From cd8a5673e9abb3fde0a1c25ee63a60fe1908c6f5 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Fri, 3 Mar 2006 17:11:40 +1100 Subject: powerpc: Fix might-sleep warning in program check exception handler On 32-bit, the exception prolog for the program check exception doesn't enable interrupts early on. If it is an illegal instruction exception, we read the instruction in order to emulate certain instructions, and the get_user of the instruction triggers a WARN_ON since interrupts are still disabled. This adds a local_irq_enable() to enable interrupts before reading the instruction. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/traps.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c index 7509aa6474f2..98660aedeeb7 100644 --- a/arch/powerpc/kernel/traps.c +++ b/arch/powerpc/kernel/traps.c @@ -814,6 +814,8 @@ void __kprobes program_check_exception(struct pt_regs *regs) return; } + local_irq_enable(); + /* Try to emulate it if we should. */ if (reason & (REASON_ILLEGAL | REASON_PRIVILEGED)) { switch (emulate_instruction(regs)) { -- cgit v1.2.2 From aa5cb02143123289bd37c30c0ad60339f8da0bad Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 1 Mar 2006 15:07:07 +1100 Subject: [PATCH] powerpc: Expose SMT and L1 icache snoop userland features This patch makes userland aware of the icache snoop capability of the POWER5 (and possibly others in the future) and of SMT capabilities. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/cputable.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c index 10696456a4c6..e4e81374cb9a 100644 --- a/arch/powerpc/kernel/cputable.c +++ b/arch/powerpc/kernel/cputable.c @@ -53,8 +53,10 @@ extern void __setup_cpu_ppc970(unsigned long offset, struct cpu_spec* spec); PPC_FEATURE_HAS_MMU) #define COMMON_USER_PPC64 (COMMON_USER | PPC_FEATURE_64) #define COMMON_USER_POWER4 (COMMON_USER_PPC64 | PPC_FEATURE_POWER4) -#define COMMON_USER_POWER5 (COMMON_USER_PPC64 | PPC_FEATURE_POWER5) -#define COMMON_USER_POWER5_PLUS (COMMON_USER_PPC64 | PPC_FEATURE_POWER5_PLUS) +#define COMMON_USER_POWER5 (COMMON_USER_PPC64 | PPC_FEATURE_POWER5 |\ + PPC_FEATURE_SMT | PPC_FEATURE_ICACHE_SNOOP) +#define COMMON_USER_POWER5_PLUS (COMMON_USER_PPC64 | PPC_FEATURE_POWER5_PLUS|\ + PPC_FEATURE_SMT | PPC_FEATURE_ICACHE_SNOOP) #define COMMON_USER_BOOKE (PPC_FEATURE_32 | PPC_FEATURE_HAS_MMU | \ PPC_FEATURE_BOOKE) @@ -267,7 +269,8 @@ struct cpu_spec cpu_specs[] = { .cpu_name = "Cell Broadband Engine", .cpu_features = CPU_FTRS_CELL, .cpu_user_features = COMMON_USER_PPC64 | - PPC_FEATURE_CELL | PPC_FEATURE_HAS_ALTIVEC_COMP, + PPC_FEATURE_CELL | PPC_FEATURE_HAS_ALTIVEC_COMP | + PPC_FEATURE_SMT, .icache_bsize = 128, .dcache_bsize = 128, .cpu_setup = __setup_cpu_be, -- cgit v1.2.2 From ab1b55e21f6977e420341727e9f4a50691057b5e Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Fri, 3 Mar 2006 10:35:40 +1100 Subject: [PATCH] powerpc: incorrect rmo_top handling in prom_init On Thu, 2006-03-02 at 19:55 +0100, Olaf Hering wrote: > My iBook1 has 2 memory regions in reg. Depending on how I boot it > (vmlinux+initrd) or zImage.initrd, it will not boot with current Linus > tree. > rmo_top should be 160MB instead of 32MB. On logically-partitioned machines the first element of the reg property in the memory node is defined to be the "RMO" region, i.e. the memory that the processor can access in real mode. On other machines the first element has no special meaning, so only take it to be the RMO region on LPAR machines. