Merge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86, asm: Clean up desc.h a bit
  x86, amd: Do not enable ARAT feature on AMD processors below family 0x12
  x86: Move do_page_fault()'s error path under unlikely()
  x86, efi: Retain boot service code until after switching to virtual mode
  x86: Remove unnecessary check in detect_ht()
  x86: Reorder mm_context_t to remove x86_64 alignment padding and thus shrink mm_struct
  x86, UV: Clean up uv_tlb.c
  x86, UV: Add support for SGI UV2 hub chip
  x86, cpufeature: Update CPU feature RDRND to RDRAND

15 files changed, 2283 insertions, 1194 deletions

diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h
index 5dc6acc98dbd..71cc3800712c 100644
--- a/arch/x86/include/asm/cpufeature.h
+++ b/arch/x86/include/asm/cpufeature.h
@@ -125,7 +125,7 @@
 #define X86_FEATURE_OSXSAVE	(4*32+27) /* "" XSAVE enabled in the OS */
 #define X86_FEATURE_AVX		(4*32+28) /* Advanced Vector Extensions */
 #define X86_FEATURE_F16C	(4*32+29) /* 16-bit fp conversions */
-#define X86_FEATURE_RDRND	(4*32+30) /* The RDRAND instruction */
+#define X86_FEATURE_RDRAND	(4*32+30) /* The RDRAND instruction */
 #define X86_FEATURE_HYPERVISOR	(4*32+31) /* Running on a hypervisor */
 
 /* VIA/Cyrix/Centaur-defined CPU features, CPUID level 0xC0000001, word 5 */
diff --git a/arch/x86/include/asm/desc.h b/arch/x86/include/asm/desc.h
index 617bd56b3070..7b439d9aea2a 100644
--- a/arch/x86/include/asm/desc.h
+++ b/arch/x86/include/asm/desc.h
@@ -4,30 +4,33 @@
 #include <asm/desc_defs.h>
 #include <asm/ldt.h>
 #include <asm/mmu.h>
+
 #include <linux/smp.h>
 
-static inline void fill_ldt(struct desc_struct *desc,
-			    const struct user_desc *info)
-{
-	desc->limit0 = info->limit & 0x0ffff;
-	desc->base0 = info->base_addr & 0x0000ffff;
-
-	desc->base1 = (info->base_addr & 0x00ff0000) >> 16;
-	desc->type = (info->read_exec_only ^ 1) << 1;
-	desc->type |= info->contents << 2;
-	desc->s = 1;
-	desc->dpl = 0x3;
-	desc->p = info->seg_not_present ^ 1;
-	desc->limit = (info->limit & 0xf0000) >> 16;
-	desc->avl = info->useable;
-	desc->d = info->seg_32bit;
-	desc->g = info->limit_in_pages;
-	desc->base2 = (info->base_addr & 0xff000000) >> 24;
+static inline void fill_ldt(struct desc_struct *desc, const struct user_desc *info)
+{
+	desc->limit0		= info->limit & 0x0ffff;
+
+	desc->base0		= (info->base_addr & 0x0000ffff);
+	desc->base1		= (info->base_addr & 0x00ff0000) >> 16;
+
+	desc->type		= (info->read_exec_only ^ 1) << 1;
+	desc->type	       |= info->contents << 2;
+
+	desc->s			= 1;
+	desc->dpl		= 0x3;
+	desc->p			= info->seg_not_present ^ 1;
+	desc->limit		= (info->limit & 0xf0000) >> 16;
+	desc->avl		= info->useable;
+	desc->d			= info->seg_32bit;
+	desc->g			= info->limit_in_pages;
+
+	desc->base2		= (info->base_addr & 0xff000000) >> 24;
 	/*
 	 * Don't allow setting of the lm bit. It is useless anyway
 	 * because 64bit system calls require __USER_CS:
 	 */
-	desc->l = 0;
+	desc->l			= 0;
 }
 
 extern struct desc_ptr idt_descr;
@@ -36,6 +39,7 @@ extern gate_desc idt_table[];
 struct gdt_page {
 	struct desc_struct gdt[GDT_ENTRIES];
 } __attribute__((aligned(PAGE_SIZE)));
+
 DECLARE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page);
 
 static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu)
@@ -48,16 +52,16 @@ static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu)
 static inline void pack_gate(gate_desc *gate, unsigned type, unsigned long func,
 			     unsigned dpl, unsigned ist, unsigned seg)
 {
-	gate->offset_low = PTR_LOW(func);
-	gate->segment = __KERNEL_CS;
-	gate->ist = ist;
-	gate->p = 1;
-	gate->dpl = dpl;
-	gate->zero0 = 0;
-	gate->zero1 = 0;
-	gate->type = type;
-	gate->offset_middle = PTR_MIDDLE(func);
-	gate->offset_high = PTR_HIGH(func);
+	gate->offset_low	= PTR_LOW(func);
+	gate->segment		= __KERNEL_CS;
+	gate->ist		= ist;
+	gate->p			= 1;
+	gate->dpl		= dpl;
+	gate->zero0		= 0;
+	gate->zero1		= 0;
+	gate->type		= type;
+	gate->offset_middle	= PTR_MIDDLE(func);
+	gate->offset_high	= PTR_HIGH(func);
 }
 
