From be0c37c985eddc46d0d67543898c086f60460e2e Mon Sep 17 00:00:00 2001
From: "Steven J. Hill" <Steven.Hill@imgtec.com>
Date: Thu, 26 Feb 2015 18:16:37 -0600
Subject: MIPS: Rearrange PTE bits into fixed positions.

This patch rearranges the PTE bits into fixed positions for R2
and later cores. In the past, the TLB handling code did runtime
checking of RI/XI and adjusted the shifts and rotates in order
to fit the largest PFN value into the PTE. The checking now
occurs when building the TLB handler, thus eliminating those
checks. These new arrangements also define the largest possible
PFN value that can fit in the PTE. HUGE page support is only
available for 64-bit cores. Layouts of the PTE bits are now:

   64-bit, R1 or earlier:     CCC D V G [S H] M A W R P
   32-bit, R1 or earler:      CCC D V G M A W R P
   64-bit, R2 or later:       CCC D V G RI/R XI [S H] M A W P
   32-bit, R2 or later:       CCC D V G RI/R XI M A W P

[ralf@linux-mips.org: Fix another build error *rant* *rant*]

Signed-off-by: Steven J. Hill <Steven.Hill@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/9353/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
 arch/mips/mm/tlbex.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c
index d75ff73a2012..20d985901e44 100644
--- a/arch/mips/mm/tlbex.c
+++ b/arch/mips/mm/tlbex.c
@@ -231,14 +231,14 @@ static void output_pgtable_bits_defines(void)
 	pr_define("_PAGE_HUGE_SHIFT %d\n", _PAGE_HUGE_SHIFT);
 	pr_define("_PAGE_SPLITTING_SHIFT %d\n", _PAGE_SPLITTING_SHIFT);
 #endif
+#ifdef CONFIG_CPU_MIPSR2
 	if (cpu_has_rixi) {
 #ifdef _PAGE_NO_EXEC_SHIFT
 		pr_define("_PAGE_NO_EXEC_SHIFT %d\n", _PAGE_NO_EXEC_SHIFT);
-#endif
-#ifdef _PAGE_NO_READ_SHIFT
 		pr_define("_PAGE_NO_READ_SHIFT %d\n", _PAGE_NO_READ_SHIFT);
 #endif
 	}
+#endif
 	pr_define("_PAGE_GLOBAL_SHIFT %d\n", _PAGE_GLOBAL_SHIFT);
 	pr_define("_PAGE_VALID_SHIFT %d\n", _PAGE_VALID_SHIFT);
 	pr_define("_PAGE_DIRTY_SHIFT %d\n", _PAGE_DIRTY_SHIFT);
-- 
cgit v1.2.3-59-g8ed1b


From c5b367835cfc7a8ef53b9670a409ffcc95194344 Mon Sep 17 00:00:00 2001
From: "Steven J. Hill" <Steven.Hill@imgtec.com>
Date: Thu, 26 Feb 2015 18:16:38 -0600
Subject: MIPS: Add support for XPA.

Add support for extended physical addressing (XPA) so that
32-bit platforms can access equal to or greater than 40 bits
of physical addresses.

NOTE:
      1) XPA and EVA are not the same and cannot be used
         simultaneously.
      2) If you configure your kernel for XPA, the PTEs
         and all address sizes become 64-bit.
      3) Your platform MUST have working HIGHMEM support.

Signed-off-by: Steven J. Hill <Steven.Hill@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/9355/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
 arch/mips/Kconfig                    | 35 ++++++++++++++
 arch/mips/include/asm/cpu-features.h |  3 ++
 arch/mips/include/asm/cpu.h          |  1 +
 arch/mips/include/asm/pgtable-32.h   | 15 +++---
 arch/mips/include/asm/pgtable-bits.h | 13 +++++-
 arch/mips/include/asm/pgtable.h      | 36 ++++++---------
 arch/mips/kernel/cpu-probe.c         |  4 ++
 arch/mips/kernel/proc.c              |  1 +
 arch/mips/mm/init.c                  |  7 ++-
 arch/mips/mm/tlb-r4k.c               | 12 +++++
 arch/mips/mm/tlbex.c                 | 90 ++++++++++++++++++++++++++++++------
 11 files changed, 173 insertions(+), 44 deletions(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index c7a16904cd03..69a3b0fab926 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -377,6 +377,7 @@ config MIPS_MALTA
 	select SYS_HAS_CPU_MIPS32_R1
 	select SYS_HAS_CPU_MIPS32_R2
 	select SYS_HAS_CPU_MIPS32_R3_5
+	select SYS_HAS_CPU_MIPS32_R5
 	select SYS_HAS_CPU_MIPS32_R6
 	select SYS_HAS_CPU_MIPS64_R1
 	select SYS_HAS_CPU_MIPS64_R2
@@ -386,6 +387,7 @@ config MIPS_MALTA
 	select SYS_SUPPORTS_32BIT_KERNEL
 	select SYS_SUPPORTS_64BIT_KERNEL
 	select SYS_SUPPORTS_BIG_ENDIAN
+	select SYS_SUPPORTS_HIGHMEM
 	select SYS_SUPPORTS_LITTLE_ENDIAN
 	select SYS_SUPPORTS_MICROMIPS
 	select SYS_SUPPORTS_MIPS_CMP
@@ -1596,6 +1598,33 @@ config CPU_MIPS32_3_5_EVA
 	  One of its primary benefits is an increase in the maximum size
 	  of lowmem (up to 3GB). If unsure, say 'N' here.
 
+config CPU_MIPS32_R5_FEATURES
+	bool "MIPS32 Release 5 Features"
+	depends on SYS_HAS_CPU_MIPS32_R5
+	depends on CPU_MIPS32_R2
+	help
+	  Choose this option to build a kernel for release 2 or later of the
+	  MIPS32 architecture including features from release 5 such as
+	  support for Extended Physical Addressing (XPA).
+
+config CPU_MIPS32_R5_XPA
+	bool "Extended Physical Addressing (XPA)"
+	depends on CPU_MIPS32_R5_FEATURES
+	depends on !EVA
+	depends on !PAGE_SIZE_4KB
+	depends on SYS_SUPPORTS_HIGHMEM
+	select XPA
+	select HIGHMEM
+	select ARCH_PHYS_ADDR_T_64BIT
+	default n
+	help
+	  Choose this option if you want to enable the Extended Physical
+	  Addressing (XPA) on your MIPS32 core (such as P5600 series). The
+	  benefit is to increase physical addressing equal to or greater
+	  than 40 bits. Note that this has the side effect of turning on
+	  64-bit addressing which in turn makes the PTEs 64-bit in size.
+	  If unsure, say 'N' here.
+
 if CPU_LOONGSON2F
 config CPU_NOP_WORKAROUNDS
 	bool
@@ -1699,6 +1728,9 @@ config SYS_HAS_CPU_MIPS32_R2
 config SYS_HAS_CPU_MIPS32_R3_5
 	bool
 
+config SYS_HAS_CPU_MIPS32_R5
+	bool
+
 config SYS_HAS_CPU_MIPS32_R6
 	bool
 
@@ -1836,6 +1868,9 @@ config CPU_MIPSR6
 config EVA
 	bool
 
+config XPA
+	bool
+
 config SYS_SUPPORTS_32BIT_KERNEL
 	bool
 config SYS_SUPPORTS_64BIT_KERNEL
diff --git a/arch/mips/include/asm/cpu-features.h b/arch/mips/include/asm/cpu-features.h
index 0d8208de9a3f..a324751b02ff 100644
--- a/arch/mips/include/asm/cpu-features.h
+++ b/arch/mips/include/asm/cpu-features.h
@@ -139,6 +139,9 @@
 # endif
 #endif
 
+#ifndef cpu_has_xpa
+#define cpu_has_xpa		(cpu_data[0].options & MIPS_CPU_XPA)
+#endif
 #ifndef cpu_has_vtag_icache
 #define cpu_has_vtag_icache	(cpu_data[0].icache.flags & MIPS_CACHE_VTAG)
 #endif
diff --git a/arch/mips/include/asm/cpu.h b/arch/mips/include/asm/cpu.h
index 15687234d70a..e492c740bb94 100644
--- a/arch/mips/include/asm/cpu.h
+++ b/arch/mips/include/asm/cpu.h
@@ -377,6 +377,7 @@ enum cpu_type_enum {
 #define MIPS_CPU_MAAR		0x400000000ull /* MAAR(I) registers are present */
 #define MIPS_CPU_FRE		0x800000000ull /* FRE & UFE bits implemented */
 #define MIPS_CPU_RW_LLB		0x1000000000ull /* LLADDR/LLB writes are allowed */
+#define MIPS_CPU_XPA		0x2000000000ull /* CPU supports Extended Physical Addressing */
 
