16 files changed, 206 insertions, 117 deletions

diff --git a/arch/i386/Kconfig b/arch/i386/Kconfig
index f17bd1d2707e..18ec9fe6deb6 100644
--- a/arch/i386/Kconfig
+++ b/arch/i386/Kconfig
@@ -53,6 +53,35 @@ source "init/Kconfig"
 
 menu "Processor type and features"
 
+config SMP
+        bool "Symmetric multi-processing support"
+        ---help---
+          This enables support for systems with more than one CPU. If you have
+          a system with only one CPU, like most personal computers, say N. If
+          you have a system with more than one CPU, say Y.
+
+          If you say N here, the kernel will run on single and multiprocessor
+          machines, but will use only one CPU of a multiprocessor machine. If
+          you say Y here, the kernel will run on many, but not all,
+          singleprocessor machines. On a singleprocessor machine, the kernel
+          will run faster if you say N here.
+
+          Note that if you say Y here and choose architecture "586" or
+          "Pentium" under "Processor family", the kernel will not work on 486
+          architectures. Similarly, multiprocessor kernels for the "PPro"
+          architecture may not work on all Pentium based boards.
+
+          People using multiprocessor machines who say Y here should also say
+          Y to "Enhanced Real Time Clock Support", below. The "Advanced Power
+          Management" code will be disabled if you say Y here.
+
+          See also the <file:Documentation/smp.txt>,
+          <file:Documentation/i386/IO-APIC.txt>,
+          <file:Documentation/nmi_watchdog.txt> and the SMP-HOWTO available at
+          <http://www.tldp.org/docs.html#howto>.
+
+          If you don't know what to do here, say N.
+
 choice
         prompt "Subarchitecture Type"
         default X86_PC
@@ -178,35 +207,6 @@ config HPET_EMULATE_RTC
         depends on HPET_TIMER && RTC=y
         default y
 
-config SMP
-        bool "Symmetric multi-processing support"
-        ---help---
-          This enables support for systems with more than one CPU. If you have
-          a system with only one CPU, like most personal computers, say N. If
-          you have a system with more than one CPU, say Y.
-
-          If you say N here, the kernel will run on single and multiprocessor
-          machines, but will use only one CPU of a multiprocessor machine. If
-          you say Y here, the kernel will run on many, but not all,
-          singleprocessor machines. On a singleprocessor machine, the kernel
-          will run faster if you say N here.
-
-          Note that if you say Y here and choose architecture "586" or
-          "Pentium" under "Processor family", the kernel will not work on 486
-          architectures. Similarly, multiprocessor kernels for the "PPro"
-          architecture may not work on all Pentium based boards.
-
-          People using multiprocessor machines who say Y here should also say
-          Y to "Enhanced Real Time Clock Support", below. The "Advanced Power
-          Management" code will be disabled if you say Y here.
-
-          See also the <file:Documentation/smp.txt>,
-          <file:Documentation/i386/IO-APIC.txt>,
-          <file:Documentation/nmi_watchdog.txt> and the SMP-HOWTO available at
-          <http://www.tldp.org/docs.html#howto>.
-
-          If you don't know what to do here, say N.
-
 config NR_CPUS
         int "Maximum number of CPUs (2-255)"
         range 2 255
@@ -522,6 +522,12 @@ config NUMA
 comment "NUMA (Summit) requires SMP, 64GB highmem support, ACPI"
         depends on X86_SUMMIT && (!HIGHMEM64G || !ACPI)
 
+config NODES_SHIFT
+        int
+        default "4" if X86_NUMAQ
+        default "3"
+        depends on NEED_MULTIPLE_NODES
+
 config HAVE_ARCH_BOOTMEM_NODE
         bool
         depends on NUMA
@@ -757,15 +763,6 @@ config HOTPLUG_CPU
 
           Say N.
 