 #else
@@ -66,8 +70,7 @@ static inline void pack_gate(gate_desc *gate, unsigned char type,
 			     unsigned short seg)
 {
 	gate->a = (seg << 16) | (base & 0xffff);
-	gate->b = (base & 0xffff0000) |
-		  (((0x80 | type | (dpl << 5)) & 0xff) << 8);
+	gate->b = (base & 0xffff0000) | (((0x80 | type | (dpl << 5)) & 0xff) << 8);
 }
 
 #endif
@@ -75,31 +78,29 @@ static inline void pack_gate(gate_desc *gate, unsigned char type,
 static inline int desc_empty(const void *ptr)
 {
 	const u32 *desc = ptr;
+
 	return !(desc[0] | desc[1]);
 }
 
 #ifdef CONFIG_PARAVIRT
 #include <asm/paravirt.h>
 #else
-#define load_TR_desc() native_load_tr_desc()
-#define load_gdt(dtr) native_load_gdt(dtr)
-#define load_idt(dtr) native_load_idt(dtr)
-#define load_tr(tr) asm volatile("ltr %0"::"m" (tr))
-#define load_ldt(ldt) asm volatile("lldt %0"::"m" (ldt))
-
-#define store_gdt(dtr) native_store_gdt(dtr)
-#define store_idt(dtr) native_store_idt(dtr)
-#define store_tr(tr) (tr = native_store_tr())
-
-#define load_TLS(t, cpu) native_load_tls(t, cpu)
-#define set_ldt native_set_ldt
-
-#define write_ldt_entry(dt, entry, desc)	\
-	native_write_ldt_entry(dt, entry, desc)
-#define write_gdt_entry(dt, entry, desc, type)		\
-	native_write_gdt_entry(dt, entry, desc, type)
-#define write_idt_entry(dt, entry, g)		\
-	native_write_idt_entry(dt, entry, g)
+#define load_TR_desc()				native_load_tr_desc()
+#define load_gdt(dtr)				native_load_gdt(dtr)
+#define load_idt(dtr)				native_load_idt(dtr)
+#define load_tr(tr)				asm volatile("ltr %0"::"m" (tr))
+#define load_ldt(ldt)				asm volatile("lldt %0"::"m" (ldt))
+
+#define store_gdt(dtr)				native_store_gdt(dtr)
+#define store_idt(dtr)				native_store_idt(dtr)
+#define store_tr(tr)				(tr = native_store_tr())
+
+#define load_TLS(t, cpu)			native_load_tls(t, cpu)
+#define set_ldt					native_set_ldt
+
+#define write_ldt_entry(dt, entry, desc)	native_write_ldt_entry(dt, entry, desc)
+#define write_gdt_entry(dt, entry, desc, type)	native_write_gdt_entry(dt, entry, desc, type)
+#define write_idt_entry(dt, entry, g)		native_write_idt_entry(dt, entry, g)
 
 static inline void paravirt_alloc_ldt(struct desc_struct *ldt, unsigned entries)
 {
@@ -112,33 +113,27 @@ static inline void paravirt_free_ldt(struct desc_struct *ldt, unsigned entries)
 
 #define store_ldt(ldt) asm("sldt %0" : "=m"(ldt))
 
-static inline void native_write_idt_entry(gate_desc *idt, int entry,
-					  const gate_desc *gate)
+static inline void native_write_idt_entry(gate_desc *idt, int entry, const gate_desc *gate)
 {
 	memcpy(&idt[entry], gate, sizeof(*gate));
 }
 
-static inline void native_write_ldt_entry(struct desc_struct *ldt, int entry,
-					  const void *desc)
+static inline void native_write_ldt_entry(struct desc_struct *ldt, int entry, const void *desc)
 {
 	memcpy(&ldt[entry], desc, 8);
 }
 
-static inline void native_write_gdt_entry(struct desc_struct *gdt, int entry,
-					  const void *desc, int type)
+static inline void
+native_write_gdt_entry(struct desc_struct *gdt, int entry, const void *desc, int type)
 {
 	unsigned int size;
+
 	switch (type) {
-	case DESC_TSS:
-		size = sizeof(tss_desc);
-		break;
-	case DESC_LDT:
-		size = sizeof(ldt_desc);
-		break;
-	default:
-		size = sizeof(struct desc_struct);
-		break;
+	case DESC_TSS:	size = sizeof(tss_desc);	break;
+	case DESC_LDT:	size = sizeof(ldt_desc);	break;
+	default:	size = sizeof(*gdt);		break;
 	}
+
 	memcpy(&gdt[entry], desc, size);
 }
 
@@ -154,20 +149,21 @@ static inline void pack_descriptor(struct desc_struct *desc, unsigned long base,
 }
 
