Merge tag 'csky-for-linus-5.3-rc1' of git://github.com/c-sky/csky-linux

Pull arch/csky pupdates from Guo Ren:
 "This round of csky subsystem gives two features (ASID algorithm
  update, Perf pmu record support) and some fixups.

  ASID updates:
   - Revert mmu ASID mechanism
   - Add new asid lib code from arm
   - Use generic asid algorithm to implement switch_mm
   - Improve tlb operation with help of asid

  Perf pmu record support:
   - Init pmu as a device
   - Add count-width property for csky pmu
   - Add pmu interrupt support
   - Fix perf record in kernel/user space
   - dt-bindings: Add csky PMU bindings

  Fixes:
   - Fixup no panic in kernel for some traps
   - Fixup some error count in 810 & 860.
   - Fixup abiv1 memset error"

* tag 'csky-for-linus-5.3-rc1' of git://github.com/c-sky/csky-linux:
  csky: Fixup abiv1 memset error
  csky: Improve tlb operation with help of asid
  csky: Use generic asid algorithm to implement switch_mm
  csky: Add new asid lib code from arm
  csky: Revert mmu ASID mechanism
  dt-bindings: csky: Add csky PMU bindings
  dt-bindings: interrupt-controller: Update csky mpintc
  csky: Fixup some error count in 810 & 860.
  csky: Fix perf record in kernel/user space
  csky: Add pmu interrupt support
  csky: Add count-width property for csky pmu
  csky: Init pmu as a device
  csky: Fixup no panic in kernel for some traps
  csky: Select intc & timer drivers