-config DOUBLEFAULT
-        default y
-        bool "Enable doublefault exception handler" if EMBEDDED
-        help
-          This option allows trapping of rare doublefault exceptions that
-          would otherwise cause a system to silently reboot. Disabling this
-          option saves about 4k and might cause you much additional grey
-          hair.
-
 endmenu
 
 
diff --git a/arch/i386/Kconfig.cpu b/arch/i386/Kconfig.cpu
index 79603b3471f9..eb130482ba18 100644
--- a/arch/i386/Kconfig.cpu
+++ b/arch/i386/Kconfig.cpu
@@ -311,5 +311,5 @@ config X86_OOSTORE
 
 config X86_TSC
         bool
-        depends on (MWINCHIP3D || MWINCHIP2 || MCRUSOE || MEFFICEON || MCYRIXIII || MK7 || MK6 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || MK8 || MVIAC3_2 || MGEODEGX1) && !X86_NUMAQ
+        depends on (MWINCHIP3D || MWINCHIP2 || MCRUSOE || MEFFICEON || MCYRIXIII || MK7 || MK6 || MPENTIUM4 || MPENTIUMM || MPENTIUMIII || MPENTIUMII || M686 || M586MMX || M586TSC || MK8 || MVIAC3_2 || MGEODEGX1 || MGEODE_LX) && !X86_NUMAQ
         default y
diff --git a/arch/i386/boot/video.S b/arch/i386/boot/video.S
index 0000a2674537..c9343c3a8082 100644
--- a/arch/i386/boot/video.S
+++ b/arch/i386/boot/video.S
@@ -97,6 +97,7 @@
 #define PARAM_VESAPM_OFF        0x30
 #define PARAM_LFB_PAGES         0x32
 #define PARAM_VESA_ATTRIB       0x34
+#define PARAM_CAPABILITIES      0x36
 
 /* Define DO_STORE according to CONFIG_VIDEO_RETAIN */
 #ifdef CONFIG_VIDEO_RETAIN
@@ -233,6 +234,10 @@ mopar_gr:
         movw    18(%di), %ax
         movl    %eax, %fs:(PARAM_LFB_SIZE)
 
+# store mode capabilities
+        movl 10(%di), %eax
+        movl %eax, %fs:(PARAM_CAPABILITIES)
+
 # switching the DAC to 8-bit is for <= 8 bpp only
         movw    %fs:(PARAM_LFB_DEPTH), %ax
         cmpw    $8, %ax
diff --git a/arch/i386/kernel/acpi/boot.c b/arch/i386/kernel/acpi/boot.c
index 033066176b3e..030a0007a4e0 100644
--- a/arch/i386/kernel/acpi/boot.c
+++ b/arch/i386/kernel/acpi/boot.c
@@ -168,7 +168,7 @@ int __init acpi_parse_mcfg(unsigned long phys_addr, unsigned long size)
         unsigned long i;
         int config_size;
 
-        if (!phys_addr || !size)
+        if (!phys_addr || !size || !cpu_has_apic)
                 return -EINVAL;
 
         mcfg = (struct acpi_table_mcfg *)__acpi_map_table(phys_addr, size);
@@ -215,7 +215,7 @@ static int __init acpi_parse_madt(unsigned long phys_addr, unsigned long size)
 {
         struct acpi_table_madt *madt = NULL;
 
-        if (!phys_addr || !size)
+        if (!phys_addr || !size || !cpu_has_apic)
                 return -EINVAL;
 
         madt = (struct acpi_table_madt *)__acpi_map_table(phys_addr, size);
@@ -693,6 +693,9 @@ static int __init acpi_parse_madt_lapic_entries(void)
 {
         int count;
 
+        if (!cpu_has_apic)
+                return -ENODEV;
+
         /* 
          * Note that the LAPIC address is obtained from the MADT (32-bit value)
          * and (optionally) overriden by a LAPIC_ADDR_OVR entry (64-bit value).
@@ -751,6 +754,9 @@ static int __init acpi_parse_madt_ioapic_entries(void)
                 return -ENODEV;
         }
 