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/prom_init.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index d34fe537400e..813c2cd194c2 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -978,7 +978,7 @@ static void __init prom_init_mem(void) if (size == 0) continue; prom_debug(" %x %x\n", base, size); - if (base == 0) + if (base == 0 && (RELOC(of_platform) & PLATFORM_LPAR)) RELOC(rmo_top) = size; if ((base + size) > RELOC(ram_top)) RELOC(ram_top) = base + size; -- cgit v1.2.2 From 1bd79336a426c5e4f3bab142407059ceb12cadf9 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Wed, 8 Mar 2006 13:24:22 +1100 Subject: powerpc: Fix various syscall/signal/swapcontext bugs A careful reading of the recent changes to the system call entry/exit paths revealed several problems, plus some things that could be simplified and improved: * 32-bit wasn't testing the _TIF_NOERROR bit in the syscall fast exit path, so it was only doing anything with it once it saw some other bit being set. In other words, the noerror behaviour would apply to the next system call where we had to reschedule or deliver a signal, which is not necessarily the current system call. * 32-bit wasn't doing the call to ptrace_notify in the syscall exit path when the _TIF_SINGLESTEP bit was set. * _TIF_RESTOREALL was in both _TIF_USER_WORK_MASK and _TIF_PERSYSCALL_MASK, which is odd since _TIF_RESTOREALL is only set by system calls. I took it out of _TIF_USER_WORK_MASK. * On 64-bit, _TIF_RESTOREALL wasn't causing the non-volatile registers to be restored (unless perhaps a signal was delivered or the syscall was traced or single-stepped). Thus the non-volatile registers weren't restored on exit from a signal handler. We probably got away with it mostly because signal handlers written in C wouldn't alter the non-volatile registers. * On 32-bit I simplified the code and made it more like 64-bit by making the syscall exit path jump to ret_from_except to handle preemption and signal delivery. * 32-bit was calling do_signal unnecessarily when _TIF_RESTOREALL was set - but I think because of that 32-bit was actually restoring the non-volatile registers on exit from a signal handler. * I changed the order of enabling interrupts and saving the non-volatile registers before calling do_syscall_trace_leave; now we enable interrupts first. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/asm-offsets.c | 1 - arch/powerpc/kernel/entry_32.S | 95 ++++++++++----------------------------- arch/powerpc/kernel/entry_64.S | 94 +++++++++----------------------------- arch/powerpc/kernel/ptrace.c | 5 +-- arch/powerpc/kernel/signal_32.c | 19 ++------ arch/powerpc/kernel/signal_64.c | 9 +--- arch/powerpc/kernel/systbl.S | 2 +- 7 files changed, 52 insertions(+), 173 deletions(-) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c index 840aad43a98b..c9a660e4c2db 100644 --- a/arch/powerpc/kernel/asm-offsets.c +++ b/arch/powerpc/kernel/asm-offsets.c @@ -92,7 +92,6 @@ int main(void) DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); - DEFINE(TI_SIGFRAME, offsetof(struct thread_info, nvgprs_frame)); DEFINE(TI_TASK, offsetof(struct thread_info, task)); #ifdef CONFIG_PPC32 DEFINE(TI_EXECDOMAIN, offsetof(struct thread_info, exec_domain)); diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S index f20a67261ec7..4827ca1ec89b 100644 --- a/arch/powerpc/kernel/entry_32.S +++ b/arch/powerpc/kernel/entry_32.S @@ -227,7 +227,7 @@ ret_from_syscall: MTMSRD(r10) lwz r9,TI_FLAGS(r12) li r8,-_LAST_ERRNO - andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SIGPENDING|_TIF_NEED_RESCHED|_TIF_RESTOREALL|_TIF_RESTORE_SIGMASK) + andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP|_TIF_USER_WORK_MASK|_TIF_PERSYSCALL_MASK) bne- syscall_exit_work cmplw 0,r3,r8 blt+ syscall_exit_cont @@ -287,8 +287,10 @@ syscall_dotrace: syscall_exit_work: andi. r0,r9,_TIF_RESTOREALL - bne- 2f - cmplw 0,r3,r8 + beq+ 0f + REST_NVGPRS(r1) + b 2f +0: cmplw 0,r3,r8 blt+ 1f andi. r0,r9,_TIF_NOERROR bne- 1f @@ -302,9 +304,7 @@ syscall_exit_work: 2: andi. r0,r9,(_TIF_PERSYSCALL_MASK) beq 4f - /* Clear per-syscall TIF flags if any are set, but _leave_ - _TIF_SAVE_NVGPRS set in r9 since we haven't dealt with that - yet. */ + /* Clear per-syscall TIF flags if any are set. */ li r11,_TIF_PERSYSCALL_MASK addi r12,r12,TI_FLAGS @@ -318,8 +318,13 @@ syscall_exit_work: subi r12,r12,TI_FLAGS 4: /* Anything which requires enabling interrupts? */ - andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP|_TIF_SAVE_NVGPRS) - beq 7f + andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP) + beq ret_from_except + + /* Re-enable interrupts */ + ori r10,r10,MSR_EE + SYNC + MTMSRD(r10) /* Save NVGPRS if they're not saved already */ lwz r4,_TRAP(r1) @@ -328,71 +333,11 @@ syscall_exit_work: SAVE_NVGPRS(r1) li r4,0xc00 stw r4,_TRAP(r1) - - /* Re-enable interrupts */ -5: ori r10,r10,MSR_EE - SYNC - MTMSRD(r10) - - andi. r0,r9,_TIF_SAVE_NVGPRS - bne save_user_nvgprs - -save_user_nvgprs_cont: - andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP) - beq 7f - +5: addi r3,r1,STACK_FRAME_OVERHEAD bl do_syscall_trace_leave - REST_NVGPRS(r1) - -6: lwz r3,GPR3(r1) - LOAD_MSR_KERNEL(r10,MSR_KERNEL) /* doesn't include MSR_EE */ - SYNC - MTMSRD(r10) /* disable interrupts again */ - rlwinm r12,r1,0,0,(31-THREAD_SHIFT) /* current_thread_info() */ - lwz r9,TI_FLAGS(r12) -7: - andi. r0,r9,_TIF_NEED_RESCHED - bne 8f - lwz r5,_MSR(r1) - andi. r5,r5,MSR_PR - beq ret_from_except - andi. r0,r9,_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK - beq ret_from_except - b do_user_signal -8: - ori r10,r10,MSR_EE - SYNC - MTMSRD(r10) /* re-enable interrupts */ - bl schedule - b 6b - -save_user_nvgprs: - lwz r8,TI_SIGFRAME(r12) - -.macro savewords start, end - 1: stw \start,4*(\start)(r8) - .section __ex_table,"a" - .align 2 - .long 1b,save_user_nvgprs_fault - .previous - .if \end - \start - savewords "(\start+1)",\end - .endif -.endm - savewords 14,31 - b save_user_nvgprs_cont - - -save_user_nvgprs_fault: - li r3,11 /* SIGSEGV */ - lwz r4,TI_TASK(r12) - bl force_sigsegv + b ret_from_except_full - rlwinm r12,r1,0,0,(31-THREAD_SHIFT) /* current_thread_info() */ - lwz r9,TI_FLAGS(r12) - b save_user_nvgprs_cont - #ifdef SHOW_SYSCALLS do_show_syscall: #ifdef SHOW_SYSCALLS_TASK @@ -490,6 +435,14 @@ ppc_clone: stw r0,_TRAP(r1) /* register set saved */ b sys_clone + .globl ppc_swapcontext +ppc_swapcontext: + SAVE_NVGPRS(r1) + lwz r0,_TRAP(r1) + rlwinm r0,r0,0,0,30 /* clear LSB to indicate full */ + stw r0,_TRAP(r1) /* register set saved */ + b sys_swapcontext + /* * Top-level page fault handling. * This is in assembler because if do_page_fault tells us that @@ -683,7 +636,7 @@ user_exc_return: /* r10 contains MSR_KERNEL here */ /* Check current_thread_info()->flags */ rlwinm r9,r1,0,0,(31-THREAD_SHIFT) lwz r9,TI_FLAGS(r9) - andi. r0,r9,(_TIF_SIGPENDING|_TIF_NEED_RESCHED|_TIF_RESTOREALL|_TIF_RESTORE_SIGMASK) + andi. r0,r9,(_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK|_TIF_NEED_RESCHED) bne