 
-static inline void set_tssldt_descriptor(void *d, unsigned long addr,
-					 unsigned type, unsigned size)
+static inline void set_tssldt_descriptor(void *d, unsigned long addr, unsigned type, unsigned size)
 {
 #ifdef CONFIG_X86_64
 	struct ldttss_desc64 *desc = d;
+
 	memset(desc, 0, sizeof(*desc));
-	desc->limit0 = size & 0xFFFF;
-	desc->base0 = PTR_LOW(addr);
-	desc->base1 = PTR_MIDDLE(addr) & 0xFF;
-	desc->type = type;
-	desc->p = 1;
-	desc->limit1 = (size >> 16) & 0xF;
-	desc->base2 = (PTR_MIDDLE(addr) >> 8) & 0xFF;
-	desc->base3 = PTR_HIGH(addr);
+
+	desc->limit0		= size & 0xFFFF;
+	desc->base0		= PTR_LOW(addr);
+	desc->base1		= PTR_MIDDLE(addr) & 0xFF;
+	desc->type		= type;
+	desc->p			= 1;
+	desc->limit1		= (size >> 16) & 0xF;
+	desc->base2		= (PTR_MIDDLE(addr) >> 8) & 0xFF;
+	desc->base3		= PTR_HIGH(addr);
 #else
 	pack_descriptor((struct desc_struct *)d, addr, size, 0x80 | type, 0);
 #endif
@@ -237,14 +233,16 @@ static inline void native_store_idt(struct desc_ptr *dtr)
 static inline unsigned long native_store_tr(void)
 {
 	unsigned long tr;
+
 	asm volatile("str %0":"=r" (tr));
+
 	return tr;
 }
 
 static inline void native_load_tls(struct thread_struct *t, unsigned int cpu)
 {
-	unsigned int i;
 	struct desc_struct *gdt = get_cpu_gdt_table(cpu);
+	unsigned int i;
 
 	for (i = 0; i < GDT_ENTRY_TLS_ENTRIES; i++)
 		gdt[GDT_ENTRY_TLS_MIN + i] = t->tls_array[i];
@@ -313,6 +311,7 @@ static inline void _set_gate(int gate, unsigned type, void *addr,
 			     unsigned dpl, unsigned ist, unsigned seg)
 {
 	gate_desc s;
+
 	pack_gate(&s, type, (unsigned long)addr, dpl, ist, seg);
 	/*
 	 * does not need to be atomic because it is only done once at
@@ -343,8 +342,9 @@ static inline void alloc_system_vector(int vector)
 		set_bit(vector, used_vectors);
 		if (first_system_vector > vector)
 			first_system_vector = vector;
-	} else
+	} else {
 		BUG();
+	}
 }
 
 static inline void alloc_intr_gate(unsigned int n, void *addr)
diff --git a/arch/x86/include/asm/mmu.h b/arch/x86/include/asm/mmu.h
index aeff3e89b222..5f55e6962769 100644
--- a/arch/x86/include/asm/mmu.h
+++ b/arch/x86/include/asm/mmu.h
@@ -11,14 +11,14 @@
 typedef struct {
 	void *ldt;
 	int size;
-	struct mutex lock;
-	void *vdso;
 
 #ifdef CONFIG_X86_64
 	/* True if mm supports a task running in 32 bit compatibility mode. */
 	unsigned short ia32_compat;
 #endif
 
+	struct mutex lock;
+	void *vdso;
 } mm_context_t;
 
 #ifdef CONFIG_SMP
diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h
index 130f1eeee5fe..a291c40efd43 100644
--- a/arch/x86/include/asm/uv/uv_bau.h
+++ b/arch/x86/include/asm/uv/uv_bau.h
@@ -5,7 +5,7 @@
  *
  * SGI UV Broadcast Assist Unit definitions
  *
- * Copyright (C) 2008 Silicon Graphics, Inc. All rights reserved.
+ * Copyright (C) 2008-2011 Silicon Graphics, Inc. All rights reserved.
  */
 
 #ifndef _ASM_X86_UV_UV_BAU_H
@@ -35,17 +35,20 @@
 
 #define MAX_CPUS_PER_UVHUB		64
 #define MAX_CPUS_PER_SOCKET		32
-#define UV_ADP_SIZE			64 /* hardware-provided max. */
-#define UV_CPUS_PER_ACT_STATUS		32 /* hardware-provided max. */
-#define UV_ITEMS_PER_DESCRIPTOR		8
+#define ADP_SZ				64 /* hardware-provided max. */
+#define UV_CPUS_PER_AS			32 /* hardware-provided max. */
+#define ITEMS_PER_DESC			8
 /* the 'throttle' to prevent the hardware stay-busy bug */
 #define MAX_BAU_CONCURRENT		3
 #define UV_ACT_STATUS_MASK		0x3
 #define UV_ACT_STATUS_SIZE		2
 #define UV_DISTRIBUTION_SIZE		256
 #define UV_SW_ACK_NPENDING		8
-#define UV_NET_ENDPOINT_INTD		0x38
-#define UV_DESC_BASE_PNODE_SHIFT	49
+#define UV1_NET_ENDPOINT_INTD		0x38
+#define UV2_NET_ENDPOINT_INTD		0x28
+#define UV_NET_ENDPOINT_INTD		(is_uv1_hub() ?			\
+			UV1_NET_ENDPOINT_INTD : UV2_NET_ENDPOINT_INTD)
+#define UV_DESC_PSHIFT			49
 #define UV_PAYLOADQ_PNODE_SHIFT		49
 #define UV_PTC_BASENAME			"sgi_uv/ptc_statistics"
 #define UV_BAU_BASENAME			"sgi_uv/bau_tunables"
@@ -53,29 +56,64 @@
 #define UV_BAU_TUNABLES_FILE		"bau_tunables"
 #define WHITESPACE			" \t\n"
 #define uv_physnodeaddr(x)		((__pa((unsigned long)(x)) & uv_mmask))
-#define UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT 15
-#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT 16
-#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD 0x0000000009UL
+#define cpubit_isset(cpu, bau_local_cpumask) \
+	test_bit((cpu), (bau_local_cpumask).bits)
+
 /* [19:16] SOFT_ACK timeout period  19: 1 is urgency 7  17:16 1 is multiplier */
-#define BAU_MISC_CONTROL_MULT_MASK 3
+/*
+ * UV2: Bit 19 selects between
+ *  (0): 10 microsecond timebase and
+ *  (1): 80 microseconds
+ *  we're using 655us, similar to UV1: 65 units of 10us
+ */
+#define UV1_INTD_SOFT_ACK_TIMEOUT_PERIOD (9UL)
+#define UV2_INTD_SOFT_ACK_TIMEOUT_PERIOD (65*10UL)
+
+#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD	(is_uv1_hub() ?			\
+		UV1_INTD_SOFT_ACK_TIMEOUT_PERIOD :			\
+		UV2_INTD_SOFT_ACK_TIMEOUT_PERIOD)
 