 /*
  * CPU ASE encodings
diff --git a/arch/mips/include/asm/pgtable-32.h b/arch/mips/include/asm/pgtable-32.h
index a6be006b6f75..7d56686c0e62 100644
--- a/arch/mips/include/asm/pgtable-32.h
+++ b/arch/mips/include/asm/pgtable-32.h
@@ -105,13 +105,16 @@ static inline void pmd_clear(pmd_t *pmdp)
 
 #if defined(CONFIG_PHYS_ADDR_T_64BIT) && defined(CONFIG_CPU_MIPS32)
 #define pte_page(x)		pfn_to_page(pte_pfn(x))
-#define pte_pfn(x)		((unsigned long)((x).pte_high >> 6))
+#define pte_pfn(x)		(((unsigned long)((x).pte_high >> _PFN_SHIFT)) | (unsigned long)((x).pte_low << _PAGE_PRESENT_SHIFT))
 static inline pte_t
 pfn_pte(unsigned long pfn, pgprot_t prot)
 {
 	pte_t pte;
-	pte.pte_high = (pfn << 6) | (pgprot_val(prot) & 0x3f);
-	pte.pte_low = pgprot_val(prot);
+
+	pte.pte_low = (pfn >> _PAGE_PRESENT_SHIFT) |
+				(pgprot_val(prot) & ~_PFNX_MASK);
+	pte.pte_high = (pfn << _PFN_SHIFT) |
+				(pgprot_val(prot) & ~_PFN_MASK);
 	return pte;
 }
 
@@ -166,9 +169,9 @@ pfn_pte(unsigned long pfn, pgprot_t prot)
 #if defined(CONFIG_PHYS_ADDR_T_64BIT) && defined(CONFIG_CPU_MIPS32)
 
 /* Swap entries must have VALID and GLOBAL bits cleared. */
-#define __swp_type(x)			(((x).val >> 2) & 0x1f)
-#define __swp_offset(x)			 ((x).val >> 7)
-#define __swp_entry(type,offset)	((swp_entry_t)	{ ((type) << 2) | ((offset) << 7) })
+#define __swp_type(x)			(((x).val >> 4) & 0x1f)
+#define __swp_offset(x)			 ((x).val >> 9)
+#define __swp_entry(type,offset)	((swp_entry_t)  { ((type) << 4) | ((offset) << 9) })
 #define __pte_to_swp_entry(pte)		((swp_entry_t) { (pte).pte_high })
 #define __swp_entry_to_pte(x)		((pte_t) { 0, (x).val })
 
diff --git a/arch/mips/include/asm/pgtable-bits.h b/arch/mips/include/asm/pgtable-bits.h
index 8e432a8ec4fe..18ae5ddef118 100644
--- a/arch/mips/include/asm/pgtable-bits.h
+++ b/arch/mips/include/asm/pgtable-bits.h
@@ -37,7 +37,11 @@
 /*
  * The following bits are implemented by the TLB hardware
  */
-#define _PAGE_GLOBAL_SHIFT	0
+#define _PAGE_NO_EXEC_SHIFT	0
+#define _PAGE_NO_EXEC		(1 << _PAGE_NO_EXEC_SHIFT)
+#define _PAGE_NO_READ_SHIFT	(_PAGE_NO_EXEC_SHIFT + 1)
+#define _PAGE_NO_READ		(1 << _PAGE_NO_READ_SHIFT)
+#define _PAGE_GLOBAL_SHIFT	(_PAGE_NO_READ_SHIFT + 1)
 #define _PAGE_GLOBAL		(1 << _PAGE_GLOBAL_SHIFT)
 #define _PAGE_VALID_SHIFT	(_PAGE_GLOBAL_SHIFT + 1)
 #define _PAGE_VALID		(1 << _PAGE_VALID_SHIFT)
@@ -49,7 +53,7 @@
 /*
  * The following bits are implemented in software
  */
-#define _PAGE_PRESENT_SHIFT	(_CACHE_SHIFT + 3)
+#define _PAGE_PRESENT_SHIFT	(24)
 #define _PAGE_PRESENT		(1 << _PAGE_PRESENT_SHIFT)
 #define _PAGE_READ_SHIFT	(_PAGE_PRESENT_SHIFT + 1)
 #define _PAGE_READ		(1 << _PAGE_READ_SHIFT)
@@ -62,6 +66,11 @@
 
 #define _PFN_SHIFT		(PAGE_SHIFT - 12 + _CACHE_SHIFT + 3)
 
+/*
+ * Bits for extended EntryLo0/EntryLo1 registers
+ */
+#define _PFNX_MASK		0xffffff
+
 #elif defined(CONFIG_CPU_R3000) || defined(CONFIG_CPU_TX39XX)
 
 /*
diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h
index e1fec0237ce2..bffd46ca3694 100644
--- a/arch/mips/include/asm/pgtable.h
+++ b/arch/mips/include/asm/pgtable.h
@@ -133,7 +133,7 @@ extern void set_pte_at(struct mm_struct *mm, unsigned long addr, pte_t *ptep,
 
 #if defined(CONFIG_PHYS_ADDR_T_64BIT) && defined(CONFIG_CPU_MIPS32)
 
-#define pte_none(pte)		(!(((pte).pte_low | (pte).pte_high) & ~_PAGE_GLOBAL))
+#define pte_none(pte)		(!(((pte).pte_high) & ~_PAGE_GLOBAL))
 #define pte_present(pte)	((pte).pte_low & _PAGE_PRESENT)
 
 static inline void set_pte(pte_t *ptep, pte_t pte)
@@ -142,16 +142,14 @@ static inline void set_pte(pte_t *ptep, pte_t pte)
 	smp_wmb();
 	ptep->pte_low = pte.pte_low;
 
-	if (pte.pte_low & _PAGE_GLOBAL) {
+	if (pte.pte_high & _PAGE_GLOBAL) {
 		pte_t *buddy = ptep_buddy(ptep);
 		/*
 		 * Make sure the buddy is global too (if it's !none,
 		 * it better already be global)
 		 */
-		if (pte_none(*buddy)) {
-			buddy->pte_low	|= _PAGE_GLOBAL;
+		if (pte_none(*buddy))
 			buddy->pte_high |= _PAGE_GLOBAL;
-		}
 	}
 }
 
@@ -161,8 +159,8 @@ static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *pt
 
 	htw_stop();
 	/* Preserve global status for the pair */
-	if (ptep_buddy(ptep)->pte_low & _PAGE_GLOBAL)
-		null.pte_low = null.pte_high = _PAGE_GLOBAL;
+	if (ptep_buddy(ptep)->pte_high & _PAGE_GLOBAL)
+		null.pte_high = _PAGE_GLOBAL;
 
 	set_pte_at(mm, addr, ptep, null);
 	htw_start();
@@ -242,21 +240,21 @@ static inline int pte_young(pte_t pte)	{ return pte.pte_low & _PAGE_ACCESSED; }
 
 static inline pte_t pte_wrprotect(pte_t pte)
 {
-	pte.pte_low  &= ~(_PAGE_WRITE | _PAGE_SILENT_WRITE);
+	pte.pte_low  &= ~_PAGE_WRITE;
 	pte.pte_high &= ~_PAGE_SILENT_WRITE;
 	return pte;
 }
 
 static inline pte_t pte_mkclean(pte_t pte)
 {
-	pte.pte_low  &= ~(_PAGE_MODIFIED | _PAGE_SILENT_WRITE);
+	pte.pte_low  &= ~_PAGE_MODIFIED;
 	pte.pte_high &= ~_PAGE_SILENT_WRITE;
 	return pte;
 }
 
