Merge tag 'arc-v3.12-rc1' of git://git.kernel.org/pub/scm/linux/kerne…

…l/git/vgupta/arc

Pull ARC changes from Vineet Gupta:

 - ARC MM changes:
    - preparation for MMUv4 (accomodate new PTE bits, new cmds)
    - Rework the ASID allocation algorithm to remove asid-mm reverse map
 - Boilerplate code consolidation in Exception Handlers
 - Disable FRAME_POINTER for ARC
 - Unaligned Access Emulation for Big-Endian from Noam
 - Bunch of fixes (udelay, missing accessors) from Mischa

* tag 'arc-v3.12-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
  ARC: fix new Section mismatches in build (post __cpuinit cleanup)
  Kconfig.debug: Add FRAME_POINTER anti-dependency for ARC
  ARC: Fix __udelay calculation
  ARC: remove console_verbose() from setup_arch()
  ARC: Add read*_relaxed to asm/io.h
  ARC: Handle un-aligned user space access in BE.
  ARC: [ASID] Track ASID allocation cycles/generations
  ARC: [ASID] activate_mm() == switch_mm()
  ARC: [ASID] get_new_mmu_context() to conditionally allocate new ASID
  ARC: [ASID] Refactor the TLB paranoid debug code
  ARC: [ASID] Remove legacy/unused debug code
  ARC: No need to flush the TLB in early boot
  ARC: MMUv4 preps/3 - Abstract out TLB Insert/Delete
  ARC: MMUv4 preps/2 - Reshuffle PTE bits
  ARC: MMUv4 preps/1 - Fold PTE K/U access flags
  ARC: Code cosmetics (Nothing semantical)
  ARC: Entry Handler tweaks: Optimize away redundant IRQ_DISABLE_SAVE
  ARC: Exception Handlers Code consolidation
  ARC: Add some .gitignore entries

arch/arc/boot/.gitignore

@@ -0,0 +1 @@
+*.dtb*

arch/arc/include/asm/cache.h

@@ -57,7 +57,7 @@
 
 extern void arc_cache_init(void);
 extern char *arc_cache_mumbojumbo(int cpu_id, char *buf, int len);
-extern void __init read_decode_cache_bcr(void);
+extern void read_decode_cache_bcr(void);
 
 #endif	/* !__ASSEMBLY__ */
 

arch/arc/include/asm/delay.h

@@ -53,11 +53,10 @@ static inline void __udelay(unsigned long usecs)
 {
 	unsigned long loops;
 
-	/* (long long) cast ensures 64 bit MPY - real or emulated
+	/* (u64) cast ensures 64 bit MPY - real or emulated
 	 * HZ * 4295 is pre-evaluated by gcc - hence only 2 mpy ops
 	 */
-	loops = ((long long)(usecs * 4295 * HZ) *
-		 (long long)(loops_per_jiffy)) >> 32;
+	loops = ((u64) usecs * 4295 * HZ * loops_per_jiffy) >> 32;
 
 	__delay(loops);
 }

arch/arc/include/asm/entry.h

@@ -365,7 +365,7 @@
  * it to memory (non-SMP case) or SCRATCH0 Aux Reg (SMP).
  *
  * Before saving the full regfile - this reg is restored back, only
- * to be saved again on kernel mode stack, as part of ptregs.
+ * to be saved again on kernel mode stack, as part of pt_regs.
  *-------------------------------------------------------------*/
 .macro EXCPN_PROLOG_FREEUP_REG	reg
 #ifdef CONFIG_SMP
@@ -383,6 +383,28 @@
 #endif
 .endm
 
+/*--------------------------------------------------------------
+ * Exception Entry prologue
+ * -Switches stack to K mode (if not already)
+ * -Saves the register file
+ *
+ * After this it is safe to call the "C" handlers
+ *-------------------------------------------------------------*/
+.macro EXCEPTION_PROLOGUE
+
+	/* Need at least 1 reg to code the early exception prologue */
+	EXCPN_PROLOG_FREEUP_REG r9
+
+	/* U/K mode at time of exception (stack not switched if already K) */
+	lr  r9, [erstatus]
+
+	/* ARC700 doesn't provide auto-stack switching */
+	SWITCH_TO_KERNEL_STK
+
+	/* save the regfile */
+	SAVE_ALL_SYS
+.endm
+
 /*--------------------------------------------------------------
  * Save all registers used by Exceptions (TLB Miss, Prot-V, Mem err etc)
  * Requires SP to be already switched to kernel mode Stack

arch/arc/include/asm/io.h

@@ -100,6 +100,10 @@ static inline void __raw_writel(u32 w, volatile void __iomem *addr)
 
