INITRD_START, INITRD_SIZE); + memmove_early((void *) min_initrd_addr, (void *) INITRD_START, INITRD_SIZE); INITRD_START = min_initrd_addr; #endif } @@ -467,7 +510,8 @@ void __init startup_init(void) ipl_save_parameters(); rescue_initrd(); clear_bss_section(); - ptff_init(); + ipl_verify_parameters(); + time_early_init(); init_kernel_storage_key(); lockdep_off(); setup_lowcore_early(); diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S index 49a30737adde..97298c58b2be 100644 --- a/arch/s390/kernel/entry.S +++ b/arch/s390/kernel/entry.S @@ -42,7 +42,7 @@ __PT_R13 = __PT_GPRS + 104 __PT_R14 = __PT_GPRS + 112 __PT_R15 = __PT_GPRS + 120 -STACK_SHIFT = PAGE_SHIFT + THREAD_ORDER +STACK_SHIFT = PAGE_SHIFT + THREAD_SIZE_ORDER STACK_SIZE = 1 << STACK_SHIFT STACK_INIT = STACK_SIZE - STACK_FRAME_OVERHEAD - __PT_SIZE @@ -123,8 +123,14 @@ _PIF_WORK = (_PIF_PER_TRAP) .macro LAST_BREAK scratch srag \scratch,%r10,23 +#ifdef CONFIG_HAVE_MARCH_Z990_FEATURES jz .+10 - stg %r10,__TI_last_break(%r12) + stg %r10,__TASK_thread+__THREAD_last_break(%r12) +#else + jz .+14 + lghi \scratch,__TASK_thread + stg %r10,__THREAD_last_break(\scratch,%r12) +#endif .endm .macro REENABLE_IRQS @@ -186,14 +192,13 @@ ENTRY(__switch_to) stmg %r6,%r15,__SF_GPRS(%r15) # store gprs of prev task lgr %r1,%r2 aghi %r1,__TASK_thread # thread_struct of prev task - lg %r5,__TASK_thread_info(%r3) # get thread_info of next + lg %r5,__TASK_stack(%r3) # start of kernel stack of next stg %r15,__THREAD_ksp(%r1) # store kernel stack of prev lgr %r1,%r3 aghi %r1,__TASK_thread # thread_struct of next task lgr %r15,%r5 aghi %r15,STACK_INIT # end of kernel stack of next stg %r3,__LC_CURRENT # store task struct of next - stg %r5,__LC_THREAD_INFO # store thread info of next stg %r15,__LC_KERNEL_STACK # store end of kernel stack lg %r15,__THREAD_ksp(%r1) # load kernel stack of next /* c4 is used in guest detection: arch/s390/kernel/perf_cpum_sf.c */ @@ -274,7 +279,7 @@ ENTRY(system_call) .Lsysc_stmg: stmg %r8,%r15,__LC_SAVE_AREA_SYNC lg %r10,__LC_LAST_BREAK - lg %r12,__LC_THREAD_INFO + lg %r12,__LC_CURRENT lghi %r14,_PIF_SYSCALL .Lsysc_per: lg %r15,__LC_KERNEL_STACK @@ -288,7 +293,13 @@ ENTRY(system_call) mvc __PT_INT_CODE(4,%r11),__LC_SVC_ILC stg %r14,__PT_FLAGS(%r11) .Lsysc_do_svc: - lg %r10,__TI_sysc_table(%r12) # address of system call table + # load address of system call table +#ifdef CONFIG_HAVE_MARCH_Z990_FEATURES + lg %r10,__TASK_thread+__THREAD_sysc_table(%r12) +#else + lghi %r13,__TASK_thread + lg %r10,__THREAD_sysc_table(%r13,%r12) +#endif llgh %r8,__PT_INT_CODE+2(%r11) slag %r8,%r8,2 # shift and test for svc 0 jnz .Lsysc_nr_ok @@ -389,7 +400,6 @@ ENTRY(system_call) TSTMSK __PT_FLAGS(%r11),_PIF_SYSCALL jno .Lsysc_return lmg %r2,%r7,__PT_R2(%r11) # load svc arguments - lg %r10,__TI_sysc_table(%r12) # address of system call table lghi %r8,0 # svc 0 returns -ENOSYS llgh %r1,__PT_INT_CODE+2(%r11) # load new svc number cghi %r1,NR_syscalls @@ -457,7 +467,7 @@ ENTRY(system_call) # ENTRY(ret_from_fork) la %r11,STACK_FRAME_OVERHEAD(%r15) - lg %r12,__LC_THREAD_INFO + lg %r12,__LC_CURRENT brasl %r14,schedule_tail TRACE_IRQS_ON ssm __LC_SVC_NEW_PSW # reenable interrupts @@ -478,7 +488,7 @@ ENTRY(pgm_check_handler) stpt __LC_SYNC_ENTER_TIMER stmg %r8,%r15,__LC_SAVE_AREA_SYNC lg %r10,__LC_LAST_BREAK - lg %r12,__LC_THREAD_INFO + lg %r12,__LC_CURRENT larl %r13,cleanup_critical lmg %r8,%r9,__LC_PGM_OLD_PSW tmhh %r8,0x0001 # test problem state bit @@ -501,7 +511,7 @@ ENTRY(pgm_check_handler) 2: LAST_BREAK %r14 UPDATE_VTIME %r14,%r15,__LC_SYNC_ENTER_TIMER lg %r15,__LC_KERNEL_STACK - lg %r14,__TI_task(%r12) + lgr %r14,%r12 aghi %r14,__TASK_thread # pointer to thread_struct lghi %r13,__LC_PGM_TDB tm __LC_PGM_ILC+2,0x02 # check for transaction abort @@ -567,7 +577,7 @@ ENTRY(io_int_handler) stpt __LC_ASYNC_ENTER_TIMER stmg %r8,%r15,__LC_SAVE_AREA_ASYNC lg %r10,__LC_LAST_BREAK - lg %r12,__LC_THREAD_INFO + lg %r12,__LC_CURRENT larl %r13,cleanup_critical lmg %r8,%r9,__LC_IO_OLD_PSW SWITCH_ASYNC __LC_SAVE_AREA_ASYNC,__LC_ASYNC_ENTER_TIMER @@ -626,7 +636,7 @@ ENTRY(io_int_handler) jo .Lio_work_user # yes -> do resched & signal #ifdef CONFIG_PREEMPT # check for preemptive scheduling - icm %r0,15,__TI_precount(%r12) + icm %r0,15,__LC_PREEMPT_COUNT jnz .Lio_restore # preemption is disabled TSTMSK __TI_flags(%r12),_TIF_NEED_RESCHED jno .Lio_restore @@ -741,7 +751,7 @@ ENTRY(ext_int_handler) stpt __LC_ASYNC_ENTER_TIMER stmg %r8,%r15,__LC_SAVE_AREA_ASYNC lg %r10,__LC_LAST_BREAK - lg %r12,__LC_THREAD_INFO + lg %r12,__LC_CURRENT larl %r13,cleanup_critical lmg %r8,%r9,__LC_EXT_OLD_PSW SWITCH_ASYNC __LC_SAVE_AREA_ASYNC,__LC_ASYNC_ENTER_TIMER @@ -798,13 +808,10 @@ ENTRY(save_fpu_regs) TSTMSK __LC_CPU_FLAGS,_CIF_FPU bor %r14 stfpc __THREAD_FPU_fpc(%r2) -.Lsave_fpu_regs_fpc_end: lg %r3,__THREAD_FPU_regs(%r2) TSTMSK __LC_MACHINE_FLAGS,MACHINE_FLAG_VX jz .Lsave_fpu_regs_fp # no -> store FP regs -.Lsave_fpu_regs_vx_low: VSTM %v0,%v15,0,%r3 # vstm 0,15,0(3) -.Lsave_fpu_regs_vx_high: VSTM %v16,%v31,256,%r3 # vstm 16,31,256(3) j .Lsave_fpu_regs_done # -> set CIF_FPU flag .Lsave_fpu_regs_fp: @@ -851,9 +858,7 @@ load_fpu_regs: TSTMSK __LC_MACHINE_FLAGS,MACHINE_FLAG_VX lg %r4,__THREAD_FPU_regs(%r4) # %r4 <- reg save area jz .Lload_fpu_regs_fp # -> no VX, load FP regs -.Lload_fpu_regs_vx: VLM %v0,%v15,0,%r4 -.Lload_fpu_regs_vx_high: VLM %v16,%v31,256,%r4 j .Lload_fpu_regs_done .Lload_fpu_regs_fp: @@ -889,7 +894,7 @@ ENTRY(mcck_int_handler) spt __LC_CPU_TIMER_SAVE_AREA-4095(%r1) # revalidate cpu timer lmg %r0,%r15,__LC_GPREGS_SAVE_AREA-4095(%r1)# revalidate gprs lg %r10,__LC_LAST_BREAK - lg %r12,__LC_THREAD_INFO + lg %r12,__LC_CURRENT larl %r13,cleanup_critical lmg %r8,%r9,__LC_MCK_OLD_PSW TSTMSK __LC_MCCK_CODE,MCCK_CODE_SYSTEM_DAMAGE @@ -948,7 +953,7 @@ ENTRY(mcck_int_handler) .Lmcck_panic: lg %r15,__LC_PANIC_STACK - aghi %r15,-(STACK_FRAME_OVERHEAD + __PT_SIZE) + la %r11,STACK_FRAME_OVERHEAD(%r15) j .Lmcck_skip # @@ -1085,7 +1090,7 @@ cleanup_critical: jhe 0f # set up saved registers r10 and r12 stg %r10,16(%r11) # r10 last break - stg %r12,32(%r11) # r12 thread-info pointer + stg %r12,32(%r11) # r12 task struct pointer 0: # check if the user time update has been done clg %r9,BASED(.Lcleanup_system_call_insn+24) jh 0f @@ -1106,7 +1111,9 @@ cleanup_critical: lg %r9,16(%r11) srag %r9,%r9,23 jz 0f - mvc __TI_last_break(8,%r12),16(%r11) + lgr %r9,%r12 + aghi %r9,__TASK_thread + mvc __THREAD_last_break(8,%r9),16(%r11) 0: # set up saved register r11 lg %r15,__LC_KERNEL_STACK la %r9,STACK_FRAME_OVERHEAD(%r15) diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index 4431905f8cfa..0b5ebf8a3d30 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -315,7 +315,7 @@ ENTRY(startup_kdump) jg startup_continue .Lstack: - .long 0x8000 + (1<<(PAGE_SHIFT+THREAD_ORDER)) + .long 0x8000 + (1<<(PAGE_SHIFT+THREAD_SIZE_ORDER)) .align 8 6: .long 0x7fffffff,0xffffffff diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S index 03c2b469c472..482d3526e32b 100644 --- a/arch/s390/kernel/head64.S +++ b/arch/s390/kernel/head64.S @@ -32,11 +32,10 @@ ENTRY(startup_continue) # # Setup stack # - larl %r15,init_thread_union - stg %r15,__LC_THREAD_INFO # cache thread info in lowcore - lg %r14,__TI_task(%r15) # cache current in lowcore + larl %r14,init_task stg %r14,__LC_CURRENT - aghi %r15,1<<(PAGE_SHIFT+THREAD_ORDER) # init_task_union + THREAD_SIZE + larl %r15,init_thread_union + aghi %r15,1<<(PAGE_SHIFT+THREAD_SIZE_ORDER) # init_task_union + THREAD_SIZE stg %r15,__LC_KERNEL_STACK # set end of kernel stack aghi %r15,-160 # diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index 295bfb7124bc..ff3364a067ff 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -1991,10 +1991,9 @@ void __init ipl_update_parameters(void) diag308_set_works = 1; } -void __init ipl_save_parameters(void) +void __init ipl_verify_parameters(void) { struct cio_iplinfo iplinfo; - void *src, *dst; if (cio_get_iplinfo(&iplinfo)) return; @@ -2005,10 +2004,6 @@ void __init ipl_save_parameters(void) if (!iplinfo.is_qdio) return; ipl_flags |= IPL_PARMBLOCK_VALID; - src = (void *)(unsigned long)S390_lowcore.ipl_parmblock_ptr; - dst = (void *)IPL_PARMBLOCK_ORIGIN; - memmove(dst, src, PAGE_SIZE); - S390_lowcore.ipl_parmblock_ptr = IPL_PARMBLOCK_ORIGIN; } static LIST_HEAD(rcall); diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c index 285d6561076d..ef60f4177331 100644 --- a/arch/s390/kernel/irq.c +++ b/arch/s390/kernel/irq.c @@ -168,7 +168,7 @@ void do_softirq_own_stack(void) old = current_stack_pointer(); /* Check against async. stack address range. */ new = S390_lowcore.async_stack; - if (((new - old) >> (PAGE_SHIFT + THREAD_ORDER)) != 0) { + if (((new - old) >> (PAGE_SHIFT + THREAD_SIZE_ORDER)) != 0) { /* Need to switch to the async. stack. */ new -= STACK_FRAME_OVERHEAD; ((struct stack_frame *) new)->back_chain = old; diff --git a/arch/s390/kernel/lgr.c b/arch/s390/kernel/lgr.c index 6ea6d69339b5..ae7dff110054 100644 --- a/arch/s390/kernel/lgr.c +++ b/arch/s390/kernel/lgr.c @@ -5,7 +5,8 @@ * Author(s): Michael Holzheu <holzheu@linux.vnet.ibm.com> */ -#include <linux/module.h> +#include <linux/init.h> +#include <linux/export.h> #include <linux/timer.h> #include <linux/slab.h> #include <asm/facility.h> @@ -183,4 +184,4 @@ static int __init lgr_init(void) lgr_timer_set(); return 0; } -module_init(lgr_init); +device_initcall(lgr_init); diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_sf.c index fcc634c1479a..763dec18edcd 100644 --- a/arch/s390/kernel/perf_cpum_sf.c +++ b/arch/s390/kernel/perf_cpum_sf.c @@ -995,39 +995,36 @@ static int perf_push_sample(struct perf_event *event, struct sf_raw_sample *sfr) regs.int_parm = CPU_MF_INT_SF_PRA; sde_regs = (struct perf_sf_sde_regs *) ®s.int_parm_long; - regs.psw.addr = sfr->basic.ia; - if (sfr->basic.T) - regs.psw.mask |= PSW_MASK_DAT; - if (sfr->basic.W) - regs.psw.mask |= PSW_MASK_WAIT; - if (sfr->basic.P) - regs.psw.mask |= PSW_MASK_PSTATE; - switch (sfr->basic.AS) { - case 0x0: - regs.psw.mask |= PSW_ASC_PRIMARY; - break; - case 0x1: - regs.psw.mask |= PSW_ASC_ACCREG; - break; - case 0x2: - regs.psw.mask |= PSW_ASC_SECONDARY; - break; - case 0x3: - regs.psw.mask |= PSW_ASC_HOME; - break; - } + psw_bits(regs.psw).ia = sfr->basic.ia; + psw_bits(regs.psw).t = sfr->basic.T; + psw_bits(regs.psw).w = sfr->basic.W; + psw_bits(regs.psw).p = sfr->basic.P; + psw_bits(regs.psw).as = sfr->basic.AS; /* - * A non-zero guest program parameter indicates a guest - * sample. - * Note that some early samples or samples from guests without + * Use the hardware provided configuration level to decide if the + * sample belongs to a guest or host. If that is not available, + * fall back to the following heuristics: + * A non-zero guest program parameter always indicates a guest + * sample. Some early samples or samples from guests without * lpp usage would be misaccounted to the host. We use the asn - * value as a heuristic to detect most of these guest samples. - * If the value differs from the host hpp value, we assume - * it to be a KVM guest. + * value as an addon heuristic to