 static inline pte_t pte_mkold(pte_t pte)
 {
-	pte.pte_low  &= ~(_PAGE_ACCESSED | _PAGE_SILENT_READ);
+	pte.pte_low  &= ~_PAGE_ACCESSED;
 	pte.pte_high &= ~_PAGE_SILENT_READ;
 	return pte;
 }
@@ -264,30 +262,24 @@ static inline pte_t pte_mkold(pte_t pte)
 static inline pte_t pte_mkwrite(pte_t pte)
 {
 	pte.pte_low |= _PAGE_WRITE;
-	if (pte.pte_low & _PAGE_MODIFIED) {
-		pte.pte_low  |= _PAGE_SILENT_WRITE;
+	if (pte.pte_low & _PAGE_MODIFIED)
 		pte.pte_high |= _PAGE_SILENT_WRITE;
-	}
 	return pte;
 }
 
 static inline pte_t pte_mkdirty(pte_t pte)
 {
 	pte.pte_low |= _PAGE_MODIFIED;
-	if (pte.pte_low & _PAGE_WRITE) {
-		pte.pte_low  |= _PAGE_SILENT_WRITE;
+	if (pte.pte_low & _PAGE_WRITE)
 		pte.pte_high |= _PAGE_SILENT_WRITE;
-	}
 	return pte;
 }
 
 static inline pte_t pte_mkyoung(pte_t pte)
 {
 	pte.pte_low |= _PAGE_ACCESSED;
-	if (pte.pte_low & _PAGE_READ) {
-		pte.pte_low  |= _PAGE_SILENT_READ;
+	if (pte.pte_low & _PAGE_READ)
 		pte.pte_high |= _PAGE_SILENT_READ;
-	}
 	return pte;
 }
 #else
@@ -391,10 +383,10 @@ static inline pgprot_t pgprot_writecombine(pgprot_t _prot)
 #if defined(CONFIG_PHYS_ADDR_T_64BIT) && defined(CONFIG_CPU_MIPS32)
 static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
 {
-	pte.pte_low  &= _PAGE_CHG_MASK;
+	pte.pte_low  &= (_PAGE_MODIFIED | _PAGE_ACCESSED | _PFNX_MASK);
 	pte.pte_high &= (_PFN_MASK | _CACHE_MASK);
-	pte.pte_low  |= pgprot_val(newprot);
-	pte.pte_high |= pgprot_val(newprot) & ~(_PFN_MASK | _CACHE_MASK);
+	pte.pte_low  |= pgprot_val(newprot) & ~_PFNX_MASK;
+	pte.pte_high |= pgprot_val(newprot) & ~_PFN_MASK;
 	return pte;
 }
 #else
diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c
index 48dfb9de853d..ac96817d1f99 100644
--- a/arch/mips/kernel/cpu-probe.c
+++ b/arch/mips/kernel/cpu-probe.c
@@ -516,6 +516,10 @@ static inline unsigned int decode_config5(struct cpuinfo_mips *c)
 		c->options |= MIPS_CPU_MAAR;
 	if (config5 & MIPS_CONF5_LLB)
 		c->options |= MIPS_CPU_RW_LLB;
+#ifdef CONFIG_XPA
+	if (config5 & MIPS_CONF5_MVH)
+		c->options |= MIPS_CPU_XPA;
+#endif
 
 	return config5 & MIPS_CONF_M;
 }
diff --git a/arch/mips/kernel/proc.c b/arch/mips/kernel/proc.c
index 130af7d26a9c..298b2b773d12 100644
--- a/arch/mips/kernel/proc.c
+++ b/arch/mips/kernel/proc.c
@@ -120,6 +120,7 @@ static int show_cpuinfo(struct seq_file *m, void *v)
 	if (cpu_has_msa)	seq_printf(m, "%s", " msa");
 	if (cpu_has_eva)	seq_printf(m, "%s", " eva");
 	if (cpu_has_htw)	seq_printf(m, "%s", " htw");
+	if (cpu_has_xpa)	seq_printf(m, "%s", " xpa");
 	seq_printf(m, "\n");
 
 	if (cpu_has_mmips) {
diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c
index 448cde372af0..faa5c9822ecc 100644
--- a/arch/mips/mm/init.c
+++ b/arch/mips/mm/init.c
@@ -96,7 +96,7 @@ static void *__kmap_pgprot(struct page *page, unsigned long addr, pgprot_t prot)
 	vaddr = __fix_to_virt(FIX_CMAP_END - idx);
 	pte = mk_pte(page, prot);
 #if defined(CONFIG_PHYS_ADDR_T_64BIT) && defined(CONFIG_CPU_MIPS32)
-	entrylo = pte.pte_high;
+	entrylo = pte_to_entrylo(pte.pte_high);
 #else
 	entrylo = pte_to_entrylo(pte_val(pte));
 #endif
@@ -106,6 +106,11 @@ static void *__kmap_pgprot(struct page *page, unsigned long addr, pgprot_t prot)
 	write_c0_entryhi(vaddr & (PAGE_MASK << 1));
 	write_c0_entrylo0(entrylo);
 	write_c0_entrylo1(entrylo);
+#ifdef CONFIG_XPA
+	entrylo = (pte.pte_low & _PFNX_MASK);
+	writex_c0_entrylo0(entrylo);
+	writex_c0_entrylo1(entrylo);
+#endif
 	tlbidx = read_c0_wired();
 	write_c0_wired(tlbidx + 1);
 	write_c0_index(tlbidx);
diff --git a/arch/mips/mm/tlb-r4k.c b/arch/mips/mm/tlb-r4k.c
index b2afa49beab0..c2500f4cb1d1 100644
--- a/arch/mips/mm/tlb-r4k.c
+++ b/arch/mips/mm/tlb-r4k.c
@@ -333,9 +333,17 @@ void __update_tlb(struct vm_area_struct * vma, unsigned long address, pte_t pte)
 		ptep = pte_offset_map(pmdp, address);
 
 #if defined(CONFIG_PHYS_ADDR_T_64BIT) && defined(CONFIG_CPU_MIPS32)
+#ifdef CONFIG_XPA
+		write_c0_entrylo0(pte_to_entrylo(ptep->pte_high));
+		writex_c0_entrylo0(ptep->pte_low & _PFNX_MASK);
+		ptep++;
+		write_c0_entrylo1(pte_to_entrylo(ptep->pte_high));
+		writex_c0_entrylo1(ptep->pte_low & _PFNX_MASK);
+#else
 		write_c0_entrylo0(ptep->pte_high);
 		ptep++;
 		write_c0_entrylo1(ptep->pte_high);
+#endif
 #else
 		write_c0_entrylo0(pte_to_entrylo(pte_val(*ptep++)));
 		write_c0_entrylo1(pte_to_entrylo(pte_val(*ptep)));
@@ -355,6 +363,9 @@ void __update_tlb(struct vm_area_struct * vma, unsigned long address, pte_t pte)
 void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1,
 		     unsigned long entryhi, unsigned long pagemask)
 {
+#ifdef CONFIG_XPA
+	panic("Broken for XPA kernels");
+#else
 	unsigned long flags;
 	unsigned long wired;
 	unsigned long old_pagemask;
@@ -383,6 +394,7 @@ void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1,
 	write_c0_pagemask(old_pagemask);
 	local_flush_tlb_all();
 	local_irq_restore(flags);
+#endif
 }
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c
index 20d985901e44..7709920e0cef 100644
--- a/arch/mips/mm/tlbex.c
+++ b/arch/mips/mm/tlbex.c
@@ -35,6 +35,17 @@
 #include <asm/uasm.h>
 #include <asm/setup.h>
 
+static int __cpuinitdata mips_xpa_disabled;
+
+static int __init xpa_disable(char *s)
+{
+	mips_xpa_disabled = 1;
+
+	return 1;
+}
+
+__setup("noxpa", xpa_disable);
+
 /*
  * TLB load/store/modify handlers.
  *
@@ -1027,12 +1038,27 @@ static void build_update_entries(u32 **p, unsigned int tmp, unsigned int ptep)
 	} else {
 		int pte_off_even = sizeof(pte_t) / 2;
 		int pte_off_odd = pte_off_even + sizeof(pte_t);
+#ifdef CONFIG_XPA
+		const int scratch = 1; /* Our extra working register */
 
