10 files changed, 108 insertions, 66 deletions

diff --git a/arch/mips/mm/Makefile b/arch/mips/mm/Makefile
index fd6203f14f1f..90ceb963aaf1 100644
--- a/arch/mips/mm/Makefile
+++ b/arch/mips/mm/Makefile
@@ -11,27 +11,12 @@ obj-$(CONFIG_64BIT)		+= pgtable-64.o
 obj-$(CONFIG_HIGHMEM)		+= highmem.o
 obj-$(CONFIG_HUGETLB_PAGE)	+= hugetlbpage.o
 
-obj-$(CONFIG_CPU_LOONGSON2)	+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_MIPS32)	+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_MIPS64)	+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_NEVADA)	+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_R10000)	+= c-r4k.o cex-gen.o tlb-r4k.o
+obj-$(CONFIG_CPU_R4K_CACHE_TLB) += c-r4k.o cex-gen.o tlb-r4k.o
 obj-$(CONFIG_CPU_R3000)		+= c-r3k.o tlb-r3k.o
-obj-$(CONFIG_CPU_R4300)		+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_R4X00)		+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_R5000)		+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_R5432)		+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_R5500)		+= c-r4k.o cex-gen.o tlb-r4k.o
 obj-$(CONFIG_CPU_R8000)		+= c-r4k.o cex-gen.o tlb-r8k.o
-obj-$(CONFIG_CPU_RM7000)	+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_RM9000)	+= c-r4k.o cex-gen.o tlb-r4k.o
 obj-$(CONFIG_CPU_SB1)		+= c-r4k.o cerr-sb1.o cex-sb1.o tlb-r4k.o
 obj-$(CONFIG_CPU_TX39XX)	+= c-tx39.o tlb-r3k.o
-obj-$(CONFIG_CPU_TX49XX)	+= c-r4k.o cex-gen.o tlb-r4k.o
-obj-$(CONFIG_CPU_VR41XX)	+= c-r4k.o cex-gen.o tlb-r4k.o
 obj-$(CONFIG_CPU_CAVIUM_OCTEON)	+= c-octeon.o cex-oct.o tlb-r4k.o
-obj-$(CONFIG_CPU_XLR)		+= c-r4k.o tlb-r4k.o cex-gen.o
-obj-$(CONFIG_CPU_XLP)		+= c-r4k.o tlb-r4k.o cex-gen.o
 
 obj-$(CONFIG_IP22_CPU_SCACHE)	+= sc-ip22.o
 obj-$(CONFIG_R5000_CPU_SCACHE)  += sc-r5k.o
diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c
index f092c265dc63..4c32ede464b5 100644
--- a/arch/mips/mm/c-r4k.c
+++ b/arch/mips/mm/c-r4k.c
@@ -786,6 +786,25 @@ static inline void rm7k_erratum31(void)
 	}
 }
 
+static inline void alias_74k_erratum(struct cpuinfo_mips *c)
+{
+	/*
+	 * Early versions of the 74K do not update the cache tags on a
+	 * vtag miss/ptag hit which can occur in the case of KSEG0/KUSEG
+	 * aliases. In this case it is better to treat the cache as always
+	 * having aliases.
+	 */
+	if ((c->processor_id & 0xff) <= PRID_REV_ENCODE_332(2, 4, 0))
+		c->dcache.flags |= MIPS_CACHE_VTAG;
+	if ((c->processor_id & 0xff) == PRID_REV_ENCODE_332(2, 4, 0))
+		write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND);
+	if (((c->processor_id & 0xff00) == PRID_IMP_1074K) &&
+	    ((c->processor_id & 0xff) <= PRID_REV_ENCODE_332(1, 1, 0))) {
+		c->dcache.flags |= MIPS_CACHE_VTAG;
+		write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND);
+	}
+}
+
 static char *way_string[] __cpuinitdata = { NULL, "direct mapped", "2-way",
 	"3-way", "4-way", "5-way", "6-way", "7-way", "8-way"
 };
@@ -1056,6 +1075,8 @@ static void __cpuinit probe_pcache(void)
 	case CPU_34K:
 	case CPU_74K:
 	case CPU_1004K:
+		if (c->cputype == CPU_74K)
+			alias_74k_erratum(c);
 		if ((read_c0_config7() & (1 << 16))) {
 			/* effectively physically indexed dcache,
 			   thus no virtual aliases. */
diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c
index 829320c7b175..07cec4407b0c 100644
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -142,7 +142,7 @@ EXPORT_SYMBOL(_page_cachable_default);
 