5 files changed, 332 insertions, 145 deletions

diff --git a/arch/csky/mm/Makefile b/arch/csky/mm/Makefile
index 4eebebdcd1bf..c94ef6481098 100644
--- a/arch/csky/mm/Makefile
+++ b/arch/csky/mm/Makefile
@@ -12,3 +12,5 @@ obj-y +=			init.o
 obj-y +=			ioremap.o
 obj-y +=			syscache.o
 obj-y +=			tlb.o
+obj-y +=			asid.o
+obj-y +=			context.o
diff --git a/arch/csky/mm/asid.c b/arch/csky/mm/asid.c
new file mode 100644
index 000000000000..b2e914745c1d
--- /dev/null
+++ b/arch/csky/mm/asid.c
@@ -0,0 +1,189 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Generic ASID allocator.
+ *
+ * Based on arch/arm/mm/context.c
+ *
+ * Copyright (C) 2002-2003 Deep Blue Solutions Ltd, all rights reserved.
+ * Copyright (C) 2012 ARM Ltd.
+ */
+
+#include <linux/slab.h>
+#include <linux/mm_types.h>
+
+#include <asm/asid.h>
+
+#define reserved_asid(info, cpu) *per_cpu_ptr((info)->reserved, cpu)
+
+#define ASID_MASK(info)			(~GENMASK((info)->bits - 1, 0))
+#define ASID_FIRST_VERSION(info)	(1UL << ((info)->bits))
+
+#define asid2idx(info, asid)		(((asid) & ~ASID_MASK(info)) >> (info)->ctxt_shift)
+#define idx2asid(info, idx)		(((idx) << (info)->ctxt_shift) & ~ASID_MASK(info))
+
+static void flush_context(struct asid_info *info)
+{
+	int i;
+	u64 asid;
+
+	/* Update the list of reserved ASIDs and the ASID bitmap. */
+	bitmap_clear(info->map, 0, NUM_CTXT_ASIDS(info));
+
+	for_each_possible_cpu(i) {
+		asid = atomic64_xchg_relaxed(&active_asid(info, i), 0);
+		/*
+		 * If this CPU has already been through a
+		 * rollover, but hasn't run another task in
+		 * the meantime, we must preserve its reserved
+		 * ASID, as this is the only trace we have of
+		 * the process it is still running.
+		 */
+		if (asid == 0)
+			asid = reserved_asid(info, i);
+		__set_bit(asid2idx(info, asid), info->map);
+		reserved_asid(info, i) = asid;
+	}
+
+	/*
+	 * Queue a TLB invalidation for each CPU to perform on next
+	 * context-switch
+	 */
+	cpumask_setall(&info->flush_pending);
+}
+
+static bool check_update_reserved_asid(struct asid_info *info, u64 asid,
+				       u64 newasid)
+{
+	int cpu;
+	bool hit = false;
+
+	/*
+	 * Iterate over the set of reserved ASIDs looking for a match.
+	 * If we find one, then we can update our mm to use newasid
+	 * (i.e. the same ASID in the current generation) but we can't
+	 * exit the loop early, since we need to ensure that all copies
+	 * of the old ASID are updated to reflect the mm. Failure to do
+	 * so could result in us missing the reserved ASID in a future
+	 * generation.
+	 */
+	for_each_possible_cpu(cpu) {
+		if (reserved_asid(info, cpu) == asid) {
+			hit = true;
+			reserved_asid(info, cpu) = newasid;
+		}
+	}
+
+	return hit;
+}
+
+static u64 new_context(struct asid_info *info, atomic64_t *pasid,
+		       struct mm_struct *mm)
+{
+	static u32 cur_idx = 1;
+	u64 asid = atomic64_read(pasid);
+	u64 generation = atomic64_read(&info->generation);
+
+	if (asid != 0) {
+		u64 newasid = generation | (asid & ~ASID_MASK(info));
+
+		/*
+		 * If our current ASID was active during a rollover, we
+		 * can continue to use it and this was just a false alarm.
+		 */
+		if (check_update_reserved_asid(info, asid, newasid))
+			return newasid;
+
+		/*
+		 * We had a valid ASID in a previous life, so try to re-use
+		 * it if possible.
+		 */
+		if (!__test_and_set_bit(asid2idx(info, asid), info->map))
+			return newasid;
+	}
+
+	/*
+	 * Allocate a free ASID. If we can't find one, take a note of the
+	 * currently active ASIDs and mark the TLBs as requiring flushes.  We
+	 * always count from ASID #2 (index 1), as we use ASID #0 when setting
+	 * a reserved TTBR0 for the init_mm and we allocate ASIDs in even/odd
+	 * pairs.
+	 */
+	asid = find_next_zero_bit(info->map, NUM_CTXT_ASIDS(info), cur_idx);
+	if (asid != NUM_CTXT_ASIDS(info))
+		goto set_asid;
+
+	/* We're out of ASIDs, so increment the global generation count */
+	generation = atomic64_add_return_relaxed(ASID_FIRST_VERSION(info),
+						 &info->generation);
+	flush_context(info);
+
+	/* We have more ASIDs than CPUs, so this will always succeed */
+	asid = find_next_zero_bit(info->map, NUM_CTXT_ASIDS(info), 1);
+
+set_asid:
+	__set_bit(asid, info->map);
+	cur_idx = asid;
+	cpumask_clear(mm_cpumask(mm));
+	return idx2asid(info, asid) | generation;
+}
+
+/*
+ * Generate a new ASID for the context.
+ *
+ * @pasid: Pointer to the current ASID batch allocated. It will be updated
+ * with the new ASID batch.
+ * @cpu: current CPU ID. Must have been acquired through get_cpu()
+ */
+void asid_new_context(struct asid_info *info, atomic64_t *pasid,
+		      unsigned int cpu, struct mm_struct *mm)
+{
+	unsigned long flags;
+	u64 asid;
+
+	raw_spin_lock_irqsave(&info->lock, flags);
+	/* Check that our ASID belongs to the current generation. */
+	asid = atomic64_read(pasid);
+	if ((asid ^ atomic64_read(&info->generation)) >> info->bits) {
+		asid = new_context(info, pasid, mm);
+		atomic64_set(pasid, asid);
+	}
+
+	if (cpumask_test_and_clear_cpu(cpu, &info->flush_pending))
+		info->flush_cpu_ctxt_cb();
+
+	atomic64_set(&active_asid(info, cpu), asid);
+	cpumask_set_cpu(cpu, mm_cpumask(mm));
+	raw_spin_unlock_irqrestore(&info->lock, flags);
+}
+
+/*
+ * Initialize the ASID allocator
+ *
+ * @info: Pointer to the asid allocator structure
+ * @bits: Number of ASIDs available
+ * @asid_per_ctxt: Number of ASIDs to allocate per-context. ASIDs are
+ * allocated contiguously for a given context. This value should be a power of
+ * 2.
+ */
+int asid_allocator_init(struct asid_info *info,
+			u32 bits, unsigned int asid_per_ctxt,
+			void (*flush_cpu_ctxt_cb)(void))
+{
+	info->bits = bits;
+	info->ctxt_shift = ilog2(asid_per_ctxt);
+	info->flush_cpu_ctxt_cb = flush_cpu_ctxt_cb;
+	/*
+	 * Expect allocation after rollover to fail if we don't have at least
+	 * one more ASID than CPUs. ASID #0 is always reserved.
+	 */
+	WARN_ON(NUM_CTXT_ASIDS(info) - 1 <= num_possible_cpus());
+	atomic64_set(&info->generation, ASID_FIRST_VERSION(info));
+	info->map = kcalloc(BITS_TO_LONGS(NUM_CTXT_ASIDS(info)),
+			    sizeof(*info->map), GFP_KERNEL);
+	if (!info->map)
+		return -ENOMEM;
+
+	raw_spin_lock_init(&info->lock);
+
+	return 0;
+}
diff --git a/arch/csky/mm/context.c b/arch/csky/mm/context.c
new file mode 100644
index 000000000000..0d95bdd93846
--- /dev/null
+++ b/arch/csky/mm/context.c
@@ -0,0 +1,46 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2018 Hangzhou C-SKY Microsystems co.,ltd.
+
+#include <linux/bitops.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/mm.h>
+
+#include <asm/asid.h>
+#include <asm/mmu_context.h>
+#include <asm/smp.h>
+#include <asm/tlbflush.h>
+
+static DEFINE_PER_CPU(atomic64_t, active_asids);
+static DEFINE_PER_CPU(u64, reserved_asids);
+
+struct asid_info asid_info;
+
+void check_and_switch_context(struct mm_struct *mm, unsigned int cpu)
+{
+	asid_check_context(&asid_info, &mm->context.asid, cpu, mm);
+}
+
+static void asid_flush_cpu_ctxt(void)
+{
+	local_tlb_invalid_all();
+}
+
+static int asids_init(void)
+{
+	BUG_ON(((1 << CONFIG_CPU_ASID_BITS) - 1) <= num_possible_cpus());
+
+	if (asid_allocator_init(&asid_info, CONFIG_CPU_ASID_BITS, 1,
+				asid_flush_cpu_ctxt))
+		panic("Unable to initialize ASID allocator for %lu ASIDs\n",
+		      NUM_ASIDS(&asid_info));
+
+	asid_info.active = &active_asids;
+	asid_info.reserved = &reserved_asids;
+
+	pr_info("ASID allocator initialised with %lu entries\n",
+		NUM_CTXT_ASIDS(&asid_info));
+
+	return 0;
+}
+early_initcall(asids_init);
diff --git a/arch/csky/mm/init.c b/arch/csky/mm/init.c
index 66e597053488..eb0dc9e5065f 100644
--- a/arch/csky/mm/init.c
+++ b/arch/csky/mm/init.c
@@ -114,8 +114,6 @@ void __init pre_mmu_init(void)
 	TLBMISS_HANDLER_SETUP_PGD(swapper_pg_dir);
 	TLBMISS_HANDLER_SETUP_PGD_KERNEL(swapper_pg_dir);
 