-		/* The pte entries are pre-shifted */
-		uasm_i_lw(p, tmp, pte_off_even, ptep); /* get even pte */
-		UASM_i_MTC0(p, tmp, C0_ENTRYLO0); /* load it */
-		uasm_i_lw(p, ptep, pte_off_odd, ptep); /* get odd pte */
-		UASM_i_MTC0(p, ptep, C0_ENTRYLO1); /* load it */
+		uasm_i_addu(p, scratch, 0, ptep);
+#endif
+		uasm_i_lw(p, tmp, pte_off_even, ptep); /* even pte */
+		uasm_i_lw(p, ptep, pte_off_odd, ptep); /* odd pte */
+		UASM_i_ROTR(p, tmp, tmp, ilog2(_PAGE_GLOBAL));
+		UASM_i_ROTR(p, ptep, ptep, ilog2(_PAGE_GLOBAL));
+		UASM_i_MTC0(p, tmp, C0_ENTRYLO0);
+		UASM_i_MTC0(p, ptep, C0_ENTRYLO1);
+#ifdef CONFIG_XPA
+		uasm_i_lw(p, tmp, 0, scratch);
+		uasm_i_lw(p, ptep, sizeof(pte_t), scratch);
+		uasm_i_lui(p, scratch, 0xff);
+		uasm_i_ori(p, scratch, scratch, 0xffff);
+		uasm_i_and(p, tmp, scratch, tmp);
+		uasm_i_and(p, ptep, scratch, ptep);
+		uasm_i_mthc0(p, tmp, C0_ENTRYLO0);
+		uasm_i_mthc0(p, ptep, C0_ENTRYLO1);
+#endif
 	}
 #else
 	UASM_i_LW(p, tmp, 0, ptep); /* get even pte */
@@ -1533,8 +1559,14 @@ iPTE_SW(u32 **p, struct uasm_reloc **r, unsigned int pte, unsigned int ptr,
 {
 #ifdef CONFIG_PHYS_ADDR_T_64BIT
 	unsigned int hwmode = mode & (_PAGE_VALID | _PAGE_DIRTY);
-#endif
 
+	if (!cpu_has_64bits) {
+		const int scratch = 1; /* Our extra working register */
+
+		uasm_i_lui(p, scratch, (mode >> 16));
+		uasm_i_or(p, pte, pte, scratch);
+	} else
+#endif
 	uasm_i_ori(p, pte, pte, mode);
 #ifdef CONFIG_SMP
 # ifdef CONFIG_PHYS_ADDR_T_64BIT
@@ -1598,15 +1630,17 @@ build_pte_present(u32 **p, struct uasm_reloc **r,
 			uasm_il_bbit0(p, r, pte, ilog2(_PAGE_PRESENT), lid);
 			uasm_i_nop(p);
 		} else {
-			uasm_i_andi(p, t, pte, _PAGE_PRESENT);
+			uasm_i_srl(p, t, pte, _PAGE_PRESENT_SHIFT);
+			uasm_i_andi(p, t, t, 1);
 			uasm_il_beqz(p, r, t, lid);
 			if (pte == t)
 				/* You lose the SMP race :-(*/
 				iPTE_LW(p, pte, ptr);
 		}
 	} else {
-		uasm_i_andi(p, t, pte, _PAGE_PRESENT | _PAGE_READ);
-		uasm_i_xori(p, t, t, _PAGE_PRESENT | _PAGE_READ);
+		uasm_i_srl(p, t, pte, _PAGE_PRESENT_SHIFT);
+		uasm_i_andi(p, t, t, 3);
+		uasm_i_xori(p, t, t, 3);
 		uasm_il_bnez(p, r, t, lid);
 		if (pte == t)
 			/* You lose the SMP race :-(*/
@@ -1635,8 +1669,9 @@ build_pte_writable(u32 **p, struct uasm_reloc **r,
 {
 	int t = scratch >= 0 ? scratch : pte;
 
-	uasm_i_andi(p, t, pte, _PAGE_PRESENT | _PAGE_WRITE);
-	uasm_i_xori(p, t, t, _PAGE_PRESENT | _PAGE_WRITE);
+	uasm_i_srl(p, t, pte, _PAGE_PRESENT_SHIFT);
+	uasm_i_andi(p, t, t, 5);
+	uasm_i_xori(p, t, t, 5);
 	uasm_il_bnez(p, r, t, lid);
 	if (pte == t)
 		/* You lose the SMP race :-(*/
@@ -1672,7 +1707,8 @@ build_pte_modifiable(u32 **p, struct uasm_reloc **r,
 		uasm_i_nop(p);
 	} else {
 		int t = scratch >= 0 ? scratch : pte;
-		uasm_i_andi(p, t, pte, _PAGE_WRITE);
+		uasm_i_srl(p, t, pte, _PAGE_WRITE_SHIFT);
+		uasm_i_andi(p, t, t, 1);
 		uasm_il_beqz(p, r, t, lid);
 		if (pte == t)
 			/* You lose the SMP race :-(*/
@@ -2285,6 +2321,11 @@ static void config_htw_params(void)
 
 	pwsize = ilog2(PTRS_PER_PGD) << MIPS_PWSIZE_GDW_SHIFT;
 	pwsize |= ilog2(PTRS_PER_PTE) << MIPS_PWSIZE_PTW_SHIFT;
+
+	/* If XPA has been enabled, PTEs are 64-bit in size. */
+	if (read_c0_pagegrain() & PG_ELPA)
+		pwsize |= 1;
+
 	write_c0_pwsize(pwsize);
 
 	/* Make sure everything is set before we enable the HTW */
@@ -2298,6 +2339,28 @@ static void config_htw_params(void)
 	print_htw_config();
 }
 
+static void config_xpa_params(void)
+{
+#ifdef CONFIG_XPA
+	unsigned int pagegrain;
+
+	if (mips_xpa_disabled) {
+		pr_info("Extended Physical Addressing (XPA) disabled\n");
+		return;
+	}
+
+	pagegrain = read_c0_pagegrain();
+	write_c0_pagegrain(pagegrain | PG_ELPA);
+	back_to_back_c0_hazard();
+	pagegrain = read_c0_pagegrain();
+
+	if (pagegrain & PG_ELPA)
+		pr_info("Extended Physical Addressing (XPA) enabled\n");
+	else
+		panic("Extended Physical Addressing (XPA) disabled");
+#endif
+}
+
 void build_tlb_refill_handler(void)
 {
 	/*
@@ -2362,8 +2425,9 @@ void build_tlb_refill_handler(void)
 		}
 		if (cpu_has_local_ebase)
 			build_r4000_tlb_refill_handler();
+		if (cpu_has_xpa)
+			config_xpa_params();
 		if (cpu_has_htw)
 			config_htw_params();
-
 	}
 }
-- 
cgit v1.2.3-59-g8ed1b


From 5b9593f3bccb9904f260f9ad7f184e1d2921bd1e Mon Sep 17 00:00:00 2001
From: Lars Persson <lars.persson@axis.com>
Date: Thu, 26 Feb 2015 14:16:02 +0100
Subject: Revert "MIPS: Remove race window in page fault handling"

Revert commit 2a4a8b1e5d9d ("MIPS: Remove race window in page fault
handling") because it increased the number of flushed dcache pages and
became a performance problem for some workloads.

Signed-off-by: Lars Persson <larper@axis.com>
Cc: linux-mips@linux-mips.org
Cc: paul.burton@imgtec.com
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/9345/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
 arch/mips/include/asm/pgtable.h |  9 +++++----
 arch/mips/mm/cache.c            | 27 ++++++++-------------------
 2 files changed, 13 insertions(+), 23 deletions(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h
index bffd46ca3694..819af9d057a8 100644
--- a/arch/mips/include/asm/pgtable.h
+++ b/arch/mips/include/asm/pgtable.h
@@ -127,10 +127,6 @@ do {									\
 	}								\
 } while(0)
 
-
-extern void set_pte_at(struct mm_struct *mm, unsigned long addr, pte_t *ptep,
-	pte_t pteval);
-
 #if defined(CONFIG_PHYS_ADDR_T_64BIT) && defined(CONFIG_CPU_MIPS32)
 