 static inline void setup_protection_map(void)
 {
-	if (kernel_uses_smartmips_rixi) {
+	if (cpu_has_rixi) {
 		protection_map[0]  = __pgprot(_page_cachable_default | _PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
 		protection_map[1]  = __pgprot(_page_cachable_default | _PAGE_PRESENT | _PAGE_NO_EXEC);
 		protection_map[2]  = __pgprot(_page_cachable_default | _PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c
index c14f6dfed995..ddcec1e1a0cd 100644
--- a/arch/mips/mm/fault.c
+++ b/arch/mips/mm/fault.c
@@ -114,7 +114,7 @@ good_area:
 		if (!(vma->vm_flags & VM_WRITE))
 			goto bad_area;
 	} else {
-		if (kernel_uses_smartmips_rixi) {
+		if (cpu_has_rixi) {
 			if (address == regs->cp0_epc && !(vma->vm_flags & VM_EXEC)) {
 #if 0
 				pr_notice("Cpu%d[%s:%d:%0*lx:%ld:%0*lx] XI violation\n",
@@ -171,6 +171,7 @@ good_area:
 		}
 		if (fault & VM_FAULT_RETRY) {
 			flags &= ~FAULT_FLAG_ALLOW_RETRY;
+			flags |= FAULT_FLAG_TRIED;
 
 			/*
 			 * No need to up_read(&mm->mmap_sem) as we would
diff --git a/arch/mips/mm/gup.c b/arch/mips/mm/gup.c
index 33aadbcf170b..dcfd573871c1 100644
--- a/arch/mips/mm/gup.c
+++ b/arch/mips/mm/gup.c
@@ -152,6 +152,8 @@ static int gup_huge_pud(pud_t pud, unsigned long addr, unsigned long end,
 	do {
 		VM_BUG_ON(compound_head(page) != head);
 		pages[*nr] = page;
+		if (PageTail(page))
+			get_huge_page_tail(page);
 		(*nr)++;
 		page++;
 		refs++;
diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c
index 1a85ba92eb5c..be9acb2b959d 100644
--- a/arch/mips/mm/init.c
+++ b/arch/mips/mm/init.c
@@ -469,19 +469,20 @@ void __init_refok free_initmem(void)
 #ifndef CONFIG_MIPS_PGD_C0_CONTEXT
 unsigned long pgd_current[NR_CPUS];
 #endif
-/*
- * On 64-bit we've got three-level pagetables with a slightly
- * different layout ...
- */
-#define __page_aligned(order) __attribute__((__aligned__(PAGE_SIZE<<order)))
 
