X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=arch%2Fsparc64%2Fmm%2Ffault.c;h=a7372dbb98545b83b6524d44033448e3a0d407a4;hb=97bf2856c6014879bd04983a3e9dfcdac1e7fe85;hp=49c3dd29a5b69ed39ed1b7fe73c3a22b0221c689;hpb=9bf4aaab3e101692164d49b7ca357651eb691cb6;p=linux-2.6.git diff --git a/arch/sparc64/mm/fault.c b/arch/sparc64/mm/fault.c index 49c3dd29a..a7372dbb9 100644 --- a/arch/sparc64/mm/fault.c +++ b/arch/sparc64/mm/fault.c @@ -18,6 +18,8 @@ #include #include #include +#include +#include #include #include @@ -27,26 +29,42 @@ #include #include #include +#include +#include -#define ELEMENTS(arr) (sizeof (arr)/sizeof (arr[0])) +#ifdef CONFIG_KPROBES +ATOMIC_NOTIFIER_HEAD(notify_page_fault_chain); -extern struct sparc_phys_banks sp_banks[SPARC_PHYS_BANKS]; +/* Hook to register for page fault notifications */ +int register_page_fault_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_register(¬ify_page_fault_chain, nb); +} -/* - * To debug kernel during syscall entry. - */ -void syscall_trace_entry(struct pt_regs *regs) +int unregister_page_fault_notifier(struct notifier_block *nb) { - printk("scall entry: %s[%d]/cpu%d: %d\n", current->comm, current->pid, smp_processor_id(), (int) regs->u_regs[UREG_G1]); + return atomic_notifier_chain_unregister(¬ify_page_fault_chain, nb); } -/* - * To debug kernel during syscall exit. - */ -void syscall_trace_exit(struct pt_regs *regs) +static inline int notify_page_fault(enum die_val val, const char *str, + struct pt_regs *regs, long err, int trap, int sig) +{ + struct die_args args = { + .regs = regs, + .str = str, + .err = err, + .trapnr = trap, + .signr = sig + }; + return atomic_notifier_call_chain(¬ify_page_fault_chain, val, &args); +} +#else +static inline int notify_page_fault(enum die_val val, const char *str, + struct pt_regs *regs, long err, int trap, int sig) { - printk("scall exit: %s[%d]/cpu%d: %d\n", current->comm, current->pid, smp_processor_id(), (int) regs->u_regs[UREG_G1]); + return NOTIFY_DONE; } +#endif /* * To debug kernel to catch accesses to certain virtual/physical addresses. @@ -85,55 +103,9 @@ void set_brkpt(unsigned long addr, unsigned char mask, int flags, int mode) : "memory"); } -/* Nice, simple, prom library does all the sweating for us. ;) */ -unsigned long __init prom_probe_memory (void) -{ - register struct linux_mlist_p1275 *mlist; - register unsigned long bytes, base_paddr, tally; - register int i; - - i = 0; - mlist = *prom_meminfo()->p1275_available; - bytes = tally = mlist->num_bytes; - base_paddr = mlist->start_adr; - - sp_banks[0].base_addr = base_paddr; - sp_banks[0].num_bytes = bytes; - - while (mlist->theres_more != (void *) 0) { - i++; - mlist = mlist->theres_more; - bytes = mlist->num_bytes; - tally += bytes; - if (i >= SPARC_PHYS_BANKS-1) { - printk ("The machine has more banks than " - "this kernel can support\n" - "Increase the SPARC_PHYS_BANKS " - "setting (currently %d)\n", - SPARC_PHYS_BANKS); - i = SPARC_PHYS_BANKS-1; - break; - } - - sp_banks[i].base_addr = mlist->start_adr; - sp_banks[i].num_bytes = mlist->num_bytes; - } - - i++; - sp_banks[i].base_addr = 0xdeadbeefbeefdeadUL; - sp_banks[i].num_bytes = 0; - - /* Now mask all bank sizes on a page boundary, it is all we can - * use anyways. - */ - for (i = 0; sp_banks[i].num_bytes != 0; i++) - sp_banks[i].num_bytes &= PAGE_MASK; - - return tally; -} - -static void unhandled_fault(unsigned long address, struct task_struct *tsk, - struct pt_regs *regs) +static void __kprobes unhandled_fault(unsigned long address, + struct task_struct *tsk, + struct pt_regs *regs) { if ((unsigned long) address < PAGE_SIZE) { printk(KERN_ALERT "Unable to handle kernel NULL " @@ -143,19 +115,27 @@ static void unhandled_fault(unsigned long address, struct task_struct *tsk, "at virtual address %016lx\n", (unsigned long)address); } printk(KERN_ALERT "tsk->{mm,active_mm}->context = %016lx\n", - (tsk->mm ? tsk->mm->context : tsk->active_mm->context)); + (tsk->mm ? + CTX_HWBITS(tsk->mm->context) : + CTX_HWBITS(tsk->active_mm->context))); printk(KERN_ALERT "tsk->{mm,active_mm}->pgd = %016lx\n", (tsk->mm ? (unsigned long) tsk->mm->pgd : (unsigned long) tsk->active_mm->pgd)); + if (notify_die(DIE_GPF, "general protection fault", regs, + 0, 0, SIGSEGV) == NOTIFY_STOP) + return; die_if_kernel("Oops", regs); } -static void bad_kernel_pc(struct pt_regs *regs) +static void bad_kernel_pc(struct pt_regs *regs, unsigned long vaddr) { unsigned long *ksp; printk(KERN_CRIT "OOPS: Bogus kernel PC [%016lx] in fault handler\n", regs->tpc); + printk(KERN_CRIT "OOPS: RPC [%016lx]\n", regs->u_regs[15]); + print_symbol("RPC: <%s>\n", regs->u_regs[15]); + printk(KERN_CRIT "OOPS: Fault was to vaddr[%lx]\n", vaddr); __asm__("mov %%sp, %0" : "=r" (ksp)); show_stack(current, ksp); unhandled_fault(regs->tpc, current, regs); @@ -166,11 +146,12 @@ static void bad_kernel_pc(struct pt_regs *regs) * this. Additionally, to prevent kswapd from ripping ptes from * under us, raise interrupts around the time that we look at the * pte, kswapd will have to wait to get his smp ipi response from - * us. This saves us having to get page_table_lock. + * us. vmtruncate likewise. This saves us having to get pte lock. */ static unsigned int get_user_insn(unsigned long tpc) { pgd_t *pgdp = pgd_offset(current->mm, tpc); + pud_t *pudp; pmd_t *pmdp; pte_t *ptep, pte; unsigned long pa; @@ -179,7 +160,10 @@ static unsigned int get_user_insn(unsigned long tpc) if (pgd_none(*pgdp)) goto outret; - pmdp = pmd_offset(pgdp, tpc); + pudp = pud_offset(pgdp, tpc); + if (pud_none(*pudp)) + goto outret; + pmdp = pmd_offset(pudp, tpc); if (pmd_none(*pmdp)) goto outret; @@ -192,7 +176,7 @@ static unsigned int get_user_insn(unsigned long tpc) if (!pte_present(pte)) goto out; - pa = (pte_val(pte) & _PAGE_PADDR); + pa = (pte_pfn(pte) << PAGE_SHIFT); pa += (tpc & ~PAGE_MASK); /* Use phys bypass so we don't pollute dtlb/dcache. */ @@ -246,7 +230,6 @@ static unsigned int get_fault_insn(struct pt_regs *regs, unsigned int insn) static void do_kernel_fault(struct pt_regs *regs, int si_code, int fault_code, unsigned int insn, unsigned long address) { - unsigned long g2; unsigned char asi = ASI_P; if ((!insn) && (regs->tstate & TSTATE_PRIV)) @@ -277,11 +260,9 @@ static void do_kernel_fault(struct pt_regs *regs, int si_code, int fault_code, } } - g2 = regs->u_regs[UREG_G2]; - /* Is this in ex_table? */ if (regs->tstate & TSTATE_PRIV) { - unsigned long fixup; + const struct exception_table_entry *entry; if (asi == ASI_P && (insn & 0xc0800000) == 0xc0800000) { if (insn & 0x2000) @@ -292,10 +273,9 @@ static void do_kernel_fault(struct pt_regs *regs, int si_code, int fault_code, /* Look in asi.h: All _S asis have LS bit set */ if ((asi & 0x1) && - (fixup = search_extables_range(regs->tpc, &g2))) { - regs->tpc = fixup; + (entry = search_exception_tables(regs->tpc))) { + regs->tpc = entry->fixup; regs->tnpc = regs->tpc + 4; - regs->u_regs[UREG_G2] = g2; return; } } else { @@ -310,16 +290,21 @@ cannot_handle: unhandled_fault (address, current, regs); } -asmlinkage void do_sparc64_fault(struct pt_regs *regs) +asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs) { struct mm_struct *mm = current->mm; struct vm_area_struct *vma; unsigned