detect most of these guest samples. + * If the value differs from the host hpp value, we assume to be a + * KVM guest. */ - if (sfr->basic.gpp || sfr->basic.prim_asn != (u16) sfr->basic.hpp) + switch (sfr->basic.CL) { + case 1: /* logical partition */ + sde_regs->in_guest = 0; + break; + case 2: /* virtual machine */ sde_regs->in_guest = 1; + break; + default: /* old machine, use heuristics */ + if (sfr->basic.gpp || + sfr->basic.prim_asn != (u16)sfr->basic.hpp) + sde_regs->in_guest = 1; + break; + } overflow = 0; if (perf_exclude_event(event, ®s, sde_regs)) diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index bba4fa74b321..400d14f0b9f5 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -103,7 +103,6 @@ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src) int copy_thread(unsigned long clone_flags, unsigned long new_stackp, unsigned long arg, struct task_struct *p) { - struct thread_info *ti; struct fake_frame { struct stack_frame sf; @@ -121,9 +120,8 @@ int copy_thread(unsigned long clone_flags, unsigned long new_stackp, memset(&p->thread.per_event, 0, sizeof(p->thread.per_event)); clear_tsk_thread_flag(p, TIF_SINGLE_STEP); /* Initialize per thread user and system timer values */ - ti = task_thread_info(p); - ti->user_timer = 0; - ti->system_timer = 0; + p->thread.user_timer = 0; + p->thread.system_timer = 0; frame->sf.back_chain = 0; /* new return point is ret_from_fork */ diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 9336e824e2db..b81ab8882e2e 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -461,7 +461,7 @@ long arch_ptrace(struct task_struct *child, long request, } return 0; case PTRACE_GET_LAST_BREAK: - put_user(task_thread_info(child)->last_break, + put_user(child->thread.last_break, (unsigned long __user *) data); return 0; case PTRACE_ENABLE_TE: @@ -811,7 +811,7 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request, } return 0; case PTRACE_GET_LAST_BREAK: - put_user(task_thread_info(child)->last_break, + put_user(child->thread.last_break, (unsigned int __user *) data); return 0; } @@ -997,10 +997,10 @@ static int s390_last_break_get(struct task_struct *target, if (count > 0) { if (kbuf) { unsigned long *k = kbuf; - *k = task_thread_info(target)->last_break; + *k = target->thread.last_break; } else { unsigned long __user *u = ubuf; - if (__put_user(task_thread_info(target)->last_break, u)) + if (__put_user(target->thread.last_break, u)) return -EFAULT; } } @@ -1113,7 +1113,7 @@ static int s390_system_call_get(struct task_struct *target, unsigned int pos, unsigned int count, void *kbuf, void __user *ubuf) { - unsigned int *data = &task_thread_info(target)->system_call; + unsigned int *data = &target->thread.system_call; return user_regset_copyout(&pos, &count, &kbuf, &ubuf, data, 0, sizeof(unsigned int)); } @@ -1123,7 +1123,7 @@ static int s390_system_call_set(struct task_struct *target, unsigned int pos, unsigned int count, const void *kbuf, const void __user *ubuf) { - unsigned int *data = &task_thread_info(target)->system_call; + unsigned int *data = &target->thread.system_call; return user_regset_copyin(&pos, &count, &kbuf, &ubuf, data, 0, sizeof(unsigned int)); } @@ -1327,7 +1327,7 @@ static int s390_compat_last_break_get(struct task_struct *target, compat_ulong_t last_break; if (count > 0) { - last_break = task_thread_info(target)->last_break; + last_break = target->thread.last_break; if (kbuf) { unsigned long *k = kbuf; *k = last_break; diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 7f7ba5f23f13..adfac9f0a89f 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -35,6 +35,7 @@ #include <linux/root_dev.h> #include <linux/console.h> #include <linux/kernel_stat.h> +#include <linux/dma-contiguous.h> #include <linux/device.h> #include <linux/notifier.h> #include <linux/pfn.h> @@ -303,7 +304,7 @@ static void __init setup_lowcore(void) * Setup lowcore for boot cpu */ BUILD_BUG_ON(sizeof(struct lowcore) != LC_PAGES * 4096); - lc = __alloc_bootmem_low(LC_PAGES * PAGE_SIZE, LC_PAGES * PAGE_SIZE, 0); + lc = memblock_virt_alloc_low(sizeof(*lc), sizeof(*lc)); lc->restart_psw.mask = PSW_KERNEL_BITS; lc->restart_psw.addr = (unsigned long) restart_int_handler; lc->external_new_psw.mask = PSW_KERNEL_BITS | @@ -324,15 +325,15 @@ static void __init setup_lowcore(void) lc->kernel_stack = ((unsigned long) &init_thread_union) + THREAD_SIZE - STACK_FRAME_OVERHEAD - sizeof(struct pt_regs); lc->async_stack = (unsigned long) - __alloc_bootmem(ASYNC_SIZE, ASYNC_SIZE, 0) + memblock_virt_alloc(ASYNC_SIZE, ASYNC_SIZE) + ASYNC_SIZE - STACK_FRAME_OVERHEAD - sizeof(struct pt_regs); lc->panic_stack = (unsigned long) - __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0) + memblock_virt_alloc(PAGE_SIZE, PAGE_SIZE) + PAGE_SIZE - STACK_FRAME_OVERHEAD - sizeof(struct pt_regs); - lc->current_task = (unsigned long) init_thread_union.thread_info.task; - lc->thread_info = (unsigned long) &init_thread_union; + lc->current_task = (unsigned long)&init_task; lc->lpp = LPP_MAGIC; lc->machine_flags = S390_lowcore.machine_flags; + lc->preempt_count = S390_lowcore.preempt_count; lc->stfl_fac_list = S390_lowcore.stfl_fac_list; memcpy(lc->stfle_fac_list, S390_lowcore.stfle_fac_list, MAX_FACILITY_BIT/8); @@ -349,7 +350,7 @@ static void __init setup_lowcore(void) lc->last_update_timer = S390_lowcore.last_update_timer; lc->last_update_clock = S390_lowcore.last_update_clock; - restart_stack = __alloc_bootmem(ASYNC_SIZE, ASYNC_SIZE, 0); + restart_stack = memblock_virt_alloc(ASYNC_SIZE, ASYNC_SIZE); restart_stack += ASYNC_SIZE; /* @@ -412,7 +413,7 @@ static void __init setup_resources(void) bss_resource.end = (unsigned long) &__bss_stop - 1; for_each_memblock(memory, reg) { - res = alloc_bootmem_low(sizeof(*res)); + res = memblock_virt_alloc(sizeof(*res), 8); res->flags = IORESOURCE_BUSY | IORESOURCE_SYSTEM_RAM; res->name = "System RAM"; @@ -426,7 +427,7 @@ static void __init setup_resources(void) std_res->start > res->end) continue; if (std_res->end > res->end) { - sub_res = alloc_bootmem_low(sizeof(*sub_res)); + sub_res = memblock_virt_alloc(sizeof(*sub_res), 8); *sub_res = *std_res; sub_res->end = res->end; std_res->start = res->end + 1; @@ -445,7 +446,7 @@ static void __init setup_resources(void) * part of the System RAM resource. */ if (crashk_res.end) { - memblock_add(crashk_res.start, resource_size(&crashk_res)); + memblock_add_node(crashk_res.start, resource_size(&crashk_res), 0); memblock_reserve(crashk_res.start, resource_size(&crashk_res)); insert_resource(&iomem_resource, &crashk_res); } @@ -903,6 +904,7 @@ void __init setup_arch(char **cmdline_p) setup_memory_end(); setup_memory(); + dma_contiguous_reserve(memory_end); check_initrd(); reserve_crashkernel(); @@ -921,6 +923,8 @@ void __init setup_arch(char **cmdline_p) cpu_detect_mhz_feature(); cpu_init(); numa_setup(); + smp_detect_cpus(); + topology_init_early(); /* * Create kernel page tables and switch to virtual addressing. diff --git a/arch/s390/kernel/signal.c b/arch/s390/kernel/signal.c index d82562cf0a0e..9f241d1efeda 100644 --- a/arch/s390/kernel/signal.c +++ b/arch/s390/kernel/signal.c @@ -359,7 +359,7 @@ static int setup_frame(int sig, struct k_sigaction *ka, /* set extra registers only for synchronous signals */ regs->gprs[4] = regs->int_code & 127; regs->gprs[5] = regs->int_parm_long; - regs->gprs[6] = task_thread_info(current)->last_break; + regs->gprs[6] = current->thread.last_break; } return 0; } @@ -430,7 +430,7 @@ static int setup_rt_frame(struct ksignal *ksig, sigset_t *set, regs->gprs[2] = ksig->sig; regs->gprs[3] = (unsigned long) &frame->info; regs->gprs[4] = (unsigned long) &frame->uc; - regs->gprs[5] = task_thread_info(current)->last_break; + regs->gprs[5] = current->thread.last_break; return 0; } @@ -467,13 +467,13 @@ void do_signal(struct pt_regs *regs) * the debugger may change all our registers, including the system * call information. */ - current_thread_info()->system_call = + current->thread.system_call = test_pt_regs_flag(regs, PIF_SYSCALL) ? regs->int_code : 0; if (get_signal(&ksig)) { /* Whee! Actually deliver the signal. */ - if (current_thread_info()->system_call) { - regs->int_code = current_thread_info()->system_call; + if (current->thread.system_call) { + regs->int_code = current->thread.system_call; /* Check for system call restarting. */ switch (regs->gprs[2]) { case -ERESTART_RESTARTBLOCK: @@ -506,8 +506,8 @@ void do_signal(struct pt_regs *regs) /* No handlers present - check for system call restart */ clear_pt_regs_flag(regs, PIF_SYSCALL); - if (current_thread_info()->system_call) { - regs->int_code = current_thread_info()->system_call; + if (current->thread.system_call) { + regs->int_code = current->thread.system_call; switch (regs->gprs[2]) { case -ERESTART_RESTARTBLOCK: /* Restart with sys_restart_syscall */ diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index df4a508ff35c..e49f61aadaf9 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -19,6 +19,7 @@ #define pr_fmt(fmt) KMSG_COMPONENT ": " fmt #include <linux/workqueue.h> +#include <linux/bootmem.h> #include <linux/module.h> #include <linux/init.h> #include <linux/mm.h> @@ -259,16 +260,14 @@ static void pcpu_prepare_secondary(struct pcpu *pcpu, int cpu) static void pcpu_attach_task(struct pcpu *pcpu, struct task_struct *tsk) { struct lowcore *lc = pcpu->lowcore; - struct thread_info *ti = task_thread_info(tsk); lc->kernel_stack = (unsigned long) task_stack_page(tsk) + THREAD_SIZE - STACK_FRAME_OVERHEAD - sizeof(struct pt_regs); - lc->thread_info = (unsigned long) task_thread_info(tsk); lc->current_task = (unsigned long) tsk; lc->lpp = LPP_MAGIC; lc->current_pid = tsk->pid; - lc->user_timer = ti->user_timer; - lc->system_timer = ti->system_timer; + lc->user_timer = tsk->thread.user_timer; + lc->system_timer = tsk->thread.system_timer; lc->steal_timer = 0; } @@ -662,14 +661,12 @@ int smp_cpu_get_polarization(int cpu) return pcpu_devices[cpu].polarization; } -static struct sclp_core_info *smp_get_core_info(void) +static void __ref smp_get_core_info(struct sclp_core_info *info, int early) { static int use_sigp_detection; - struct sclp_core_info *info; int address; - info = kzalloc(sizeof(*info), GFP_KERNEL); - if (info && (use_sigp_detection || sclp_get_core_info(info))) { + if (use_sigp_detection || sclp_get_core_info(info, early)) { use_sigp_detection = 1; for (address = 0; address < (SCLP_MAX_CORES << smp_cpu_mt_shift); @@ -683,7 +680,6 @@ static struct sclp_core_info *smp_get_core_info(void) } info->combined = info->configured; } - return info; } static int smp_add_present_cpu(int cpu); @@ -724,17 +720,15 @@ static int __smp_rescan_cpus(struct sclp_core_info *info, int sysfs_add) return nr; } -static void __init smp_detect_cpus(void) +void __init smp_detect_cpus(void) { unsigned int cpu, mtid, c_cpus, s_cpus; struct sclp_core_info *info; u16 address; /* Get CPU information */ - info = smp_get_core_info(); - if (!info) - panic("smp_detect_cpus failed to allocate memory\n"); - + info = memblock_virt_alloc(sizeof(*info), 8); + smp_get_core_info(info, 1); /* Find boot CPU type */ if (sclp.has_core_type) { address = stap(); @@ -770,7 +764,7 @@ static void __init smp_detect_cpus(void) get_online_cpus(); __smp_rescan_cpus(info, 0); put_online_cpus(); - kfree(info); + memblock_free_early((unsigned long)info, sizeof(*info)); } /* @@ -807,7 +801,7 @@ int __cpu_up(unsigned int cpu, struct task_struct *tidle) pcpu = pcpu_devices + cpu; if (pcpu->state != CPU_STATE_CONFIGURED) return -EIO; - base = cpu - (cpu % (smp_cpu_mtid + 1)); + base = smp_get_base_cpu(cpu); for (i = 0; i <= smp_cpu_mtid; i++) { if (base + i < nr_cpu_ids) if (cpu_online(base + i)) @@ -907,7 +901,6 @@ void __init smp_prepare_cpus(unsigned int max_cpus) /* request the 0x1202 external call external interrupt */ if (register_external_irq(EXT_IRQ_EXTERNAL_CALL, do_ext_call_interrupt)) panic("Couldn't request external interrupt 0x1202"); - smp_detect_cpus(); } void __init smp_prepare_boot_cpu(void) @@ -973,7 +966,7 @@ static ssize_t cpu_configure_store(struct device *dev, rc = -EBUSY; /* disallow configuration changes of online cpus and cpu 0 */ cpu = dev->id; - cpu -= cpu % (smp_cpu_mtid + 1); + cpu = smp_get_base_cpu(cpu); if (cpu == 0) goto out; for (i = 0; i <= smp_cpu_mtid; i++) @@ -1106,9 +1099,10 @@ int __ref smp_rescan_cpus(void) struct sclp_core_info *info; int nr; - info = smp_get_core_info(); + info = kzalloc(sizeof(*info), GFP_KERNEL); if (!info) return -ENOMEM; + smp_get_core_info(info, 0); get_online_cpus(); mutex_lock(&smp_cpu_state_mutex); nr = __smp_rescan_cpus(info, 1); diff --git a/arch/s390/kernel/swsusp.S b/arch/s390/kernel/swsusp.S index 2d6b6e81f812..1ff21f05d7dd 100644 --- a/arch/s390/kernel/swsusp.S +++ b/arch/s390/kernel/swsusp.S @@ -194,7 +194,7 @@ pgm_check_entry: /* Suspend CPU not available -> panic */ larl %r15,init_thread_union - ahi %r15,1<<(PAGE_SHIFT+THREAD_ORDER) + ahi %r15,1<<(PAGE_SHIFT+THREAD_SIZE_ORDER) larl %r2,.Lpanic_string larl %r3,_sclp_print_early lghi %r1,0 diff --git a/arch/s390/kernel/sysinfo.c b/arch/s390/kernel/sysinfo.c index bfda6aa40280..24021c1e3ecb 100644 --- a/arch/s390/kernel/sysinfo.c +++ b/arch/s390/kernel/sysinfo.c @@ -56,6 +56,20 @@ int stsi(void *sysinfo, int fc, int sel1, int sel2) } EXPORT_SYMBOL(stsi); +static bool convert_ext_name(unsigned char encoding, char *name, size_t len) +{ + switch (encoding) { + case 1: /* EBCDIC */ + EBCASC(name, len); + break; + case 2: /* UTF-8 */ + break; + default: + return false; + } + return true; +} + static void stsi_1_1_1(struct seq_file *m, struct sysinfo_1_1_1 *info) { int i; @@ -207,24 +221,19 @@ static void stsi_2_2_2(struct seq_file *m, struct sysinfo_2_2_2 *info) seq_printf(m, "LPAR CPUs S-MTID: %d\n", info->mt_stid); seq_printf(m, "LPAR CPUs PS-MTID: %d\n", info->mt_psmtid); } + if (convert_ext_name(info->vsne, info->ext_name, sizeof(info->ext_name))) { + seq_printf(m, "LPAR Extended Name: %-.256s\n", info->ext_name); + seq_printf(m, "LPAR UUID: %pUb\n", &info->uuid); + } } static void print_ext_name(struct seq_file *m, int lvl, struct sysinfo_3_2_2 *info) { - if (info->vm[lvl].ext_name_encoding == 0) - return; - if (info->ext_names[lvl][0] == 0) - return; - switch (info->vm[lvl].ext_name_encoding) { - case 1: /* EBCDIC */ - EBCASC(info->ext_names[lvl], sizeof(info->ext_names[lvl])); - break; - case 2: /* UTF-8 */ - break; - default: + size_t len = sizeof(info->ext_names[lvl]); + + if (!convert_ext_name(info->vm[lvl].evmne, info->ext_names[lvl], len)) return; - } seq_printf(m, "VM%02d Extended Name: %-.256s\n", lvl, info->ext_names[lvl]); } diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index 0bfcc492987e..867d0a057046 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -59,19 +59,27 @@ ATOMIC_NOTIFIER_HEAD(s390_epoch_delta_notifier); EXPORT_SYMBOL(s390_epoch_delta_notifier); unsigned char ptff_function_mask[16]; -unsigned long lpar_offset; -unsigned long initial_leap_seconds; + +static unsigned long long lpar_offset; +static unsigned long long initial_leap_seconds; +static unsigned long long tod_steering_end; +static long long tod_steering_delta; /* * Get time offsets with PTFF */ -void __init ptff_init(void) +void __init time_early_init(void) { struct ptff_qto qto; struct ptff_qui qui; + /* Initialize TOD steering parameters */ + tod_steering_end = sched_clock_base_cc; + vdso_data->ts_end = tod_steering_end; + if (!test_facility(28)) return; + ptff(&ptff_function_mask, sizeof(ptff_function_mask), PTFF_QAF); /* get LPAR offset */ @@ -80,7 +88,7 @@ void __init ptff_init(void) /* get initial leap seconds */ if (ptff_query(PTFF_QUI) && ptff(&qui, sizeof(qui), PTFF_QUI) == 0) - initial_leap_seconds = (unsigned long) + initial_leap_seconds = (unsigned long long) ((long) qui.old_leap * 4096000000L); } @@ -123,18 +131,6 @@ void clock_comparator_work(void) cd->event_handler(cd); } -/* - * Fixup the clock comparator. - */ -static void fixup_clock_comparator(unsigned long long delta) -{ - /* If nobody is waiting there's nothing to fix. */ - if (S390_lowcore.clock_comparator == -1ULL) - return; - S390_lowcore.clock_comparator += delta; - set_clock_comparator(S390_lowcore.clock_comparator); -} - static int s390_next_event(unsigned long delta, struct clock_event_device *evt) { @@ -215,7 +211,21 @@ void read_boot_clock64(struct timespec64 *ts) static cycle_t read_tod_clock(struct clocksource *cs) { - return get_tod_clock(); + unsigned long long now, adj; + + preempt_disable(); /* protect from changes to steering parameters */ + now = get_tod_clock(); + adj = tod_steering_end - now; + if (unlikely((s64) adj >= 0)) + /* + * manually steer by 1 cycle every 2^16 cycles. This + * corresponds to shifting the tod delta by 15. 