-	asid_cache(smp_processor_id()) = ASID_FIRST_VERSION;
-
 	/* Setup page mask to 4k */
 	write_mmu_pagemask(0);
 }
diff --git a/arch/csky/mm/tlb.c b/arch/csky/mm/tlb.c
index 08b8394e5b8f..eb3ba6c9c927 100644
--- a/arch/csky/mm/tlb.c
+++ b/arch/csky/mm/tlb.c
@@ -10,7 +10,12 @@
 #include <asm/pgtable.h>
 #include <asm/setup.h>
 
-#define CSKY_TLB_SIZE CONFIG_CPU_TLB_SIZE
+/*
+ * One C-SKY MMU TLB entry contain two PFN/page entry, ie:
+ * 1VPN -> 2PFN
+ */
+#define TLB_ENTRY_SIZE		(PAGE_SIZE * 2)
+#define TLB_ENTRY_SIZE_MASK	(PAGE_MASK << 1)
 
 void flush_tlb_all(void)
 {
@@ -19,201 +24,148 @@ void flush_tlb_all(void)
 
 void flush_tlb_mm(struct mm_struct *mm)
 {
-	int cpu = smp_processor_id();
-
-	if (cpu_context(cpu, mm) != 0)
-		drop_mmu_context(mm, cpu);
-
+#ifdef CONFIG_CPU_HAS_TLBI
+	asm volatile("tlbi.asids %0"::"r"(cpu_asid(mm)));
+#else
 	tlb_invalid_all();
+#endif
 }
 