int insn = 0; int si_code, fault_code; - unsigned long address; + unsigned long address, mm_rss; - si_code = SEGV_MAPERR; fault_code = get_thread_fault_code(); + + if (notify_page_fault(DIE_PAGE_FAULT, "page_fault", regs, + fault_code, 0, SIGSEGV) == NOTIFY_STOP) + return; + + si_code = SEGV_MAPERR; address = current_thread_info()->fault_address; if ((fault_code & FAULT_CODE_ITLB) && @@ -334,7 +319,7 @@ asmlinkage void do_sparc64_fault(struct pt_regs *regs) (tpc >= MODULES_VADDR && tpc < MODULES_END)) { /* Valid, no problems... */ } else { - bad_kernel_pc(regs); + bad_kernel_pc(regs, address); return; } } @@ -343,7 +328,7 @@ asmlinkage void do_sparc64_fault(struct pt_regs *regs) * If we're in an interrupt or have no user * context, we must not take the fault.. */ - if (in_interrupt() || !mm) + if (in_atomic() || !mm) goto intr_or_no_mm; if (test_thread_flag(TIF_32BIT)) { @@ -352,7 +337,15 @@ asmlinkage void do_sparc64_fault(struct pt_regs *regs) address &= 0xffffffff; } - down_read(&mm->mmap_sem); + if (!down_read_trylock(&mm->mmap_sem)) { + if ((regs->tstate & TSTATE_PRIV) && + !search_exception_tables(regs->tpc)) { + insn = get_fault_insn(regs, insn); + goto handle_kernel_fault; + } + down_read(&mm->mmap_sem); + } + vma = find_vma(mm, address); if (!vma) goto bad_area; @@ -371,8 +364,12 @@ asmlinkage void do_sparc64_fault(struct pt_regs *regs) insn = get_fault_insn(regs, 0); if (!insn) goto continue_fault; + /* All loads, stores and atomics have bits 30 and 31 both set + * in the instruction. Bit 21 is set in all stores, but we + * have to avoid prefetches which also have bit 21 set. + */ if ((insn & 0xc0200000) == 0xc0200000 && - (insn & 0x1780000) != 0x1680000) { + (insn & 0x01780000) != 0x01680000) { /* Don't bother updating thread struct value, * because update_mmu_cache only cares which tlb * the access came from. @@ -437,20 +434,36 @@ good_area: } switch (handle_mm_fault(mm, vma, address, (fault_code & FAULT_CODE_WRITE))) { - case 1: + case VM_FAULT_MINOR: current->min_flt++; break; - case 2: + case VM_FAULT_MAJOR: current->maj_flt++; break; - case 0: + case VM_FAULT_SIGBUS: goto do_sigbus; - default: + case VM_FAULT_OOM: goto out_of_memory; + default: + BUG(); } up_read(&mm->mmap_sem); - goto fault_done; + + mm_rss = get_mm_rss(mm); +#ifdef CONFIG_HUGETLB_PAGE + mm_rss -= (mm->context.huge_pte_count * (HPAGE_SIZE / PAGE_SIZE)); +#endif + if (unlikely(mm_rss > + mm->context.tsb_block[MM_TSB_BASE].tsb_rss_limit)) + tsb_grow(mm, MM_TSB_BASE, mm_rss); +#ifdef CONFIG_HUGETLB_PAGE + mm_rss = mm->context.huge_pte_count; + if (unlikely(mm_rss > + mm->context.tsb_block[MM_TSB_HUGE].tsb_rss_limit)) + tsb_grow(mm, MM_TSB_HUGE, mm_rss); +#endif + return; /* * Something tried to access memory that isn't in our memory map.. @@ -462,8 +475,7 @@ bad_area: handle_kernel_fault: do_kernel_fault(regs, si_code, fault_code, insn, address); - - goto fault_done; + return; /* * We ran out of memory, or some other thing happened to us that made @@ -472,7 +484,8 @@ handle_kernel_fault: out_of_memory: insn = get_fault_insn(regs, insn); up_read(&mm->mmap_sem); - printk("VM: killing process %s\n", current->comm); + printk("VM: killing process %s(%d:#%u)\n", + current->comm, current->pid, current->xid); if (!(regs->tstate & TSTATE_PRIV)) do_exit(SIGKILL); goto handle_kernel_fault; @@ -494,9 +507,4 @@ do_sigbus: /* Kernel mode? Handle exceptions or die */ if (regs->tstate & TSTATE_PRIV) goto handle_kernel_fault; - -fault_done: - /* These values are no longer needed, clear them. */ - set_thread_fault_code(0); - current_thread_info()->fault_address = 0; }