 #define pte_none(pte)		(!(((pte).pte_high) & ~_PAGE_GLOBAL))
@@ -152,6 +148,7 @@ static inline void set_pte(pte_t *ptep, pte_t pte)
 			buddy->pte_high |= _PAGE_GLOBAL;
 	}
 }
+#define set_pte_at(mm, addr, ptep, pteval) set_pte(ptep, pteval)
 
 static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
 {
@@ -190,6 +187,7 @@ static inline void set_pte(pte_t *ptep, pte_t pteval)
 	}
 #endif
 }
+#define set_pte_at(mm, addr, ptep, pteval) set_pte(ptep, pteval)
 
 static inline void pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
 {
@@ -399,12 +397,15 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
 
 extern void __update_tlb(struct vm_area_struct *vma, unsigned long address,
 	pte_t pte);
+extern void __update_cache(struct vm_area_struct *vma, unsigned long address,
+	pte_t pte);
 
 static inline void update_mmu_cache(struct vm_area_struct *vma,
 	unsigned long address, pte_t *ptep)
 {
 	pte_t pte = *ptep;
 	__update_tlb(vma, address, pte);
+	__update_cache(vma, address, pte);
 }
 
 static inline void update_mmu_cache_pmd(struct vm_area_struct *vma,
diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c
index 7e3ea7766822..f7b91d3a371d 100644
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -119,36 +119,25 @@ void __flush_anon_page(struct page *page, unsigned long vmaddr)
 
 EXPORT_SYMBOL(__flush_anon_page);
 
-static void mips_flush_dcache_from_pte(pte_t pteval, unsigned long address)
+void __update_cache(struct vm_area_struct *vma, unsigned long address,
+	pte_t pte)
 {
 	struct page *page;
-	unsigned long pfn = pte_pfn(pteval);
+	unsigned long pfn, addr;
+	int exec = (vma->vm_flags & VM_EXEC) && !cpu_has_ic_fills_f_dc;
 
+	pfn = pte_pfn(pte);
 	if (unlikely(!pfn_valid(pfn)))
 		return;
-
 	page = pfn_to_page(pfn);
 	if (page_mapping(page) && Page_dcache_dirty(page)) {
-		unsigned long page_addr = (unsigned long) page_address(page);
-
-		if (!cpu_has_ic_fills_f_dc ||
-		    pages_do_alias(page_addr, address & PAGE_MASK))
-			flush_data_cache_page(page_addr);
+		addr = (unsigned long) page_address(page);
+		if (exec || pages_do_alias(addr, address & PAGE_MASK))
+			flush_data_cache_page(addr);
 		ClearPageDcacheDirty(page);
 	}
 }
 
-void set_pte_at(struct mm_struct *mm, unsigned long addr,
-        pte_t *ptep, pte_t pteval)
-{
-        if (cpu_has_dc_aliases || !cpu_has_ic_fills_f_dc) {
-                if (pte_present(pteval))
-                        mips_flush_dcache_from_pte(pteval, addr);
-        }
-
-        set_pte(ptep, pteval);
-}
-
 unsigned long _page_cachable_default;
 EXPORT_SYMBOL(_page_cachable_default);
 
-- 
cgit v1.2.3-59-g8ed1b


From 4d46a67a3eb827ccf1125959936fd51ba318dabc Mon Sep 17 00:00:00 2001
From: Lars Persson <lars.persson@axis.com>
Date: Thu, 26 Feb 2015 14:16:03 +0100
Subject: MIPS: Fix race condition in lazy cache flushing.

The lazy cache flushing implemented in the MIPS kernel suffers from a
race condition that is exposed by do_set_pte() in mm/memory.c.

A pre-condition is a file-system that writes to the page from the CPU
in its readpage method and then calls flush_dcache_page(). One example
is ubifs. Another pre-condition is that the dcache flush is postponed
in __flush_dcache_page().

Upon a page fault for an executable mapping not existing in the
page-cache, the following will happen:
1. Write to the page
2. flush_dcache_page
3. flush_icache_page
4. set_pte_at
5. update_mmu_cache (commits the flush of a dcache-dirty page)

Between steps 4 and 5 another thread can hit the same page and it will
encounter a valid pte. Because the data still is in the L1 dcache the CPU
will fetch stale data from L2 into the icache and execute garbage.

This fix moves the commit of the cache flush to step 3 to close the
race window. It also reduces the amount of flushes on non-executable
mappings because we never enter __flush_dcache_page() for non-aliasing
CPUs.

Regressions can occur in drivers that mistakenly relies on the
flush_dcache_page() in get_user_pages() for DMA operations.

[ralf@linux-mips.org: Folded in patch 9346 to fix highmem issue.]

Signed-off-by: Lars Persson <larper@axis.com>
Cc: linux-mips@linux-mips.org
Cc: paul.burton@imgtec.com
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/9346/
Patchwork: https://patchwork.linux-mips.org/patch/9738/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
 arch/mips/include/asm/cacheflush.h | 38 +++++++++++++++++++++++---------------
 arch/mips/mm/cache.c               | 12 ++++++++++++
 2 files changed, 35 insertions(+), 15 deletions(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/include/asm/cacheflush.h b/arch/mips/include/asm/cacheflush.h
index e08381a37f8b..723229f4cf27 100644
--- a/arch/mips/include/asm/cacheflush.h
+++ b/arch/mips/include/asm/cacheflush.h
@@ -29,6 +29,20 @@
  *  - flush_icache_all() flush the entire instruction cache
  *  - flush_data_cache_page() flushes a page from the data cache
  */
+
+ /*
+ * This flag is used to indicate that the page pointed to by a pte
+ * is dirty and requires cleaning before returning it to the user.
+ */
+#define PG_dcache_dirty			PG_arch_1
+
+#define Page_dcache_dirty(page)		\
+	test_bit(PG_dcache_dirty, &(page)->flags)
+#define SetPageDcacheDirty(page)	\
+	set_bit(PG_dcache_dirty, &(page)->flags)
+#define ClearPageDcacheDirty(page)	\
+	clear_bit(PG_dcache_dirty, &(page)->flags)
+
 extern void (*flush_cache_all)(void);
 extern void (*__flush_cache_all)(void);
 extern void (*flush_cache_mm)(struct mm_struct *mm);
@@ -37,13 +51,15 @@ extern void (*flush_cache_range)(struct vm_area_struct *vma,
 	unsigned long start, unsigned long end);
 extern void (*flush_cache_page)(struct vm_area_struct *vma, unsigned long page, unsigned long pfn);
 extern void __flush_dcache_page(struct page *page);
+extern void __flush_icache_page(struct vm_area_struct *vma, struct page *page);
 
 #define ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE 1
 static inline void flush_dcache_page(struct page *page)
 {
-	if (cpu_has_dc_aliases || !cpu_has_ic_fills_f_dc)
+	if (cpu_has_dc_aliases)
 		__flush_dcache_page(page);
-
+	else if (!cpu_has_ic_fills_f_dc)
+		SetPageDcacheDirty(page);
 }
 
 #define flush_dcache_mmap_lock(mapping)		do { } while (0)
@@ -61,6 +77,11 @@ static inline void flush_anon_page(struct vm_area_struct *vma,
 static inline void flush_icache_page(struct vm_area_struct *vma,
 	struct page *page)
 {
+	if (!cpu_has_ic_fills_f_dc && (vma->vm_flags & VM_EXEC) &&
+	    Page_dcache_dirty(page)) {
+		__flush_icache_page(vma, page);
+		ClearPageDcacheDirty(page);
+	}
 }
 
 extern void (*flush_icache_range)(unsigned long start, unsigned long end);
@@ -95,19 +116,6 @@ extern void (*flush_icache_all)(void);
 extern void (*local_flush_data_cache_page)(void * addr);
 extern void (*flush_data_cache_page)(unsigned long addr);
 
-/*
- * This flag is used to indicate that the page pointed to by a pte
- * is dirty and requires cleaning before returning it to the user.
- */
-#define PG_dcache_dirty			PG_arch_1
-
-#define Page_dcache_dirty(page)		\
-	test_bit(PG_dcache_dirty, &(page)->flags)
-#define SetPageDcacheDirty(page)	\
-	set_bit(PG_dcache_dirty, &(page)->flags)
-#define ClearPageDcacheDirty(page)	\
-	clear_bit(PG_dcache_dirty, &(page)->flags)
-
 /* Run kernel code uncached, useful for cache probing functions. */
 unsigned long run_uncached(void *func);
 
diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c
index f7b91d3a371d..77d96db8253c 100644
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -119,6 +119,18 @@ void __flush_anon_page(struct page *page, unsigned long vmaddr)
 