+        if (!cpu_has_apic)
+                return -ENODEV;
+
         /*
          * if "noapic" boot option, don't look for IO-APICs
          */
diff --git a/arch/i386/kernel/apic.c b/arch/i386/kernel/apic.c
index 6273bf74c203..254cee9f0b7b 100644
--- a/arch/i386/kernel/apic.c
+++ b/arch/i386/kernel/apic.c
@@ -62,6 +62,18 @@ int apic_verbosity;
 
 static void apic_pm_activate(void);
 
+int modern_apic(void)
+{
+        unsigned int lvr, version;
+        /* AMD systems use old APIC versions, so check the CPU */
+        if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
+                boot_cpu_data.x86 >= 0xf)
+                return 1;
+        lvr = apic_read(APIC_LVR);
+        version = GET_APIC_VERSION(lvr);
+        return version >= 0x14;
+}
+
 /*
  * 'what should we do if we get a hw irq event on an illegal vector'.
  * each architecture has to answer this themselves.
@@ -119,10 +131,7 @@ void enable_NMI_through_LVT0 (void * dummy)
 
 int get_physical_broadcast(void)
 {
-        unsigned int lvr, version;
-        lvr = apic_read(APIC_LVR);
-        version = GET_APIC_VERSION(lvr);
-        if (!APIC_INTEGRATED(version) || version >= 0x14)
+        if (modern_apic())
                 return 0xff;
         else
                 return 0xf;
@@ -349,9 +358,9 @@ int __init verify_local_APIC(void)
 
 void __init sync_Arb_IDs(void)
 {
-        /* Unsupported on P4 - see Intel Dev. Manual Vol. 3, Ch. 8.6.1 */
-        unsigned int ver = GET_APIC_VERSION(apic_read(APIC_LVR));
-        if (ver >= 0x14)        /* P4 or higher */
+        /* Unsupported on P4 - see Intel Dev. Manual Vol. 3, Ch. 8.6.1
+           And not needed on AMD */
+        if (modern_apic())
                 return;
         /*
          * Wait for idle.
diff --git a/arch/i386/kernel/cpu/amd.c b/arch/i386/kernel/cpu/amd.c
index 0810f81f2a05..ff2b2154ac1b 100644
--- a/arch/i386/kernel/cpu/amd.c
+++ b/arch/i386/kernel/cpu/amd.c
@@ -212,8 +212,6 @@ static void __init init_amd(struct cpuinfo_x86 *c)
 
         if (cpuid_eax(0x80000000) >= 0x80000008) {
                 c->x86_max_cores = (cpuid_ecx(0x80000008) & 0xff) + 1;
-                if (c->x86_max_cores & (c->x86_max_cores - 1))
-                        c->x86_max_cores = 1;
         }
 
         if (cpuid_eax(0x80000000) >= 0x80000007) {
diff --git a/arch/i386/kernel/cpu/cpufreq/powernow-k8.c b/arch/i386/kernel/cpu/cpufreq/powernow-k8.c
index 712a26bd4457..7c0e160a2145 100644
--- a/arch/i386/kernel/cpu/cpufreq/powernow-k8.c
+++ b/arch/i386/kernel/cpu/cpufreq/powernow-k8.c
@@ -46,7 +46,7 @@
 
 #define PFX "powernow-k8: "
 #define BFX PFX "BIOS error: "
-#define VERSION "version 1.60.1"
+#define VERSION "version 1.60.2"
 #include "powernow-k8.h"
 
 /* serialize freq changes  */
@@ -55,7 +55,7 @@ static DEFINE_MUTEX(fidvid_mutex);
 static struct powernow_k8_data *powernow_data[NR_CPUS];
 