 /*
  * gcc 3.3 and older have trouble determining that PTRS_PER_PGD and PGD_ORDER
  * are constants.  So we use the variants from asm-offset.h until that gcc
  * will officially be retired.
+ *
+ * Align swapper_pg_dir in to 64K, allows its address to be loaded
+ * with a single LUI instruction in the TLB handlers.  If we used
+ * __aligned(64K), its size would get rounded up to the alignment
+ * size, and waste space.  So we place it in its own section and align
+ * it in the linker script.
  */
-pgd_t swapper_pg_dir[_PTRS_PER_PGD] __page_aligned(_PGD_ORDER);
+pgd_t swapper_pg_dir[_PTRS_PER_PGD] __section(.bss..swapper_pg_dir);
 #ifndef __PAGETABLE_PMD_FOLDED
-pmd_t invalid_pmd_table[PTRS_PER_PMD] __page_aligned(PMD_ORDER);
+pmd_t invalid_pmd_table[PTRS_PER_PMD] __page_aligned_bss;
 #endif
-pte_t invalid_pte_table[PTRS_PER_PTE] __page_aligned(PTE_ORDER);
+pte_t invalid_pte_table[PTRS_PER_PTE] __page_aligned_bss;
diff --git a/arch/mips/mm/pgtable-64.c b/arch/mips/mm/pgtable-64.c
index cda4e300eb0a..25407794edb4 100644
--- a/arch/mips/mm/pgtable-64.c
+++ b/arch/mips/mm/pgtable-64.c
@@ -26,17 +26,17 @@ void pgd_init(unsigned long page)
  	p = (unsigned long *) page;
 	end = p + PTRS_PER_PGD;
 
-	while (p < end) {
+	do {
 		p[0] = entry;
 		p[1] = entry;
 		p[2] = entry;
 		p[3] = entry;
 		p[4] = entry;
-		p[5] = entry;
-		p[6] = entry;
-		p[7] = entry;
 		p += 8;
-	}
+		p[-3] = entry;
+		p[-2] = entry;
+		p[-1] = entry;
+	} while (p != end);
 }
 
 #ifndef __PAGETABLE_PMD_FOLDED
@@ -47,17 +47,17 @@ void pmd_init(unsigned long addr, unsigned long pagetable)
  	p = (unsigned long *) addr;
 	end = p + PTRS_PER_PMD;
 
-	while (p < end) {
+	do {
 		p[0] = pagetable;
 		p[1] = pagetable;
 		p[2] = pagetable;
 		p[3] = pagetable;
 		p[4] = pagetable;
-		p[5] = pagetable;
-		p[6] = pagetable;
-		p[7] = pagetable;
 		p += 8;
-	}
+		p[-3] = pagetable;
+		p[-2] = pagetable;
+		p[-1] = pagetable;
+	} while (p != end);
 }
 #endif
 
diff --git a/arch/mips/mm/tlb-r4k.c b/arch/mips/mm/tlb-r4k.c
index d2572cb232db..87b9cfcc30ff 100644
--- a/arch/mips/mm/tlb-r4k.c
+++ b/arch/mips/mm/tlb-r4k.c
@@ -401,7 +401,7 @@ void __cpuinit tlb_init(void)
 	    current_cpu_type() == CPU_R14000)
 		write_c0_framemask(0);
 
-	if (kernel_uses_smartmips_rixi) {
+	if (cpu_has_rixi) {
 		/*
 		 * Enable the no read, no exec bits, and enable large virtual
 		 * address.
diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c
index 03eb0ef91580..658a520364ce 100644
--- a/arch/mips/mm/tlbex.c
+++ b/arch/mips/mm/tlbex.c
@@ -449,8 +449,20 @@ static void __cpuinit build_tlb_write_entry(u32 **p, struct uasm_label **l,
 	}
 