do_work restore_user: diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S index 388f861b8ed1..24be0cf86d7f 100644 --- a/arch/powerpc/kernel/entry_64.S +++ b/arch/powerpc/kernel/entry_64.S @@ -160,7 +160,7 @@ syscall_exit: mtmsrd r10,1 ld r9,TI_FLAGS(r12) li r11,-_LAST_ERRNO - andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP|_TIF_SIGPENDING|_TIF_NEED_RESCHED|_TIF_RESTOREALL|_TIF_SAVE_NVGPRS|_TIF_NOERROR|_TIF_RESTORE_SIGMASK) + andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP|_TIF_USER_WORK_MASK|_TIF_PERSYSCALL_MASK) bne- syscall_exit_work cmpld r3,r11 ld r5,_CCR(r1) @@ -216,8 +216,10 @@ syscall_exit_work: If TIF_NOERROR is set, just save r3 as it is. */ andi. r0,r9,_TIF_RESTOREALL - bne- 2f - cmpld r3,r11 /* r10 is -LAST_ERRNO */ + beq+ 0f + REST_NVGPRS(r1) + b 2f +0: cmpld r3,r11 /* r10 is -LAST_ERRNO */ blt+ 1f andi. r0,r9,_TIF_NOERROR bne- 1f @@ -229,9 +231,7 @@ syscall_exit_work: 2: andi. r0,r9,(_TIF_PERSYSCALL_MASK) beq 4f - /* Clear per-syscall TIF flags if any are set, but _leave_ - _TIF_SAVE_NVGPRS set in r9 since we haven't dealt with that - yet. */ + /* Clear per-syscall TIF flags if any are set. */ li r11,_TIF_PERSYSCALL_MASK addi r12,r12,TI_FLAGS @@ -240,10 +240,9 @@ syscall_exit_work: stdcx. r10,0,r12 bne- 3b subi r12,r12,TI_FLAGS - -4: bl .save_nvgprs - /* Anything else left to do? */ - andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP|_TIF_SAVE_NVGPRS) + +4: /* Anything else left to do? */ + andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP) beq .ret_from_except_lite /* Re-enable interrupts */ @@ -251,26 +250,10 @@ syscall_exit_work: ori r10,r10,MSR_EE mtmsrd r10,1 - andi. r0,r9,_TIF_SAVE_NVGPRS - bne save_user_nvgprs - - /* If tracing, re-enable interrupts and do it */ -save_user_nvgprs_cont: - andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SINGLESTEP) - beq 5f - + bl .save_nvgprs addi r3,r1,STACK_FRAME_OVERHEAD bl .do_syscall_trace_leave - REST_NVGPRS(r1) - clrrdi r12,r1,THREAD_SHIFT - - /* Disable interrupts again and handle other work if any */ -5: mfmsr r10 - rldicl r10,r10,48,1 - rotldi r10,r10,16 - mtmsrd r10,1 - - b .ret_from_except_lite + b .ret_from_except /* Save non-volatile GPRs, if not already saved. */ _GLOBAL(save_nvgprs) @@ -282,51 +265,6 @@ _GLOBAL(save_nvgprs) std r0,_TRAP(r1) blr - -save_user_nvgprs: - ld r10,TI_SIGFRAME(r12) - andi. r0,r9,_TIF_32BIT - beq- save_user_nvgprs_64 - - /* 32-bit save to userspace */ - -.macro savewords start, end - 1: stw \start,4*(\start)(r10) - .section __ex_table,"a" - .align 3 - .llong 1b,save_user_nvgprs_fault - .previous - .if \end - \start - savewords "(\start+1)",\end - .endif -.endm - savewords 14,31 - b save_user_nvgprs_cont - -save_user_nvgprs_64: - /* 64-bit save to userspace */ - -.macro savelongs start, end - 1: std \start,8*(\start)(r10) - .section __ex_table,"a" - .align 3 - .llong 1b,save_user_nvgprs_fault - .previous - .if \end - \start - savelongs "(\start+1)",\end - .endif -.endm - savelongs 14,31 - b save_user_nvgprs_cont - -save_user_nvgprs_fault: - li r3,11 /* SIGSEGV */ - ld r4,TI_TASK(r12) - bl .force_sigsegv - - clrrdi r12,r1,THREAD_SHIFT - ld r9,TI_FLAGS(r12) - b save_user_nvgprs_cont /* * The sigsuspend and rt_sigsuspend system calls can call do_signal @@ -352,6 +290,16 @@ _GLOBAL(ppc_clone) bl .sys_clone b syscall_exit +_GLOBAL(ppc32_swapcontext) + bl .save_nvgprs + bl .compat_sys_swapcontext + b syscall_exit + +_GLOBAL(ppc64_swapcontext) + bl .save_nvgprs + bl .sys_swapcontext + b syscall_exit + _GLOBAL(ret_from_fork) bl .schedule_tail REST_NVGPRS(r1) diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c index 400793c71304..bcb83574335b 100644 --- a/arch/powerpc/kernel/ptrace.c +++ b/arch/powerpc/kernel/ptrace.c @@ -561,10 +561,7 @@ void do_syscall_trace_leave(struct pt_regs *regs) regs->result); if ((test_thread_flag(TIF_SYSCALL_TRACE) -#ifdef CONFIG_PPC64 - || test_thread_flag(TIF_SINGLESTEP) -#endif - ) + || test_thread_flag(TIF_SINGLESTEP)) && (current->ptrace & PT_PTRACED)) do_syscall_trace(); } diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c index bd837b5dbf06..d7a4e814974d 100644 --- a/arch/powerpc/kernel/signal_32.c +++ b/arch/powerpc/kernel/signal_32.c @@ -151,10 +151,7 @@ static inline int save_general_regs(struct pt_regs *regs, elf_greg_t64 *gregs = (elf_greg_t64 *)regs; int i; - if (!FULL_REGS(regs)) { - set_thread_flag(TIF_SAVE_NVGPRS); - current_thread_info()->nvgprs_frame = frame->mc_gregs; - } + WARN_ON(!FULL_REGS(regs)); for (i = 0; i <= PT_RESULT; i ++) { if (i == 14 && !FULL_REGS(regs)) @@ -215,15 +212,7 @@ static inline int get_old_sigaction(struct k_sigaction *new_ka, static inline int save_general_regs(struct pt_regs *regs, struct mcontext __user *frame) { - if (!FULL_REGS(regs)) { - /* Zero out the unsaved GPRs to avoid information - leak, and set TIF_SAVE_NVGPRS to ensure that the - registers do actually get saved later. */ - memset(®s->gpr[14], 0, 18 * sizeof(unsigned long)); - current_thread_info()->nvgprs_frame = &frame->mc_gregs; - set_thread_flag(TIF_SAVE_NVGPRS); - } - + WARN_ON(!FULL_REGS(regs)); return __copy_to_user(&frame->mc_gregs, regs, GP_REGS_SIZE); } @@ -826,8 +815,8 @@ static int do_setcontext(struct ucontext __user *ucp, struct pt_regs *regs, int } long sys_swapcontext(struct ucontext __user *old_ctx, - struct ucontext __user *new_ctx, - int ctx_size, int r6, int r7, int r8, struct pt_regs *regs) + struct ucontext __user *new_ctx, + int ctx_size, int r6, int r7, int r8, struct pt_regs *regs) { unsigned char tmp; diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c index 497a5d3df359..4324f8a8ba24 100644 --- a/arch/powerpc/kernel/signal_64.c +++ b/arch/powerpc/kernel/signal_64.c @@ -118,14 +118,7 @@ static long setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, err |= __put_user(0, &sc->v_regs); #endif /* CONFIG_ALTIVEC */ err |= __put_user(&sc->gp_regs, &sc->regs); - if (!FULL_REGS(regs)) { - /* Zero out the unsaved GPRs to avoid information - leak, and set TIF_SAVE_NVGPRS to ensure that the - registers do actually get saved later. */ - memset(®s->gpr[14], 0, 18 * sizeof(unsigned long)); - set_thread_flag(TIF_SAVE_NVGPRS); - current_thread_info()->nvgprs_frame = &sc->gp_regs; - } + WARN_ON(!FULL_REGS(regs)); err |= __copy_to_user(&sc->gp_regs, regs, GP_REGS_SIZE); err |= __copy_to_user(&sc->fp_regs, ¤t->thread.fpr, FP_REGS_SIZE); err |= __put_user(signr, &sc->signal); diff --git a/arch/powerpc/kernel/systbl.S b/arch/powerpc/kernel/systbl.S index 8a9f994ed917..1ad55f0466fd 100644 --- a/arch/powerpc/kernel/systbl.S +++ b/arch/powerpc/kernel/systbl.S @@ -288,7 +288,7 @@ COMPAT_SYS(clock_settime) COMPAT_SYS(clock_gettime) COMPAT_SYS(clock_getres) COMPAT_SYS(clock_nanosleep) -COMPAT_SYS(swapcontext) +SYSX(ppc64_swapcontext,ppc32_swapcontext,ppc_swapcontext) COMPAT_SYS(tgkill) COMPAT_SYS(utimes) COMPAT_SYS(statfs64) -- cgit