 #ifndef CONFIG_SMP
-static cpumask_t cpu_core_map[1] = { CPU_MASK_ALL };
+static cpumask_t cpu_core_map[1];
 #endif
 
 /* Return a frequency in MHz, given an input fid */
@@ -910,6 +910,9 @@ static int powernowk8_target(struct cpufreq_policy *pol, unsigned targfreq, unsi
         unsigned int newstate;
         int ret = -EIO;
 
+        if (!data)
+                return -EINVAL;
+
         /* only run on specific CPU from here on */
         oldmask = current->cpus_allowed;
         set_cpus_allowed(current, cpumask_of_cpu(pol->cpu));
@@ -969,6 +972,9 @@ static int powernowk8_verify(struct cpufreq_policy *pol)
 {
         struct powernow_k8_data *data = powernow_data[pol->cpu];
 
+        if (!data)
+                return -EINVAL;
+
         return cpufreq_frequency_table_verify(pol, data->powernow_table);
 }
 
@@ -977,7 +983,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
 {
         struct powernow_k8_data *data;
         cpumask_t oldmask = CPU_MASK_ALL;
-        int rc, i;
+        int rc;
 
         if (!cpu_online(pol->cpu))
                 return -ENODEV;
@@ -1063,8 +1069,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol)
         printk("cpu_init done, current fid 0x%x, vid 0x%x\n",
                data->currfid, data->currvid);
 
-        for_each_cpu_mask(i, cpu_core_map[pol->cpu])
-                powernow_data[i] = data;
+        powernow_data[pol->cpu] = data;
 
         return 0;
 
@@ -1104,6 +1109,9 @@ static unsigned int powernowk8_get (unsigned int cpu)
         if (!data)
                 return -EINVAL;
 
+        if (!data)
+                return -EINVAL;
+
         set_cpus_allowed(current, cpumask_of_cpu(cpu));
         if (smp_processor_id() != cpu) {
                 printk(KERN_ERR PFX "limiting to CPU %d failed in powernowk8_get\n", cpu);
diff --git a/arch/i386/kernel/i386_ksyms.c b/arch/i386/kernel/i386_ksyms.c
index 055325056a74..036a9857936f 100644
--- a/arch/i386/kernel/i386_ksyms.c
+++ b/arch/i386/kernel/i386_ksyms.c
@@ -19,7 +19,6 @@ EXPORT_SYMBOL(__put_user_2);
 EXPORT_SYMBOL(__put_user_4);
 EXPORT_SYMBOL(__put_user_8);
 
-EXPORT_SYMBOL(strpbrk);
 EXPORT_SYMBOL(strstr);
 
 #ifdef CONFIG_SMP
diff --git a/arch/i386/kernel/mpparse.c b/arch/i386/kernel/mpparse.c
index 8d8aa9d1796d..34d21e21e012 100644
--- a/arch/i386/kernel/mpparse.c
+++ b/arch/i386/kernel/mpparse.c
@@ -38,12 +38,6 @@
 int smp_found_config;
 unsigned int __initdata maxcpus = NR_CPUS;
 
-#ifdef CONFIG_HOTPLUG_CPU
-#define CPU_HOTPLUG_ENABLED     (1)
-#else
-#define CPU_HOTPLUG_ENABLED     (0)
-#endif
-
 /*
  * Various Linux-internal data structures created from the
  * MP-table.
@@ -110,21 +104,6 @@ static int __init mpf_checksum(unsigned char *mp, int len)
 static int mpc_record; 
 static struct mpc_config_translation *translation_table[MAX_MPC_ENTRY] __initdata;
 
-#ifdef CONFIG_X86_NUMAQ
-static int MP_valid_apicid(int apicid, int version)
-{
-        return hweight_long(apicid & 0xf) == 1 && (apicid >> 4) != 0xf;
-}
-#else
-static int MP_valid_apicid(int apicid, int version)
-{
-        if (version >= 0x14)
-                return apicid < 0xff;
-        else
-                return apicid < 0xf;
-}
-#endif
-
 static void __devinit MP_processor_info (struct mpc_config_processor *m)
 {
         int ver, apicid;
@@ -190,12 +169,6 @@ static void __devinit MP_processor_info (struct mpc_config_processor *m)
 
         ver = m->mpc_apicver;
 
