diff options
Diffstat (limited to 'arch/ia64')
40 files changed, 1071 insertions, 497 deletions
diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 75d8397..f1d2899 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -11,6 +11,8 @@ menu "Processor type and features" config IA64 bool + select PCI if (!IA64_HP_SIM) + select ACPI if (!IA64_HP_SIM) default y help The Itanium Processor Family is Intel's 64-bit successor to @@ -28,12 +30,19 @@ config MMU config SWIOTLB bool - default y config RWSEM_XCHGADD_ALGORITHM bool default y +config ARCH_HAS_ILOG2_U32 + bool + default n + +config ARCH_HAS_ILOG2_U64 + bool + default n + config GENERIC_FIND_NEXT_BIT bool default y @@ -76,10 +85,9 @@ choice config IA64_GENERIC bool "generic" - select ACPI - select PCI select NUMA select ACPI_NUMA + select SWIOTLB help This selects the system type of your hardware. A "generic" kernel will run on any supported IA-64 system. However, if you configure @@ -96,6 +104,7 @@ config IA64_GENERIC config IA64_DIG bool "DIG-compliant" + select SWIOTLB config IA64_HP_ZX1 bool "HP-zx1/sx1000" @@ -105,6 +114,7 @@ config IA64_HP_ZX1 config IA64_HP_ZX1_SWIOTLB bool "HP-zx1/sx1000 with software I/O TLB" + select SWIOTLB help Build a kernel that runs on HP zx1 and sx1000 systems even when they have broken PCI devices which cannot DMA to full 32 bits. Apart @@ -123,6 +133,7 @@ config IA64_SGI_SN2 config IA64_HP_SIM bool "Ski-simulator" + select SWIOTLB endchoice diff --git a/arch/ia64/hp/common/hwsw_iommu.c b/arch/ia64/hp/common/hwsw_iommu.c index a5a5637..2153bca 100644 --- a/arch/ia64/hp/common/hwsw_iommu.c +++ b/arch/ia64/hp/common/hwsw_iommu.c @@ -192,3 +192,7 @@ EXPORT_SYMBOL(hwsw_unmap_sg); EXPORT_SYMBOL(hwsw_dma_supported); EXPORT_SYMBOL(hwsw_alloc_coherent); EXPORT_SYMBOL(hwsw_free_coherent); +EXPORT_SYMBOL(hwsw_sync_single_for_cpu); +EXPORT_SYMBOL(hwsw_sync_single_for_device); +EXPORT_SYMBOL(hwsw_sync_sg_for_cpu); +EXPORT_SYMBOL(hwsw_sync_sg_for_device); diff --git a/arch/ia64/hp/sim/simserial.c b/arch/ia64/hp/sim/simserial.c index 1f16ebb9..324ea75 100644 --- a/arch/ia64/hp/sim/simserial.c +++ b/arch/ia64/hp/sim/simserial.c @@ -488,7 +488,7 @@ static int rs_ioctl(struct tty_struct *tty, struct file * file, #define RELEVANT_IFLAG(iflag) (iflag & (IGNBRK|BRKINT|IGNPAR|PARMRK|INPCK)) -static void rs_set_termios(struct tty_struct *tty, struct termios *old_termios) +static void rs_set_termios(struct tty_struct *tty, struct ktermios *old_termios) { unsigned int cflag = tty->termios->c_cflag; diff --git a/arch/ia64/ia32/sys_ia32.c b/arch/ia64/ia32/sys_ia32.c index a4a6e14..957681c 100644 --- a/arch/ia64/ia32/sys_ia32.c +++ b/arch/ia64/ia32/sys_ia32.c @@ -235,7 +235,7 @@ mmap_subpage (struct file *file, unsigned long start, unsigned long end, int pro if (!(flags & MAP_ANONYMOUS)) { /* read the file contents */ - inode = file->f_dentry->d_inode; + inode = file->f_path.dentry->d_inode; if (!inode->i_fop || !file->f_op->read || ((*file->f_op->read)(file, (char __user *) start, end - start, &off) < 0)) { @@ -837,7 +837,7 @@ emulate_mmap (struct file *file, unsigned long start, unsigned long len, int pro if (!is_congruent) { /* read the file contents */ - inode = file->f_dentry->d_inode; + inode = file->f_path.dentry->d_inode; if (!inode->i_fop || !file->f_op->read || ((*file->f_op->read)(file, (char __user *) pstart, pend - pstart, &poff) < 0)) diff --git a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile index 8ae384e..098ee60 100644 --- a/arch/ia64/kernel/Makefile +++ b/arch/ia64/kernel/Makefile @@ -29,6 +29,7 @@ obj-$(CONFIG_CPU_FREQ) += cpufreq/ obj-$(CONFIG_IA64_MCA_RECOVERY) += mca_recovery.o obj-$(CONFIG_KPROBES) += kprobes.o jprobes.o obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o +obj-$(CONFIG_CRASH_DUMP) += crash_dump.o obj-$(CONFIG_IA64_UNCACHED_ALLOCATOR) += uncached.o obj-$(CONFIG_AUDIT) += audit.o obj-$(CONFIG_PCI_MSI) += msi_ia64.o diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c index 73ef4a85..9197d7b 100644 --- a/arch/ia64/kernel/acpi.c +++ b/arch/ia64/kernel/acpi.c @@ -55,7 +55,7 @@ #define BAD_MADT_ENTRY(entry, end) ( \ (!entry) || (unsigned long)entry + sizeof(*entry) > end || \ - ((acpi_table_entry_header *)entry)->length < sizeof(*entry)) + ((struct acpi_subtable_header *)entry)->length < sizeof(*entry)) #define PREFIX "ACPI: " @@ -67,16 +67,11 @@ EXPORT_SYMBOL(pm_power_off); unsigned int acpi_cpei_override; unsigned int acpi_cpei_phys_cpuid; -#define MAX_SAPICS 256 -u16 ia64_acpiid_to_sapicid[MAX_SAPICS] = {[0 ... MAX_SAPICS - 1] = -1 }; - -EXPORT_SYMBOL(ia64_acpiid_to_sapicid); - const char *acpi_get_sysname(void) { #ifdef CONFIG_IA64_GENERIC unsigned long rsdp_phys; - struct acpi20_table_rsdp *rsdp; + struct acpi_table_rsdp *rsdp; struct acpi_table_xsdt *xsdt; struct acpi_table_header *hdr; @@ -87,16 +82,16 @@ const char *acpi_get_sysname(void) return "dig"; } - rsdp = (struct acpi20_table_rsdp *)__va(rsdp_phys); - if (strncmp(rsdp->signature, RSDP_SIG, sizeof(RSDP_SIG) - 1)) { + rsdp = (struct acpi_table_rsdp *)__va(rsdp_phys); + if (strncmp(rsdp->signature, ACPI_SIG_RSDP, sizeof(ACPI_SIG_RSDP) - 1)) { printk(KERN_ERR "ACPI 2.0 RSDP signature incorrect, default to \"dig\"\n"); return "dig"; } - xsdt = (struct acpi_table_xsdt *)__va(rsdp->xsdt_address); + xsdt = (struct acpi_table_xsdt *)__va(rsdp->xsdt_physical_address); hdr = &xsdt->header; - if (strncmp(hdr->signature, XSDT_SIG, sizeof(XSDT_SIG) - 1)) { + if (strncmp(hdr->signature, ACPI_SIG_XSDT, sizeof(ACPI_SIG_XSDT) - 1)) { printk(KERN_ERR "ACPI 2.0 XSDT signature incorrect, default to \"dig\"\n"); return "dig"; @@ -169,12 +164,12 @@ struct acpi_table_madt *acpi_madt __initdata; static u8 has_8259; static int __init -acpi_parse_lapic_addr_ovr(acpi_table_entry_header * header, +acpi_parse_lapic_addr_ovr(struct acpi_subtable_header * header, const unsigned long end) { - struct acpi_table_lapic_addr_ovr *lapic; + struct acpi_madt_local_apic_override *lapic; - lapic = (struct acpi_table_lapic_addr_ovr *)header; + lapic = (struct acpi_madt_local_apic_override *)header; if (BAD_MADT_ENTRY(lapic, end)) return -EINVAL; @@ -187,22 +182,19 @@ acpi_parse_lapic_addr_ovr(acpi_table_entry_header * header, } static int __init -acpi_parse_lsapic(acpi_table_entry_header * header, const unsigned long end) +acpi_parse_lsapic(struct acpi_subtable_header * header, const unsigned long end) { - struct acpi_table_lsapic *lsapic; + struct acpi_madt_local_sapic *lsapic; - lsapic = (struct acpi_table_lsapic *)header; + lsapic = (struct acpi_madt_local_sapic *)header; - if (BAD_MADT_ENTRY(lsapic, end)) - return -EINVAL; + /*Skip BAD_MADT_ENTRY check, as lsapic size could vary */ - if (lsapic->flags.enabled) { + if (lsapic->lapic_flags & ACPI_MADT_ENABLED) { #ifdef CONFIG_SMP smp_boot_data.cpu_phys_id[available_cpus] = (lsapic->id << 8) | lsapic->eid; #endif - ia64_acpiid_to_sapicid[lsapic->acpi_id] = - (lsapic->id << 8) | lsapic->eid; ++available_cpus; } @@ -211,11 +203,11 @@ acpi_parse_lsapic(acpi_table_entry_header * header, const unsigned long end) } static int __init -acpi_parse_lapic_nmi(acpi_table_entry_header * header, const unsigned long end) +acpi_parse_lapic_nmi(struct acpi_subtable_header * header, const unsigned long end) { - struct acpi_table_lapic_nmi *lacpi_nmi; + struct acpi_madt_local_apic_nmi *lacpi_nmi; - lacpi_nmi = (struct acpi_table_lapic_nmi *)header; + lacpi_nmi = (struct acpi_madt_local_apic_nmi *)header; if (BAD_MADT_ENTRY(lacpi_nmi, end)) return -EINVAL; @@ -225,11 +217,11 @@ acpi_parse_lapic_nmi(acpi_table_entry_header * header, const unsigned long end) } static int __init -acpi_parse_iosapic(acpi_table_entry_header * header, const unsigned long end) +acpi_parse_iosapic(struct acpi_subtable_header * header, const unsigned long end) { - struct acpi_table_iosapic *iosapic; + struct acpi_madt_io_sapic *iosapic; - iosapic = (struct acpi_table_iosapic *)header; + iosapic = (struct acpi_madt_io_sapic *)header; if (BAD_MADT_ENTRY(iosapic, end)) return -EINVAL; @@ -240,13 +232,13 @@ acpi_parse_iosapic(acpi_table_entry_header * header, const unsigned long end) static unsigned int __initdata acpi_madt_rev; static int __init -acpi_parse_plat_int_src(acpi_table_entry_header * header, +acpi_parse_plat_int_src(struct acpi_subtable_header * header, const unsigned long end) { - struct acpi_table_plat_int_src *plintsrc; + struct acpi_madt_interrupt_source *plintsrc; int vector; - plintsrc = (struct acpi_table_plat_int_src *)header; + plintsrc = (struct acpi_madt_interrupt_source *)header; if (BAD_MADT_ENTRY(plintsrc, end)) return -EINVAL; @@ -257,19 +249,19 @@ acpi_parse_plat_int_src(acpi_table_entry_header * header, */ vector = iosapic_register_platform_intr(plintsrc->type, plintsrc->global_irq, - plintsrc->iosapic_vector, + plintsrc->io_sapic_vector, plintsrc->eid, plintsrc->id, - (plintsrc->flags.polarity == - 1) ? IOSAPIC_POL_HIGH : - IOSAPIC_POL_LOW, - (plintsrc->flags.trigger == - 1) ? IOSAPIC_EDGE : - IOSAPIC_LEVEL); + ((plintsrc->inti_flags & ACPI_MADT_POLARITY_MASK) == + ACPI_MADT_POLARITY_ACTIVE_HIGH) ? + IOSAPIC_POL_HIGH : IOSAPIC_POL_LOW, + ((plintsrc->inti_flags & ACPI_MADT_TRIGGER_MASK) == + ACPI_MADT_TRIGGER_EDGE) ? + IOSAPIC_EDGE : IOSAPIC_LEVEL); platform_intr_list[plintsrc->type] = vector; if (acpi_madt_rev > 1) { - acpi_cpei_override = plintsrc->plint_flags.cpei_override_flag; + acpi_cpei_override = plintsrc->flags & ACPI_MADT_CPEI_OVERRIDE; } /* @@ -324,30 +316,32 @@ unsigned int get_cpei_target_cpu(void) } static int __init -acpi_parse_int_src_ovr(acpi_table_entry_header * header, +acpi_parse_int_src_ovr(struct acpi_subtable_header * header, const unsigned long end) { - struct acpi_table_int_src_ovr *p; + struct acpi_madt_interrupt_override *p; - p = (struct acpi_table_int_src_ovr *)header; + p = (struct acpi_madt_interrupt_override *)header; if (BAD_MADT_ENTRY(p, end)) return -EINVAL; - iosapic_override_isa_irq(p->bus_irq, p->global_irq, - (p->flags.polarity == - 1) ? IOSAPIC_POL_HIGH : IOSAPIC_POL_LOW, - (p->flags.trigger == - 1) ? IOSAPIC_EDGE : IOSAPIC_LEVEL); + iosapic_override_isa_irq(p->source_irq, p->global_irq, + ((p->inti_flags & ACPI_MADT_POLARITY_MASK) == + ACPI_MADT_POLARITY_ACTIVE_HIGH) ? + IOSAPIC_POL_HIGH : IOSAPIC_POL_LOW, + ((p->inti_flags & ACPI_MADT_TRIGGER_MASK) == + ACPI_MADT_TRIGGER_EDGE) ? + IOSAPIC_EDGE : IOSAPIC_LEVEL); return 0; } static int __init -acpi_parse_nmi_src(acpi_table_entry_header * header, const unsigned long end) +acpi_parse_nmi_src(struct acpi_subtable_header * header, const unsigned long end) { - struct acpi_table_nmi_src *nmi_src; + struct acpi_madt_nmi_source *nmi_src; - nmi_src = (struct acpi_table_nmi_src *)header; + nmi_src = (struct acpi_madt_nmi_source *)header; if (BAD_MADT_ENTRY(nmi_src, end)) return -EINVAL; @@ -371,12 +365,12 @@ static void __init acpi_madt_oem_check(char *oem_id, char *oem_table_id) } } -static int __init acpi_parse_madt(unsigned long phys_addr, unsigned long size) +static int __init acpi_parse_madt(struct acpi_table_header *table) { - if (!phys_addr || !size) + if (!table) return -EINVAL; - acpi_madt = (struct acpi_table_madt *)__va(phys_addr); + acpi_madt = (struct acpi_table_madt *)table; acpi_madt_rev = acpi_madt->header.revision; @@ -384,14 +378,14 @@ static int __init acpi_parse_madt(unsigned long phys_addr, unsigned long size) #ifdef CONFIG_ITANIUM has_8259 = 1; /* Firmware on old Itanium systems is broken */ #else - has_8259 = acpi_madt->flags.pcat_compat; + has_8259 = acpi_madt->flags & ACPI_MADT_PCAT_COMPAT; #endif iosapic_system_init(has_8259); /* Get base address of IPI Message Block */ - if (acpi_madt->lapic_address) - ipi_base_addr = ioremap(acpi_madt->lapic_address, 0); + if (acpi_madt->address) + ipi_base_addr = ioremap(acpi_madt->address, 0); printk(KERN_INFO PREFIX "Local APIC address %p\n", ipi_base_addr); @@ -413,23 +407,24 @@ static u32 __devinitdata pxm_flag[PXM_FLAG_LEN]; #define pxm_bit_test(bit) (test_bit(bit,(void *)pxm_flag)) static struct acpi_table_slit __initdata *slit_table; -static int get_processor_proximity_domain(struct acpi_table_processor_affinity *pa) +static int get_processor_proximity_domain(struct acpi_srat_cpu_affinity *pa) { int pxm; - pxm = pa->proximity_domain; + pxm = pa->proximity_domain_lo; if (ia64_platform_is("sn2")) - pxm += pa->reserved[0] << 8; + pxm += pa->proximity_domain_hi[0] << 8; return pxm; } -static int get_memory_proximity_domain(struct acpi_table_memory_affinity *ma) +static int get_memory_proximity_domain(struct acpi_srat_mem_affinity *ma) { int pxm; pxm = ma->proximity_domain; - if (ia64_platform_is("sn2")) - pxm += ma->reserved1[0] << 8; + if (!ia64_platform_is("sn2")) + pxm &= 0xff; + return pxm; } @@ -442,7 +437,7 @@ void __init acpi_numa_slit_init(struct acpi_table_slit *slit) u32 len; len = sizeof(struct acpi_table_header) + 8 - + slit->localities * slit->localities; + + slit->locality_count * slit->locality_count; if (slit->header.length != len) { printk(KERN_ERR "ACPI 2.0 SLIT: size mismatch: %d expected, %d actual\n", @@ -454,11 +449,11 @@ void __init acpi_numa_slit_init(struct acpi_table_slit *slit) } void __init -acpi_numa_processor_affinity_init(struct acpi_table_processor_affinity *pa) +acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) { int pxm; - if (!pa->flags.enabled) + if (!(pa->flags & ACPI_SRAT_CPU_ENABLED)) return; pxm = get_processor_proximity_domain(pa); @@ -467,14 +462,14 @@ acpi_numa_processor_affinity_init(struct acpi_table_processor_affinity *pa) pxm_bit_set(pxm); node_cpuid[srat_num_cpus].phys_id = - (pa->apic_id << 8) | (pa->lsapic_eid); + (pa->apic_id << 8) | (pa->local_sapic_eid); /* nid should be overridden as logical node id later */ node_cpuid[srat_num_cpus].nid = pxm; srat_num_cpus++; } void __init -acpi_numa_memory_affinity_init(struct acpi_table_memory_affinity *ma) +acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma) { unsigned long paddr, size; int pxm; @@ -483,13 +478,11 @@ acpi_numa_memory_affinity_init(struct acpi_table_memory_affinity *ma) pxm = get_memory_proximity_domain(ma); /* fill node memory chunk structure */ - paddr = ma->base_addr_hi; - paddr = (paddr << 32) | ma->base_addr_lo; - size = ma->length_hi; - size = (size << 32) | ma->length_lo; + paddr = ma->base_address; + size = ma->length; /* Ignore disabled entries */ - if (!ma->flags.enabled) + if (!(ma->flags & ACPI_SRAT_MEM_ENABLED)) return; /* record this node in proximity bitmap */ @@ -560,16 +553,16 @@ void __init acpi_numa_arch_fixup(void) if (!slit_table) return; memset(numa_slit, -1, sizeof(numa_slit)); - for (i = 0; i < slit_table->localities; i++) { + for (i = 0; i < slit_table->locality_count; i++) { if (!pxm_bit_test(i)) continue; node_from = pxm_to_node(i); - for (j = 0; j < slit_table->localities; j++) { + for (j = 0; j < slit_table->locality_count; j++) { if (!pxm_bit_test(j)) continue; node_to = pxm_to_node(j); node_distance(node_from, node_to) = - slit_table->entry[i * slit_table->localities + j]; + slit_table->entry[i * slit_table->locality_count + j]; } } @@ -590,6 +583,9 @@ void __init acpi_numa_arch_fixup(void) */ int acpi_register_gsi(u32 gsi, int triggering, int polarity) { + if (acpi_irq_model == ACPI_IRQ_MODEL_PLATFORM) + return gsi; + if (has_8259 && gsi < 16) return isa_irq_to_vector(gsi); @@ -606,26 +602,29 @@ EXPORT_SYMBOL(acpi_register_gsi); void acpi_unregister_gsi(u32 gsi) { + if (acpi_irq_model == ACPI_IRQ_MODEL_PLATFORM) + return; + iosapic_unregister_intr(gsi); } EXPORT_SYMBOL(acpi_unregister_gsi); -static int __init acpi_parse_fadt(unsigned long phys_addr, unsigned long size) +static int __init acpi_parse_fadt(struct acpi_table_header *table) { struct acpi_table_header *fadt_header; - struct fadt_descriptor *fadt; + struct acpi_table_fadt *fadt; - if (!phys_addr || !size) + if (!table) return -EINVAL; - fadt_header = (struct acpi_table_header *)__va(phys_addr); + fadt_header = (struct acpi_table_header *)table; if (fadt_header->revision != 3) return -ENODEV; /* Only deal with ACPI 2.0 FADT */ - fadt = (struct fadt_descriptor *)fadt_header; + fadt = (struct acpi_table_fadt *)fadt_header; - acpi_register_gsi(fadt->sci_int, ACPI_LEVEL_SENSITIVE, ACPI_ACTIVE_LOW); + acpi_register_gsi(fadt->sci_interrupt, ACPI_LEVEL_SENSITIVE, ACPI_ACTIVE_LOW); return 0; } @@ -652,7 +651,7 @@ int __init acpi_boot_init(void) * information -- the successor to MPS tables. */ - if (acpi_table_parse(ACPI_APIC, acpi_parse_madt) < 1) { + if (acpi_table_parse(ACPI_SIG_MADT, acpi_parse_madt) < 1) { printk(KERN_ERR PREFIX "Can't find MADT\n"); goto skip_madt; } @@ -660,40 +659,40 @@ int __init acpi_boot_init(void) /* Local APIC */ if (acpi_table_parse_madt - (ACPI_MADT_LAPIC_ADDR_OVR, acpi_parse_lapic_addr_ovr, 0) < 0) + (ACPI_MADT_TYPE_LOCAL_APIC_OVERRIDE, acpi_parse_lapic_addr_ovr, 0) < 0) printk(KERN_ERR PREFIX "Error parsing LAPIC address override entry\n"); - if (acpi_table_parse_madt(ACPI_MADT_LSAPIC, acpi_parse_lsapic, NR_CPUS) + if (acpi_table_parse_madt(ACPI_MADT_TYPE_LOCAL_SAPIC, acpi_parse_lsapic, NR_CPUS) < 1) printk(KERN_ERR PREFIX "Error parsing MADT - no LAPIC entries\n"); - if (acpi_table_parse_madt(ACPI_MADT_LAPIC_NMI, acpi_parse_lapic_nmi, 0) + if (acpi_table_parse_madt(ACPI_MADT_TYPE_LOCAL_APIC_NMI, acpi_parse_lapic_nmi, 0) < 0) printk(KERN_ERR PREFIX "Error parsing LAPIC NMI entry\n"); /* I/O APIC */ if (acpi_table_parse_madt - (ACPI_MADT_IOSAPIC, acpi_parse_iosapic, NR_IOSAPICS) < 1) + (ACPI_MADT_TYPE_IO_SAPIC, acpi_parse_iosapic, NR_IOSAPICS) < 1) printk(KERN_ERR PREFIX "Error parsing MADT - no IOSAPIC entries\n"); /* System-Level Interrupt Routing */ if (acpi_table_parse_madt - (ACPI_MADT_PLAT_INT_SRC, acpi_parse_plat_int_src, + (ACPI_MADT_TYPE_INTERRUPT_SOURCE, acpi_parse_plat_int_src, ACPI_MAX_PLATFORM_INTERRUPTS) < 0) printk(KERN_ERR PREFIX "Error parsing platform interrupt source entry\n"); if (acpi_table_parse_madt - (ACPI_MADT_INT_SRC_OVR, acpi_parse_int_src_ovr, 0) < 0) + (ACPI_MADT_TYPE_INTERRUPT_OVERRIDE, acpi_parse_int_src_ovr, 0) < 0) printk(KERN_ERR PREFIX "Error parsing interrupt source overrides entry\n"); - if (acpi_table_parse_madt(ACPI_MADT_NMI_SRC, acpi_parse_nmi_src, 0) < 0) + if (acpi_table_parse_madt(ACPI_MADT_TYPE_NMI_SOURCE, acpi_parse_nmi_src, 0) < 0) printk(KERN_ERR PREFIX "Error parsing NMI SRC entry\n"); skip_madt: @@ -703,7 +702,7 @@ int __init acpi_boot_init(void) * gets interrupts such as power and sleep buttons. If it's not * on a Legacy interrupt, it needs to be setup. */ - if (acpi_table_parse(ACPI_FADT, acpi_parse_fadt) < 1) + if (acpi_table_parse(ACPI_SIG_FADT, acpi_parse_fadt) < 1) printk(KERN_ERR PREFIX "Can't find FADT\n"); #ifdef CONFIG_SMP @@ -836,7 +835,7 @@ int acpi_map_lsapic(acpi_handle handle, int *pcpu) { struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; union acpi_object *obj; - struct acpi_table_lsapic *lsapic; + struct acpi_madt_local_sapic *lsapic; cpumask_t tmp_map; long physid; int cpu; @@ -848,16 +847,16 @@ int acpi_map_lsapic(acpi_handle handle, int *pcpu) return -EINVAL; obj = buffer.pointer; - if (obj->type != ACPI_TYPE_BUFFER || - obj->buffer.length < sizeof(*lsapic)) { + if (obj->type != ACPI_TYPE_BUFFER) + { kfree(buffer.pointer); return -EINVAL; } - lsapic = (struct acpi_table_lsapic *)obj->buffer.pointer; + lsapic = (struct acpi_madt_local_sapic *)obj->buffer.pointer; - if ((lsapic->header.type != ACPI_MADT_LSAPIC) || - (!lsapic->flags.enabled)) { + if ((lsapic->header.type != ACPI_MADT_TYPE_LOCAL_SAPIC) || + (!lsapic->lapic_flags & ACPI_MADT_ENABLED)) { kfree(buffer.pointer); return -EINVAL; } @@ -877,7 +876,6 @@ int acpi_map_lsapic(acpi_handle handle, int *pcpu) cpu_set(cpu, cpu_present_map); ia64_cpu_to_sapicid[cpu] = physid; - ia64_acpiid_to_sapicid[lsapic->acpi_id] = ia64_cpu_to_sapicid[cpu]; *pcpu = cpu; return (0); @@ -887,14 +885,6 @@ EXPORT_SYMBOL(acpi_map_lsapic); int acpi_unmap_lsapic(int cpu) { - int i; - - for (i = 0; i < MAX_SAPICS; i++) { - if (ia64_acpiid_to_sapicid[i] == ia64_cpu_to_sapicid[cpu]) { - ia64_acpiid_to_sapicid[i] = -1; - break; - } - } ia64_cpu_to_sapicid[cpu] = -1; cpu_clear(cpu, cpu_present_map); @@ -914,7 +904,7 @@ acpi_map_iosapic(acpi_handle handle, u32 depth, void *context, void **ret) { struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; union acpi_object *obj; - struct acpi_table_iosapic *iosapic; + struct acpi_madt_io_sapic *iosapic; unsigned int gsi_base; int pxm, node; @@ -932,9 +922,9 @@ acpi_map_iosapic(acpi_handle handle, u32 depth, void *context, void **ret) return AE_OK; } - iosapic = (struct acpi_table_iosapic *)obj->buffer.pointer; + iosapic = (struct acpi_madt_io_sapic *)obj->buffer.pointer; - if (iosapic->header.type != ACPI_MADT_IOSAPIC) { + if (iosapic->header.type != ACPI_MADT_TYPE_IO_SAPIC) { kfree(buffer.pointer); return AE_OK; } diff --git a/arch/ia64/kernel/cpufreq/acpi-cpufreq.c b/arch/ia64/kernel/cpufreq/acpi-cpufreq.c index 088f130..15c08d5 100644 --- a/arch/ia64/kernel/cpufreq/acpi-cpufreq.c +++ b/arch/ia64/kernel/cpufreq/acpi-cpufreq.c @@ -276,12 +276,10 @@ acpi_cpufreq_cpu_init ( dprintk("acpi_cpufreq_cpu_init\n"); - data = kmalloc(sizeof(struct cpufreq_acpi_io), GFP_KERNEL); + data = kzalloc(sizeof(struct cpufreq_acpi_io), GFP_KERNEL); if (!data) return (-ENOMEM); - memset(data, 0, sizeof(struct cpufreq_acpi_io)); - acpi_io_data[cpu] = data; result = acpi_processor_register_performance(&data->acpi_data, cpu); diff --git a/arch/ia64/kernel/crash.c b/arch/ia64/kernel/crash.c index 0aabedf..37bb16f 100644 --- a/arch/ia64/kernel/crash.c +++ b/arch/ia64/kernel/crash.c @@ -19,29 +19,11 @@ #include <asm/kdebug.h> #include <asm/mca.h> -#include <asm/uaccess.h> int kdump_status[NR_CPUS]; atomic_t kdump_cpu_freezed; atomic_t kdump_in_progress; int kdump_on_init = 1; -ssize_t -copy_oldmem_page(unsigned long pfn, char *buf, - size_t csize, unsigned long offset, int userbuf) -{ - void *vaddr; - - if (!csize) - return 0; - vaddr = __va(pfn<<PAGE_SHIFT); - if (userbuf) { - if (copy_to_user(buf, (vaddr + offset), csize)) { - return -EFAULT; - } - } else - memcpy(buf, (vaddr + offset), csize); - return csize; -} static inline Elf64_Word *append_elf_note(Elf64_Word *buf, char *name, unsigned type, void *data, @@ -70,7 +52,7 @@ extern void ia64_dump_cpu_regs(void *); static DEFINE_PER_CPU(struct elf_prstatus, elf_prstatus); void -crash_save_this_cpu() +crash_save_this_cpu(void) { void *buf; unsigned long cfm, sof, sol; @@ -97,6 +79,7 @@ crash_save_this_cpu() final_note(buf); } +#ifdef CONFIG_SMP static int kdump_wait_cpu_freeze(void) { @@ -109,6 +92,7 @@ kdump_wait_cpu_freeze(void) } return 1; } +#endif void machine_crash_shutdown(struct pt_regs *pt) @@ -134,6 +118,11 @@ machine_crash_shutdown(struct pt_regs *pt) static void machine_kdump_on_init(void) { + if (!ia64_kimage) { + printk(KERN_NOTICE "machine_kdump_on_init(): " + "kdump not configured\n"); + return; + } local_irq_disable(); kexec_disable_iosapic(); machine_kexec(ia64_kimage); @@ -150,11 +139,12 @@ kdump_cpu_freeze(struct unw_frame_info *info, void *arg) atomic_inc(&kdump_cpu_freezed); kdump_status[cpuid] = 1; mb(); - if (cpuid == 0) { - for (;;) - cpu_relax(); - } else +#ifdef CONFIG_HOTPLUG_CPU + if (cpuid != 0) ia64_jump_to_sal(&sal_boot_rendez_state[cpuid]); +#endif + for (;;) + cpu_relax(); } static int @@ -225,14 +215,10 @@ static ctl_table sys_table[] = { static int machine_crash_setup(void) { - char *from = strstr(saved_command_line, "elfcorehdr="); static struct notifier_block kdump_init_notifier_nb = { .notifier_call = kdump_init_notifier, }; int ret; - if (from) - elfcorehdr_addr = memparse(from+11, &from); - saved_max_pfn = (unsigned long)-1; if((ret = register_die_notifier(&kdump_init_notifier_nb)) != 0) return ret; #ifdef CONFIG_SYSCTL diff --git a/arch/ia64/kernel/crash_dump.c b/arch/ia64/kernel/crash_dump.c new file mode 100644 index 0000000..da60e90 --- /dev/null +++ b/arch/ia64/kernel/crash_dump.c @@ -0,0 +1,49 @@ +/* + * kernel/crash_dump.c - Memory preserving reboot related code. + * + * Created by: Simon Horman <horms@verge.net.au> + * Original code moved from kernel/crash.c + * Original code comment copied from the i386 version of this file + */ + +#include <linux/errno.h> +#include <linux/types.h> + +#include <asm/page.h> +#include <asm/uaccess.h> + +/** + * copy_oldmem_page - copy one page from "oldmem" + * @pfn: page frame number to be copied + * @buf: target memory address for the copy; this can be in kernel address + * space or user address space (see @userbuf) + * @csize: number of bytes to copy + * @offset: offset in bytes into the page (based on pfn) to begin the copy + * @userbuf: if set, @buf is in user address space, use copy_to_user(), + * otherwise @buf is in kernel address space, use memcpy(). + * + * Copy a page from "oldmem". For this page, there is no pte mapped + * in the current kernel. We stitch up a pte, similar to kmap_atomic. + * + * Calling copy_to_user() in atomic context is not desirable. Hence first + * copying the data to a pre-allocated kernel page and then copying to user + * space in non-atomic context. + */ +ssize_t +copy_oldmem_page(unsigned long pfn, char *buf, + size_t csize, unsigned long offset, int userbuf) +{ + void *vaddr; + + if (!csize) + return 0; + vaddr = __va(pfn<<PAGE_SHIFT); + if (userbuf) { + if (copy_to_user(buf, (vaddr + offset), csize)) { + return -EFAULT; + } + } else + memcpy(buf, (vaddr + offset), csize); + return csize; +} + diff --git a/arch/ia64/kernel/efi.c b/arch/ia64/kernel/efi.c index 0b25a7d..6c03928 100644 --- a/arch/ia64/kernel/efi.c +++ b/arch/ia64/kernel/efi.c @@ -380,7 +380,7 @@ efi_get_pal_addr (void) #endif return __va(md->phys_addr); } - printk(KERN_WARNING "%s: no PAL-code memory-descriptor found", + printk(KERN_WARNING "%s: no PAL-code memory-descriptor found\n", __FUNCTION__); return NULL; } diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S index 15234ed..e7873ee 100644 --- a/arch/ia64/kernel/entry.S +++ b/arch/ia64/kernel/entry.S @@ -1610,5 +1610,7 @@ sys_call_table: data8 sys_sync_file_range // 1300 data8 sys_tee data8 sys_vmsplice + data8 sys_ni_syscall // reserved for move_pages + data8 sys_getcpu .org sys_call_table + 8*NR_syscalls // guard against failures to increase NR_syscalls diff --git a/arch/ia64/kernel/iosapic.c b/arch/ia64/kernel/iosapic.c index 0fc5fb7..d6aab40 100644 --- a/arch/ia64/kernel/iosapic.c +++ b/arch/ia64/kernel/iosapic.c @@ -925,6 +925,11 @@ iosapic_unregister_intr (unsigned int gsi) /* Clear the interrupt controller descriptor */ idesc->chip = &no_irq_type; +#ifdef CONFIG_SMP + /* Clear affinity */ + cpus_setall(idesc->affinity); +#endif + /* Clear the interrupt information */ memset(&iosapic_intr_info[vector], 0, sizeof(struct iosapic_intr_info)); diff --git a/arch/ia64/kernel/irq.c b/arch/ia64/kernel/irq.c index 54d55e4..ce49c85 100644 --- a/arch/ia64/kernel/irq.c +++ b/arch/ia64/kernel/irq.c @@ -122,6 +122,9 @@ static void migrate_irqs(void) for (irq=0; irq < NR_IRQS; irq++) { desc = irq_desc + irq; + if (desc->status == IRQ_DISABLED) + continue; + /* * No handling for now. * TBD: Implement a disable function so we can now diff --git a/arch/ia64/kernel/jprobes.S b/arch/ia64/kernel/jprobes.S index 5cd6226f..6216302 100644 --- a/arch/ia64/kernel/jprobes.S +++ b/arch/ia64/kernel/jprobes.S @@ -45,13 +45,14 @@ * to the correct location. */ #include <asm/asmmacro.h> +#include <asm-ia64/break.h> /* * void jprobe_break(void) */ .section .kprobes.text, "ax" ENTRY(jprobe_break) - break.m 0x80300 + break.m __IA64_BREAK_JPROBE END(jprobe_break) /* diff --git a/arch/ia64/kernel/kprobes.c b/arch/ia64/kernel/kprobes.c index 76e7789..6cb56dd 100644 --- a/arch/ia64/kernel/kprobes.c +++ b/arch/ia64/kernel/kprobes.c @@ -88,6 +88,7 @@ static void __kprobes update_kprobe_inst_flag(uint template, uint slot, { p->ainsn.inst_flag = 0; p->ainsn.target_br_reg = 0; + p->ainsn.slot = slot; /* Check for Break instruction * Bits 37:40 Major opcode to be zero @@ -129,48 +130,6 @@ static void __kprobes update_kprobe_inst_flag(uint template, uint slot, /* * In this function we check to see if the instruction - * on which we are inserting kprobe is supported. - * Returns 0 if supported - * Returns -EINVAL if unsupported - */ -static int __kprobes unsupported_inst(uint template, uint slot, - uint major_opcode, - unsigned long kprobe_inst, - unsigned long addr) -{ - if (bundle_encoding[template][slot] == I) { - switch (major_opcode) { - case 0x0: //I_UNIT_MISC_OPCODE: - /* - * Check for Integer speculation instruction - * - Bit 33-35 to be equal to 0x1 - */ - if (((kprobe_inst >> 33) & 0x7) == 1) { - printk(KERN_WARNING - "Kprobes on speculation inst at <0x%lx> not supported\n", - addr); - return -EINVAL; - } - - /* - * IP relative mov instruction - * - Bit 27-35 to be equal to 0x30 - */ - if (((kprobe_inst >> 27) & 0x1FF) == 0x30) { - printk(KERN_WARNING - "Kprobes on \"mov r1=ip\" at <0x%lx> not supported\n", - addr); - return -EINVAL; - - } - } - } - return 0; -} - - -/* - * In this function we check to see if the instruction * (qp) cmpx.crel.ctype p1,p2=r2,r3 * on which we are inserting kprobe is cmp instruction * with ctype as unc. @@ -206,26 +165,136 @@ out: } /* + * In this function we check to see if the instruction + * on which we are inserting kprobe is supported. + * Returns qp value if supported + * Returns -EINVAL if unsupported + */ +static int __kprobes unsupported_inst(uint template, uint slot, + uint major_opcode, + unsigned long kprobe_inst, + unsigned long addr) +{ + int qp; + + qp = kprobe_inst & 0x3f; + if (is_cmp_ctype_unc_inst(template, slot, major_opcode, kprobe_inst)) { + if (slot == 1 && qp) { + printk(KERN_WARNING "Kprobes on cmp unc" + "instruction on slot 1 at <0x%lx>" + "is not supported\n", addr); + return -EINVAL; + + } + qp = 0; + } + else if (bundle_encoding[template][slot] == I) { + if (major_opcode == 0) { + /* + * Check for Integer speculation instruction + * - Bit 33-35 to be equal to 0x1 + */ + if (((kprobe_inst >> 33) & 0x7) == 1) { + printk(KERN_WARNING + "Kprobes on speculation inst at <0x%lx> not supported\n", + addr); + return -EINVAL; + } + /* + * IP relative mov instruction + * - Bit 27-35 to be equal to 0x30 + */ + if (((kprobe_inst >> 27) & 0x1FF) == 0x30) { + printk(KERN_WARNING + "Kprobes on \"mov r1=ip\" at <0x%lx> not supported\n", + addr); + return -EINVAL; + + } + } + else if ((major_opcode == 5) && !(kprobe_inst & (0xFUl << 33)) && + (kprobe_inst & (0x1UL << 12))) { + /* test bit instructions, tbit,tnat,tf + * bit 33-36 to be equal to 0 + * bit 12 to be equal to 1 + */ + if (slot == 1 && qp) { + printk(KERN_WARNING "Kprobes on test bit" + "instruction on slot at <0x%lx>" + "is not supported\n", addr); + return -EINVAL; + } + qp = 0; + } + } + else if (bundle_encoding[template][slot] == B) { + if (major_opcode == 7) { + /* IP-Relative Predict major code is 7 */ + printk(KERN_WARNING "Kprobes on IP-Relative" + "Predict is not supported\n"); + return -EINVAL; + } + else if (major_opcode == 2) { + /* Indirect Predict, major code is 2 + * bit 27-32 to be equal to 10 or 11 + */ + int x6=(kprobe_inst >> 27) & 0x3F; + if ((x6 == 0x10) || (x6 == 0x11)) { + printk(KERN_WARNING "Kprobes on" + "Indirect Predict is not supported\n"); + return -EINVAL; + } + } + } + /* kernel does not use float instruction, here for safety kprobe + * will judge whether it is fcmp/flass/float approximation instruction + */ + else if (unlikely(bundle_encoding[template][slot] == F)) { + if ((major_opcode == 4 || major_opcode == 5) && + (kprobe_inst & (0x1 << 12))) { + /* fcmp/fclass unc instruction */ + if (slot == 1 && qp) { + printk(KERN_WARNING "Kprobes on fcmp/fclass " + "instruction on slot at <0x%lx> " + "is not supported\n", addr); + return -EINVAL; + + } + qp = 0; + } + if ((major_opcode == 0 || major_opcode == 1) && + (kprobe_inst & (0x1UL << 33))) { + /* float Approximation instruction */ + if (slot == 1 && qp) { + printk(KERN_WARNING "Kprobes on float Approx " + "instr at <0x%lx> is not supported\n", + addr); + return -EINVAL; + } + qp = 0; + } + } + return qp; +} + +/* * In this function we override the bundle with * the break instruction at the given slot. */ static void __kprobes prepare_break_inst(uint template, uint slot, uint major_opcode, unsigned long kprobe_inst, - struct kprobe *p) + struct kprobe *p, + int qp) { unsigned long break_inst = BREAK_INST; bundle_t *bundle = &p->opcode.bundle; /* * Copy the original kprobe_inst qualifying predicate(qp) - * to the break instruction iff !is_cmp_ctype_unc_inst - * because for cmp instruction with ctype equal to unc, - * which is a special instruction always needs to be - * executed regradless of qp + * to the break instruction */ - if (!is_cmp_ctype_unc_inst(template, slot, major_opcode, kprobe_inst)) - break_inst |= (0x3f & kprobe_inst); + break_inst |= qp; switch (slot) { case 0: @@ -296,12 +365,6 @@ static int __kprobes valid_kprobe_addr(int template, int slot, return -EINVAL; } - if (slot == 1 && bundle_encoding[template][1] != L) { - printk(KERN_WARNING "Inserting kprobes on slot #1 " - "is not supported\n"); - return -EINVAL; - } - return 0; } @@ -427,6 +490,7 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p) unsigned long kprobe_inst=0; unsigned int slot = addr & 0xf, template, major_opcode = 0; bundle_t *bundle; + int qp; bundle = &((kprobe_opcode_t *)kprobe_addr)->bundle; template = bundle->quad0.template; @@ -441,9 +505,9 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p) /* Get kprobe_inst and major_opcode from the bundle */ get_kprobe_inst(bundle, slot, &kprobe_inst, &major_opcode); - if (unsupported_inst(template, slot, major_opcode, kprobe_inst, addr)) - return -EINVAL; - + qp = unsupported_inst(template, slot, major_opcode, kprobe_inst, addr); + if (qp < 0) + return -EINVAL; p->ainsn.insn = get_insn_slot(); if (!p->ainsn.insn) @@ -451,30 +515,56 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p) memcpy(&p->opcode, kprobe_addr, sizeof(kprobe_opcode_t)); memcpy(p->ainsn.insn, kprobe_addr, sizeof(kprobe_opcode_t)); - prepare_break_inst(template, slot, major_opcode, kprobe_inst, p); + prepare_break_inst(template, slot, major_opcode, kprobe_inst, p, qp); return 0; } void __kprobes arch_arm_kprobe(struct kprobe *p) { - unsigned long addr = (unsigned long)p->addr; - unsigned long arm_addr = addr & ~0xFULL; + unsigned long arm_addr; + bundle_t *src, *dest; + + arm_addr = ((unsigned long)p->addr) & ~0xFUL; + dest = &((kprobe_opcode_t *)arm_addr)->bundle; + src = &p->opcode.bundle; flush_icache_range((unsigned long)p->ainsn.insn, (unsigned long)p->ainsn.insn + sizeof(kprobe_opcode_t)); - memcpy((char *)arm_addr, &p->opcode, sizeof(kprobe_opcode_t)); + switch (p->ainsn.slot) { + case 0: + dest->quad0.slot0 = src->quad0.slot0; + break; + case 1: + dest->quad1.slot1_p1 = src->quad1.slot1_p1; + break; + case 2: + dest->quad1.slot2 = src->quad1.slot2; + break; + } flush_icache_range(arm_addr, arm_addr + sizeof(kprobe_opcode_t)); } void __kprobes arch_disarm_kprobe(struct kprobe *p) { - unsigned long addr = (unsigned long)p->addr; - unsigned long arm_addr = addr & ~0xFULL; + unsigned long arm_addr; + bundle_t *src, *dest; + arm_addr = ((unsigned long)p->addr) & ~0xFUL; + dest = &((kprobe_opcode_t *)arm_addr)->bundle; /* p->ainsn.insn contains the original unaltered kprobe_opcode_t */ - memcpy((char *) arm_addr, (char *) p->ainsn.insn, - sizeof(kprobe_opcode_t)); + src = &p->ainsn.insn->bundle; + switch (p->ainsn.slot) { + case 0: + dest->quad0.slot0 = src->quad0.slot0; + break; + case 1: + dest->quad1.slot1_p1 = src->quad1.slot1_p1; + break; + case 2: + dest->quad1.slot2 = src->quad1.slot2; + break; + } flush_icache_range(arm_addr, arm_addr + sizeof(kprobe_opcode_t)); } @@ -807,7 +897,9 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self, switch(val) { case DIE_BREAK: /* err is break number from ia64_bad_break() */ - if (args->err == 0x80200 || args->err == 0x80300 || args->err == 0) + if ((args->err >> 12) == (__IA64_BREAK_KPROBE >> 12) + || args->err == __IA64_BREAK_JPROBE + || args->err == 0) if (pre_kprobes_handler(args)) ret = NOTIFY_STOP; break; diff --git a/arch/ia64/kernel/machine_kexec.c b/arch/ia64/kernel/machine_kexec.c index 468233f..4f0f3b8 100644 --- a/arch/ia64/kernel/machine_kexec.c +++ b/arch/ia64/kernel/machine_kexec.c @@ -14,13 +14,17 @@ #include <linux/kexec.h> #include <linux/cpu.h> #include <linux/irq.h> +#include <linux/efi.h> #include <asm/mmu_context.h> #include <asm/setup.h> #include <asm/delay.h> #include <asm/meminit.h> -typedef void (*relocate_new_kernel_t)(unsigned long, unsigned long, - struct ia64_boot_param *, unsigned long); +typedef NORET_TYPE void (*relocate_new_kernel_t)( + unsigned long indirection_page, + unsigned long start_address, + struct ia64_boot_param *boot_param, + unsigned long pal_addr) ATTRIB_NORET; struct kimage *ia64_kimage; @@ -65,22 +69,10 @@ void machine_kexec_cleanup(struct kimage *image) { } -void machine_shutdown(void) -{ - int cpu; - - for_each_online_cpu(cpu) { - if (cpu != smp_processor_id()) - cpu_down(cpu); - } - kexec_disable_iosapic(); -} - /* * Do not allocate memory (or fail in any way) in machine_kexec(). * We are past the point of no return, committed to rebooting now. */ -extern void *efi_get_pal_addr(void); static void ia64_machine_kexec(struct unw_frame_info *info, void *arg) { struct kimage *image = arg; @@ -90,6 +82,7 @@ static void ia64_machine_kexec(struct unw_frame_info *info, void *arg) unsigned long vector; int ii; + BUG_ON(!image); if (image->type == KEXEC_TYPE_CRASH) { crash_save_this_cpu(); current->thread.ksp = (__u64)info->sw - 16; @@ -128,6 +121,7 @@ static void ia64_machine_kexec(struct unw_frame_info *info, void *arg) void machine_kexec(struct kimage *image) { + BUG_ON(!image); unw_init_running(ia64_machine_kexec, image); for(;;); } diff --git a/arch/ia64/kernel/mca.c b/arch/ia64/kernel/mca.c index 87c1c4f..a76add3 100644 --- a/arch/ia64/kernel/mca.c +++ b/arch/ia64/kernel/mca.c @@ -1239,7 +1239,7 @@ ia64_mca_handler(struct pt_regs *regs, struct switch_stack *sw, } else { /* Dump buffered message to console */ ia64_mlogbuf_finish(1); -#ifdef CONFIG_CRASH_DUMP +#ifdef CONFIG_KEXEC atomic_set(&kdump_in_progress, 1); monarch_cpu = -1; #endif diff --git a/arch/ia64/kernel/msi_ia64.c b/arch/ia64/kernel/msi_ia64.c index 822e59a..0d05450 100644 --- a/arch/ia64/kernel/msi_ia64.c +++ b/arch/ia64/kernel/msi_ia64.c @@ -64,12 +64,17 @@ static void ia64_set_msi_irq_affinity(unsigned int irq, cpumask_t cpu_mask) } #endif /* CONFIG_SMP */ -int ia64_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) +int ia64_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *desc) { struct msi_msg msg; unsigned long dest_phys_id; - unsigned int vector; + unsigned int irq, vector; + irq = create_irq(); + if (irq < 0) + return irq; + + set_irq_msi(irq, desc); dest_phys_id = cpu_physical_id(first_cpu(cpu_online_map)); vector = irq; @@ -89,12 +94,12 @@ int ia64_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) write_msi_msg(irq, &msg); set_irq_chip_and_handler(irq, &ia64_msi_chip, handle_edge_irq); - return 0; + return irq; } void ia64_teardown_msi_irq(unsigned int irq) { - return; /* no-op */ + destroy_irq(irq); } static void ia64_ack_msi_irq(unsigned int irq) @@ -126,12 +131,12 @@ static struct irq_chip ia64_msi_chip = { }; -int arch_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) +int arch_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *desc) { if (platform_setup_msi_irq) - return platform_setup_msi_irq(irq, pdev); + return platform_setup_msi_irq(pdev, desc); - return ia64_setup_msi_irq(irq, pdev); + return ia64_setup_msi_irq(pdev, desc); } void arch_teardown_msi_irq(unsigned int irq) diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c index dbb28164..aa94f60 100644 --- a/arch/ia64/kernel/perfmon.c +++ b/arch/ia64/kernel/perfmon.c @@ -2188,13 +2188,13 @@ pfm_alloc_fd(struct file **cfile) /* * allocate a new dcache entry */ - file->f_dentry = d_alloc(pfmfs_mnt->mnt_sb->s_root, &this); - if (!file->f_dentry) goto out; + file->f_path.dentry = d_alloc(pfmfs_mnt->mnt_sb->s_root, &this); + if (!file->f_path.dentry) goto out; - file->f_dentry->d_op = &pfmfs_dentry_operations; + file->f_path.dentry->d_op = &pfmfs_dentry_operations; - d_add(file->f_dentry, inode); - file->f_vfsmnt = mntget(pfmfs_mnt); + d_add(file->f_path.dentry, inode); + file->f_path.mnt = mntget(pfmfs_mnt); file->f_mapping = inode->i_mapping; file->f_op = &pfm_file_ops; diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c index 51922b9..ae96d41 100644 --- a/arch/ia64/kernel/process.c +++ b/arch/ia64/kernel/process.c @@ -34,6 +34,7 @@ #include <asm/ia32.h> #include <asm/irq.h> #include <asm/kdebug.h> +#include <asm/kexec.h> #include <asm/pgalloc.h> #include <asm/processor.h> #include <asm/sal.h> @@ -268,10 +269,16 @@ cpu_idle (void) /* endless idle loop with no priority at all */ while (1) { - if (can_do_pal_halt) + if (can_do_pal_halt) { current_thread_info()->status &= ~TS_POLLING; - else + /* + * TS_POLLING-cleared state must be visible before we + * test NEED_RESCHED: + */ + smp_mb(); + } else { current_thread_info()->status |= TS_POLLING; + } if (!need_resched()) { void (*idle)(void); @@ -797,6 +804,21 @@ cpu_halt (void) ia64_pal_halt(min_power_state); } +void machine_shutdown(void) +{ +#ifdef CONFIG_HOTPLUG_CPU + int cpu; + + for_each_online_cpu(cpu) { + if (cpu != smp_processor_id()) + cpu_down(cpu); + } +#endif +#ifdef CONFIG_KEXEC + kexec_disable_iosapic(); +#endif +} + void machine_restart (char *restart_cmd) { diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index aa705e4..3f89187 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -607,7 +607,7 @@ find_thread_for_addr (struct task_struct *child, unsigned long addr) */ list_for_each_safe(this, next, ¤t->children) { p = list_entry(this, struct task_struct, sibling); - if (p->mm != mm) + if (p->tgid != child->tgid) continue; if (thread_matches(p, addr)) { child = p; @@ -1405,6 +1405,7 @@ ptrace_disable (struct task_struct *child) struct ia64_psr *child_psr = ia64_psr(task_pt_regs(child)); /* make sure the single step/taken-branch trap bits are not set: */ + clear_tsk_thread_flag(child, TIF_SINGLESTEP); child_psr->ss = 0; child_psr->tb = 0; } @@ -1525,6 +1526,7 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data) * Make sure the single step/taken-branch trap bits * are not set: */ + clear_tsk_thread_flag(child, TIF_SINGLESTEP); ia64_psr(pt)->ss = 0; ia64_psr(pt)->tb = 0; @@ -1556,6 +1558,7 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data) goto out_tsk; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); + set_tsk_thread_flag(child, TIF_SINGLESTEP); if (request == PTRACE_SINGLESTEP) { ia64_psr(pt)->ss = 1; } else { @@ -1595,13 +1598,9 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data) } -void +static void syscall_trace (void) { - if (!test_thread_flag(TIF_SYSCALL_TRACE)) - return; - if (!(current->ptrace & PT_PTRACED)) - return; /* * The 0x80 provides a way for the tracing parent to * distinguish between a syscall stop and SIGTRAP delivery. @@ -1664,7 +1663,8 @@ syscall_trace_leave (long arg0, long arg1, long arg2, long arg3, audit_syscall_exit(success, result); } - if (test_thread_flag(TIF_SYSCALL_TRACE) + if ((test_thread_flag(TIF_SYSCALL_TRACE) + || test_thread_flag(TIF_SINGLESTEP)) && (current->ptrace & PT_PTRACED)) syscall_trace(); } diff --git a/arch/ia64/kernel/salinfo.c b/arch/ia64/kernel/salinfo.c index fd607ca..e375a2f 100644 --- a/arch/ia64/kernel/salinfo.c +++ b/arch/ia64/kernel/salinfo.c @@ -302,7 +302,7 @@ salinfo_event_open(struct inode *inode, struct file *file) static ssize_t salinfo_event_read(struct file *file, char __user *buffer, size_t count, loff_t *ppos) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file->f_path.dentry->d_inode; struct proc_dir_entry *entry = PDE(inode); struct salinfo_data *data = entry->data; char cmd[32]; @@ -464,7 +464,7 @@ retry: static ssize_t salinfo_log_read(struct file *file, char __user *buffer, size_t count, loff_t *ppos) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file->f_path.dentry->d_inode; struct proc_dir_entry *entry = PDE(inode); struct salinfo_data *data = entry->data; u8 *buf; @@ -525,7 +525,7 @@ salinfo_log_clear(struct salinfo_data *data, int cpu) static ssize_t salinfo_log_write(struct file *file, const char __user *buffer, size_t count, loff_t *ppos) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file->f_path.dentry->d_inode; struct proc_dir_entry *entry = PDE(inode); struct salinfo_data *data = entry->data; char cmd[32]; diff --git a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c index 14e1200..83c2629 100644 --- a/arch/ia64/kernel/setup.c +++ b/arch/ia64/kernel/setup.c @@ -256,7 +256,7 @@ reserve_memory (void) #ifdef CONFIG_KEXEC /* crashkernel=size@offset specifies the size to reserve for a crash - * kernel.