1s is + * therefore steered in ~9h. The adjust will decrease + * over time, until it finally reaches 0. + */ + now += (tod_steering_delta < 0) ? (adj >> 15) : -(adj >> 15); + preempt_enable(); + return now; } static struct clocksource clocksource_tod = { @@ -384,6 +394,55 @@ static inline int check_sync_clock(void) return rc; } +/* + * Apply clock delta to the global data structures. + * This is called once on the CPU that performed the clock sync. + */ +static void clock_sync_global(unsigned long long delta) +{ + unsigned long now, adj; + struct ptff_qto qto; + + /* Fixup the monotonic sched clock. */ + sched_clock_base_cc += delta; + /* Adjust TOD steering parameters. */ + vdso_data->tb_update_count++; + now = get_tod_clock(); + adj = tod_steering_end - now; + if (unlikely((s64) adj >= 0)) + /* Calculate how much of the old adjustment is left. */ + tod_steering_delta = (tod_steering_delta < 0) ? + -(adj >> 15) : (adj >> 15); + tod_steering_delta += delta; + if ((abs(tod_steering_delta) >> 48) != 0) + panic("TOD clock sync offset %lli is too large to drift\n", + tod_steering_delta); + tod_steering_end = now + (abs(tod_steering_delta) << 15); + vdso_data->ts_dir = (tod_steering_delta < 0) ? 0 : 1; + vdso_data->ts_end = tod_steering_end; + vdso_data->tb_update_count++; + /* Update LPAR offset. */ + if (ptff_query(PTFF_QTO) && ptff(&qto, sizeof(qto), PTFF_QTO) == 0) + lpar_offset = qto.tod_epoch_difference; + /* Call the TOD clock change notifier. */ + atomic_notifier_call_chain(&s390_epoch_delta_notifier, 0, &delta); +} + +/* + * Apply clock delta to the per-CPU data structures of this CPU. + * This is called for each online CPU after the call to clock_sync_global. + */ +static void clock_sync_local(unsigned long long delta) +{ + /* Add the delta to the clock comparator. */ + if (S390_lowcore.clock_comparator != -1ULL) { + S390_lowcore.clock_comparator += delta; + set_clock_comparator(S390_lowcore.clock_comparator); + } + /* Adjust the last_update_clock time-stamp. */ + S390_lowcore.last_update_clock += delta; +} + /* Single threaded workqueue used for stp sync events */ static struct workqueue_struct *time_sync_wq; @@ -397,31 +456,9 @@ static void __init time_init_wq(void) struct clock_sync_data { atomic_t cpus; int in_sync; - unsigned long long fixup_cc; + unsigned long long clock_delta; }; -static void clock_sync_cpu(struct clock_sync_data *sync) -{ - atomic_dec(&sync->cpus); - enable_sync_clock(); - while (sync->in_sync == 0) { - __udelay(1); - /* - * A different cpu changes *in_sync. Therefore use - * barrier() to force memory access. - */ - barrier(); - } - if (sync->in_sync != 1) - /* Didn't work. Clear per-cpu in sync bit again. */ - disable_sync_clock(NULL); - /* - * This round of TOD syncing is done. Set the clock comparator - * to the next tick and let the processor continue. - */ - fixup_clock_comparator(sync->fixup_cc); -} - /* * Server Time Protocol (STP) code. */ @@ -523,54 +560,46 @@ void stp_queue_work(void) static int stp_sync_clock(void *data) { - static int first; + struct clock_sync_data *sync = data; unsigned long long clock_delta; - struct clock_sync_data *stp_sync; - struct ptff_qto qto; + static int first; int rc; - stp_sync = data; - - if (xchg(&first, 1) == 1) { - /* Slave */ - clock_sync_cpu(stp_sync); - return 0; - } - - /* Wait until all other cpus entered the sync function. */ - while (atomic_read(&stp_sync->cpus) != 0) - cpu_relax(); - enable_sync_clock(); - - rc = 0; - if (stp_info.todoff[0] || stp_info.todoff[1] || - stp_info.todoff[2] || stp_info.todoff[3] || - stp_info.tmd != 2) { - rc = chsc_sstpc(stp_page, STP_OP_SYNC, 0, &clock_delta); - if (rc == 0) { - /* fixup the monotonic sched clock */ - sched_clock_base_cc += clock_delta; - if (ptff_query(PTFF_QTO) && - ptff(&qto, sizeof(qto), PTFF_QTO) == 0) - /* Update LPAR offset */ - lpar_offset = qto.tod_epoch_difference; - atomic_notifier_call_chain(&s390_epoch_delta_notifier, - 0, &clock_delta); - stp_sync->fixup_cc = clock_delta; - fixup_clock_comparator(clock_delta); - rc = chsc_sstpi(stp_page, &stp_info, - sizeof(struct stp_sstpi)); - if (rc == 0 && stp_info.tmd != 2) - rc = -EAGAIN; + if (xchg(&first, 1) == 0) { + /* Wait until all other cpus entered the sync function. */ + while (atomic_read(&sync->cpus) != 0) + cpu_relax(); + rc = 0; + if (stp_info.todoff[0] || stp_info.todoff[1] || + stp_info.todoff[2] || stp_info.todoff[3] || + stp_info.tmd != 2) { + rc = chsc_sstpc(stp_page, STP_OP_SYNC, 0, + &clock_delta); + if (rc == 0) { + sync->clock_delta = clock_delta; + clock_sync_global(clock_delta); + rc = chsc_sstpi(stp_page, &stp_info, + sizeof(struct stp_sstpi)); + if (rc == 0 && stp_info.tmd != 2) + rc = -EAGAIN; + } } + sync->in_sync = rc ? -EAGAIN : 1; + xchg(&first, 0); + } else { + /* Slave */ + atomic_dec(&sync->cpus); + /* Wait for in_sync to be set. */ + while (READ_ONCE(sync->in_sync) == 0) + __udelay(1); } - if (rc) { + if (sync->in_sync != 1) + /* Didn't work. Clear per-cpu in sync bit again. */ disable_sync_clock(NULL); - stp_sync->in_sync = -EAGAIN; - } else - stp_sync->in_sync = 1; - xchg(&first, 0); + /* Apply clock delta to per-CPU fields of this CPU. */ + clock_sync_local(sync->clock_delta); + return 0; } diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c index e959c02e0cac..93dcbae1e98d 100644 --- a/arch/s390/kernel/topology.c +++ b/arch/s390/kernel/topology.c @@ -7,6 +7,7 @@ #define pr_fmt(fmt) KMSG_COMPONENT ": " fmt #include <linux/workqueue.h> +#include <linux/bootmem.h> #include <linux/cpuset.h> #include <linux/device.h> #include <linux/export.h> @@ -41,15 +42,17 @@ static bool topology_enabled = true; static DECLARE_WORK(topology_work, topology_work_fn); /* - * Socket/Book linked lists and per_cpu(cpu_topology) updates are + * Socket/Book linked lists and cpu_topology updates are * protected by "sched_domains_mutex". */ static struct mask_info socket_info; static struct mask_info book_info; static struct mask_info drawer_info; -DEFINE_PER_CPU(struct cpu_topology_s390, cpu_topology); -EXPORT_PER_CPU_SYMBOL_GPL(cpu_topology); +struct cpu_topology_s390 cpu_topology[NR_CPUS]; +EXPORT_SYMBOL_GPL(cpu_topology); + +cpumask_t cpus_with_topology; static cpumask_t cpu_group_map(struct mask_info *info, unsigned int cpu) { @@ -97,7 +100,7 @@ static void add_cpus_to_mask(struct topology_core *tl_core, if (lcpu < 0) continue; for (i = 0; i <= smp_cpu_mtid; i++) { - topo = &per_cpu(cpu_topology, lcpu + i); + topo = &cpu_topology[lcpu + i]; topo->drawer_id = drawer->id; topo->book_id = book->id; topo->socket_id = socket->id; @@ -106,6 +109,7 @@ static void add_cpus_to_mask(struct topology_core *tl_core, cpumask_set_cpu(lcpu + i, &drawer->mask); cpumask_set_cpu(lcpu + i, &book->mask); cpumask_set_cpu(lcpu + i, &socket->mask); + cpumask_set_cpu(lcpu + i, &cpus_with_topology); smp_cpu_set_polarization(lcpu + i, tl_core->pp); } } @@ -220,7 +224,7 @@ static void update_cpu_masks(void) int cpu; for_each_possible_cpu(cpu) { - topo = &per_cpu(cpu_topology, cpu); + topo = &cpu_topology[cpu]; topo->thread_mask = cpu_thread_map(cpu); topo->core_mask = cpu_group_map(&socket_info, cpu); topo->book_mask = cpu_group_map(&book_info, cpu); @@ -231,6 +235,8 @@ static void update_cpu_masks(void) topo->socket_id = cpu; topo->book_id = cpu; topo->drawer_id = cpu; + if (cpu_present(cpu)) + cpumask_set_cpu(cpu, &cpus_with_topology); } } numa_update_cpu_topology(); @@ -241,12 +247,12 @@ void store_topology(struct sysinfo_15_1_x *info) stsi(info, 15, 1, min(topology_max_mnest, 4)); } -int arch_update_cpu_topology(void) +static int __arch_update_cpu_topology(void) { struct sysinfo_15_1_x *info = tl_info; - struct device *dev; - int cpu, rc = 0; + int rc = 0; + cpumask_clear(&cpus_with_topology); if (MACHINE_HAS_TOPOLOGY) { rc = 1; store_topology(info); @@ -255,6 +261,15 @@ int arch_update_cpu_topology(void) update_cpu_masks(); if (!MACHINE_HAS_TOPOLOGY) topology_update_polarization_simple(); + return rc; +} + +int arch_update_cpu_topology(void) +{ + struct device *dev; + int cpu, rc; + + rc = __arch_update_cpu_topology(); for_each_online_cpu(cpu) { dev = get_cpu_device(cpu); kobject_uevent(&dev->kobj, KOBJ_CHANGE); @@ -394,23 +409,23 @@ int topology_cpu_init(struct cpu *cpu) static const struct cpumask *cpu_thread_mask(int cpu) { - return &per_cpu(cpu_topology, cpu).thread_mask; + return &cpu_topology[cpu].thread_mask; } const struct cpumask *cpu_coregroup_mask(int cpu) { - return &per_cpu(cpu_topology, cpu).core_mask; + return &cpu_topology[cpu].core_mask; } static const struct cpumask *cpu_book_mask(int cpu) { - return &per_cpu(cpu_topology, cpu).book_mask; + return &cpu_topology[cpu].book_mask; } static const struct cpumask *cpu_drawer_mask(int cpu) { - return &per_cpu(cpu_topology, cpu).drawer_mask; + return &cpu_topology[cpu].drawer_mask; } static int __init early_parse_topology(char *p) @@ -438,19 +453,20 @@ static void __init alloc_masks(struct sysinfo_15_1_x *info, nr_masks *= info->mag[TOPOLOGY_NR_MAG - offset - 1 - i]; nr_masks = max(nr_masks, 1); for (i = 0; i < nr_masks; i++) { - mask->next = kzalloc(sizeof(*mask->next), GFP_KERNEL); + mask->next = memblock_virt_alloc(sizeof(*mask->next), 8); mask = mask->next; } } -static int __init s390_topology_init(void) +void __init topology_init_early(void) { struct sysinfo_15_1_x *info; int i; + set_sched_topology(s390_topology); if (!MACHINE_HAS_TOPOLOGY) - return 0; - tl_info = (struct sysinfo_15_1_x *)__get_free_page(GFP_KERNEL); + goto out; + tl_info = memblock_virt_alloc(sizeof(*tl_info), PAGE_SIZE); info = tl_info; store_topology(info); pr_info("The CPU configuration topology of the machine is:"); @@ -460,10 +476,9 @@ static int __init s390_topology_init(void) alloc_masks(info, &socket_info, 1); alloc_masks(info, &book_info, 2); alloc_masks(info, &drawer_info, 3); - set_sched_topology(s390_topology); - return 0; +out: + __arch_update_cpu_topology(); } -early_initcall(s390_topology_init); static int __init topology_init(void) { diff --git a/arch/s390/kernel/vdso32/clock_gettime.S b/arch/s390/kernel/vdso32/clock_gettime.S index 5eec9afbb5b5..a5769b83d90e 100644 --- a/arch/s390/kernel/vdso32/clock_gettime.S +++ b/arch/s390/kernel/vdso32/clock_gettime.S @@ -99,8 +99,27 @@ __kernel_clock_gettime: tml %r4,0x0001 /* pending update ? loop */ jnz 11b stcke 0(%r15) /* Store TOD clock */ - lm %r0,%r1,1(%r15) - s %r0,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ + lm %r0,%r1,__VDSO_TS_END(%r5) /* TOD steering end time */ + s %r0,1(%r15) /* no - ts_steering_end */ + sl %r1,5(%r15) + brc 3,22f + ahi %r0,-1 +22: ltr %r0,%r0 /* past end of steering? */ + jm 24f + srdl %r0,15 /* 1 per 2^16 */ + tm __VDSO_TS_DIR+3(%r5),0x01 /* steering direction? */ + jz 23f + lcr %r0,%r0 /* negative TOD offset */ + lcr %r1,%r1 + je 23f + ahi %r0,-1 +23: a %r0,1(%r15) /* add TOD timestamp */ + al %r1,5(%r15) + brc 12,25f + ahi %r0,1 + j 25f +24: lm %r0,%r1,1(%r15) /* load TOD timestamp */ +25: s %r0,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ sl %r1,__VDSO_XTIME_STAMP+4(%r5) brc 