+/*
+ * MMU operation regs only could invalid tlb entry in jtlb and we
+ * need change asid field to invalid I-utlb & D-utlb.
+ */
+#ifndef CONFIG_CPU_HAS_TLBI
 #define restore_asid_inv_utlb(oldpid, newpid) \
 do { \
-	if ((oldpid & ASID_MASK) == newpid) \
+	if (oldpid == newpid) \
 		write_mmu_entryhi(oldpid + 1); \
 	write_mmu_entryhi(oldpid); \
 } while (0)
+#endif
 
 void flush_tlb_range(struct vm_area_struct *vma, unsigned long start,
-			   unsigned long end)
+			unsigned long end)
 {
-	struct mm_struct *mm = vma->vm_mm;
-	int cpu = smp_processor_id();
-
-	if (cpu_context(cpu, mm) != 0) {
-		unsigned long size, flags;
-		int newpid = cpu_asid(cpu, mm);
-
-		local_irq_save(flags);
-		size = (end - start + (PAGE_SIZE - 1)) >> PAGE_SHIFT;
-		size = (size + 1) >> 1;
-		if (size <= CSKY_TLB_SIZE/2) {
-			start &= (PAGE_MASK << 1);
-			end += ((PAGE_SIZE << 1) - 1);
-			end &= (PAGE_MASK << 1);
-#ifdef CONFIG_CPU_HAS_TLBI
-			while (start < end) {
-				asm volatile("tlbi.vaas %0"
-					     ::"r"(start | newpid));
-				start += (PAGE_SIZE << 1);
-			}
-			sync_is();
-#else
-			{
-			int oldpid = read_mmu_entryhi();
-
-			while (start < end) {
-				int idx;
-
-				write_mmu_entryhi(start | newpid);
-				start += (PAGE_SIZE << 1);
-				tlb_probe();
-				idx = read_mmu_index();
-				if (idx >= 0)
-					tlb_invalid_indexed();
-			}
-			restore_asid_inv_utlb(oldpid, newpid);
-			}
-#endif
-		} else {
-			drop_mmu_context(mm, cpu);
-		}
-		local_irq_restore(flags);
-	}
-}
+	unsigned long newpid = cpu_asid(vma->vm_mm);
 
-void flush_tlb_kernel_range(unsigned long start, unsigned long end)
-{
-	unsigned long size, flags;
+	start &= TLB_ENTRY_SIZE_MASK;
+	end   += TLB_ENTRY_SIZE - 1;
+	end   &= TLB_ENTRY_SIZE_MASK;
 
-	local_irq_save(flags);
-	size = (end - start + (PAGE_SIZE - 1)) >> PAGE_SHIFT;
-	if (size <= CSKY_TLB_SIZE) {
-		start &= (PAGE_MASK << 1);
-		end += ((PAGE_SIZE << 1) - 1);
-		end &= (PAGE_MASK << 1);
 #ifdef CONFIG_CPU_HAS_TLBI
-		while (start < end) {
-			asm volatile("tlbi.vaas %0"::"r"(start));
-			start += (PAGE_SIZE << 1);
-		}
-		sync_is();
-#else
-		{
-		int oldpid = read_mmu_entryhi();
-
-		while (start < end) {
-			int idx;
-
-			write_mmu_entryhi(start);
-			start += (PAGE_SIZE << 1);
-			tlb_probe();
-			idx = read_mmu_index();
-			if (idx >= 0)
-				tlb_invalid_indexed();
-		}
-		restore_asid_inv_utlb(oldpid, 0);
-		}
-#endif
-	} else {
-		flush_tlb_all();
+	while (start < end) {
+		asm volatile("tlbi.vas %0"::"r"(start | newpid));
+		start += 2*PAGE_SIZE;
 	}
+	sync_is();
+#else
+	{
+	unsigned long flags, oldpid;
+
+	local_irq_save(flags);
+	oldpid = read_mmu_entryhi() & ASID_MASK;
+	while (start < end) {
+		int idx;
 
+		write_mmu_entryhi(start | newpid);
+		start += 2*PAGE_SIZE;
+		tlb_probe();
+		idx = read_mmu_index();
+		if (idx >= 0)
+			tlb_invalid_indexed();
+	}
+	restore_asid_inv_utlb(oldpid, newpid);
 	local_irq_restore(flags);
+	}
+#endif
 }
 
-void flush_tlb_page(struct vm_area_struct *vma, unsigned long page)
+void flush_tlb_kernel_range(unsigned long start, unsigned long end)
 {
-	int cpu = smp_processor_id();
-	int newpid = cpu_asid(cpu, vma->vm_mm);
-
-	if (!vma || cpu_context(cpu, vma->vm_mm) != 0) {
-		page &= (PAGE_MASK << 1);
+	start &= TLB_ENTRY_SIZE_MASK;
+	end   += TLB_ENTRY_SIZE - 1;
+	end   &= TLB_ENTRY_SIZE_MASK;
 