(offset is ingored for keep compatibility with other archs) + * kernel. If offset is 0, then it is determined automatically. * By reserving this memory we guarantee that linux never set's it * up as a DMA target.Useful for holding code to do something * appropriate after a kernel panic. @@ -266,10 +266,16 @@ reserve_memory (void) unsigned long base, size; if (from) { size = memparse(from + 12, &from); + if (*from == '@') + base = memparse(from+1, &from); + else + base = 0; if (size) { - sort_regions(rsvd_region, n); - base = kdump_find_rsvd_region(size, - rsvd_region, n); + if (!base) { + sort_regions(rsvd_region, n); + base = kdump_find_rsvd_region(size, + rsvd_region, n); + } if (base != ~0UL) { rsvd_region[n].start = (unsigned long)__va(base); @@ -434,6 +440,21 @@ static __init int setup_nomca(char *s) } early_param("nomca", setup_nomca); +#ifdef CONFIG_PROC_VMCORE +/* elfcorehdr= specifies the location of elf core header + * stored by the crashed kernel. + */ +static int __init parse_elfcorehdr(char *arg) +{ + if (!arg) + return -EINVAL; + + elfcorehdr_addr = memparse(arg, &arg); + return 0; +} +early_param("elfcorehdr", parse_elfcorehdr); +#endif /* CONFIG_PROC_VMCORE */ + void __init setup_arch (char **cmdline_p) { @@ -548,34 +569,31 @@ show_cpuinfo (struct seq_file *m, void *v) { 1UL << 1, "spontaneous deferral"}, { 1UL << 2, "16-byte atomic ops" } }; - char features[128], *cp, sep; + char features[128], *cp, *sep; struct cpuinfo_ia64 *c = v; unsigned long mask; unsigned long proc_freq; - int i; + int i, size; mask = c->features; /* build the feature string: */ - memcpy(features, " standard", 10); + memcpy(features, "standard", 9); cp = features; - sep = 0; - for (i = 0; i < (int) ARRAY_SIZE(feature_bits); ++i) { + size = sizeof(features); + sep = ""; + for (i = 0; i < ARRAY_SIZE(feature_bits) && size > 1; ++i) { if (mask & feature_bits[i].mask) { - if (sep) - *cp++ = sep; - sep = ','; - *cp++ = ' '; - strcpy(cp, feature_bits[i].feature_name); - cp += strlen(feature_bits[i].feature_name); + cp += snprintf(cp, size, "%s%s", sep, + feature_bits[i].feature_name), + sep = ", "; mask &= ~feature_bits[i].mask; + size = sizeof(features) - (cp - features); } } - if (mask) { - /* print unknown features as a hex value: */ - if (sep) - *cp++ = sep; - sprintf(cp, " 0x%lx", mask); + if (mask && size > 1) { + /* print unknown features as a hex value */ + snprintf(cp, size, "%s0x%lx", sep, mask); } proc_freq = cpufreq_quick_get(cpunum); @@ -591,7 +609,7 @@ show_cpuinfo (struct seq_file *m, void *v) "model name : %s\n" "revision : %u\n" "archrev : %u\n" - "features :%s\n" /* don't change this---it _is_ right! */ + "features : %s\n" "cpu number : %lu\n" "cpu regs : %u\n" "cpu MHz : %lu.%06lu\n" @@ -653,6 +671,7 @@ get_model_name(__u8 family, __u8 model) { char brand[128]; + memcpy(brand, "Unknown", 8); if (ia64_pal_get_brand_info(brand)) { if (family == 0x7) memcpy(brand, "Merced", 7); @@ -660,8 +679,7 @@ get_model_name(__u8 family, __u8 model) case 0: memcpy(brand, "McKinley", 9); break; case 1: memcpy(brand, "Madison", 8); break; case 2: memcpy(brand, "Madison up to 9M cache", 23); break; - } else - memcpy(brand, "Unknown", 8); + } } if (brandname[0] == '\0') return strcpy(brandname, brand); diff --git a/arch/ia64/kernel/smp.c b/arch/ia64/kernel/smp.c index b1b9aa4..55ddd80 100644 --- a/arch/ia64/kernel/smp.c +++ b/arch/ia64/kernel/smp.c @@ -157,7 +157,7 @@ handle_IPI (int irq, void *dev_id) case IPI_CPU_STOP: stop_this_cpu(); break; -#ifdef CONFIG_CRASH_DUMP +#ifdef CONFIG_KEXEC case IPI_KDUMP_CPU_STOP: unw_init_running(kdump_cpu_freeze, NULL); break; @@ -219,15 +219,15 @@ send_IPI_self (int op) send_IPI_single(smp_processor_id(), op); } -#ifdef CONFIG_CRASH_DUMP +#ifdef CONFIG_KEXEC void -kdump_smp_send_stop() +kdump_smp_send_stop(void) { send_IPI_allbutself(IPI_KDUMP_CPU_STOP); } void -kdump_smp_send_init() +kdump_smp_send_init(void) { unsigned int cpu, self_cpu; self_cpu = smp_processor_id(); diff --git a/arch/ia64/kernel/traps.c b/arch/ia64/kernel/traps.c index fffa9e0..ab68474 100644 --- a/arch/ia64/kernel/traps.c +++ b/arch/ia64/kernel/traps.c @@ -307,6 +307,15 @@ fp_emulate (int fp_fault, void *bundle, long *ipsr, long *fpsr, long *isr, long return ret.status; } +struct fpu_swa_msg { + unsigned long count; + unsigned long time; +}; +static DEFINE_PER_CPU(struct fpu_swa_msg, cpulast); +DECLARE_PER_CPU(struct fpu_swa_msg, cpulast); +static struct fpu_swa_msg last __cacheline_aligned; + + /* * Handle floating-point assist faults and traps. */ @@ -316,8 +325,6 @@ handle_fpu_swa (int fp_fault, struct pt_regs *regs, unsigned long isr) long exception, bundle[2]; unsigned long fault_ip; struct siginfo siginfo; - static int fpu_swa_count = 0; - static unsigned long last_time; fault_ip = regs->cr_iip; if (!fp_fault && (ia64_psr(regs)->ri == 0)) @@ -325,14 +332,37 @@ handle_fpu_swa (int fp_fault, struct pt_regs *regs, unsigned long isr) if (copy_from_user(bundle, (void __user *) fault_ip, sizeof(bundle))) return -1; - if (jiffies - last_time > 5*HZ) - fpu_swa_count = 0; - if ((fpu_swa_count < 4) && !(current->thread.flags & IA64_THREAD_FPEMU_NOPRINT)) { - last_time = jiffies; - ++fpu_swa_count; - printk(KERN_WARNING - "%s(%d): floating-point assist fault at ip %016lx, isr %016lx\n", - current->comm, current->pid, regs->cr_iip + ia64_psr(regs)->ri, isr); + if (!(current->thread.flags & IA64_THREAD_FPEMU_NOPRINT)) { + unsigned long count, current_jiffies = jiffies; + struct fpu_swa_msg *cp = &__get_cpu_var(cpulast); + + if (unlikely(current_jiffies > cp->time)) + cp->count = 0; + if (unlikely(cp->count < 5)) { + cp->count++; + cp->time = current_jiffies + 5 * HZ; + + /* minimize races by grabbing a copy of count BEFORE checking last.time. */ + count = last.count; + barrier(); + + /* + * Lower 4 bits are used as a count. Upper bits are a sequence + * number that is updated when count is reset. The cmpxchg will + * fail is seqno has changed. This minimizes mutiple cpus + * reseting the count. + */ + if (current_jiffies > last.time) + (void) cmpxchg_acq(&last.count, count, 16 + (count & ~15)); + + /* used fetchadd to atomically update the count */ + if ((last.count & 15) < 5 && (ia64_fetchadd(1, &last.count, acq) & 15) < 5) { + last.time = current_jiffies + 5 * HZ; + printk(KERN_WARNING + "%s(%d): floating-point assist fault at ip %016lx, isr %016lx\n", + current->comm, current->pid, regs->cr_iip + ia64_psr(regs)->ri, isr); + } + } } exception = fp_emulate(fp_fault, bundle, ®s->cr_ipsr, ®s->ar_fpsr, &isr, ®s->pr, diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S index d6083a0..8f3d006 100644 --- a/arch/ia64/kernel/vmlinux.lds.S +++ b/arch/ia64/kernel/vmlinux.lds.S @@ -157,6 +157,7 @@ SECTIONS } #endif + . = ALIGN(8); __con_initcall_start = .; .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) { *(.con_initcall.init) } diff --git a/arch/ia64/mm/contig.c b/arch/ia64/mm/contig.c index 82deaa3..63e6d49 100644 --- a/arch/ia64/mm/contig.c +++ b/arch/ia64/mm/contig.c @@ -30,47 +30,69 @@ static unsigned long max_gap; #endif /** - * show_mem - display a memory statistics summary + * show_mem - give short summary of memory stats * - * Just walks the pages in the system and describes where they're allocated. + * Shows a simple page count of reserved and used pages in the system. + * For discontig machines, it does this on a per-pgdat basis. */ -void -show_mem (void) +void show_mem(void) { - int i, total = 0, reserved = 0; - int shared = 0, cached = 0; + int i, total_reserved = 0; + int total_shared = 0, total_cached = 0; + unsigned long total_present = 0; + pg_data_t *pgdat; printk(KERN_INFO "Mem-info:\n"); show_free_areas(); - printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); - i = max_mapnr; - for (i = 0; i < max_mapnr; i++) { - if (!pfn_valid(i)) { + printk(KERN_INFO "Node memory in pages:\n"); + for_each_online_pgdat(pgdat) { + unsigned long present; + unsigned long flags; + int shared = 0, cached = 0, reserved = 0; + + pgdat_resize_lock(pgdat, &flags); + present = pgdat->node_present_pages; + for(i = 0; i < pgdat->node_spanned_pages; i++) { + struct page *page; + if (pfn_valid(pgdat->node_start_pfn + i)) + page = pfn_to_page(pgdat->node_start_pfn + i); + else { #ifdef CONFIG_VIRTUAL_MEM_MAP - if (max_gap < LARGE_GAP) - continue; - i = vmemmap_find_next_valid_pfn(0, i) - 1; + if (max_gap < LARGE_GAP) + continue; #endif - continue; + i = vmemmap_find_next_valid_pfn(pgdat->node_id, + i) - 1; + continue; + } + if (PageReserved(page)) + reserved++; + else if (PageSwapCache(page)) + cached++; + else if (page_count(page)) + shared += page_count(page)-1; } - total++; - if (PageReserved(mem_map+i)) - reserved++; - else if (PageSwapCache(mem_map+i)) - cached++; - else if (page_count(mem_map + i)) - shared += page_count(mem_map + i) - 1; + pgdat_resize_unlock(pgdat, &flags); + total_present += present; + total_reserved += reserved; + total_cached += cached; + total_shared += shared; + printk(KERN_INFO "Node %4d: RAM: %11ld, rsvd: %8d, " + "shrd: %10d, swpd: %10d\n", pgdat->node_id, + present, reserved, shared, cached); } - printk(KERN_INFO "%d pages of RAM\n", total); - printk(KERN_INFO "%d reserved pages\n", reserved); - printk(KERN_INFO "%d pages shared\n", shared); - printk(KERN_INFO "%d pages swap cached\n", cached); - printk(KERN_INFO "%ld pages in page table cache\n", + printk(KERN_INFO "%ld pages of RAM\n", total_present); + printk(KERN_INFO "%d reserved pages\n", total_reserved); + printk(KERN_INFO "%d pages shared\n", total_shared); + printk(KERN_INFO "%d pages swap cached\n", total_cached); + printk(KERN_INFO "Total of %ld pages in page table cache\n", pgtable_quicklist_total_size()); + printk(KERN_INFO "%d free buffer pages\n", nr_free_buffer_pages()); } + /* physical address where the bootmem map is located */ unsigned long bootmap_start; @@ -174,6 +196,12 @@ find_memory (void) reserve_bootmem(bootmap_start, bootmap_size); find_initrd(); + +#ifdef CONFIG_CRASH_DUMP + /* If we are doing a crash dump, we still need to know the real mem + * size before original memory map is reset. */ + saved_max_pfn = max_pfn; +#endif } #ifdef CONFIG_SMP @@ -226,7 +254,6 @@ void __init paging_init (void) { unsigned long max_dma; - unsigned long nid = 0; unsigned long max_zone_pfns[MAX_NR_ZONES]; num_physpages = 0; @@ -238,7 +265,7 @@ paging_init (void) max_zone_pfns[ZONE_NORMAL] = max_low_pfn; #ifdef CONFIG_VIRTUAL_MEM_MAP - efi_memmap_walk(register_active_ranges, &nid); + efi_memmap_walk(register_active_ranges, NULL); efi_memmap_walk(find_largest_hole, (u64 *)&max_gap); if (max_gap < LARGE_GAP) { vmem_map = (struct page *) 0; diff --git a/arch/ia64/mm/discontig.c b/arch/ia64/mm/discontig.c index 96722cb..6eae596 100644 --- a/arch/ia64/mm/discontig.c +++ b/arch/ia64/mm/discontig.c @@ -412,37 +412,6 @@ static void __init memory_less_nodes(void) return; } -#ifdef CONFIG_SPARSEMEM -/** - * register_sparse_mem - notify SPARSEMEM that this memory range exists. - * @start: physical start of range - * @end: physical end of range - * @arg: unused - * - * Simply calls SPARSEMEM to register memory section(s). - */ -static int __init register_sparse_mem(unsigned long start, unsigned long end, - void *arg) -{ - int nid; - - start = __pa(start) >> PAGE_SHIFT; - end = __pa(end) >> PAGE_SHIFT; - nid = early_pfn_to_nid(start); - memory_present(nid, start, end); - - return 0; -} - -static void __init arch_sparse_init(void) -{ - efi_memmap_walk(register_sparse_mem, NULL); - sparse_init(); -} -#else -#define arch_sparse_init() do {} while (0) -#endif - /** * find_memory - walk the EFI memory map and setup the bootmem allocator * @@ -473,6 +442,9 @@ void __init find_memory(void) node_clear(node, memory_less_mask); mem_data[node].min_pfn = ~0UL; } + + efi_memmap_walk(register_active_ranges, NULL); + /* * Initialize the boot memory maps in reverse order since that's * what the bootmem allocator expects @@ -506,6 +478,12 @@ void __init find_memory(void) max_pfn = max_low_pfn; find_initrd(); + +#ifdef CONFIG_CRASH_DUMP + /* If we are doing a crash dump, we still need to know the real mem + * size before original memory map is reset. */ + saved_max_pfn = max_pfn; +#endif } #ifdef CONFIG_SMP @@ -654,7 +632,6 @@ static __init int count_node_pages(unsigned long start, unsigned long len, int n { unsigned long end = start + len; - add_active_range(node, start >> PAGE_SHIFT, end >> PAGE_SHIFT); mem_data[node].num_physpages += len >> PAGE_SHIFT; if (start <= __pa(MAX_DMA_ADDRESS)) mem_data[node].num_dma_physpages += @@ -686,10 +663,11 @@ void __init paging_init(void) max_dma = virt_to_phys((void *) MAX_DMA_ADDRESS) >> PAGE_SHIFT; - arch_sparse_init(); - efi_memmap_walk(filter_rsvd_memory, count_node_pages); + sparse_memory_present_with_active_regions(MAX_NUMNODES); + sparse_init(); + #ifdef CONFIG_VIRTUAL_MEM_MAP vmalloc_end -= PAGE_ALIGN(ALIGN(max_low_pfn, MAX_ORDER_NR_PAGES) * sizeof(struct page)); diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 56dc2024..faaca21 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -19,6 +19,7 @@ #include <linux/swap.h> #include <linux/proc_fs.h> #include <linux/bitops.h> +#include <linux/kexec.h> #include <asm/a.out.h> #include <asm/dma.h> @@ -128,6 +129,25 @@ lazy_mmu_prot_update (pte_t pte) set_bit(PG_arch_1, &page->flags); /* mark page as clean */ } +/* + * Since DMA is i-cache coherent, any (complete) pages that were written via + * DMA can be marked as "clean" so that lazy_mmu_prot_update() doesn't have to + * flush them when they get mapped into an executable vm-area. + */ +void +dma_mark_clean(void *addr, size_t size) +{ + unsigned long pg_addr, end; + + pg_addr = PAGE_ALIGN((unsigned long) addr); + end = (unsigned long) addr + size; + while (pg_addr + PAGE_SIZE <= end) { + struct page *page = virt_to_page(pg_addr); + set_bit(PG_arch_1, &page->flags); + pg_addr += PAGE_SIZE; + } +} + inline void ia64_set_rbs_bot (void) { @@ -543,7 +563,8 @@ virtual_memmap_init (u64 start, u64 end, void *arg) if (map_start < map_end) memmap_init_zone((unsigned long)(map_end - map_start), - args->nid, args->zone, page_to_pfn(map_start)); + args->nid, args->zone, page_to_pfn(map_start), + MEMMAP_EARLY); return 0; } @@ -552,7 +573,7 @@ memmap_init (unsigned long size, int nid, unsigned long zone, unsigned long start_pfn) { if (!vmem_map) - memmap_init_zone(size, nid, zone, start_pfn); + memmap_init_zone(size, nid, zone, start_pfn, MEMMAP_EARLY); else { struct page *start; struct memmap_init_callback_data args; @@ -594,18 +615,27 @@ find_largest_hole (u64 start, u64 end, void *arg) return 0; } +#endif /* CONFIG_VIRTUAL_MEM_MAP */ + int __init -register_active_ranges(u64 start, u64 end, void *nid) +register_active_ranges(u64 start, u64 end, void *arg) { - BUG_ON(nid == NULL); - BUG_ON(*(unsigned long *)nid >= MAX_NUMNODES); + int nid = paddr_to_nid(__pa(start)); + + if (nid < 0) + nid = 0; +#ifdef CONFIG_KEXEC + if (start > crashk_res.start && start < crashk_res.end) + start = crashk_res.end; + if (end > crashk_res.start && end < crashk_res.end) + end = crashk_res.start; +#endif - add_active_range(*(unsigned long *)nid, - __pa(start) >> PAGE_SHIFT, - __pa(end) >> PAGE_SHIFT); + if (start < end) + add_active_range(nid, __pa(start) >> PAGE_SHIFT, + __pa(end) >> PAGE_SHIFT); return 0; } -#endif /* CONFIG_VIRTUAL_MEM_MAP */ static int __init count_reserved_pages (u64 start, u64 end, void *arg) diff --git a/arch/ia64/sn/kernel/huberror.c b/arch/ia64/sn/kernel/huberror.c index abca6bd..fcf7f93 100644 --- a/arch/ia64/sn/kernel/huberror.c +++ b/arch/ia64/sn/kernel/huberror.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (C) 1992 - 1997, 2000,2002-2005 Silicon Graphics, Inc. All rights reserved. + * Copyright (C) 1992 - 1997, 2000,2002-2007 Silicon Graphics, Inc. All rights reserved. */ #include <linux/types.h> @@ -38,12 +38,20 @@ static irqreturn_t hub_eint_handler(int irq, void *arg) (u64) nasid, 0, 0, 0, 0, 0, 0); if ((int)ret_stuff.v0) - panic("hubii_eint_handler(): Fatal TIO Error"); + panic("%s: Fatal %s Error", __FUNCTION__, + ((nasid & 1) ? "TIO" : "HUBII")); if (!(nasid & 1)) /* Not a TIO, handle CRB errors */ (void)hubiio_crb_error_handler(hubdev_info); - } else - bte_error_handler((unsigned long)NODEPDA(nasid_to_cnodeid(nasid))); + } else + if (nasid & 1) { /* TIO errors */ + SAL_CALL_NOLOCK(ret_stuff, SN_SAL_HUB_ERROR_INTERRUPT, + (u64) nasid, 0, 0, 0, 0, 0, 0); + + if ((int)ret_stuff.v0) + panic("%s: Fatal TIO Error", __FUNCTION__); + } else + bte_error_handler((unsigned long)NODEPDA(nasid_to_cnodeid(nasid))); return IRQ_HANDLED; } diff --git a/arch/ia64/sn/kernel/io_acpi_init.c b/arch/ia64/sn/kernel/io_acpi_init.c index 99d7f27..8c331ca 100644 --- a/arch/ia64/sn/kernel/io_acpi_init.c +++ b/arch/ia64/sn/kernel/io_acpi_init.c @@ -13,6 +13,7 @@ #include <asm/sn/sn_sal.h> #include "xtalk/hubdev.h" #include <linux/acpi.h> +#include <acpi/acnamesp.h> /* @@ -31,6 +32,12 @@ struct acpi_vendor_uuid sn_uuid = { 0xa2, 0x7c, 0x08, 0x00, 0x69, 0x13, 0xea, 0x51 }, }; +struct sn_pcidev_match { + u8 bus; + unsigned int devfn; + acpi_handle handle; +}; + /* * Perform the early IO init in PROM. */ @@ -119,9 +126,11 @@ sn_get_bussoft_ptr(struct pci_bus *bus) status = acpi_get_vendor_resource(handle, METHOD_NAME__CRS, &sn_uuid, &buffer); if (ACPI_FAILURE(status)) { - printk(KERN_ERR "get_acpi_pcibus_ptr: " - "get_acpi_bussoft_info() failed: %d\n", - status); + printk(KERN_ERR "%s: " + "acpi_get_vendor_resource() failed (0x%x) for: ", + __FUNCTION__, status); + acpi_ns_print_node_pathname(handle, NULL); + printk("\n"); return NULL; } resource = buffer.pointer; @@ -130,8 +139,8 @@ sn_get_bussoft_ptr(struct pci_bus *bus) if ((vendor->byte_length - sizeof(struct acpi_vendor_uuid)) != sizeof(struct pcibus_bussoft *)) { printk(KERN_ERR - "get_acpi_bussoft_ptr: Invalid vendor data " - "length %d\n", vendor->byte_length); + "%s: Invalid vendor data length %d\n", + __FUNCTION__, vendor->byte_length); kfree(buffer.pointer); return NULL; } @@ -143,34 +152,254 @@ sn_get_bussoft_ptr(struct pci_bus *bus) } /* - * sn_acpi_bus_fixup + * sn_extract_device_info - Extract the pcidev_info and the sn_irq_info + * pointers from the vendor resource using the + * provided acpi handle, and copy the structures + * into the argument buffers. */ -void -sn_acpi_bus_fixup(struct pci_bus *bus) +static int +sn_extract_device_info(acpi_handle handle, struct pcidev_info **pcidev_info, + struct sn_irq_info **sn_irq_info) { - struct pci_dev *pci_dev = NULL; - struct pcibus_bussoft *prom_bussoft_ptr; - extern void sn_common_bus_fixup(struct pci_bus *, - struct pcibus_bussoft *); + u64 addr; + struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; + struct sn_irq_info *irq_info, *irq_info_prom; + struct pcidev_info *pcidev_ptr, *pcidev_prom_ptr; + struct acpi_resource *resource; + int ret = 0; + acpi_status status; + struct acpi_resource_vendor_typed *vendor; - if (!bus->parent) { /* If root bus */ - prom_bussoft_ptr = sn_get_bussoft_ptr(bus); - if (prom_bussoft_ptr == NULL) { + /* + * The pointer to this device's pcidev_info structure in + * the PROM, is in the vendor resource. + */ + status = acpi_get_vendor_resource(handle, METHOD_NAME__CRS, + &sn_uuid, &buffer); + if (ACPI_FAILURE(status)) { + printk(KERN_ERR + "%s: acpi_get_vendor_resource() failed (0x%x) for: ", + __FUNCTION__, status); + acpi_ns_print_node_pathname(handle, NULL); + printk("\n"); + return 1; + } + + resource = buffer.pointer; + vendor = &resource->data.vendor_typed; + if ((vendor->byte_length - sizeof(struct acpi_vendor_uuid)) != + sizeof(struct pci_devdev_info *)) { + printk(KERN_ERR + "%s: Invalid vendor data length: %d for: ", + __FUNCTION__, vendor->byte_length); + acpi_ns_print_node_pathname(handle, NULL); + printk("\n"); + ret = 1; + goto exit; + } + + pcidev_ptr = kzalloc(sizeof(struct pcidev_info), GFP_KERNEL); + if (!pcidev_ptr) + panic("%s: Unable to alloc memory for pcidev_info", __FUNCTION__); + + memcpy(&addr, vendor->byte_data, sizeof(struct pcidev_info *)); + pcidev_prom_ptr = __va(addr); + memcpy(pcidev_ptr, pcidev_prom_ptr, sizeof(struct pcidev_info)); + + /* Get the IRQ info */ + irq_info = kzalloc(sizeof(struct sn_irq_info), GFP_KERNEL); + if (!irq_info) + panic("%s: Unable to alloc memory for sn_irq_info", __FUNCTION__); + + if (pcidev_ptr->pdi_sn_irq_info) { + irq_info_prom = __va(pcidev_ptr->pdi_sn_irq_info); + memcpy(irq_info, irq_info_prom, sizeof(struct sn_irq_info)); + } + + *pcidev_info = pcidev_ptr; + *sn_irq_info = irq_info; + +exit: + kfree(buffer.pointer); + return ret; +} + +static unsigned int +get_host_devfn(acpi_handle device_handle, acpi_handle rootbus_handle) +{ + unsigned long adr; + acpi_handle child; + unsigned int devfn; + int function; + acpi_handle parent; + int slot; + acpi_status status; + + /* + * Do an upward search to find the root bus device, and + * obtain the host devfn from the previous child device. + */ + child = device_handle; + while (child) { + status = acpi_get_parent(child, &parent); + if (ACPI_FAILURE(status)) { + printk(KERN_ERR "%s: acpi_get_parent() failed " + "(0x%x) for: ", __FUNCTION__, status); + acpi_ns_print_node_pathname(child, NULL); + printk("\n"); + panic("%s: Unable to find host devfn\n", __FUNCTION__); + } + if (parent == rootbus_handle) + break; + child = parent; + } + if (!child) { + printk(KERN_ERR "%s: Unable to find root bus for: ", + __FUNCTION__); + acpi_ns_print_node_pathname(device_handle, NULL); + printk("\n"); + BUG(); + } + + status = acpi_evaluate_integer(child, METHOD_NAME__ADR, NULL, &adr); + if (ACPI_FAILURE(status)) { + printk(KERN_ERR "%s: Unable to get _ADR (0x%x) for: ", + __FUNCTION__, status); + acpi_ns_print_node_pathname(child, NULL); + printk("\n"); + panic("%s: Unable to find host devfn\n", __FUNCTION__); + } + + slot = (adr >> 16) & 0xffff; + function = adr & 0xffff; + devfn = PCI_DEVFN(slot, function); + return devfn; +} + +/* + * find_matching_device - Callback routine to find the ACPI device + * that matches up with our pci_dev device. + * Matching is done on bus number and devfn. + * To find the bus number for a particular + * ACPI device, we must look at the _BBN method + * of its parent. + */ +static acpi_status +find_matching_device(acpi_handle handle, u32 lvl, void *context, void **rv) +{ + unsigned long bbn = -1; + unsigned long adr; + acpi_handle parent = NULL; + acpi_status status; + unsigned int devfn; + int function; + int slot; + struct sn_pcidev_match *info = context; + + status = acpi_evaluate_integer(handle, METHOD_NAME__ADR, NULL, + &adr); + if (ACPI_SUCCESS(status)) { + status = acpi_get_parent(handle, &parent); + if (ACPI_FAILURE(status)) { printk(KERN_ERR - "sn_pci_fixup_bus: 0x%04x:0x%02x Unable to " - "obtain prom_bussoft_ptr\n", - pci_domain_nr(bus), bus->number); - return; + "%s: acpi_get_parent() failed (0x%x) for: ", + __FUNCTION__, status); + acpi_ns_print_node_pathname(handle, NULL); + printk("\n"); + return AE_OK; + } + status = acpi_evaluate_integer(parent, METHOD_NAME__BBN, + NULL, &bbn); + if (ACPI_FAILURE(status)) { + printk(KERN_ERR + "%s: Failed to find _BBN in parent of: ", + __FUNCTION__); + acpi_ns_print_node_pathname(handle, NULL); + printk("\n"); + return AE_OK; + } + + slot = (adr >> 16) & 0xffff; + function = adr & 0xffff; + devfn = PCI_DEVFN(slot, function); + if ((info->devfn == devfn) && (info->bus == bbn)) { + /* We have a match! */ + info->handle = handle; + return 1; } - sn_common_bus_fixup(bus, prom_bussoft_ptr); } - list_for_each_entry(pci_dev, &bus->devices, bus_list) { - sn_pci_fixup_slot(pci_dev); + return AE_OK; +} + +/* + * sn_acpi_get_pcidev_info - Search ACPI namespace for the acpi + * device matching the specified pci_dev, + * and return the pcidev info and irq info. + */ +int +sn_acpi_get_pcidev_info(struct pci_dev *dev, struct pcidev_info **pcidev_info, + struct sn_irq_info **sn_irq_info) +{ + unsigned int host_devfn; + struct sn_pcidev_match pcidev_match; + acpi_handle rootbus_handle; + unsigned long segment; + acpi_status status; + + rootbus_handle = PCI_CONTROLLER(dev)->acpi_handle; + status = acpi_evaluate_integer(rootbus_handle, METHOD_NAME__SEG, NULL, + &segment); + if (ACPI_SUCCESS(status)) { + if (segment != pci_domain_nr(dev)) { + printk(KERN_ERR + "%s: Segment number mismatch, 0x%lx vs 0x%x for: ", + __FUNCTION__, segment, pci_domain_nr(dev)); + acpi_ns_print_node_pathname(rootbus_handle, NULL); + printk("\n"); + return 1; + } + } else { + printk(KERN_ERR "%s: Unable to get __SEG from: ", + __FUNCTION__); + acpi_ns_print_node_pathname(rootbus_handle, NULL); + printk("\n"); + return 1; } + + /* + * We want to search all devices in this segment/domain + * of the ACPI namespace for the matching ACPI device, + * which holds the pcidev_info pointer in its vendor resource. + */ + pcidev_match.bus = dev->bus->number; + pcidev_match.devfn = dev->devfn; + pcidev_match.handle = NULL; + + acpi_walk_namespace(ACPI_TYPE_DEVICE, rootbus_handle, ACPI_UINT32_MAX, + find_matching_device, &pcidev_match, NULL); + + if (!pcidev_match.handle) { + printk(KERN_ERR + "%s: Could not find matching ACPI device for %s.