3,12f ahi %r0,-1 diff --git a/arch/s390/kernel/vdso32/gettimeofday.S b/arch/s390/kernel/vdso32/gettimeofday.S index 719de6186b20..63b86dceb0bf 100644 --- a/arch/s390/kernel/vdso32/gettimeofday.S +++ b/arch/s390/kernel/vdso32/gettimeofday.S @@ -31,8 +31,27 @@ __kernel_gettimeofday: tml %r4,0x0001 /* pending update ? loop */ jnz 1b stcke 0(%r15) /* Store TOD clock */ - lm %r0,%r1,1(%r15) - s %r0,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ + lm %r0,%r1,__VDSO_TS_END(%r5) /* TOD steering end time */ + s %r0,1(%r15) + sl %r1,5(%r15) + brc 3,14f + ahi %r0,-1 +14: ltr %r0,%r0 /* past end of steering? */ + jm 16f + srdl %r0,15 /* 1 per 2^16 */ + tm __VDSO_TS_DIR+3(%r5),0x01 /* steering direction? */ + jz 15f + lcr %r0,%r0 /* negative TOD offset */ + lcr %r1,%r1 + je 15f + ahi %r0,-1 +15: a %r0,1(%r15) /* add TOD timestamp */ + al %r1,5(%r15) + brc 12,17f + ahi %r0,1 + j 17f +16: lm %r0,%r1,1(%r15) /* load TOD timestamp */ +17: s %r0,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ sl %r1,__VDSO_XTIME_STAMP+4(%r5) brc 3,3f ahi %r0,-1 diff --git a/arch/s390/kernel/vdso64/clock_gettime.S b/arch/s390/kernel/vdso64/clock_gettime.S index 61541fb93dc6..9c3b12626dba 100644 --- a/arch/s390/kernel/vdso64/clock_gettime.S +++ b/arch/s390/kernel/vdso64/clock_gettime.S @@ -83,8 +83,17 @@ __kernel_clock_gettime: tmll %r4,0x0001 /* pending update ? loop */ jnz 5b stcke 0(%r15) /* Store TOD clock */ - lgf %r2,__VDSO_TK_SHIFT(%r5) /* Timekeeper shift */ lg %r1,1(%r15) + lg %r0,__VDSO_TS_END(%r5) /* TOD steering end time */ + slgr %r0,%r1 /* now - ts_steering_end */ + ltgr %r0,%r0 /* past end of steering ? */ + jm 17f + srlg %r0,%r0,15 /* 1 per 2^16 */ + tm __VDSO_TS_DIR+3(%r5),0x01 /* steering direction? */ + jz 18f + lcgr %r0,%r0 /* negative TOD offset */ +18: algr %r1,%r0 /* add steering offset */ +17: lgf %r2,__VDSO_TK_SHIFT(%r5) /* Timekeeper shift */ sg %r1,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ msgf %r1,__VDSO_TK_MULT(%r5) /* * tk->mult */ alg %r1,__VDSO_XTIME_NSEC(%r5) /* + tk->xtime_nsec */ diff --git a/arch/s390/kernel/vdso64/gettimeofday.S b/arch/s390/kernel/vdso64/gettimeofday.S index 6ce46707663c..b02e62f3bc12 100644 --- a/arch/s390/kernel/vdso64/gettimeofday.S +++ b/arch/s390/kernel/vdso64/gettimeofday.S @@ -31,7 +31,16 @@ __kernel_gettimeofday: jnz 0b stcke 0(%r15) /* Store TOD clock */ lg %r1,1(%r15) - sg %r1,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ + lg %r0,__VDSO_TS_END(%r5) /* TOD steering end time */ + slgr %r0,%r1 /* now - ts_steering_end */ + ltgr %r0,%r0 /* past end of steering ? */ + jm 6f + srlg %r0,%r0,15 /* 1 per 2^16 */ + tm __VDSO_TS_DIR+3(%r5),0x01 /* steering direction? */ + jz 7f + lcgr %r0,%r0 /* negative TOD offset */ +7: algr %r1,%r0 /* add steering offset */ +6: sg %r1,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ msgf %r1,__VDSO_TK_MULT(%r5) /* * tk->mult */ alg %r1,__VDSO_XTIME_NSEC(%r5) /* + tk->xtime_nsec */ lg %r0,__VDSO_XTIME_SEC(%r5) /* tk->xtime_sec */ diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c index 1bd5dde2d5a9..6b246aadf311 100644 --- a/arch/s390/kernel/vtime.c +++ b/arch/s390/kernel/vtime.c @@ -96,7 +96,6 @@ static void update_mt_scaling(void) */ static int do_account_vtime(struct task_struct *tsk, int hardirq_offset) { - struct thread_info *ti = task_thread_info(tsk); u64 timer, clock, user, system, steal; u64 user_scaled, system_scaled; @@ -119,13 +118,13 @@ static int do_account_vtime(struct task_struct *tsk, int hardirq_offset) time_after64(jiffies_64, this_cpu_read(mt_scaling_jiffies))) update_mt_scaling(); - user = S390_lowcore.user_timer - ti->user_timer; + user = S390_lowcore.user_timer - tsk->thread.user_timer; S390_lowcore.steal_timer -= user; - ti->user_timer = S390_lowcore.user_timer; + tsk->thread.user_timer = S390_lowcore.user_timer; - system = S390_lowcore.system_timer - ti->system_timer; + system = S390_lowcore.system_timer - tsk->thread.system_timer; S390_lowcore.steal_timer -= system; - ti->system_timer = S390_lowcore.system_timer; + tsk->thread.system_timer = S390_lowcore.system_timer; user_scaled = user; system_scaled = system; @@ -153,15 +152,11 @@ static int do_account_vtime(struct task_struct *tsk, int hardirq_offset) void vtime_task_switch(struct task_struct *prev) { - struct thread_info *ti; - do_account_vtime(prev, 0); - ti = task_thread_info(prev); - ti->user_timer = S390_lowcore.user_timer; - ti->system_timer = S390_lowcore.system_timer; - ti = task_thread_info(current); - S390_lowcore.user_timer = ti->user_timer; - S390_lowcore.system_timer = ti->system_timer; + prev->thread.user_timer = S390_lowcore.user_timer; + prev->thread.system_timer = S390_lowcore.system_timer; + S390_lowcore.user_timer = current->thread.user_timer; + S390_lowcore.system_timer = current->thread.system_timer; } /* @@ -181,7 +176,6 @@ void vtime_account_user(struct task_struct *tsk) */ void vtime_account_irq_enter(struct task_struct *tsk) { - struct thread_info *ti = task_thread_info(tsk); u64 timer, system, system_scaled; timer = S390_lowcore.last_update_timer; @@ -193,9 +187,9 @@ void vtime_account_irq_enter(struct task_struct *tsk) time_after64(jiffies_64, this_cpu_read(mt_scaling_jiffies))) update_mt_scaling(); - system = S390_lowcore.system_timer - ti->system_timer; + system = S390_lowcore.system_timer - tsk->thread.system_timer; S390_lowcore.steal_timer -= system; - ti->system_timer = S390_lowcore.system_timer; + tsk->thread.system_timer = S390_lowcore.system_timer; system_scaled = system; /* Do MT utilization scaling */ if (smp_cpu_mtid) { |