 }
 
+#define readb_relaxed readb
+#define readw_relaxed readw
+#define readl_relaxed readl
+
 #include <asm-generic/io.h>
 
 #endif /* _ASM_ARC_IO_H */

arch/arc/include/asm/irqflags.h

@@ -157,13 +157,6 @@ static inline void arch_unmask_irq(unsigned int irq)
 	flag	\scratch
 .endm
 
-.macro IRQ_DISABLE_SAVE  scratch, save
-	lr	\scratch, [status32]
-	mov	\save, \scratch		/* Make a copy */
-	bic	\scratch, \scratch, (STATUS_E1_MASK | STATUS_E2_MASK)
-	flag	\scratch
-.endm
-
 .macro IRQ_ENABLE  scratch
 	lr	\scratch, [status32]
 	or	\scratch, \scratch, (STATUS_E1_MASK | STATUS_E2_MASK)

arch/arc/include/asm/mmu.h

@@ -32,6 +32,8 @@
 /* Error code if probe fails */
 #define TLB_LKUP_ERR		0x80000000
 
+#define TLB_DUP_ERR	(TLB_LKUP_ERR | 0x00000001)
+
 /* TLB Commands */
 #define TLBWrite    0x1
 #define TLBRead     0x2
@@ -46,21 +48,18 @@
 #ifndef __ASSEMBLY__
 
 typedef struct {
-	unsigned long asid;	/* Pvt Addr-Space ID for mm */
-#ifdef CONFIG_ARC_TLB_DBG
-	struct task_struct *tsk;
-#endif
+	unsigned long asid;	/* 8 bit MMU PID + Generation cycle */
 } mm_context_t;
 
 #ifdef CONFIG_ARC_DBG_TLB_PARANOIA
-void tlb_paranoid_check(unsigned int pid_sw, unsigned long address);
+void tlb_paranoid_check(unsigned int mm_asid, unsigned long address);
 #else
 #define tlb_paranoid_check(a, b)
 #endif
 
 void arc_mmu_init(void);
 extern char *arc_mmu_mumbojumbo(int cpu_id, char *buf, int len);
-void __init read_decode_mmu_bcr(void);
+void read_decode_mmu_bcr(void);
 
 #endif	/* !__ASSEMBLY__ */
 

arch/arc/include/asm/mmu_context.h

@@ -34,95 +34,65 @@
  * When it reaches max 255, the allocation cycle starts afresh by flushing
  * the entire TLB and wrapping ASID back to zero.
  *
- * For book-keeping, Linux uses a couple of data-structures:
- *  -mm_struct has an @asid field to keep a note of task's ASID (needed at the
- *   time of say switch_mm( )
- *  -An array of mm structs @asid_mm_map[] for asid->mm the reverse mapping,
- *  given an ASID, finding the mm struct associated.
- *
- * The round-robin allocation algorithm allows for ASID stealing.
- * If asid tracker is at "x-1", a new req will allocate "x", even if "x" was
- * already assigned to another (switched-out) task. Obviously the prev owner
- * is marked with an invalid ASID to make it request for a new ASID when it
- * gets scheduled next time. However its TLB entries (with ASID "x") could
- * exist, which must be cleared before the same ASID is used by the new owner.
- * Flushing them would be plausible but costly solution. Instead we force a
- * allocation policy quirk, which ensures that a stolen ASID won't have any
- * TLB entries associates, alleviating the need to flush.
- * The quirk essentially is not allowing ASID allocated in prev cycle
- * to be used past a roll-over in the next cycle.
- * When this happens (i.e. task ASID > asid tracker), task needs to refresh
- * its ASID, aligning it to current value of tracker. If the task doesn't get
- * scheduled past a roll-over, hence its ASID is not yet realigned with
- * tracker, such ASID is anyways safely reusable because it is
- * gauranteed that TLB entries with that ASID wont exist.
+ * A new allocation cycle, post rollover, could potentially reassign an ASID
+ * to a different task. Thus the rule is to refresh the ASID in a new cycle.
+ * The 32 bit @asid_cache (and mm->asid) have 8 bits MMU PID and rest 24 bits
+ * serve as cycle/generation indicator and natural 32 bit unsigned math
+ * automagically increments the generation when lower 8 bits rollover.
  */
 