-        if (!MP_valid_apicid(apicid, ver)) {
-                printk(KERN_WARNING "Processor #%d INVALID. (Max ID: %d).\n",
-                        m->mpc_apicid, MAX_APICS);
-                return;
-        }
-
         /*
          * Validate version
          */
@@ -225,7 +198,14 @@ static void __devinit MP_processor_info (struct mpc_config_processor *m)
         cpu_set(num_processors, cpu_possible_map);
         num_processors++;
 
-        if (CPU_HOTPLUG_ENABLED || (num_processors > 8)) {
+        /*
+         * Would be preferable to switch to bigsmp when CONFIG_HOTPLUG_CPU=y
+         * but we need to work other dependencies like SMP_SUSPEND etc
+         * before this can be done without some confusion.
+         * if (CPU_HOTPLUG_ENABLED || num_processors > 8)
+         *       - Ashok Raj <ashok.raj@intel.com>
+         */
+        if (num_processors > 8) {
                 switch (boot_cpu_data.x86_vendor) {
                 case X86_VENDOR_INTEL:
                         if (!APIC_XAPIC(ver)) {
@@ -249,6 +229,13 @@ static void __init MP_bus_info (struct mpc_config_bus *m)
 
         mpc_oem_bus_info(m, str, translation_table[mpc_record]);
 
+        if (m->mpc_busid >= MAX_MP_BUSSES) {
+                printk(KERN_WARNING "MP table busid value (%d) for bustype %s "
+                        " is too large, max. supported is %d\n",
+                        m->mpc_busid, str, MAX_MP_BUSSES - 1);
+                return;
+        }
+
         if (strncmp(str, BUSTYPE_ISA, sizeof(BUSTYPE_ISA)-1) == 0) {
                 mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA;
         } else if (strncmp(str, BUSTYPE_EISA, sizeof(BUSTYPE_EISA)-1) == 0) {
diff --git a/arch/i386/kernel/reboot_fixups.c b/arch/i386/kernel/reboot_fixups.c
index 10e21a4773dd..99aab41a05b0 100644
--- a/arch/i386/kernel/reboot_fixups.c
+++ b/arch/i386/kernel/reboot_fixups.c
@@ -51,7 +51,5 @@ void mach_reboot_fixups(void)
 
                 cur->reboot_fixup(dev);
         }
-
-        printk(KERN_WARNING "No reboot fixup found for your hardware\n");
 }
 
diff --git a/arch/i386/kernel/setup.c b/arch/i386/kernel/setup.c
index eacc3f0a2ea4..80cb3b2d0997 100644
--- a/arch/i386/kernel/setup.c
+++ b/arch/i386/kernel/setup.c
@@ -963,6 +963,36 @@ efi_memory_present_wrapper(unsigned long start, unsigned long end, void *arg)
         return 0;
 }
 