-#define UVH_AGING_PRESCALE_SEL 0x000000b000UL
+#define BAU_MISC_CONTROL_MULT_MASK	3
+
+#define UVH_AGING_PRESCALE_SEL		0x000000b000UL
 /* [30:28] URGENCY_7  an index into a table of times */
-#define BAU_URGENCY_7_SHIFT 28
-#define BAU_URGENCY_7_MASK 7
+#define BAU_URGENCY_7_SHIFT		28
+#define BAU_URGENCY_7_MASK		7
 
-#define UVH_TRANSACTION_TIMEOUT 0x000000b200UL
+#define UVH_TRANSACTION_TIMEOUT		0x000000b200UL
 /* [45:40] BAU - BAU transaction timeout select - a multiplier */
-#define BAU_TRANS_SHIFT 40
-#define BAU_TRANS_MASK 0x3f
+#define BAU_TRANS_SHIFT			40
+#define BAU_TRANS_MASK			0x3f
+
+/*
+ * shorten some awkward names
+ */
+#define AS_PUSH_SHIFT UVH_LB_BAU_SB_ACTIVATION_CONTROL_PUSH_SHFT
+#define SOFTACK_MSHIFT UVH_LB_BAU_MISC_CONTROL_ENABLE_INTD_SOFT_ACK_MODE_SHFT
+#define SOFTACK_PSHIFT UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHFT
+#define SOFTACK_TIMEOUT_PERIOD UV_INTD_SOFT_ACK_TIMEOUT_PERIOD
+#define write_gmmr	uv_write_global_mmr64
+#define write_lmmr	uv_write_local_mmr
+#define read_lmmr	uv_read_local_mmr
+#define read_gmmr	uv_read_global_mmr64
 
 /*
  * bits in UVH_LB_BAU_SB_ACTIVATION_STATUS_0/1
  */
-#define DESC_STATUS_IDLE		0
-#define DESC_STATUS_ACTIVE		1
-#define DESC_STATUS_DESTINATION_TIMEOUT	2
-#define DESC_STATUS_SOURCE_TIMEOUT	3
+#define DS_IDLE				0
+#define DS_ACTIVE			1
+#define DS_DESTINATION_TIMEOUT		2
+#define DS_SOURCE_TIMEOUT		3
+/*
+ * bits put together from HRP_LB_BAU_SB_ACTIVATION_STATUS_0/1/2
+ * values 1 and 5 will not occur
+ */
+#define UV2H_DESC_IDLE			0
+#define UV2H_DESC_DEST_TIMEOUT		2
+#define UV2H_DESC_DEST_STRONG_NACK	3
+#define UV2H_DESC_BUSY			4
+#define UV2H_DESC_SOURCE_TIMEOUT	6
+#define UV2H_DESC_DEST_PUT_ERR		7
 
 /*
  * delay for 'plugged' timeout retries, in microseconds
@@ -86,15 +124,24 @@
  * threshholds at which to use IPI to free resources
  */
 /* after this # consecutive 'plugged' timeouts, use IPI to release resources */
-#define PLUGSB4RESET 100
+#define PLUGSB4RESET			100
 /* after this many consecutive timeouts, use IPI to release resources */
-#define TIMEOUTSB4RESET 1
+#define TIMEOUTSB4RESET			1
 /* at this number uses of IPI to release resources, giveup the request */
-#define IPI_RESET_LIMIT 1
+#define IPI_RESET_LIMIT			1
 /* after this # consecutive successes, bump up the throttle if it was lowered */
-#define COMPLETE_THRESHOLD 5
+#define COMPLETE_THRESHOLD		5
+
+#define UV_LB_SUBNODEID			0x10
 