-#define FIRST_ASID  0
-#define MAX_ASID    255			/* 8 bit PID field in PID Aux reg */
-#define NO_ASID     (MAX_ASID + 1)	/* ASID Not alloc to mmu ctxt */
-#define NUM_ASID    ((MAX_ASID - FIRST_ASID) + 1)
+#define MM_CTXT_ASID_MASK	0x000000ff /* MMU PID reg :8 bit PID */
+#define MM_CTXT_CYCLE_MASK	(~MM_CTXT_ASID_MASK)
+
+#define MM_CTXT_FIRST_CYCLE	(MM_CTXT_ASID_MASK + 1)
+#define MM_CTXT_NO_ASID		0UL
 
-/* ASID to mm struct mapping */
-extern struct mm_struct *asid_mm_map[NUM_ASID + 1];
+#define hw_pid(mm)		(mm->context.asid & MM_CTXT_ASID_MASK)
 
-extern int asid_cache;
+extern unsigned int asid_cache;
 
 /*
- * Assign a new ASID to task. If the task already has an ASID, it is
- * relinquished.
+ * Get a new ASID if task doesn't have a valid one (unalloc or from prev cycle)
+ * Also set the MMU PID register to existing/updated ASID
  */
 static inline void get_new_mmu_context(struct mm_struct *mm)
 {
-	struct mm_struct *prev_owner;
 	unsigned long flags;
 
 	local_irq_save(flags);
 
 	/*
-	 * Relinquish the currently owned ASID (if any).
-	 * Doing unconditionally saves a cmp-n-branch; for already unused
-	 * ASID slot, the value was/remains NULL
+	 * Move to new ASID if it was not from current alloc-cycle/generation.
+	 * This is done by ensuring that the generation bits in both mm->ASID
+	 * and cpu's ASID counter are exactly same.
+	 *
+	 * Note: Callers needing new ASID unconditionally, independent of
+	 * 	 generation, e.g. local_flush_tlb_mm() for forking  parent,
+	 * 	 first need to destroy the context, setting it to invalid
+	 * 	 value.
 	 */
-	asid_mm_map[mm->context.asid] = (struct mm_struct *)NULL;
+	if (!((mm->context.asid ^ asid_cache) & MM_CTXT_CYCLE_MASK))
+		goto set_hw;
+
+	/* move to new ASID and handle rollover */
+	if (unlikely(!(++asid_cache & MM_CTXT_ASID_MASK))) {
 
-	/* move to new ASID */
-	if (++asid_cache > MAX_ASID) {	/* ASID roll-over */
-		asid_cache = FIRST_ASID;
 		flush_tlb_all();
-	}
 
-	/*
-	 * Is next ASID already owned by some-one else (we are stealing it).
-	 * If so, let the orig owner be aware of this, so when it runs, it
-	 * asks for a brand new ASID. This would only happen for a long-lived
-	 * task with ASID from prev allocation cycle (before ASID roll-over).
-	 *
-	 * This might look wrong - if we are re-using some other task's ASID,
-	 * won't we use it's stale TLB entries too. Actually switch_mm( ) takes
-	 * care of such a case: it ensures that task with ASID from prev alloc
-	 * cycle, when scheduled will refresh it's ASID: see switch_mm( ) below
-	 * The stealing scenario described here will only happen if that task
-	 * didn't get a chance to refresh it's ASID - implying stale entries
-	 * won't exist.
-	 */
-	prev_owner = asid_mm_map[asid_cache];
-	if (prev_owner)
-		prev_owner->context.asid = NO_ASID;
+		/*
+		 * Above checke for rollover of 8 bit ASID in 32 bit container.
+		 * If the container itself wrapped around, set it to a non zero
+		 * "generation" to distinguish from no context
+		 */
+		if (!asid_cache)
+			asid_cache = MM_CTXT_FIRST_CYCLE;
+	}
 