+ /*
+  * This function checks if the entire range <start,end> is mapped with type.
+  *
+  * Note: this function only works correct if the e820 table is sorted and
+  * not-overlapping, which is the case
+  */
+int __init
+e820_all_mapped(unsigned long start, unsigned long end, unsigned type)
+{
+        int i;
+        for (i = 0; i < e820.nr_map; i++) {
+                struct e820entry *ei = &e820.map[i];
+                if (type && ei->type != type)
+                        continue;
+                /* is the region (part) in overlap with the current region ?*/
+                if (ei->addr >= end || ei->addr + ei->size <= start)
+                        continue;
+                /* if the region is at the beginning of <start,end> we move
+                 * start to the end of the region since it's ok until there
+                 */
+                if (ei->addr <= start)
+                        start = ei->addr + ei->size;
+                /* if start is now at or beyond end, we're done, full
+                 * coverage */
+                if (start >= end)
+                        return 1; /* we're done */
+        }
+        return 0;
+}
+
 /*
  * Find the highest page frame number we have available
  */
@@ -1317,8 +1347,8 @@ legacy_init_iomem_resources(struct resource *code_resource, struct resource *dat
 /*
  * Request address space for all standard resources
  *
- * This is called just before pcibios_assign_resources(), which is also
- * an fs_initcall, but is linked in later (in arch/i386/pci/i386.c).
+ * This is called just before pcibios_init(), which is also a
+ * subsys_initcall, but is linked in later (in arch/i386/pci/common.c).
  */
 static int __init request_standard_resources(void)
 {
@@ -1339,7 +1369,7 @@ static int __init request_standard_resources(void)
         return 0;
 }
 
-fs_initcall(request_standard_resources);
+subsys_initcall(request_standard_resources);
 
 static void __init register_memory(void)
 {
diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c
index e38527994590..2d22f5761b1d 100644
--- a/arch/i386/kernel/traps.c
+++ b/arch/i386/kernel/traps.c
@@ -365,6 +365,9 @@ void die(const char * str, struct pt_regs * regs, long err)
 
         if (++die.lock_owner_depth < 3) {
                 int nl = 0;
+                unsigned long esp;
+                unsigned short ss;
+
                 handle_BUG(regs);
                 printk(KERN_EMERG "%s: %04lx [#%d]\n", str, err & 0xffff, ++die_counter);
 #ifdef CONFIG_PREEMPT
@@ -387,8 +390,19 @@ void die(const char * str, struct pt_regs * regs, long err)
                         printk("\n");
                 if (notify_die(DIE_OOPS, str, regs, err,
                                         current->thread.trap_no, SIGSEGV) !=
-                                NOTIFY_STOP)
+                                NOTIFY_STOP) {
                         show_registers(regs);
+                        /* Executive summary in case the oops scrolled away */
+                        esp = (unsigned long) (&regs->esp);
+                        savesegment(ss, ss);
+                        if (user_mode(regs)) {
+                                esp = regs->esp;
+                                ss = regs->xss & 0xffff;
+                        }
+                        printk(KERN_EMERG "EIP: [<%08lx>] ", regs->eip);
+                        print_symbol("%s", regs->eip);
+                        printk(" SS:ESP %04x:%08lx\n", ss, esp);
+                }
                 else
                         regs = NULL;
         } else
diff --git a/arch/i386/mach-voyager/voyager_cat.c b/arch/i386/mach-voyager/voyager_cat.c
index 23967fe658d3..3039539de51e 100644
--- a/arch/i386/mach-voyager/voyager_cat.c
+++ b/arch/i386/mach-voyager/voyager_cat.c
@@ -106,9 +106,15 @@ voyager_module_t *voyager_cat_list;
 
 /* the I/O port assignments for the VIC and QIC */
 static struct resource vic_res = {
-        "Voyager Interrupt Controller", 0xFC00, 0xFC6F };
+        .name   = "Voyager Interrupt Controller",
+        .start  = 0xFC00,
+        .end    = 0xFC6F
+};
 static struct resource qic_res = {
-        "Quad Interrupt Controller", 0xFC70, 0xFCFF };
+        .name   = "Quad Interrupt Controller",
+        .start  = 0xFC70,
+        .end    = 0xFCFF
+};
 