\n", + __FUNCTION__, pci_name(dev)); + return 1; + } + + if (sn_extract_device_info(pcidev_match.handle, pcidev_info, sn_irq_info)) + return 1; + + /* Build up the pcidev_info.pdi_slot_host_handle */ + host_devfn = get_host_devfn(pcidev_match.handle, rootbus_handle); + (*pcidev_info)->pdi_slot_host_handle = + ((unsigned long) pci_domain_nr(dev) << 40) | + /* bus == 0 */ + host_devfn; + return 0; } /* - * sn_acpi_slot_fixup - Perform any SN specific slot fixup. + * sn_acpi_slot_fixup - Obtain the pcidev_info and sn_irq_info. + * Perform any SN specific slot fixup. * At present there does not appear to be * any generic way to handle a ROM image * that has been shadowed by the PROM, so @@ -179,11 +408,18 @@ sn_acpi_bus_fixup(struct pci_bus *bus) */ void -sn_acpi_slot_fixup(struct pci_dev *dev, struct pcidev_info *pcidev_info) +sn_acpi_slot_fixup(struct pci_dev *dev) { void __iomem *addr; + struct pcidev_info *pcidev_info = NULL; + struct sn_irq_info *sn_irq_info = NULL; size_t size; + if (sn_acpi_get_pcidev_info(dev, &pcidev_info, &sn_irq_info)) { + panic("%s: Failure obtaining pcidev_info for %s\n", + __FUNCTION__, pci_name(dev)); + } + if (pcidev_info->pdi_pio_mapped_addr[PCI_ROM_RESOURCE]) { /* * A valid ROM image exists and has been shadowed by the @@ -200,8 +436,11 @@ sn_acpi_slot_fixup(struct pci_dev *dev, struct pcidev_info *pcidev_info) (unsigned long) addr + size; dev->resource[PCI_ROM_RESOURCE].flags |= IORESOURCE_ROM_BIOS_COPY; } + sn_pci_fixup_slot(dev, pcidev_info, sn_irq_info); } +EXPORT_SYMBOL(sn_acpi_slot_fixup); + static struct acpi_driver acpi_sn_hubdev_driver = { .name = "SGI HUBDEV Driver", .ids = "SGIHUB,SGITIO", @@ -212,6 +451,33 @@ static struct acpi_driver acpi_sn_hubdev_driver = { /* + * sn_acpi_bus_fixup - Perform SN specific setup of software structs + * (pcibus_bussoft, pcidev_info) and hardware + * registers, for the specified bus and devices under it. + */ +void +sn_acpi_bus_fixup(struct pci_bus *bus) +{ + struct pci_dev *pci_dev = NULL; + struct pcibus_bussoft *prom_bussoft_ptr; + + if (!bus->parent) { /* If root bus */ + prom_bussoft_ptr = sn_get_bussoft_ptr(bus); + if (prom_bussoft_ptr == NULL) { + printk(KERN_ERR + "%s: 0x%04x:0x%02x Unable to " + "obtain prom_bussoft_ptr\n", + __FUNCTION__, pci_domain_nr(bus), bus->number); + return; + } + sn_common_bus_fixup(bus, prom_bussoft_ptr); + } + list_for_each_entry(pci_dev, &bus->devices, bus_list) { + sn_acpi_slot_fixup(pci_dev); + } +} + +/* * sn_io_acpi_init - PROM has ACPI support for IO, defining at a minimum the * nodes and root buses in the DSDT. As a result, bus scanning * will be initiated by the Linux ACPI code. @@ -223,6 +489,9 @@ sn_io_acpi_init(void) u64 result; s64 status; + /* SN Altix does not follow the IOSAPIC IRQ routing model */ + acpi_irq_model = ACPI_IRQ_MODEL_PLATFORM; + acpi_bus_register_driver(&acpi_sn_hubdev_driver); status = sal_ioif_init(&result); if (status || result) diff --git a/arch/ia64/sn/kernel/io_common.c b/arch/ia64/sn/kernel/io_common.c index d4dd8f4..d48bcd8 100644 --- a/arch/ia64/sn/kernel/io_common.c +++ b/arch/ia64/sn/kernel/io_common.c @@ -26,14 +26,10 @@ #include <linux/acpi.h> #include <asm/sn/sn2/sn_hwperf.h> #include <asm/sn/acpi.h> +#include "acpi/acglobal.h" extern void sn_init_cpei_timer(void); extern void register_sn_procfs(void); -extern void sn_acpi_bus_fixup(struct pci_bus *); -extern void sn_bus_fixup(struct pci_bus *); -extern void sn_acpi_slot_fixup(struct pci_dev *, struct pcidev_info *); -extern void sn_more_slot_fixup(struct pci_dev *, struct pcidev_info *); -extern void sn_legacy_pci_window_fixup(struct pci_controller *, u64, u64); extern void sn_io_acpi_init(void); extern void sn_io_init(void); @@ -48,6 +44,9 @@ struct sysdata_el { int sn_ioif_inited; /* SN I/O infrastructure initialized? */ +int sn_acpi_rev; /* SN ACPI revision */ +EXPORT_SYMBOL_GPL(sn_acpi_rev); + struct sn_pcibus_provider *sn_pci_provider[PCIIO_ASIC_MAX_TYPES]; /* indexed by asic type */ /* @@ -99,25 +98,6 @@ sal_get_device_dmaflush_list(u64 nasid, u64 widget_num, u64 device_num, } /* - * Retrieve the pci device information given the bus and device|function number. - */ -static inline u64 -sal_get_pcidev_info(u64 segment, u64 bus_number, u64 devfn, u64 pci_dev, - u64 sn_irq_info) -{ - struct ia64_sal_retval ret_stuff; - ret_stuff.status = 0; - ret_stuff.v0 = 0; - - SAL_CALL_NOLOCK(ret_stuff, - (u64) SN_SAL_IOIF_GET_PCIDEV_INFO, - (u64) segment, (u64) bus_number, (u64) devfn, - (u64) pci_dev, - sn_irq_info, 0, 0); - return ret_stuff.v0; -} - -/* * sn_pcidev_info_get() - Retrieve the pcidev_info struct for the specified * device. */ @@ -249,50 +229,25 @@ void sn_pci_unfixup_slot(struct pci_dev *dev) } /* - * sn_pci_fixup_slot() - This routine sets up a slot's resources consistent - * with the Linux PCI abstraction layer. Resources - * acquired from our PCI provider include PIO maps - * to BAR space and interrupt objects. + * sn_pci_fixup_slot() */ -void sn_pci_fixup_slot(struct pci_dev *dev) +void sn_pci_fixup_slot(struct pci_dev *dev, struct pcidev_info *pcidev_info, + struct sn_irq_info *sn_irq_info) { int segment = pci_domain_nr(dev->bus); - int status = 0; struct pcibus_bussoft *bs; - struct pci_bus *host_pci_bus; - struct pci_dev *host_pci_dev; - struct pcidev_info *pcidev_info; - struct sn_irq_info *sn_irq_info; - unsigned int bus_no, devfn; + struct pci_bus *host_pci_bus; + struct pci_dev *host_pci_dev; + unsigned int bus_no, devfn; pci_dev_get(dev); /* for the sysdata pointer */ - pcidev_info = kzalloc(sizeof(struct pcidev_info), GFP_KERNEL); - if (!pcidev_info) - BUG(); /* Cannot afford to run out of memory */ - - sn_irq_info = kzalloc(sizeof(struct sn_irq_info), GFP_KERNEL); - if (!sn_irq_info) - BUG(); /* Cannot afford to run out of memory */ - - /* Call to retrieve pci device information needed by kernel. */ - status = sal_get_pcidev_info((u64) segment, (u64) dev->bus->number, - dev->devfn, - (u64) __pa(pcidev_info), - (u64) __pa(sn_irq_info)); - if (status) - BUG(); /* Cannot get platform pci device information */ /* Add pcidev_info to list in pci_controller.platform_data */ list_add_tail(&pcidev_info->pdi_list, &(SN_PLATFORM_DATA(dev->bus)->pcidev_info)); - - if (SN_ACPI_BASE_SUPPORT()) - sn_acpi_slot_fixup(dev, pcidev_info); - else - sn_more_slot_fixup(dev, pcidev_info); /* * Using the PROMs values for the PCI host bus, get the Linux - * PCI host_pci_dev struct and set up host bus linkages + * PCI host_pci_dev struct and set up host bus linkages */ bus_no = (pcidev_info->pdi_slot_host_handle >> 32) & 0xff; @@ -489,11 +444,6 @@ void sn_generate_path(struct pci_bus *pci_bus, char *address) sprintf(address, "%s^%d", address, geo_slot(geoid)); } -/* - * sn_pci_fixup_bus() - Perform SN specific setup of software structs - * (pcibus_bussoft, pcidev_info) and hardware - * registers, for the specified bus and devices under it. - */ void __devinit sn_pci_fixup_bus(struct pci_bus *bus) { @@ -519,6 +469,15 @@ sn_io_early_init(void) if (!ia64_platform_is("sn2") || IS_RUNNING_ON_FAKE_PROM()) return 0; + /* we set the acpi revision to that of the DSDT table OEM rev. */ + { + struct acpi_table_header *header = NULL; + + acpi_get_table_by_index(ACPI_TABLE_INDEX_DSDT, &header); + BUG_ON(header == NULL); + sn_acpi_rev = header->oem_revision; + } + /* * prime sn_pci_provider[]. Individial provider init routines will * override their respective default entries. @@ -544,8 +503,12 @@ sn_io_early_init(void) register_sn_procfs(); #endif - printk(KERN_INFO "ACPI DSDT OEM Rev 0x%x\n", - acpi_gbl_DSDT->oem_revision); + { + struct acpi_table_header *header; + (void)acpi_get_table_by_index(ACPI_TABLE_INDEX_DSDT, &header); + printk(KERN_INFO "ACPI DSDT OEM Rev 0x%x\n", + header->oem_revision); + } if (SN_ACPI_BASE_SUPPORT()) sn_io_acpi_init(); else @@ -605,7 +568,6 @@ sn_io_late_init(void) fs_initcall(sn_io_late_init); -EXPORT_SYMBOL(sn_pci_fixup_slot); EXPORT_SYMBOL(sn_pci_unfixup_slot); EXPORT_SYMBOL(sn_bus_store_sysdata); EXPORT_SYMBOL(sn_bus_free_sysdata); diff --git a/arch/ia64/sn/kernel/io_init.c b/arch/ia64/sn/kernel/io_init.c index 9ad843e..600be3e 100644 --- a/arch/ia64/sn/kernel/io_init.c +++ b/arch/ia64/sn/kernel/io_init.c @@ -56,6 +56,25 @@ static inline u64 sal_get_pcibus_info(u64 segment, u64 busnum, u64 address) return ret_stuff.v0; } +/* + * Retrieve the pci device information given the bus and device|function number. + */ +static inline u64 +sal_get_pcidev_info(u64 segment, u64 bus_number, u64 devfn, u64 pci_dev, + u64 sn_irq_info) +{ + struct ia64_sal_retval ret_stuff; + ret_stuff.status = 0; + ret_stuff.v0 = 0; + + SAL_CALL_NOLOCK(ret_stuff, + (u64) SN_SAL_IOIF_GET_PCIDEV_INFO, + (u64) segment, (u64) bus_number, (u64) devfn, + (u64) pci_dev, + sn_irq_info, 0, 0); + return ret_stuff.v0; +} + /* * sn_fixup_ionodes() - This routine initializes the HUB data structure for @@ -172,18 +191,40 @@ sn_pci_window_fixup(struct pci_dev *dev, unsigned int count, } /* - * sn_more_slot_fixup() - We are not running with an ACPI capable PROM, + * sn_io_slot_fixup() - We are not running with an ACPI capable PROM, * and need to convert the pci_dev->resource * 'start' and 'end' addresses to mapped addresses, * and setup the pci_controller->window array entries. */ void -sn_more_slot_fixup(struct pci_dev *dev, struct pcidev_info *pcidev_info) +sn_io_slot_fixup(struct pci_dev *dev) { unsigned int count = 0; int idx; s64 pci_addrs[PCI_ROM_RESOURCE + 1]; unsigned long addr, end, size, start; + struct pcidev_info *pcidev_info; + struct sn_irq_info *sn_irq_info; + int status; + + pcidev_info = kzalloc(sizeof(struct pcidev_info), GFP_KERNEL); + if (!pcidev_info) + panic("%s: Unable to alloc memory for pcidev_info", __FUNCTION__); + + sn_irq_info = kzalloc(sizeof(struct sn_irq_info), GFP_KERNEL); + if (!sn_irq_info) + panic("%s: Unable to alloc memory for sn_irq_info", __FUNCTION__); + + /* Call to retrieve pci device information needed by kernel. */ + status = sal_get_pcidev_info((u64) pci_domain_nr(dev), + (u64) dev->bus->number, + dev->devfn, + (u64) __pa(pcidev_info), + (u64) __pa(sn_irq_info)); + + if (status) + BUG(); /* Cannot get platform pci device