-#define UV_LB_SUBNODEID 0x10
+/* these two are the same for UV1 and UV2: */
+#define UV_SA_SHFT UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHFT
+#define UV_SA_MASK UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_MASK
+/* 4 bits of software ack period */
+#define UV2_ACK_MASK			0x7UL
+#define UV2_ACK_UNITS_SHFT		3
+#define UV2_LEG_SHFT UV2H_LB_BAU_MISC_CONTROL_USE_LEGACY_DESCRIPTOR_FORMATS_SHFT
+#define UV2_EXT_SHFT UV2H_LB_BAU_MISC_CONTROL_ENABLE_EXTENDED_SB_STATUS_SHFT
 
 /*
  * number of entries in the destination side payload queue
@@ -115,9 +162,16 @@
 /*
  * tuning the action when the numalink network is extremely delayed
  */
-#define CONGESTED_RESPONSE_US 1000 /* 'long' response time, in microseconds */
-#define CONGESTED_REPS 10 /* long delays averaged over this many broadcasts */
-#define CONGESTED_PERIOD 30 /* time for the bau to be disabled, in seconds */
+#define CONGESTED_RESPONSE_US		1000	/* 'long' response time, in
+						   microseconds */
+#define CONGESTED_REPS			10	/* long delays averaged over
+						   this many broadcasts */
+#define CONGESTED_PERIOD		30	/* time for the bau to be
+						   disabled, in seconds */
+/* see msg_type: */
+#define MSG_NOOP			0
+#define MSG_REGULAR			1
+#define MSG_RETRY			2
 
 /*
  * Distribution: 32 bytes (256 bits) (bytes 0-0x1f of descriptor)
@@ -129,8 +183,8 @@
  * 'base_dest_nasid' field of the header corresponds to the
  * destination nodeID associated with that specified bit.
  */
-struct bau_target_uvhubmask {
-	unsigned long bits[BITS_TO_LONGS(UV_DISTRIBUTION_SIZE)];
+struct bau_targ_hubmask {
+	unsigned long		bits[BITS_TO_LONGS(UV_DISTRIBUTION_SIZE)];
 };
 
 /*
@@ -139,7 +193,7 @@ struct bau_target_uvhubmask {
  *  enough bits for max. cpu's per uvhub)
  */
 struct bau_local_cpumask {
-	unsigned long bits;
+	unsigned long		bits;
 };
 
 /*
@@ -160,14 +214,14 @@ struct bau_local_cpumask {
  * The payload is software-defined for INTD transactions
  */
 struct bau_msg_payload {
-	unsigned long address;		/* signifies a page or all TLB's
-						of the cpu */
+	unsigned long	address;		/* signifies a page or all
+						   TLB's of the cpu */
 	/* 64 bits */
-	unsigned short sending_cpu;	/* filled in by sender */
+	unsigned short	sending_cpu;		/* filled in by sender */
 	/* 16 bits */
-	unsigned short acknowledge_count;/* filled in by destination */
+	unsigned short	acknowledge_count;	/* filled in by destination */
 	/* 16 bits */
-	unsigned int reserved1:32;	/* not usable */
+	unsigned int	reserved1:32;		/* not usable */
 };
 
 
@@ -176,93 +230,96 @@ struct bau_msg_payload {
  * see table 4.2.3.0.1 in broacast_assist spec.
  */
 struct bau_msg_header {
-	unsigned int dest_subnodeid:6;	/* must be 0x10, for the LB */
+	unsigned int	dest_subnodeid:6;	/* must be 0x10, for the LB */
 	/* bits 5:0 */
-	unsigned int base_dest_nasid:15; /* nasid of the */
-	/* bits 20:6 */			  /* first bit in uvhub map */
-	unsigned int command:8;	/* message type */
+	unsigned int	base_dest_nasid:15;	/* nasid of the first bit */
+	/* bits 20:6 */				/* in uvhub map */
+	unsigned int	command:8;		/* message type */
 	/* bits 28:21 */
-				/* 0x38: SN3net EndPoint Message */
-	unsigned int rsvd_1:3;	/* must be zero */
+	/* 0x38: SN3net EndPoint Message */
+	unsigned int	rsvd_1:3;		/* must be zero */
 	/* bits 31:29 */
-				/* int will align on 32 bits */
-	unsigned int rsvd_2:9;	/* must be zero */
+	/* int will align on 32 bits */
+	unsigned int	rsvd_2:9;		/* must be zero */
 	/* bits 40:32 */
-				/* Suppl_A is 56-41 */
-	unsigned int sequence:16;/* message sequence number */
-	/* bits 56:41 */	/* becomes bytes 16-17 of msg */
-				/* Address field (96:57) is never used as an
-				   address (these are address bits 42:3) */
-
-	unsigned int rsvd_3:1;	/* must be zero */
+	/* Suppl_A is 56-41 */
+	unsigned int	sequence:16;		/* message sequence number */
+	/* bits 56:41 */			/* becomes bytes 16-17 of msg */
+						/* Address field (96:57) is
+						   never used as an address
+						   (these are address bits
+						   42:3) */
+
+	unsigned int	rsvd_3:1;		/* must be zero */
 	/* bit 57 */
-				/* address bits 27:4 are payload */
+	/* address bits 27:4 are payload */
 	/* these next 24  (58-81) bits become bytes 12-14 of msg */
-
 	/* bits 65:58 land in byte 12 */
-	unsigned int replied_to:1;/* sent as 0 by the source to byte 12 */
+	unsigned int	replied_to:1;		/* sent as 0 by the source to
+						   byte 12 */
 	/* bit 58 */
-	unsigned int msg_type:3; /* software type of the message*/
+	unsigned int	msg_type:3;		/* software type of the
+						   message */
 	/* bits 61:59 */
-	unsigned int canceled:1; /* message canceled, resource to be freed*/
+	unsigned int	canceled:1;		/* message canceled, resource
+						   is to be freed*/
 	/* bit 62 */
-	unsigned int payload_1a:1;/* not currently used */
+	unsigned int	payload_1a:1;		/* not currently used */
 	/* bit 63 */
-	unsigned int payload_1b:2;/* not currently used */
+	unsigned int	payload_1b:2;		/* not currently used */
 	/* bits 65:64 */
 