 /* This function is used to pack a data bit stream inside a message.
  * It writes num_bits of the data buffer in msg starting at start_bit.
diff --git a/arch/i386/mm/init.c b/arch/i386/mm/init.c
index 9f66ac582a8b..ae6534ad8161 100644
--- a/arch/i386/mm/init.c
+++ b/arch/i386/mm/init.c
@@ -651,6 +651,7 @@ void __init mem_init(void)
  * Specifically, in the case of x86, we will always add
  * memory to the highmem for now.
  */
+#ifdef CONFIG_HOTPLUG_MEMORY
 #ifndef CONFIG_NEED_MULTIPLE_NODES
 int add_memory(u64 start, u64 size)
 {
@@ -667,6 +668,7 @@ int remove_memory(u64 start, u64 size)
         return -EINVAL;
 }
 #endif
+#endif
 
 kmem_cache_t *pgd_cache;
 kmem_cache_t *pmd_cache;
diff --git a/arch/i386/pci/direct.c b/arch/i386/pci/direct.c
index 99012b93bd12..5d81fb510375 100644
--- a/arch/i386/pci/direct.c
+++ b/arch/i386/pci/direct.c
@@ -4,6 +4,7 @@
 
 #include <linux/pci.h>
 #include <linux/init.h>
+#include <linux/dmi.h>
 #include "pci.h"
 
 /*
@@ -18,8 +19,10 @@ int pci_conf1_read(unsigned int seg, unsigned int bus,
 {
         unsigned long flags;
 
-        if (!value || (bus > 255) || (devfn > 255) || (reg > 255))
+        if ((bus > 255) || (devfn > 255) || (reg > 255)) {
+                *value = -1;
                 return -EINVAL;
+        }
 
         spin_lock_irqsave(&pci_config_lock, flags);
 
@@ -91,8 +94,10 @@ static int pci_conf2_read(unsigned int seg, unsigned int bus,
         unsigned long flags;
         int dev, fn;
 
-        if (!value || (bus > 255) || (devfn > 255) || (reg > 255))
+        if ((bus > 255) || (devfn > 255) || (reg > 255)) {
+                *value = -1;
                 return -EINVAL;
+        }
 
         dev = PCI_SLOT(devfn);
         fn = PCI_FUNC(devfn);
@@ -188,6 +193,10 @@ static int __init pci_sanity_check(struct pci_raw_ops *o)
 
         if (pci_probe & PCI_NO_CHECKS)
                 return 1;
+        /* Assume Type 1 works for newer systems.
+           This handles machines that don't have anything on PCI Bus 0. */
+        if (dmi_get_year(DMI_BIOS_DATE) >= 2001)
+                return 1;
 
         for (devfn = 0; devfn < 0x100; devfn++) {
                 if (o->read(0, 0, devfn, PCI_CLASS_DEVICE, 2, &x))
diff --git a/arch/i386/pci/mmconfig.c b/arch/i386/pci/mmconfig.c
index 613789071f30..6b1ea0c9a570 100644
--- a/arch/i386/pci/mmconfig.c
+++ b/arch/i386/pci/mmconfig.c
@@ -12,14 +12,20 @@
 #include <linux/pci.h>
 #include <linux/init.h>
 #include <linux/acpi.h>
+#include <asm/e820.h>
 #include "pci.h"
 
+#define MMCONFIG_APER_SIZE (256*1024*1024)
+
+/* Assume systems with more busses have correct MCFG */
+#define MAX_CHECK_BUS 16
+
 #define mmcfg_virt_addr ((void __iomem *) fix_to_virt(FIX_PCIE_MCFG))
 
 /* The base address of the last MMCONFIG device accessed */
 static u32 mmcfg_last_accessed_device;
 
-static DECLARE_BITMAP(fallback_slots, 32);
+static DECLARE_BITMAP(fallback_slots, MAX_CHECK_BUS*32);
 
 /*
  * Functions for accessing PCI configuration space with MMCONFIG accesses
@@ -29,8 +35,8 @@ static u32 get_base_addr(unsigned int seg, int bus, unsigned devfn)
         int cfg_num = -1;
         struct acpi_table_mcfg_config *cfg;
 