 EXPORT_SYMBOL(__flush_anon_page);
 
+void __flush_icache_page(struct vm_area_struct *vma, struct page *page)
+{
+	unsigned long addr;
+
+	if (PageHighMem(page))
+		return;
+
+	addr = (unsigned long) page_address(page);
+	flush_data_cache_page(addr);
+}
+EXPORT_SYMBOL_GPL(__flush_icache_page);
+
 void __update_cache(struct vm_area_struct *vma, unsigned long address,
 	pte_t pte)
 {
-- 
cgit v1.2.3-59-g8ed1b


From 0acbfc66d09e97e5a01e7a23ac7e99f360ff851b Mon Sep 17 00:00:00 2001
From: Ralf Baechle <ralf@linux-mips.org>
Date: Fri, 27 Mar 2015 15:10:30 +0100
Subject: MIPS: DMA: Implement platform hook to perform post-DMA cache flushes.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
 arch/mips/include/asm/mach-ath25/dma-coherence.h         | 4 ++++
 arch/mips/include/asm/mach-bmips/dma-coherence.h         | 4 ++++
 arch/mips/include/asm/mach-cavium-octeon/dma-coherence.h | 4 ++++
 arch/mips/include/asm/mach-generic/dma-coherence.h       | 4 ++++
 arch/mips/include/asm/mach-ip27/dma-coherence.h          | 4 ++++
 arch/mips/include/asm/mach-ip32/dma-coherence.h          | 4 ++++
 arch/mips/include/asm/mach-jazz/dma-coherence.h          | 4 ++++
 arch/mips/include/asm/mach-loongson/dma-coherence.h      | 4 ++++
 arch/mips/mm/dma-default.c                               | 4 +++-
 9 files changed, 35 insertions(+), 1 deletion(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/include/asm/mach-ath25/dma-coherence.h b/arch/mips/include/asm/mach-ath25/dma-coherence.h
index 65084a079b60..d5defdde32db 100644
--- a/arch/mips/include/asm/mach-ath25/dma-coherence.h
+++ b/arch/mips/include/asm/mach-ath25/dma-coherence.h
@@ -69,4 +69,8 @@ static inline int plat_device_is_coherent(struct device *dev)
 #endif
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 #endif /* __ASM_MACH_ATH25_DMA_COHERENCE_H */
diff --git a/arch/mips/include/asm/mach-bmips/dma-coherence.h b/arch/mips/include/asm/mach-bmips/dma-coherence.h
index 65e95b03ef92..509ab72725eb 100644
--- a/arch/mips/include/asm/mach-bmips/dma-coherence.h
+++ b/arch/mips/include/asm/mach-bmips/dma-coherence.h
@@ -45,4 +45,8 @@ static inline int plat_device_is_coherent(struct device *dev)
 	return 0;
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 #endif /* __ASM_MACH_BMIPS_DMA_COHERENCE_H */
diff --git a/arch/mips/include/asm/mach-cavium-octeon/dma-coherence.h b/arch/mips/include/asm/mach-cavium-octeon/dma-coherence.h
index f9f448650505..460042ee5d6f 100644
--- a/arch/mips/include/asm/mach-cavium-octeon/dma-coherence.h
+++ b/arch/mips/include/asm/mach-cavium-octeon/dma-coherence.h
@@ -57,6 +57,10 @@ static inline int plat_device_is_coherent(struct device *dev)
 	return 1;
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 dma_addr_t phys_to_dma(struct device *dev, phys_addr_t paddr);
 phys_addr_t dma_to_phys(struct device *dev, dma_addr_t daddr);
 
diff --git a/arch/mips/include/asm/mach-generic/dma-coherence.h b/arch/mips/include/asm/mach-generic/dma-coherence.h
index 7629c35986f7..671330928e67 100644
--- a/arch/mips/include/asm/mach-generic/dma-coherence.h
+++ b/arch/mips/include/asm/mach-generic/dma-coherence.h
@@ -52,6 +52,10 @@ static inline int plat_device_is_coherent(struct device *dev)
 	return coherentio;
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 #ifdef CONFIG_SWIOTLB
 static inline dma_addr_t phys_to_dma(struct device *dev, phys_addr_t paddr)
 {
diff --git a/arch/mips/include/asm/mach-ip27/dma-coherence.h b/arch/mips/include/asm/mach-ip27/dma-coherence.h
index 4ffddfdb5062..1daa64412569 100644
--- a/arch/mips/include/asm/mach-ip27/dma-coherence.h
+++ b/arch/mips/include/asm/mach-ip27/dma-coherence.h
@@ -58,6 +58,10 @@ static inline int plat_dma_supported(struct device *dev, u64 mask)
 	return 1;
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 static inline int plat_device_is_coherent(struct device *dev)
 {
 	return 1;		/* IP27 non-cohernet mode is unsupported */
diff --git a/arch/mips/include/asm/mach-ip32/dma-coherence.h b/arch/mips/include/asm/mach-ip32/dma-coherence.h
index 104cfbc3ed63..0a0b0e2ced60 100644
--- a/arch/mips/include/asm/mach-ip32/dma-coherence.h
+++ b/arch/mips/include/asm/mach-ip32/dma-coherence.h
@@ -80,6 +80,10 @@ static inline int plat_dma_supported(struct device *dev, u64 mask)
 	return 1;
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 static inline int plat_device_is_coherent(struct device *dev)
 {
 	return 0;		/* IP32 is non-cohernet */
diff --git a/arch/mips/include/asm/mach-jazz/dma-coherence.h b/arch/mips/include/asm/mach-jazz/dma-coherence.h
index 949003ef97b3..dc347c25c343 100644
--- a/arch/mips/include/asm/mach-jazz/dma-coherence.h
+++ b/arch/mips/include/asm/mach-jazz/dma-coherence.h
@@ -48,6 +48,10 @@ static inline int plat_dma_supported(struct device *dev, u64 mask)
 	return 1;
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 static inline int plat_device_is_coherent(struct device *dev)
 {
 	return 0;
diff --git a/arch/mips/include/asm/mach-loongson/dma-coherence.h b/arch/mips/include/asm/mach-loongson/dma-coherence.h
index a90534161bd2..4bf4e19f72e8 100644
--- a/arch/mips/include/asm/mach-loongson/dma-coherence.h
+++ b/arch/mips/include/asm/mach-loongson/dma-coherence.h
@@ -78,4 +78,8 @@ static inline int plat_device_is_coherent(struct device *dev)
 #endif /* CONFIG_DMA_NONCOHERENT */
 }
 
+static inline void plat_post_dma_flush(struct device *dev)
+{
+}
+
 #endif /* __ASM_MACH_LOONGSON_DMA_COHERENCE_H */
diff --git a/arch/mips/mm/dma-default.c b/arch/mips/mm/dma-default.c
index af5f046e627e..609d1241b0c4 100644
--- a/arch/mips/mm/dma-default.c
+++ b/arch/mips/mm/dma-default.c
@@ -258,7 +258,7 @@ static void mips_dma_unmap_page(struct device *dev, dma_addr_t dma_addr,
 	if (cpu_needs_post_dma_flush(dev))
 		__dma_sync(dma_addr_to_page(dev, dma_addr),
 			   dma_addr & ~PAGE_MASK, size, direction);
-
+	plat_post_dma_flush(dev);
 	plat_unmap_dma_mem(dev, dma_addr, size, direction);
 }
 
@@ -312,6 +312,7 @@ static void mips_dma_sync_single_for_cpu(struct device *dev,
 	if (cpu_needs_post_dma_flush(dev))
 		__dma_sync(dma_addr_to_page(dev, dma_handle),
 			   dma_handle & ~PAGE_MASK, size, direction);
+	plat_post_dma_flush(dev);
 }
 
 static void mips_dma_sync_single_for_device(struct device *dev,
@@ -331,6 +332,7 @@ static void mips_dma_sync_sg_for_cpu(struct device *dev,
 		for (i = 0; i < nelems; i++, sg++)
 			__dma_sync(sg_page(sg), sg->offset, sg->length,
 				   direction);
+	plat_post_dma_flush(dev);
 }
 
 static void mips_dma_sync_sg_for_device(struct device *dev,
-- 
cgit v1.2.3-59-g8ed1b


From 3057739138eb8fbaa5154b149a864f5218898c73 Mon Sep 17 00:00:00 2001
From: Joshua Kinard <kumba@gentoo.org>
Date: Wed, 21 Jan 2015 07:59:45 -0500
Subject: MIPS: Add R16000 detection

This allows the kernel to correctly detect an R16000 MIPS CPU on systems that
have those.  Otherwise, such systems will detect the CPU as an R14000, due to
similarities in the CPU PRId value.