v1.2.2 From 0a45d4491d0f172e02126370f312405c5d473363 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Wed, 15 Mar 2006 13:47:15 +1100 Subject: powerpc: Fix problem with time going backwards The recent changes to keep gettimeofday in sync with xtime had the side effect that it was occasionally possible for the time reported by gettimeofday to go back by a microsecond. There were two reasons: (1) when we recalculated the offsets used by gettimeofday every 2^31 timebase ticks, we lost an accumulated fractional microsecond, and (2) because the update is done some time after the notional start of jiffy, if ntp is slowing the clock, it is possible to see time go backwards when the timebase factor gets reduced. This fixes it by (a) slowing the gettimeofday clock by about 1us in 2^31 timebase ticks (a factor of less than 1 in 3.7 million), and (b) adjusting the timebase offsets in the rare case that the gettimeofday result could possibly go backwards (i.e. when ntp is slowing the clock and the timer interrupt is late). In this case the adjustment will reduce to zero eventually because of (a). Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/time.c | 48 ++++++++++++++++++++++++++++++++-------------- 1 file changed, 34 insertions(+), 14 deletions(-) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index 2a7ddc579379..86f7e3d154d8 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -283,9 +283,9 @@ static inline void update_gtod(u64 new_tb_stamp, u64 new_stamp_xsec, * the two values of tb_update_count match and are even then the * tb_to_xs and stamp_xsec values are consistent. If not, then it * loops back and reads them again until this criteria is met. + * We expect the caller to have done the first increment of + * vdso_data->tb_update_count already. */ - ++(vdso_data->tb_update_count); - smp_wmb(); vdso_data->tb_orig_stamp = new_tb_stamp; vdso_data->stamp_xsec = new_stamp_xsec; vdso_data->tb_to_xs = new_tb_to_xs; @@ -310,20 +310,15 @@ static __inline__ void timer_recalc_offset(u64 cur_tb) unsigned long offset; u64 new_stamp_xsec; u64 tlen, t2x; + u64 tb, xsec_old, xsec_new; + struct gettimeofday_vars *varp; if (__USE_RTC()) return; tlen = current_tick_length(); offset = cur_tb - do_gtod.varp->tb_orig_stamp; - if (tlen == last_tick_len && offset < 0x80000000u) { - /* check that we're still in sync; if not, resync */ - struct timeval tv; - __do_gettimeofday(&tv, cur_tb); - if (tv.tv_sec <= xtime.tv_sec && - (tv.tv_sec < xtime.tv_sec || - tv.tv_usec * 1000 <= xtime.tv_nsec)) - return; - } + if (tlen == last_tick_len && offset < 0x80000000u) + return; if (tlen != last_tick_len) { t2x = mulhdu(tlen << TICKLEN_SHIFT, ticklen_to_xs); last_tick_len = tlen; @@ -332,6 +327,21 @@ static __inline__ void timer_recalc_offset(u64 cur_tb) new_stamp_xsec = (u64) xtime.tv_nsec * XSEC_PER_SEC; do_div(new_stamp_xsec, 1000000000); new_stamp_xsec += (u64) xtime.tv_sec * XSEC_PER_SEC; + + ++vdso_data->tb_update_count; + smp_mb(); + + /* + * Make sure time doesn't go backwards for userspace gettimeofday. + */ + tb = get_tb(); + varp = do_gtod.varp; + xsec_old = mulhdu(tb - varp->tb_orig_stamp, varp->tb_to_xs) + + varp->stamp_xsec; + xsec_new = mulhdu(tb - cur_tb, t2x) + new_stamp_xsec; + if (xsec_new < xsec_old) + new_stamp_xsec += xsec_old - xsec_new; + update_gtod(cur_tb, new_stamp_xsec, t2x); } @@ -564,6 +574,10 @@ int do_settimeofday(struct timespec *tv) } #endif + /* Make userspace