 	/* bits 73:66 land in byte 13 */
-	unsigned int payload_1ca:6;/* not currently used */
+	unsigned int	payload_1ca:6;		/* not currently used */
 	/* bits 71:66 */
-	unsigned int payload_1c:2;/* not currently used */
+	unsigned int	payload_1c:2;		/* not currently used */
 	/* bits 73:72 */
 
 	/* bits 81:74 land in byte 14 */
-	unsigned int payload_1d:6;/* not currently used */
+	unsigned int	payload_1d:6;		/* not currently used */
 	/* bits 79:74 */
-	unsigned int payload_1e:2;/* not currently used */
+	unsigned int	payload_1e:2;		/* not currently used */
 	/* bits 81:80 */
 
-	unsigned int rsvd_4:7;	/* must be zero */
+	unsigned int	rsvd_4:7;		/* must be zero */
 	/* bits 88:82 */
-	unsigned int sw_ack_flag:1;/* software acknowledge flag */
+	unsigned int	swack_flag:1;		/* software acknowledge flag */
 	/* bit 89 */
-				/* INTD trasactions at destination are to
-				   wait for software acknowledge */
-	unsigned int rsvd_5:6;	/* must be zero */
+						/* INTD trasactions at
+						   destination are to wait for
+						   software acknowledge */
+	unsigned int	rsvd_5:6;		/* must be zero */
 	/* bits 95:90 */
-	unsigned int rsvd_6:5;	/* must be zero */
+	unsigned int	rsvd_6:5;		/* must be zero */
 	/* bits 100:96 */
-	unsigned int int_both:1;/* if 1, interrupt both sockets on the uvhub */
+	unsigned int	int_both:1;		/* if 1, interrupt both sockets
+						   on the uvhub */
 	/* bit 101*/
-	unsigned int fairness:3;/* usually zero */
+	unsigned int	fairness:3;		/* usually zero */
 	/* bits 104:102 */
-	unsigned int multilevel:1;	/* multi-level multicast format */
+	unsigned int	multilevel:1;		/* multi-level multicast
+						   format */
 	/* bit 105 */
-				/* 0 for TLB: endpoint multi-unicast messages */
-	unsigned int chaining:1;/* next descriptor is part of this activation*/
+	/* 0 for TLB: endpoint multi-unicast messages */
+	unsigned int	chaining:1;		/* next descriptor is part of
+						   this activation*/
 	/* bit 106 */
-	unsigned int rsvd_7:21;	/* must be zero */
+	unsigned int	rsvd_7:21;		/* must be zero */
 	/* bits 127:107 */
 };
 