-        if (seg == 0 && bus == 0 &&
-            test_bit(PCI_SLOT(devfn), fallback_slots))
+        if (seg == 0 && bus < MAX_CHECK_BUS &&
+            test_bit(PCI_SLOT(devfn) + 32*bus, fallback_slots))
                 return 0;
 
         while (1) {
@@ -74,8 +80,10 @@ static int pci_mmcfg_read(unsigned int seg, unsigned int bus,
         unsigned long flags;
         u32 base;
 
-        if (!value || (bus > 255) || (devfn > 255) || (reg > 4095))
+        if ((bus > 255) || (devfn > 255) || (reg > 4095)) {
+                *value = -1;
                 return -EINVAL;
+        }
 
         base = get_base_addr(seg, bus, devfn);
         if (!base)
@@ -146,29 +154,34 @@ static struct pci_raw_ops pci_mmcfg = {
    Normally this can be expressed in the MCFG by not listing them
    and assigning suitable _SEGs, but this isn't implemented in some BIOS.
    Instead try to discover all devices on bus 0 that are unreachable using MM
-   and fallback for them.
-   We only do this for bus 0/seg 0 */
+   and fallback for them. */
 static __init void unreachable_devices(void)
 {
-        int i;
+        int i, k;
         unsigned long flags;
 
-        for (i = 0; i < 32; i++) {
-                u32 val1;
-                u32 addr;
-
-                pci_conf1_read(0, 0, PCI_DEVFN(i, 0), 0, 4, &val1);
-                if (val1 == 0xffffffff)
-                        continue;
-
-                /* Locking probably not needed, but safer */
-                spin_lock_irqsave(&pci_config_lock, flags);
-                addr = get_base_addr(0, 0, PCI_DEVFN(i, 0));
-                if (addr != 0)
-                        pci_exp_set_dev_base(addr, 0, PCI_DEVFN(i, 0));
-                if (addr == 0 || readl((u32 __iomem *)mmcfg_virt_addr) != val1)
-                        set_bit(i, fallback_slots);
-                spin_unlock_irqrestore(&pci_config_lock, flags);
+        for (k = 0; k < MAX_CHECK_BUS; k++) {
+                for (i = 0; i < 32; i++) {
+                        u32 val1;
+                        u32 addr;
+
+                        pci_conf1_read(0, k, PCI_DEVFN(i, 0), 0, 4, &val1);
+                        if (val1 == 0xffffffff)
+                                continue;
+
+                        /* Locking probably not needed, but safer */
+                        spin_lock_irqsave(&pci_config_lock, flags);
+                        addr = get_base_addr(0, k, PCI_DEVFN(i, 0));
+                        if (addr != 0)
+                                pci_exp_set_dev_base(addr, k, PCI_DEVFN(i, 0));
+                        if (addr == 0 ||
+                            readl((u32 __iomem *)mmcfg_virt_addr) != val1) {
+                                set_bit(i, fallback_slots);
+                                printk(KERN_NOTICE
+                        "PCI: No mmconfig possible on %x:%x\n", k, i);
+                        }
+                        spin_unlock_irqrestore(&pci_config_lock, flags);
+                }
         }
 }
 
@@ -183,6 +196,14 @@ void __init pci_mmcfg_init(void)
             (pci_mmcfg_config[0].base_address == 0))
                 return;
 
+        if (!e820_all_mapped(pci_mmcfg_config[0].base_address,
+                        pci_mmcfg_config[0].base_address + MMCONFIG_APER_SIZE,
+                        E820_RESERVED)) {
+                printk(KERN_ERR "PCI: BIOS Bug: MCFG area is not E820-reserved\n");
+                printk(KERN_ERR "PCI: Not using MMCONFIG.\n");
+                return;
+        }
+
         printk(KERN_INFO "PCI: Using MMCONFIG\n");
         raw_pci_ops = &pci_mmcfg;
         pci_probe = (pci_probe & ~PCI_PROBE_MASK) | PCI_PROBE_MMCONF;