 	if (cpu_has_mips_r2) {
-		if (cpu_has_mips_r2_exec_hazard)
+		/*
+		 * The architecture spec says an ehb is required here,
+		 * but a number of cores do not have the hazard and
+		 * using an ehb causes an expensive pipeline stall.
+		 */
+		switch (current_cpu_type()) {
+		case CPU_M14KC:
+		case CPU_74K:
+			break;
+
+		default:
 			uasm_i_ehb(p);
+			break;
+		}
 		tlbw(p);
 		return;
 	}
@@ -586,9 +598,8 @@ static void __cpuinit build_tlb_write_entry(u32 **p, struct uasm_label **l,
 static __cpuinit __maybe_unused void build_convert_pte_to_entrylo(u32 **p,
 								  unsigned int reg)
 {
-	if (kernel_uses_smartmips_rixi) {
-		UASM_i_SRL(p, reg, reg, ilog2(_PAGE_NO_EXEC));
-		UASM_i_ROTR(p, reg, reg, ilog2(_PAGE_GLOBAL) - ilog2(_PAGE_NO_EXEC));
+	if (cpu_has_rixi) {
+		UASM_i_ROTR(p, reg, reg, ilog2(_PAGE_GLOBAL));
 	} else {
 #ifdef CONFIG_64BIT_PHYS_ADDR
 		uasm_i_dsrl_safe(p, reg, reg, ilog2(_PAGE_GLOBAL));
@@ -921,6 +932,13 @@ build_get_pgde32(u32 **p, unsigned int tmp, unsigned int ptr)
 #endif
 	uasm_i_mfc0(p, tmp, C0_BADVADDR); /* get faulting address */
 	uasm_i_lw(p, ptr, uasm_rel_lo(pgdc), ptr);
+
+	if (cpu_has_mips_r2) {
+		uasm_i_ext(p, tmp, tmp, PGDIR_SHIFT, (32 - PGDIR_SHIFT));
+		uasm_i_ins(p, ptr, tmp, PGD_T_LOG2, (32 - PGDIR_SHIFT));
+		return;
+	}
+
 	uasm_i_srl(p, tmp, tmp, PGDIR_SHIFT); /* get pgd only bits */
 	uasm_i_sll(p, tmp, tmp, PGD_T_LOG2);
 	uasm_i_addu(p, ptr, ptr, tmp); /* add in pgd offset */
@@ -956,6 +974,15 @@ static void __cpuinit build_adjust_context(u32 **p, unsigned int ctx)
 