Signed-off-by: Joshua Kinard <kumba@gentoo.org>
Cc: Linux MIPS List <linux-mips@linux-mips.org>
Patchwork: https://patchwork.linux-mips.org/patch/9092/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
 arch/mips/include/asm/cpu-type.h     |  1 +
 arch/mips/include/asm/cpu.h          |  6 +++---
 arch/mips/kernel/cpu-probe.c         |  9 +++++++--
 arch/mips/kernel/perf_event_mipsxx.c |  1 +
 arch/mips/mm/c-r4k.c                 | 10 +++++++---
 arch/mips/mm/page.c                  |  1 +
 arch/mips/mm/tlb-r4k.c               |  3 ++-
 arch/mips/mm/tlbex.c                 |  1 +
 arch/mips/oprofile/common.c          |  1 +
 arch/mips/oprofile/op_model_mipsxx.c |  5 +++++
 10 files changed, 29 insertions(+), 9 deletions(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/include/asm/cpu-type.h b/arch/mips/include/asm/cpu-type.h
index 8245875f8b33..33f3cab9e689 100644
--- a/arch/mips/include/asm/cpu-type.h
+++ b/arch/mips/include/asm/cpu-type.h
@@ -157,6 +157,7 @@ static inline int __pure __get_cpu_type(const int cpu_type)
 	case CPU_R10000:
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 #endif
 #ifdef CONFIG_SYS_HAS_CPU_RM7000
 	case CPU_RM7000:
diff --git a/arch/mips/include/asm/cpu.h b/arch/mips/include/asm/cpu.h
index 15903cad1c6f..fd2e893e9d9f 100644
--- a/arch/mips/include/asm/cpu.h
+++ b/arch/mips/include/asm/cpu.h
@@ -67,7 +67,7 @@
 #define PRID_IMP_R4300		0x0b00
 #define PRID_IMP_VR41XX		0x0c00
 #define PRID_IMP_R12000		0x0e00
-#define PRID_IMP_R14000		0x0f00
+#define PRID_IMP_R14000		0x0f00		/* R14K && R16K */
 #define PRID_IMP_R8000		0x1000
 #define PRID_IMP_PR4450		0x1200
 #define PRID_IMP_R4600		0x2000
@@ -284,8 +284,8 @@ enum cpu_type_enum {
 	CPU_R4000PC, CPU_R4000SC, CPU_R4000MC, CPU_R4200, CPU_R4300, CPU_R4310,
 	CPU_R4400PC, CPU_R4400SC, CPU_R4400MC, CPU_R4600, CPU_R4640, CPU_R4650,
 	CPU_R4700, CPU_R5000, CPU_R5500, CPU_NEVADA, CPU_R5432, CPU_R10000,
-	CPU_R12000, CPU_R14000, CPU_VR41XX, CPU_VR4111, CPU_VR4121, CPU_VR4122,
-	CPU_VR4131, CPU_VR4133, CPU_VR4181, CPU_VR4181A, CPU_RM7000,
+	CPU_R12000, CPU_R14000, CPU_R16000, CPU_VR41XX, CPU_VR4111, CPU_VR4121,
+	CPU_VR4122, CPU_VR4131, CPU_VR4133, CPU_VR4181, CPU_VR4181A, CPU_RM7000,
 	CPU_SR71000, CPU_TX49XX,
 
 	/*
diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c
index 1bae00678b9b..cb436f585bfa 100644
--- a/arch/mips/kernel/cpu-probe.c
+++ b/arch/mips/kernel/cpu-probe.c
@@ -852,8 +852,13 @@ static inline void cpu_probe_legacy(struct cpuinfo_mips *c, unsigned int cpu)
 		c->tlbsize = 64;
 		break;
 	case PRID_IMP_R14000:
-		c->cputype = CPU_R14000;
-		__cpu_name[cpu] = "R14000";
+		if (((c->processor_id >> 4) & 0x0f) > 2) {
+			c->cputype = CPU_R16000;
+			__cpu_name[cpu] = "R16000";
+		} else {
+			c->cputype = CPU_R14000;
+			__cpu_name[cpu] = "R14000";
+		}
 		set_isa(c, MIPS_CPU_ISA_IV);
 		c->options = MIPS_CPU_TLB | MIPS_CPU_4K_CACHE | MIPS_CPU_4KEX |
 			     MIPS_CPU_FPU | MIPS_CPU_32FPR |
diff --git a/arch/mips/kernel/perf_event_mipsxx.c b/arch/mips/kernel/perf_event_mipsxx.c
index 9d90efea8bb0..192e7f59245e 100644
--- a/arch/mips/kernel/perf_event_mipsxx.c
+++ b/arch/mips/kernel/perf_event_mipsxx.c
@@ -777,6 +777,7 @@ static int n_counters(void)
 
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 		counters = 4;
 		break;
 
diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c
index 3f8059602765..58e2eeddc391 100644
--- a/arch/mips/mm/c-r4k.c
+++ b/arch/mips/mm/c-r4k.c
@@ -430,6 +430,7 @@ static inline void local_r4k___flush_cache_all(void * args)
 	case CPU_R10000:
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 		/*
 		 * These caches are inclusive caches, that is, if something
 		 * is not cached in the S-cache, we know it also won't be
@@ -506,7 +507,7 @@ static inline void local_r4k_flush_cache_mm(void * args)
 
 	/*
 	 * Kludge alert.  For obscure reasons R4000SC and R4400SC go nuts if we
-	 * only flush the primary caches but R10000 and R12000 behave sane ...
+	 * only flush the primary caches but R1x000 behave sane ...
 	 * R4000SC and R4400SC indexed S-cache ops also invalidate primary
 	 * caches, so we can bail out early.
 	 */
@@ -1012,6 +1013,7 @@ static void probe_pcache(void)
 	case CPU_R10000:
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 		icache_size = 1 << (12 + ((config & R10K_CONF_IC) >> 29));
 		c->icache.linesz = 64;
 		c->icache.ways = 2;
@@ -1223,8 +1225,8 @@ static void probe_pcache(void)
 		dcache_size / (c->dcache.linesz * c->dcache.ways) : 0;
 
 	/*
-	 * R10000 and R12000 P-caches are odd in a positive way.  They're 32kB
-	 * 2-way virtually indexed so normally would suffer from aliases.  So
+	 * R1x000 P-caches are odd in a positive way.  They're 32kB 2-way
+	 * virtually indexed so normally would suffer from aliases.  So
 	 * normally they'd suffer from aliases but magic in the hardware deals
 	 * with that for us so we don't need to take care ourselves.
 	 */
@@ -1240,6 +1242,7 @@ static void probe_pcache(void)
 	case CPU_R10000:
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 		break;
 
 	case CPU_74K:
@@ -1438,6 +1441,7 @@ static void setup_scache(void)
 	case CPU_R10000:
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 		scache_size = 0x80000 << ((config & R10K_CONF_SS) >> 16);
 		c->scache.linesz = 64 << ((config >> 13) & 1);
 		c->scache.ways = 2;
diff --git a/arch/mips/mm/page.c b/arch/mips/mm/page.c
index 3f85f921801b..885d73ffd6fb 100644
--- a/arch/mips/mm/page.c
+++ b/arch/mips/mm/page.c
@@ -157,6 +157,7 @@ static void set_prefetch_parameters(void)
 		case CPU_R10000:
 		case CPU_R12000:
 		case CPU_R14000:
+		case CPU_R16000:
 			/*
 			 * Those values have been experimentally tuned for an
 			 * Origin 200.
diff --git a/arch/mips/mm/tlb-r4k.c b/arch/mips/mm/tlb-r4k.c
index b2afa49beab0..37ad381c3e5f 100644
--- a/arch/mips/mm/tlb-r4k.c
+++ b/arch/mips/mm/tlb-r4k.c
@@ -477,7 +477,8 @@ static void r4k_tlb_configure(void)
 	write_c0_wired(0);
 	if (current_cpu_type() == CPU_R10000 ||
 	    current_cpu_type() == CPU_R12000 ||
-	    current_cpu_type() == CPU_R14000)
+	    current_cpu_type() == CPU_R14000 ||
+	    current_cpu_type() == CPU_R16000)
 		write_c0_framemask(0);
 