 #ifdef CONFIG_CPU_HAS_TLBI
-		asm volatile("tlbi.vaas %0"::"r"(page | newpid));
-		sync_is();
+	while (start < end) {
+		asm volatile("tlbi.vaas %0"::"r"(start));
+		start += 2*PAGE_SIZE;
+	}
+	sync_is();
 #else
-		{
-		int oldpid, idx;
-		unsigned long flags;
+	{
+	unsigned long flags, oldpid;
 
-		local_irq_save(flags);
-		oldpid = read_mmu_entryhi();
-		write_mmu_entryhi(page | newpid);
+	local_irq_save(flags);
+	oldpid = read_mmu_entryhi() & ASID_MASK;
+	while (start < end) {
+		int idx;
+
+		write_mmu_entryhi(start | oldpid);
+		start += 2*PAGE_SIZE;
 		tlb_probe();
 		idx = read_mmu_index();
 		if (idx >= 0)
 			tlb_invalid_indexed();
-
-		restore_asid_inv_utlb(oldpid, newpid);
-		local_irq_restore(flags);
-		}
-#endif
 	}
+	restore_asid_inv_utlb(oldpid, oldpid);
+	local_irq_restore(flags);
+	}
+#endif
 }
 
-/*
- * Remove one kernel space TLB entry.  This entry is assumed to be marked
- * global so we don't do the ASID thing.
- */
-void flush_tlb_one(unsigned long page)
+void flush_tlb_page(struct vm_area_struct *vma, unsigned long addr)
 {
-	int oldpid;
+	int newpid = cpu_asid(vma->vm_mm);
 
-	oldpid = read_mmu_entryhi();
-	page &= (PAGE_MASK << 1);
+	addr &= TLB_ENTRY_SIZE_MASK;
 
 #ifdef CONFIG_CPU_HAS_TLBI
-	page = page | (oldpid & 0xfff);
-	asm volatile("tlbi.vaas %0"::"r"(page));
+	asm volatile("tlbi.vas %0"::"r"(addr | newpid));
 	sync_is();
 #else
 	{
-	int idx;
+	int oldpid, idx;
 	unsigned long flags;
 
-	page = page | (oldpid & 0xff);
-
 	local_irq_save(flags);
-	write_mmu_entryhi(page);
+	oldpid = read_mmu_entryhi() & ASID_MASK;
+	write_mmu_entryhi(addr | newpid);
 	tlb_probe();
 	idx = read_mmu_index();
 	if (idx >= 0)
 		tlb_invalid_indexed();
-	restore_asid_inv_utlb(oldpid, oldpid);
+
+	restore_asid_inv_utlb(oldpid, newpid);
 	local_irq_restore(flags);
 	}
 #endif
 }
-EXPORT_SYMBOL(flush_tlb_one);
 
-/* show current 32 jtlbs */
-void show_jtlb_table(void)
+void flush_tlb_one(unsigned long addr)
 {
+	addr &= TLB_ENTRY_SIZE_MASK;
+
+#ifdef CONFIG_CPU_HAS_TLBI
+	asm volatile("tlbi.vaas %0"::"r"(addr));
+	sync_is();
+#else
+	{
+	int oldpid, idx;
 	unsigned long flags;
-	int entryhi, entrylo0, entrylo1;
-	int entry;
-	int oldpid;
 
 	local_irq_save(flags);
-	entry = 0;
-	pr_info("\n\n\n");
-
-	oldpid = read_mmu_entryhi();
-	while (entry < CSKY_TLB_SIZE) {
-		write_mmu_index(entry);
-		tlb_read();
-		entryhi = read_mmu_entryhi();
-		entrylo0 = read_mmu_entrylo0();
-		entrylo0 = entrylo0;
-		entrylo1 = read_mmu_entrylo1();
-		entrylo1 = entrylo1;
-		pr_info("jtlb[%d]:	entryhi - 0x%x;	entrylo0 - 0x%x;"
-			"	entrylo1 - 0x%x\n",
-			entry, entryhi, entrylo0, entrylo1);
-		entry++;
-	}
-	write_mmu_entryhi(oldpid);
+	oldpid = read_mmu_entryhi() & ASID_MASK;
+	write_mmu_entryhi(addr | oldpid);
+	tlb_probe();
+	idx = read_mmu_index();
+	if (idx >= 0)
+		tlb_invalid_indexed();
+
+	restore_asid_inv_utlb(oldpid, oldpid);
 	local_irq_restore(flags);
+	}
+#endif
 }
+EXPORT_SYMBOL(flush_tlb_one);