-/* see msg_type: */
-#define MSG_NOOP 0
-#define MSG_REGULAR 1
-#define MSG_RETRY 2
-
 /*
  * The activation descriptor:
  * The format of the message to send, plus all accompanying control
  * Should be 64 bytes
  */
 struct bau_desc {
-	struct bau_target_uvhubmask distribution;
+	struct bau_targ_hubmask	distribution;
 	/*
 	 * message template, consisting of header and payload:
 	 */
-	struct bau_msg_header header;
-	struct bau_msg_payload payload;
+	struct bau_msg_header		header;
+	struct bau_msg_payload		payload;
 };
 /*
  *   -payload--    ---------header------
@@ -281,59 +338,51 @@ struct bau_desc {
  * are 32 bytes (2 micropackets) (256 bits) in length, but contain only 17
  * bytes of usable data, including the sw ack vector in byte 15 (bits 127:120)
  * (12 bytes come from bau_msg_payload, 3 from payload_1, 2 from
- *  sw_ack_vector and payload_2)
+ *  swack_vec and payload_2)
  * "Enabling Software Acknowledgment mode (see Section 4.3.3 Software
  *  Acknowledge Processing) also selects 32 byte (17 bytes usable) payload
  *  operation."
  */
-struct bau_payload_queue_entry {
-	unsigned long address;		/* signifies a page or all TLB's
-						of the cpu */
+struct bau_pq_entry {
+	unsigned long	address;	/* signifies a page or all TLB's
+					   of the cpu */
 	/* 64 bits, bytes 0-7 */
-
-	unsigned short sending_cpu;	/* cpu that sent the message */
+	unsigned short	sending_cpu;	/* cpu that sent the message */
 	/* 16 bits, bytes 8-9 */
-
-	unsigned short acknowledge_count; /* filled in by destination */
+	unsigned short	acknowledge_count; /* filled in by destination */
 	/* 16 bits, bytes 10-11 */
-
 	/* these next 3 bytes come from bits 58-81 of the message header */
-	unsigned short replied_to:1;    /* sent as 0 by the source */
-	unsigned short msg_type:3;      /* software message type */
-	unsigned short canceled:1;      /* sent as 0 by the source */
-	unsigned short unused1:3;       /* not currently using */
+	unsigned short	replied_to:1;	/* sent as 0 by the source */
+	unsigned short	msg_type:3;	/* software message type */
+	unsigned short	canceled:1;	/* sent as 0 by the source */
+	unsigned short	unused1:3;	/* not currently using */
 	/* byte 12 */
-
-	unsigned char unused2a;		/* not currently using */
+	unsigned char	unused2a;	/* not currently using */
 	/* byte 13 */
-	unsigned char unused2;		/* not currently using */
+	unsigned char	unused2;	/* not currently using */
 	/* byte 14 */
-
-	unsigned char sw_ack_vector;	/* filled in by the hardware */
+	unsigned char	swack_vec;	/* filled in by the hardware */
 	/* byte 15 (bits 127:120) */
-
-	unsigned short sequence;	/* message sequence number */
+	unsigned short	sequence;	/* message sequence number */
 	/* bytes 16-17 */
-	unsigned char unused4[2];	/* not currently using bytes 18-19 */
+	unsigned char	unused4[2];	/* not currently using bytes 18-19 */
 	/* bytes 18-19 */
-
-	int number_of_cpus;		/* filled in at destination */
+	int		number_of_cpus;	/* filled in at destination */
 	/* 32 bits, bytes 20-23 (aligned) */
-
-	unsigned char unused5[8];       /* not using */
+	unsigned char	unused5[8];	/* not using */
 	/* bytes 24-31 */
 };
 
 struct msg_desc {
-	struct bau_payload_queue_entry *msg;
-	int msg_slot;
-	int sw_ack_slot;
-	struct bau_payload_queue_entry *va_queue_first;
-	struct bau_payload_queue_entry *va_queue_last;
+	struct bau_pq_entry	*msg;
+	int			msg_slot;
+	int			swack_slot;
+	struct bau_pq_entry	*queue_first;
+	struct bau_pq_entry	*queue_last;
 };
 
 struct reset_args {
-	int sender;
+	int			sender;
 };
 