 	if (cpu_has_rixi) {
diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c
index d75ff73a2012..7c7469f56ec3 100644
--- a/arch/mips/mm/tlbex.c
+++ b/arch/mips/mm/tlbex.c
@@ -569,6 +569,7 @@ static void build_tlb_write_entry(u32 **p, struct uasm_label **l,
 	case CPU_R10000:
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 	case CPU_4KC:
 	case CPU_4KEC:
 	case CPU_M14KC:
diff --git a/arch/mips/oprofile/common.c b/arch/mips/oprofile/common.c
index a26cbe372e06..81f58958cf08 100644
--- a/arch/mips/oprofile/common.c
+++ b/arch/mips/oprofile/common.c
@@ -98,6 +98,7 @@ int __init oprofile_arch_init(struct oprofile_operations *ops)
 	case CPU_R10000:
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 	case CPU_XLR:
 		lmodel = &op_model_mipsxx_ops;
 		break;
diff --git a/arch/mips/oprofile/op_model_mipsxx.c b/arch/mips/oprofile/op_model_mipsxx.c
index d6b9e69e7c69..6a6e2cc55b89 100644
--- a/arch/mips/oprofile/op_model_mipsxx.c
+++ b/arch/mips/oprofile/op_model_mipsxx.c
@@ -296,6 +296,7 @@ static inline int n_counters(void)
 
 	case CPU_R12000:
 	case CPU_R14000:
+	case CPU_R16000:
 		counters = 4;
 		break;
 
@@ -411,6 +412,10 @@ static int __init mipsxx_init(void)
 		op_model_mipsxx_ops.cpu_type = "mips/r12000";
 		break;
 
+	case CPU_R16000:
+		op_model_mipsxx_ops.cpu_type = "mips/r16000";
+		break;
+
 	case CPU_SB1:
 	case CPU_SB1A:
 		op_model_mipsxx_ops.cpu_type = "mips/sb1";
-- 
cgit v1.2.3-59-g8ed1b


From e2e7f29af84aa59dd8191b9f6fee80aafa4e06cd Mon Sep 17 00:00:00 2001
From: "Maciej W. Rozycki" <macro@codesourcery.com>
Date: Sun, 16 Nov 2014 01:02:29 +0000
Subject: MIPS: c-r4k.c: Fix the 74K D-cache alias erratum workaround

Fix the 74K D-cache alias erratum workaround so that it actually works.
Our current code sets MIPS_CACHE_VTAG for the D-cache, but that flag
only has any effect for the I-cache.  Additionally MIPS_CACHE_PINDEX is
set for the D-cache if CP0.Config7.AR is also set for an affected
processor, leading to confusing information in the bootstrap log (the
flag isn't used beyond that).

So delete the setting of MIPS_CACHE_VTAG and rely on MIPS_CACHE_ALIASES,
set in a common place, removing I-cache coherency issues seen in GDB
testing with software breakpoints, gdbserver and ptrace(2), on affected
systems.

While at it add a little piece of explanation of what CP0.Config6.SYND
is so that people do not have to chase documentation.

Signed-off-by: Maciej W. Rozycki <macro@codesourcery.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/8507/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---
 arch/mips/mm/c-r4k.c | 23 +++++++++++++++--------
 1 file changed, 15 insertions(+), 8 deletions(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c
index 58e2eeddc391..0dbb65a51ce5 100644
--- a/arch/mips/mm/c-r4k.c
+++ b/arch/mips/mm/c-r4k.c
@@ -889,33 +889,39 @@ static inline void rm7k_erratum31(void)
 	}
 }
 
-static inline void alias_74k_erratum(struct cpuinfo_mips *c)
+static inline int alias_74k_erratum(struct cpuinfo_mips *c)
 {
 	unsigned int imp = c->processor_id & PRID_IMP_MASK;
 	unsigned int rev = c->processor_id & PRID_REV_MASK;
+	int present = 0;
 
 	/*
 	 * 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.
+	 * aliases.  In this case it is better to treat the cache as always
+	 * having aliases.  Also disable the synonym tag update feature
+	 * where available.  In this case no opportunistic tag update will
+	 * happen where a load causes a virtual address miss but a physical
+	 * address hit during a D-cache look-up.
 	 */
 	switch (imp) {
 	case PRID_IMP_74K:
 		if (rev <= PRID_REV_ENCODE_332(2, 4, 0))
-			c->dcache.flags |= MIPS_CACHE_VTAG;
+			present = 1;
 		if (rev == PRID_REV_ENCODE_332(2, 4, 0))
 			write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND);
 		break;
 	case PRID_IMP_1074K:
 		if (rev <= PRID_REV_ENCODE_332(1, 1, 0)) {
-			c->dcache.flags |= MIPS_CACHE_VTAG;
+			present = 1;
 			write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND);
 		}
 		break;
 	default:
 		BUG();
 	}
+
+	return present;
 }
 
 static void b5k_instruction_hazard(void)
@@ -939,6 +945,7 @@ static void probe_pcache(void)
 	struct cpuinfo_mips *c = &current_cpu_data;
 	unsigned int config = read_c0_config();
 	unsigned int prid = read_c0_prid();
+	int has_74k_erratum = 0;
 	unsigned long config1;
 	unsigned int lsize;
 
@@ -1247,7 +1254,7 @@ static void probe_pcache(void)
 
 	case CPU_74K:
 	case CPU_1074K:
-		alias_74k_erratum(c);
+		has_74k_erratum = alias_74k_erratum(c);
 		/* Fall through. */
 	case CPU_M14KC:
 	case CPU_M14KEC:
@@ -1262,7 +1269,7 @@ static void probe_pcache(void)
 		if (!(read_c0_config7() & MIPS_CONF7_IAR) &&
 		    (c->icache.waysize > PAGE_SIZE))
 			c->icache.flags |= MIPS_CACHE_ALIASES;
-		if (read_c0_config7() & MIPS_CONF7_AR) {
+		if (!has_74k_erratum && (read_c0_config7() & MIPS_CONF7_AR)) {
 			/*
 			 * Effectively physically indexed dcache,
 			 * thus no virtual aliases.
@@ -1271,7 +1278,7 @@ static void probe_pcache(void)
 			break;
 		}
 	default:
-		if (c->dcache.waysize > PAGE_SIZE)
+		if (has_74k_erratum || c->dcache.waysize > PAGE_SIZE)
 			c->dcache.flags |= MIPS_CACHE_ALIASES;
 	}
 
-- 
cgit v1.2.3-59-g8ed1b


From 9eaffa84a8a46adab065c983401fc9d5949c958f Mon Sep 17 00:00:00 2001
From: Ralf Baechle <ralf@linux-mips.org>
Date: Wed, 25 Mar 2015 13:18:27 +0100
Subject: Revert "MIPS: Avoid pipeline stalls on some MIPS32R2 cores."

For a discussion, see http://patchwork.linux-mips.org/patch/9539/.

This reverts commit 625c0a21700bdb90844d926a1508a17a77e369c9.
---
 arch/mips/mm/tlbex.c | 21 ++-------------------
 1 file changed, 2 insertions(+), 19 deletions(-)

(limited to 'arch/mips/mm')

diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c
index 7709920e0cef..971b1ee51234 100644
--- a/arch/mips/mm/tlbex.c
+++ b/arch/mips/mm/tlbex.c
@@ -512,26 +512,9 @@ static void build_tlb_write_entry(u32 **p, struct uasm_label **l,
 	case tlb_indexed: tlbw = uasm_i_tlbwi; break;
 	}
 
-	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:
-		case CPU_1074K:
-		case CPU_PROAPTIV:
-		case CPU_P5600:
-		case CPU_M5150:
-		case CPU_QEMU_GENERIC:
-			break;
-
-		default:
+	if (cpu_has_mips_r2_r6) {
+		if (cpu_has_mips_r2_exec_hazard)
 			uasm_i_ehb(p);
-			break;
-		}
 		tlbw(p);
 		return;
 	}
-- 
cgit v1.2.3-59-g8ed1b