 static void __cpuinit build_get_ptep(u32 **p, unsigned int tmp, unsigned int ptr)
 {
+	if (cpu_has_mips_r2) {
+		/* PTE ptr offset is obtained from BadVAddr */
+		UASM_i_MFC0(p, tmp, C0_BADVADDR);
+		UASM_i_LW(p, ptr, 0, ptr);
+		uasm_i_ext(p, tmp, tmp, PAGE_SHIFT+1, PGDIR_SHIFT-PAGE_SHIFT-1);
+		uasm_i_ins(p, ptr, tmp, PTE_T_LOG2+1, PGDIR_SHIFT-PAGE_SHIFT-1);
+		return;
+	}
+
 	/*
 	 * Bug workaround for the Nevada. It seems as if under certain
 	 * circumstances the move from cp0_context might produce a
@@ -990,12 +1017,10 @@ static void __cpuinit build_update_entries(u32 **p, unsigned int tmp,
 	if (cpu_has_64bits) {
 		uasm_i_ld(p, tmp, 0, ptep); /* get even pte */
 		uasm_i_ld(p, ptep, sizeof(pte_t), ptep); /* get odd pte */
-		if (kernel_uses_smartmips_rixi) {
-			UASM_i_SRL(p, tmp, tmp, ilog2(_PAGE_NO_EXEC));
-			UASM_i_SRL(p, ptep, ptep, ilog2(_PAGE_NO_EXEC));
-			UASM_i_ROTR(p, tmp, tmp, ilog2(_PAGE_GLOBAL) - ilog2(_PAGE_NO_EXEC));
+		if (cpu_has_rixi) {
+			UASM_i_ROTR(p, tmp, tmp, ilog2(_PAGE_GLOBAL));
 			UASM_i_MTC0(p, tmp, C0_ENTRYLO0); /* load it */
-			UASM_i_ROTR(p, ptep, ptep, ilog2(_PAGE_GLOBAL) - ilog2(_PAGE_NO_EXEC));
+			UASM_i_ROTR(p, ptep, ptep, ilog2(_PAGE_GLOBAL));
 		} else {
 			uasm_i_dsrl_safe(p, tmp, tmp, ilog2(_PAGE_GLOBAL)); /* convert to entrylo0 */
 			UASM_i_MTC0(p, tmp, C0_ENTRYLO0); /* load it */
@@ -1017,14 +1042,12 @@ static void __cpuinit build_update_entries(u32 **p, unsigned int tmp,
 	UASM_i_LW(p, ptep, sizeof(pte_t), ptep); /* get odd pte */
 	if (r45k_bvahwbug())
 		build_tlb_probe_entry(p);
-	if (kernel_uses_smartmips_rixi) {
-		UASM_i_SRL(p, tmp, tmp, ilog2(_PAGE_NO_EXEC));
-		UASM_i_SRL(p, ptep, ptep, ilog2(_PAGE_NO_EXEC));
-		UASM_i_ROTR(p, tmp, tmp, ilog2(_PAGE_GLOBAL) - ilog2(_PAGE_NO_EXEC));
+	if (cpu_has_rixi) {
+		UASM_i_ROTR(p, tmp, tmp, ilog2(_PAGE_GLOBAL));
 		if (r4k_250MHZhwbug())
 			UASM_i_MTC0(p, 0, C0_ENTRYLO0);
 		UASM_i_MTC0(p, tmp, C0_ENTRYLO0); /* load it */
-		UASM_i_ROTR(p, ptep, ptep, ilog2(_PAGE_GLOBAL) - ilog2(_PAGE_NO_EXEC));
+		UASM_i_ROTR(p, ptep, ptep, ilog2(_PAGE_GLOBAL));
 	} else {
 		UASM_i_SRL(p, tmp, tmp, ilog2(_PAGE_GLOBAL)); /* convert to entrylo0 */
 		if (r4k_250MHZhwbug())
@@ -1183,14 +1206,10 @@ build_fast_tlb_refill_handler (u32 **p, struct uasm_label **l,
 		UASM_i_LW(p, even, 0, ptr); /* get even pte */
 		UASM_i_LW(p, odd, sizeof(pte_t), ptr); /* get odd pte */
 	}
-	if (kernel_uses_smartmips_rixi) {
-		uasm_i_dsrl_safe(p, even, even, ilog2(_PAGE_NO_EXEC));
-		uasm_i_dsrl_safe(p, odd, odd, ilog2(_PAGE_NO_EXEC));
-		uasm_i_drotr(p, even, even,
-			     ilog2(_PAGE_GLOBAL) - ilog2(_PAGE_NO_EXEC));
+	if (cpu_has_rixi) {
+		uasm_i_drotr(p, even, even, ilog2(_PAGE_GLOBAL));
 		UASM_i_MTC0(p, even, C0_ENTRYLO0); /* load it */
-		uasm_i_drotr(p, odd, odd,
-			     ilog2(_PAGE_GLOBAL) - ilog2(_PAGE_NO_EXEC));
+		uasm_i_drotr(p, odd, odd, ilog2(_PAGE_GLOBAL));
 	} else {
 		uasm_i_dsrl_safe(p, even, even, ilog2(_PAGE_GLOBAL));
 		UASM_i_MTC0(p, even, C0_ENTRYLO0); /* load it */
@@ -1545,7 +1564,7 @@ build_pte_present(u32 **p, struct uasm_reloc **r,
 {
 	int t = scratch >= 0 ? scratch : pte;
 
-	if (kernel_uses_smartmips_rixi) {
+	if (cpu_has_rixi) {
 		if (use_bbit_insns()) {
 			uasm_il_bbit0(p, r, pte, ilog2(_PAGE_PRESENT), lid);
 			uasm_i_nop(p);
@@ -1875,7 +1894,7 @@ static void __cpuinit build_r4000_tlb_load_handler(void)
 	if (m4kc_tlbp_war())
 		build_tlb_probe_entry(&p);
 
-	if (kernel_uses_smartmips_rixi) {
+	if (cpu_has_rixi) {
 		/*
 		 * If the page is not _PAGE_VALID, RI or XI could not
 		 * have triggered it.  Skip the expensive test..
@@ -1929,7 +1948,7 @@ static void __cpuinit build_r4000_tlb_load_handler(void)
 	build_pte_present(&p, &r, wr.r1, wr.r2, wr.r3, label_nopage_tlbl);
 	build_tlb_probe_entry(&p);
 