gettimeofday spin until we're done. */ + ++vdso_data->tb_update_count; + smp_mb(); + /* * Subtract off the number of nanoseconds since the * beginning of the last tick. @@ -724,10 +738,16 @@ void __init time_init(void) * It is computed as: * ticklen_to_xs = 2^N / (tb_ticks_per_jiffy * 1e9) * where N = 64 + 20 - TICKLEN_SCALE - TICKLEN_SHIFT - * so as to give the result as a 0.64 fixed-point fraction. + * which turns out to be N = 51 - SHIFT_HZ. + * This gives the result as a 0.64 fixed-point fraction. + * That value is reduced by an offset amounting to 1 xsec per + * 2^31 timebase ticks to avoid problems with time going backwards + * by 1 xsec when we do timer_recalc_offset due to losing the + * fractional xsec. That offset is equal to ppc_tb_freq/2^51 + * since there are 2^20 xsec in a second. */ - div128_by_32(1ULL << (64 + 20 - TICKLEN_SCALE - TICKLEN_SHIFT), 0, - tb_ticks_per_jiffy, &res); + div128_by_32((1ULL << 51) - ppc_tb_freq, 0, + tb_ticks_per_jiffy << SHIFT_HZ, &res); div128_by_32(res.result_high, res.result_low, NSEC_PER_SEC, &res); ticklen_to_xs = res.result_low; -- cgit v1.2.2 From 0e8ed479125a6d246e17cfe6c24e140bfeb01972 Mon Sep 17 00:00:00 2001 From: Michael Neuling Date: Tue, 14 Mar 2006 17:11:51 +1100 Subject: [PATCH] powerpc: RTC memory corruption We should be memset'ing the data we are pointing to, not the pointer itself. This is in an error path so we probably don't hit it much. Signed-off-by: Michael Neuling Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/rtas-rtc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/rtas-rtc.c b/arch/powerpc/kernel/rtas-rtc.c index 635d3b9a8811..34d073fb6091 100644 --- a/arch/powerpc/kernel/rtas-rtc.c +++ b/arch/powerpc/kernel/rtas-rtc.c @@ -52,7 +52,7 @@ void rtas_get_rtc_time(struct rtc_time *rtc_tm) error = rtas_call(rtas_token("get-time-of-day"), 0, 8, ret); if (error == RTAS_CLOCK_BUSY || rtas_is_extended_busy(error)) { if (in_interrupt() && printk_ratelimit()) { - memset(&rtc_tm, 0, sizeof(struct rtc_time)); + memset(rtc_tm, 0, sizeof(struct rtc_time)); printk(KERN_WARNING "error: reading clock" " would delay interrupt\n"); return; /* delay not allowed */ -- cgit v1.2.2 From 920573bd03bf690135967b5022362d34ede589c3 Mon Sep 17 00:00:00 2001 From: Olaf Hering Date: Tue, 14 Mar 2006 21:21:11 +0100 Subject: [PATCH] powerpc: remove duplicate EXPORT_SYMBOLS remove warnings when building a 64bit kernel. smp_call_function triggers also with 32bit kernel. WARNING: vmlinux: duplicate symbol 'smp_call_function' previous definition was in vmlinux arch/powerpc/kernel/ppc_ksyms.c:164:EXPORT_SYMBOL(smp_call_function); arch/powerpc/kernel/smp.c:300:EXPORT_SYMBOL(smp_call_function); WARNING: vmlinux: duplicate symbol 'ioremap' previous definition was in vmlinux arch/powerpc/kernel/ppc_ksyms.c:113:EXPORT_SYMBOL(ioremap); arch/powerpc/mm/pgtable_64.c:321:EXPORT_SYMBOL(ioremap); WARNING: vmlinux: duplicate symbol '__ioremap' previous definition was in vmlinux arch/powerpc/kernel/ppc_ksyms.c:117:EXPORT_SYMBOL(__ioremap); arch/powerpc/mm/pgtable_64.c:322:EXPORT_SYMBOL(__ioremap); WARNING: vmlinux: duplicate symbol 'iounmap' previous definition was in vmlinux arch/powerpc/kernel/ppc_ksyms.c:118:EXPORT_SYMBOL(iounmap); arch/powerpc/mm/pgtable_64.c:323:EXPORT_SYMBOL(iounmap); Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/ppc_ksyms.c | 10 ---------- 1 file changed, 10 deletions(-) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c index 8a731ea877b7..63ecbec05202 