 	/* Assign new ASID to tsk */
-	asid_mm_map[asid_cache] = mm;
 	mm->context.asid = asid_cache;
 
-#ifdef CONFIG_ARC_TLB_DBG
-	pr_info("ARC_TLB_DBG: NewMM=0x%x OldMM=0x%x task_struct=0x%x Task: %s,"
-	       " pid:%u, assigned asid:%lu\n",
-	       (unsigned int)mm, (unsigned int)prev_owner,
-	       (unsigned int)(mm->context.tsk), (mm->context.tsk)->comm,
-	       (mm->context.tsk)->pid, mm->context.asid);
-#endif
-
-	write_aux_reg(ARC_REG_PID, asid_cache | MMU_ENABLE);
+set_hw:
+	write_aux_reg(ARC_REG_PID, hw_pid(mm) | MMU_ENABLE);
 
 	local_irq_restore(flags);
 }
@@ -134,10 +104,7 @@ static inline void get_new_mmu_context(struct mm_struct *mm)
 static inline int
 init_new_context(struct task_struct *tsk, struct mm_struct *mm)
 {
-	mm->context.asid = NO_ASID;
-#ifdef CONFIG_ARC_TLB_DBG
-	mm->context.tsk = tsk;
-#endif
+	mm->context.asid = MM_CTXT_NO_ASID;
 	return 0;
 }
 
@@ -152,40 +119,21 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 	write_aux_reg(ARC_REG_SCRATCH_DATA0, next->pgd);
 #endif
 
-	/*
-	 * Get a new ASID if task doesn't have a valid one. Possible when
-	 *  -task never had an ASID (fresh after fork)
-	 *  -it's ASID was stolen - past an ASID roll-over.
-	 *  -There's a third obscure scenario (if this task is running for the
-	 *   first time afer an ASID rollover), where despite having a valid
-	 *   ASID, we force a get for new ASID - see comments at top.
-	 *
-	 * Both the non-alloc scenario and first-use-after-rollover can be
-	 * detected using the single condition below:  NO_ASID = 256
-	 * while asid_cache is always a valid ASID value (0-255).
-	 */
-	if (next->context.asid > asid_cache) {
-		get_new_mmu_context(next);
-	} else {
-		/*
-		 * XXX: This will never happen given the chks above
-		 * BUG_ON(next->context.asid > MAX_ASID);
-		 */
-		write_aux_reg(ARC_REG_PID, next->context.asid | MMU_ENABLE);
-	}
-
+	get_new_mmu_context(next);
 }
 
+/*
+ * Called at the time of execve() to get a new ASID
+ * Note the subtlety here: get_new_mmu_context() behaves differently here
+ * vs. in switch_mm(). Here it always returns a new ASID, because mm has
+ * an unallocated "initial" value, while in latter, it moves to a new ASID,
+ * only if it was unallocated
+ */
+#define activate_mm(prev, next)		switch_mm(prev, next, NULL)
+
 static inline void destroy_context(struct mm_struct *mm)
 {
-	unsigned long flags;
-
-	local_irq_save(flags);
-
-	asid_mm_map[mm->context.asid] = NULL;
-	mm->context.asid = NO_ASID;
-
-	local_irq_restore(flags);
+	mm->context.asid = MM_CTXT_NO_ASID;
 }
 
 /* it seemed that deactivate_mm( ) is a reasonable place to do book-keeping
@@ -197,17 +145,6 @@ static inline void destroy_context(struct mm_struct *mm)
  */
 #define deactivate_mm(tsk, mm)   do { } while (0)
 
-static inline void activate_mm(struct mm_struct *prev, struct mm_struct *next)
-{
-#ifndef CONFIG_SMP
-	write_aux_reg(ARC_REG_SCRATCH_DATA0, next->pgd);
-#endif
-
-	/* Unconditionally get a new ASID */
-	get_new_mmu_context(next);
-
-}
-
 #define enter_lazy_tlb(mm, tsk)
 
 #endif /* __ASM_ARC_MMU_CONTEXT_H */