information */ + /* Copy over PIO Mapped Addresses */ for (idx = 0; idx <= PCI_ROM_RESOURCE; idx++) { @@ -219,8 +260,12 @@ sn_more_slot_fixup(struct pci_dev *dev, struct pcidev_info *pcidev_info) */ if (count > 0) sn_pci_window_fixup(dev, count, pci_addrs); + + sn_pci_fixup_slot(dev, pcidev_info, sn_irq_info); } +EXPORT_SYMBOL(sn_io_slot_fixup); + /* * sn_pci_controller_fixup() - This routine sets up a bus's resources * consistent with the Linux PCI abstraction layer. @@ -272,9 +317,6 @@ sn_bus_fixup(struct pci_bus *bus) { struct pci_dev *pci_dev = NULL; struct pcibus_bussoft *prom_bussoft_ptr; - extern void sn_common_bus_fixup(struct pci_bus *, - struct pcibus_bussoft *); - if (!bus->parent) { /* If root bus */ prom_bussoft_ptr = PCI_CONTROLLER(bus)->platform_data; @@ -291,7 +333,7 @@ sn_bus_fixup(struct pci_bus *bus) prom_bussoft_ptr->bs_legacy_mem); } list_for_each_entry(pci_dev, &bus->devices, bus_list) { - sn_pci_fixup_slot(pci_dev); + sn_io_slot_fixup(pci_dev); } } diff --git a/arch/ia64/sn/kernel/iomv.c b/arch/ia64/sn/kernel/iomv.c index 4aa4f30..ab7e2fd 100644 --- a/arch/ia64/sn/kernel/iomv.c +++ b/arch/ia64/sn/kernel/iomv.c @@ -1,4 +1,4 @@ -/* +/* * This file is subject to the terms and conditions of the GNU General Public * License. See the file "COPYING" in the main directory of this archive * for more details. @@ -26,9 +26,10 @@ * @port: port to convert * * Legacy in/out instructions are converted to ld/st instructions - * on IA64. This routine will convert a port number into a valid + * on IA64. This routine will convert a port number into a valid * SN i/o address. Used by sn_in*() and sn_out*(). */ + void *sn_io_addr(unsigned long port) { if (!IS_RUNNING_ON_SIMULATOR()) { diff --git a/arch/ia64/sn/kernel/msi_sn.c b/arch/ia64/sn/kernel/msi_sn.c index b3a435f..ea3dc38 100644 --- a/arch/ia64/sn/kernel/msi_sn.c +++ b/arch/ia64/sn/kernel/msi_sn.c @@ -59,13 +59,12 @@ void sn_teardown_msi_irq(unsigned int irq) sn_intr_free(nasid, widget, sn_irq_info); sn_msi_info[irq].sn_irq_info = NULL; - return; + destroy_irq(irq); } -int sn_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) +int sn_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *entry) { struct msi_msg msg; - struct msi_desc *entry; int widget; int status; nasid_t nasid; @@ -73,8 +72,8 @@ int sn_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) struct sn_irq_info *sn_irq_info; struct pcibus_bussoft *bussoft = SN_PCIDEV_BUSSOFT(pdev); struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); + int irq; - entry = get_irq_data(irq); if (!entry->msi_attrib.is_64) return -EINVAL; @@ -84,6 +83,11 @@ int sn_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) if (provider == NULL || provider->dma_map_consistent == NULL) return -EINVAL; + irq = create_irq(); + if (irq < 0) + return irq; + + set_irq_msi(irq, entry); /* * Set up the vector plumbing. Let the prom (via sn_intr_alloc) * decide which cpu to direct this msi at by default. @@ -95,12 +99,15 @@ int sn_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) SWIN_WIDGETNUM(bussoft->bs_base); sn_irq_info = kzalloc(sizeof(struct sn_irq_info), GFP_KERNEL); - if (! sn_irq_info) + if (! sn_irq_info) { + destroy_irq(irq); return -ENOMEM; + } status = sn_intr_alloc(nasid, widget, sn_irq_info, irq, -1, -1); if (status) { kfree(sn_irq_info); + destroy_irq(irq); return -ENOMEM; } @@ -121,6 +128,7 @@ int sn_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) if (! bus_addr) { sn_intr_free(nasid, widget, sn_irq_info); kfree(sn_irq_info); + destroy_irq(irq); return -ENOMEM; } @@ -139,7 +147,7 @@ int sn_setup_msi_irq(unsigned int irq, struct pci_dev *pdev) write_msi_msg(irq, &msg); set_irq_chip_and_handler(irq, &sn_msi_chip, handle_edge_irq); - return 0; + return irq; } #ifdef CONFIG_SMP diff --git a/arch/ia64/sn/kernel/setup.c b/arch/ia64/sn/kernel/setup.c index a934ad0..8571e52 100644 --- a/arch/ia64/sn/kernel/setup.c +++ b/arch/ia64/sn/kernel/setup.c @@ -580,7 +580,7 @@ void __cpuinit sn_cpu_init(void) int slice; int cnode; int i; - static int wars_have_been_checked; + static int wars_have_been_checked, set_cpu0_number; cpuid = smp_processor_id(); if (cpuid == 0 && IS_MEDUSA()) { @@ -605,8 +605,16 @@ void __cpuinit sn_cpu_init(void) /* * Don't check status. The SAL call is not supported on all PROMs * but a failure is harmless. + * Architechtuallly, cpu_init is always called twice on cpu 0. We + * should set cpu_number on cpu 0 once. */ - (void) ia64_sn_set_cpu_number(cpuid); + if (cpuid == 0) { + if (!set_cpu0_number) { + (void) ia64_sn_set_cpu_number(cpuid); + set_cpu0_number = 1; + } + } else + (void) ia64_sn_set_cpu_number(cpuid); /* * The boot cpu makes this call again after platform initialization is diff --git a/arch/ia64/sn/kernel/sn2/sn_hwperf.c b/arch/ia64/sn/kernel/sn2/sn_hwperf.c index 462ea17..3336799 100644 --- a/arch/ia64/sn/kernel/sn2/sn_hwperf.c +++ b/arch/ia64/sn/kernel/sn2/sn_hwperf.c @@ -189,7 +189,7 @@ static void print_pci_topology(struct seq_file *s) int e; for (sz = PAGE_SIZE; sz < 16 * PAGE_SIZE; sz += PAGE_SIZE) { - if (!(p = (char *)kmalloc(sz, GFP_KERNEL))) + if (!(p = kmalloc(sz, GFP_KERNEL))) break; e = ia64_sn_ioif_get_pci_topology(__pa(p), sz); if (e == SALRET_OK) diff --git a/arch/ia64/sn/kernel/xpc_channel.c b/arch/ia64/sn/kernel/xpc_channel.c index 1f35408..c08db9c 100644 --- a/arch/ia64/sn/kernel/xpc_channel.c +++ b/arch/ia64/sn/kernel/xpc_channel.c @@ -632,7 +632,7 @@ xpc_process_connect(struct xpc_channel *ch, unsigned long *irq_flags) ch->number, ch->partid); spin_unlock_irqrestore(&ch->lock, *irq_flags); - xpc_create_kthreads(ch, 1); + xpc_create_kthreads(ch, 1, 0); spin_lock_irqsave(&ch->lock, *irq_flags); } @@ -754,12 +754,12 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags) /* make sure all activity has settled down first */ - if (atomic_read(&ch->references) > 0 || - ((ch->flags & XPC_C_CONNECTEDCALLOUT_MADE) && - !(ch->flags & XPC_C_DISCONNECTINGCALLOUT_MADE))) { + if (atomic_read(&ch->kthreads_assigned) > 0 || + atomic_read(&ch->references) > 0) { return; } - DBUG_ON(atomic_read(&ch->kthreads_assigned) != 0); + DBUG_ON((ch->flags & XPC_C_CONNECTEDCALLOUT_MADE) && + !(ch->flags & XPC_C_DISCONNECTINGCALLOUT_MADE)); if (part->act_state == XPC_P_DEACTIVATING) { /* can't proceed until the other side disengages from us */ @@ -1651,6 +1651,11 @@ xpc_disconnect_channel(const int line, struct xpc_channel *ch, /* wake all idle kthreads so they can exit */ if (atomic_read(&ch->kthreads_idle) > 0) { wake_up_all(&ch->idle_wq); + + } else if ((ch->flags & XPC_C_CONNECTEDCALLOUT_MADE) && + !(ch->flags & XPC_C_DISCONNECTINGCALLOUT)) { + /* start a kthread that will do the xpcDisconnecting callout */ + xpc_create_kthreads(ch, 1, 1); } /* wake those waiting to allocate an entry from the local msg queue */ diff --git a/arch/ia64/sn/kernel/xpc_main.c b/arch/ia64/sn/kernel/xpc_main.c index fa96dfc..7a387d2 100644 --- a/arch/ia64/sn/kernel/xpc_main.c +++ b/arch/ia64/sn/kernel/xpc_main.c @@ -681,7 +681,7 @@ xpc_activate_kthreads(struct xpc_channel *ch, int needed) dev_dbg(xpc_chan, "create %d new kthreads, partid=%d, channel=%d\n", needed, ch->partid, ch->number); - xpc_create_kthreads(ch, needed); + xpc_create_kthreads(ch, needed, 0); } @@ -775,26 +775,28 @@ xpc_daemonize_kthread(void *args) xpc_kthread_waitmsgs(part, ch); } - if (atomic_dec_return(&ch->kthreads_assigned) == 0) { - spin_lock_irqsave(&ch->lock, irq_flags); - if ((ch->flags & XPC_C_CONNECTEDCALLOUT_MADE) && - !(ch->flags & XPC_C_DISCONNECTINGCALLOUT)) { - ch->flags |= XPC_C_DISCONNECTINGCALLOUT; - spin_unlock_irqrestore(&ch->lock, irq_flags); + /* let registerer know that connection is disconnecting */ - xpc_disconnect_callout(ch, xpcDisconnecting); - - spin_lock_irqsave(&ch->lock, irq_flags); - ch->flags |= XPC_C_DISCONNECTINGCALLOUT_MADE; - } + spin_lock_irqsave(&ch->lock, irq_flags); + if ((ch->flags & XPC_C_CONNECTEDCALLOUT_MADE) && + !(ch->flags & XPC_C_DISCONNECTINGCALLOUT)) { + ch->flags |= XPC_C_DISCONNECTINGCALLOUT; spin_unlock_irqrestore(&ch->lock, irq_flags); + + xpc_disconnect_callout(ch, xpcDisconnecting); + + spin_lock_irqsave(&ch->lock, irq_flags); + ch->flags |= XPC_C_DISCONNECTINGCALLOUT_MADE; + } + spin_unlock_irqrestore(&ch->lock, irq_flags); + + if (atomic_dec_return(&ch->kthreads_assigned) == 0) { if (atomic_dec_return(&part->nchannels_engaged) == 0) { xpc_mark_partition_disengaged(part); xpc_IPI_send_disengage(part); } } - xpc_msgqueue_deref(ch); dev_dbg(xpc_chan, "kthread exiting, partid=%d, channel=%d\n", @@ -818,7 +820,8 @@ xpc_daemonize_kthread(void *args) * partition. */ void -xpc_create_kthreads(struct xpc_channel *ch, int needed) +xpc_create_kthreads(struct xpc_channel *ch, int needed, + int ignore_disconnecting) { unsigned long irq_flags; pid_t pid; @@ -833,16 +836,38 @@ xpc_create_kthreads(struct xpc_channel *ch, int needed) * kthread. That kthread is responsible for doing the * counterpart to the following before it exits. */ + if (ignore_disconnecting) { + if (!atomic_inc_not_zero(&ch->kthreads_assigned)) { + /* kthreads assigned had gone to zero */ + BUG_ON(!(ch->flags & + XPC_C_DISCONNECTINGCALLOUT_MADE)); + break; + } + + } else if (ch->flags & XPC_C_DISCONNECTING) { + break; + + } else if (atomic_inc_return(&ch->kthreads_assigned) == 1) { + if (atomic_inc_return(&part->nchannels_engaged) == 1) + xpc_mark_partition_engaged(part); + } (void) xpc_part_ref(part); xpc_msgqueue_ref(ch); - if (atomic_inc_return(&ch->kthreads_assigned) == 1 && - atomic_inc_return(&part->nchannels_engaged) == 1) { - xpc_mark_partition_engaged(part); - } pid = kernel_thread(xpc_daemonize_kthread, (void *) args, 0); if (pid < 0) { /* the fork failed */ + + /* + * NOTE: if (ignore_disconnecting && + * !(ch->flags & XPC_C_DISCONNECTINGCALLOUT)) is true, + * then we'll deadlock if all other kthreads assigned + * to this channel are blocked in the channel's + * registerer, because the only thing that will unblock + * them is the xpcDisconnecting callout that this + * failed kernel_thread would have made. + */ + if (atomic_dec_return(&ch->kthreads_assigned) == 0 && atomic_dec_return(&part->nchannels_engaged) == 0) { xpc_mark_partition_disengaged(part); @@ -857,9 +882,6 @@ xpc_create_kthreads(struct xpc_channel *ch, int needed) * Flag this as an error only if we have an * insufficient #of kthreads for the channel * to function. - * - * No xpc_msgqueue_ref() is needed here since - * the channel mgr is doing this. */ spin_lock_irqsave(&ch->lock, irq_flags); XPC_DISCONNECT_CHANNEL(ch, xpcLackOfResources, diff --git a/arch/ia64/sn/pci/pcibr/pcibr_provider.c b/arch/ia64/sn/pci/pcibr/pcibr_provider.c index 6846dc9..04a82560 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_provider.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_provider.c @@ -20,7 +20,8 @@ #include "xtalk/hubdev.h" int -sal_pcibr_slot_enable(struct pcibus_info *soft, int device, void *resp) +sal_pcibr_slot_enable(struct pcibus_info *soft, int device, void *resp, + char **ssdt) { struct ia64_sal_retval ret_stuff; u64 busnum; @@ -32,7 +33,8 @@ sal_pcibr_slot_enable(struct pcibus_info *soft, int device, void *resp) segment = soft->pbi_buscommon.bs_persist_segment; busnum = soft->pbi_buscommon.bs_persist_busnum; SAL_CALL_NOLOCK(ret_stuff, (u64) SN_SAL_IOIF_SLOT_ENABLE, segment, - busnum, (u64) device, (u64) resp, 0, 0, 0); + busnum, (u64) device, (u64) resp, (u64)ia64_tpa(ssdt), + 0, 0); return (int)ret_stuff.v0; } |