-	if (kernel_uses_smartmips_rixi) {
+	if (cpu_has_rixi) {
 		/*
 		 * If the page is not _PAGE_VALID, RI or XI could not
 		 * have triggered it.  Skip the expensive test..
diff --git a/arch/mips/mm/uasm.c b/arch/mips/mm/uasm.c
index 64a28e819064..39b891056227 100644
--- a/arch/mips/mm/uasm.c
+++ b/arch/mips/mm/uasm.c
@@ -63,11 +63,12 @@ enum opcode {
 	insn_bne, insn_cache, insn_daddiu, insn_daddu, insn_dins, insn_dinsm,
 	insn_dmfc0, insn_dmtc0, insn_drotr, insn_drotr32, insn_dsll,
 	insn_dsll32, insn_dsra, insn_dsrl, insn_dsrl32, insn_dsubu, insn_eret,
-	insn_j, insn_jal, insn_jr, insn_ld, insn_ldx, insn_ll, insn_lld,
-	insn_lui, insn_lw, insn_lwx, insn_mfc0, insn_mtc0, insn_or, insn_ori,
-	insn_pref, insn_rfe, insn_rotr, insn_sc, insn_scd, insn_sd, insn_sll,
-	insn_sra, insn_srl, insn_subu, insn_sw, insn_syscall, insn_tlbp,
-	insn_tlbr, insn_tlbwi, insn_tlbwr, insn_xor, insn_xori,
+	insn_ext, insn_ins, insn_j, insn_jal, insn_jr, insn_ld, insn_ldx,
+	insn_ll, insn_lld, insn_lui, insn_lw, insn_lwx, insn_mfc0, insn_mtc0,
+	insn_or, insn_ori, insn_pref, insn_rfe, insn_rotr, insn_sc, insn_scd,
+	insn_sd, insn_sll, insn_sra, insn_srl, insn_subu, insn_sw,
+	insn_syscall, insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_xor,
+	insn_xori,
 };
 
 struct insn {
@@ -115,6 +116,9 @@ static struct insn insn_table[] __uasminitdata = {
 	{ insn_dsrl, M(spec_op, 0, 0, 0, 0, dsrl_op), RT | RD | RE },
 	{ insn_dsubu, M(spec_op, 0, 0, 0, 0, dsubu_op), RS | RT | RD },
 	{ insn_eret,  M(cop0_op, cop_op, 0, 0, 0, eret_op),  0 },
+	{ insn_ext, M(spec3_op, 0, 0, 0, 0, ext_op), RS | RT | RD | RE },
+	{ insn_ins, M(spec3_op, 0, 0, 0, 0, ins_op), RS | RT | RD | RE },
+	{ insn_j,  M(j_op, 0, 0, 0, 0, 0),  JIMM },
 	{ insn_jal,  M(jal_op, 0, 0, 0, 0, 0),  JIMM },
 	{ insn_j,  M(j_op, 0, 0, 0, 0, 0),  JIMM },
 	{ insn_jr,  M(spec_op, 0, 0, 0, 0, jr_op),  RS },
@@ -341,6 +345,13 @@ Ip_u2u1msbu3(op)					\
 }							\
 UASM_EXPORT_SYMBOL(uasm_i##op);
 
+#define I_u2u1msbdu3(op) 				\
+Ip_u2u1msbu3(op)					\
+{							\
+	build_insn(buf, insn##op, b, a, d-1, c);	\
+}							\
+UASM_EXPORT_SYMBOL(uasm_i##op);
+
 #define I_u1u2(op)					\
 Ip_u1u2(op)						\
 {							\
@@ -394,6 +405,8 @@ I_u2u1u3(_drotr)
 I_u2u1u3(_drotr32)
 I_u3u1u2(_dsubu)
 I_0(_eret)
+I_u2u1msbdu3(_ext)
+I_u2u1msbu3(_ins)
 I_u1(_j)
 I_u1(_jal)
 I_u1(_jr)