100644 --- a/arch/powerpc/kernel/ppc_ksyms.c +++ b/arch/powerpc/kernel/ppc_ksyms.c @@ -110,15 +110,6 @@ EXPORT_SYMBOL(_insw_ns); EXPORT_SYMBOL(_outsw_ns); EXPORT_SYMBOL(_insl_ns); EXPORT_SYMBOL(_outsl_ns); -EXPORT_SYMBOL(ioremap); -#ifdef CONFIG_44x -EXPORT_SYMBOL(ioremap64); -#endif -EXPORT_SYMBOL(__ioremap); -EXPORT_SYMBOL(iounmap); -#ifdef CONFIG_PPC32 -EXPORT_SYMBOL(ioremap_bot); /* aka VMALLOC_END */ -#endif #if defined(CONFIG_PPC32) && (defined(CONFIG_BLK_DEV_IDE) || defined(CONFIG_BLK_DEV_IDE_MODULE)) EXPORT_SYMBOL(ppc_ide_md); @@ -161,7 +152,6 @@ EXPORT_SYMBOL(__flush_icache_range); EXPORT_SYMBOL(flush_dcache_range); #ifdef CONFIG_SMP -EXPORT_SYMBOL(smp_call_function); #ifdef CONFIG_PPC32 EXPORT_SYMBOL(smp_hw_index); #endif -- cgit v1.2.2 From 92eb4602eb5c37db86cd9d2b1f4c8ca304fbc49f Mon Sep 17 00:00:00 2001 From: John Rose Date: Tue, 14 Mar 2006 17:46:45 -0600 Subject: [PATCH] powerpc: properly configure DDR/P5IOC children devs The dynamic add path for PCI Host Bridges can fail to configure children adapters under P5IOC controllers. It fails to properly fixup bus/device resources, and it fails to properly enable EEH. Both of these steps need to occur before any children devices are enabled in pci_bus_add_devices(). Signed-off-by: John Rose Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/pci_64.c | 5 +++-- arch/powerpc/kernel/rtas_pci.c | 24 +----------------------- 2 files changed, 4 insertions(+), 25 deletions(-) (limited to 'arch/powerpc/kernel') diff --git a/arch/powerpc/kernel/pci_64.c b/arch/powerpc/kernel/pci_64.c index c367520bc1c3..ba92bab7cc2c 100644 --- a/arch/powerpc/kernel/pci_64.c +++ b/arch/powerpc/kernel/pci_64.c @@ -589,7 +589,6 @@ void __devinit scan_phb(struct pci_controller *hose) #endif /* CONFIG_PPC_MULTIPLATFORM */ if (mode == PCI_PROBE_NORMAL) hose->last_busno = bus->subordinate = pci_scan_child_bus(bus); - pci_bus_add_devices(bus); } static int __init pcibios_init(void) @@ -608,8 +607,10 @@ static int __init pcibios_init(void) printk("PCI: Probing PCI hardware\n"); /* Scan all of the recorded PCI controllers. */ - list_for_each_entry_safe(hose, tmp, &hose_list, list_node) + list_for_each_entry_safe(hose, tmp, &hose_list, list_node) { scan_phb(hose); + pci_bus_add_devices(hose->bus); + } #ifndef CONFIG_PPC_ISERIES if (pci_probe_only) diff --git a/arch/powerpc/kernel/rtas_pci.c b/arch/powerpc/kernel/rtas_pci.c index 5579f6559912..7442775ef2a1 100644 --- a/arch/powerpc/kernel/rtas_pci.c +++ b/arch/powerpc/kernel/rtas_pci.c @@ -280,8 +280,7 @@ static int phb_set_bus_ranges(struct device_node *dev, return 0; } -static int __devinit setup_phb(struct device_node *dev, - struct pci_controller *phb) +int __devinit setup_phb(struct device_node *dev, struct pci_controller *phb) { if (is_python(dev)) python_countermeasures(dev); @@ -359,27 +358,6 @@ unsigned long __init find_and_init_phbs(void) return 0; } -struct pci_controller * __devinit init_phb_dynamic(struct device_node *dn) -{ - struct pci_controller *phb; - int primary; - - primary = list_empty(&hose_list); - phb = pcibios_alloc_controller(dn); - if (!phb) - return NULL; - setup_phb(dn, phb); - pci_process_bridge_OF_ranges(phb, dn, primary); - - pci_setup_phb_io_dynamic(phb, primary); - - pci_devs_phb_init_dynamic(phb); - scan_phb(phb); - - return phb; -} -EXPORT_SYMBOL(init_phb_dynamic); - /* RPA-specific bits for removing PHBs */ int pcibios_remove_root_bus(struct pci_controller *phb) { -- cgit v1.2.2