 /*
@@ -341,112 +390,226 @@ struct reset_args {
  */
 struct ptc_stats {
 	/* sender statistics */
-	unsigned long s_giveup; /* number of fall backs to IPI-style flushes */
-	unsigned long s_requestor; /* number of shootdown requests */
-	unsigned long s_stimeout; /* source side timeouts */
-	unsigned long s_dtimeout; /* destination side timeouts */
-	unsigned long s_time; /* time spent in sending side */
-	unsigned long s_retriesok; /* successful retries */
-	unsigned long s_ntargcpu; /* total number of cpu's targeted */
-	unsigned long s_ntargself; /* times the sending cpu was targeted */
-	unsigned long s_ntarglocals; /* targets of cpus on the local blade */
-	unsigned long s_ntargremotes; /* targets of cpus on remote blades */
-	unsigned long s_ntarglocaluvhub; /* targets of the local hub */
-	unsigned long s_ntargremoteuvhub; /* remotes hubs targeted */
-	unsigned long s_ntarguvhub; /* total number of uvhubs targeted */
-	unsigned long s_ntarguvhub16; /* number of times target hubs >= 16*/
-	unsigned long s_ntarguvhub8; /* number of times target hubs >= 8 */
-	unsigned long s_ntarguvhub4; /* number of times target hubs >= 4 */
-	unsigned long s_ntarguvhub2; /* number of times target hubs >= 2 */
-	unsigned long s_ntarguvhub1; /* number of times target hubs == 1 */
-	unsigned long s_resets_plug; /* ipi-style resets from plug state */
-	unsigned long s_resets_timeout; /* ipi-style resets from timeouts */
-	unsigned long s_busy; /* status stayed busy past s/w timer */
-	unsigned long s_throttles; /* waits in throttle */
-	unsigned long s_retry_messages; /* retry broadcasts */
-	unsigned long s_bau_reenabled; /* for bau enable/disable */
-	unsigned long s_bau_disabled; /* for bau enable/disable */
+	unsigned long	s_giveup;		/* number of fall backs to
+						   IPI-style flushes */
+	unsigned long	s_requestor;		/* number of shootdown
+						   requests */
+	unsigned long	s_stimeout;		/* source side timeouts */
+	unsigned long	s_dtimeout;		/* destination side timeouts */
+	unsigned long	s_time;			/* time spent in sending side */
+	unsigned long	s_retriesok;		/* successful retries */
+	unsigned long	s_ntargcpu;		/* total number of cpu's
+						   targeted */
+	unsigned long	s_ntargself;		/* times the sending cpu was
+						   targeted */
+	unsigned long	s_ntarglocals;		/* targets of cpus on the local
+						   blade */
+	unsigned long	s_ntargremotes;		/* targets of cpus on remote
+						   blades */
+	unsigned long	s_ntarglocaluvhub;	/* targets of the local hub */
+	unsigned long	s_ntargremoteuvhub;	/* remotes hubs targeted */
+	unsigned long	s_ntarguvhub;		/* total number of uvhubs
+						   targeted */
+	unsigned long	s_ntarguvhub16;		/* number of times target
+						   hubs >= 16*/
+	unsigned long	s_ntarguvhub8;		/* number of times target
+						   hubs >= 8 */
+	unsigned long	s_ntarguvhub4;		/* number of times target
+						   hubs >= 4 */
+	unsigned long	s_ntarguvhub2;		/* number of times target
+						   hubs >= 2 */
+	unsigned long	s_ntarguvhub1;		/* number of times target
+						   hubs == 1 */
+	unsigned long	s_resets_plug;		/* ipi-style resets from plug
+						   state */
+	unsigned long	s_resets_timeout;	/* ipi-style resets from
+						   timeouts */
+	unsigned long	s_busy;			/* status stayed busy past
+						   s/w timer */
+	unsigned long	s_throttles;		/* waits in throttle */
+	unsigned long	s_retry_messages;	/* retry broadcasts */
+	unsigned long	s_bau_reenabled;	/* for bau enable/disable */
+	unsigned long	s_bau_disabled;		/* for bau enable/disable */
 	/* destination statistics */
-	unsigned long d_alltlb; /* times all tlb's on this cpu were flushed */
-	unsigned long d_onetlb; /* times just one tlb on this cpu was flushed */
-	unsigned long d_multmsg; /* interrupts with multiple messages */
-	unsigned long d_nomsg; /* interrupts with no message */
-	unsigned long d_time; /* time spent on destination side */
-	unsigned long d_requestee; /* number of messages processed */
-	unsigned long d_retries; /* number of retry messages processed */
-	unsigned long d_canceled; /* number of messages canceled by retries */
-	unsigned long d_nocanceled; /* retries that found nothing to cancel */
-	unsigned long d_resets; /* number of ipi-style requests processed */
-	unsigned long d_rcanceled; /* number of messages canceled by resets */
+	unsigned long	d_alltlb;		/* times all tlb's on this
+						   cpu were flushed */
+	unsigned long	d_onetlb;		/* times just one tlb on this
+						   cpu was flushed */
+	unsigned long	d_multmsg;		/* interrupts with multiple
+						   messages */
+	unsigned long	d_nomsg;		/* interrupts with no message */
+	unsigned long	d_time;			/* time spent on destination
+						   side */
+	unsigned long	d_requestee;		/* number of messages
+						   processed */
+	unsigned long	d_retries;		/* number of retry messages
+						   processed */
+	unsigned long	d_canceled;		/* number of messages canceled
+						   by retries */
+	unsigned long	d_nocanceled;		/* retries that found nothing
+						   to cancel */
+	unsigned long	d_resets;		/* number of ipi-style requests
+						   processed */
+	unsigned long	d_rcanceled;		/* number of messages canceled
+						   by resets */
+};
+
+struct tunables {
+	int			*tunp;
+	int			deflt;
 };
 
 struct hub_and_pnode {
-	short uvhub;
-	short pnode;
+	short			uvhub;
+	short			pnode;
 };
+
+struct socket_desc {
+	short			num_cpus;
+	short			cpu_number[MAX_CPUS_PER_SOCKET];
+};
+
+struct uvhub_desc {
+	unsigned short		socket_mask;
+	short			num_cpus;
+	short			uvhub;
+	short			pnode;
+	struct socket_desc	socket[2];
+};
+
 /*
  * one per-cpu; to locate the software tables
  */
 struct bau_control {
-	struct bau_desc *descriptor_base;
-	struct bau_payload_queue_entry *va_queue_first;
-	struct bau_payload_queue_entry *va_queue_last;
-	struct bau_payload_queue_entry *bau_msg_head;
-	struct bau_control *uvhub_master;
-	struct bau_control *socket_master;
-	struct ptc_stats *statp;
-	unsigned long timeout_interval;
-	unsigned long set_bau_on_time;
-	atomic_t active_descriptor_count;
-	int plugged_tries;
-	int timeout_tries;
-	int ipi_attempts;
-	int conseccompletes;
-	int baudisabled;
-	int set_bau_off;
-	short cpu;
-	short osnode;
-	short uvhub_cpu;
-	short uvhub;
-	short cpus_in_socket;
-	short cpus_in_uvhub;
-	short partition_base_pnode;
-	unsigned short message_number;
-	unsigned short uvhub_quiesce;
-	short socket_acknowledge_count[DEST_Q_SIZE];
-	cycles_t send_message;
-	spinlock_t uvhub_lock;
-	spinlock_t queue_lock;
+	struct bau_desc		*descriptor_base;
+	struct bau_pq_entry	*queue_first;
+	struct bau_pq_entry	*queue_last;
+	struct bau_pq_entry	*bau_msg_head;
+	struct bau_control	*uvhub_master;
+	struct bau_control	*socket_master;
+	struct ptc_stats	*statp;
+	unsigned long		timeo