From 1efc959e0b28942d69d3084ecd4e9bde3beb9866 Mon Sep 17 00:00:00 2001
From: Rob Herring <rob.herring@calxeda.com>
Date: Mon, 26 Aug 2013 11:23:27 -0500
Subject: arc: use unflatten_and_copy_device_tree

Use the common unflatten_and_copy_device_tree to copy the built-in FDT
out of init section.

Signed-off-by: Rob Herring <rob.herring@calxeda.com>
Acked-by: Vineet Gupta <vgupta@synopsys.com>
Acked-by: Grant Likely <grant.likely@linaro.org>
---
 arch/arc/kernel/devtree.c | 15 ---------------
 arch/arc/kernel/setup.c   |  3 +--
 2 files changed, 1 insertion(+), 17 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/devtree.c b/arch/arc/kernel/devtree.c
index 2340af0e1d6f..eeb613aa44ff 100644
--- a/arch/arc/kernel/devtree.c
+++ b/arch/arc/kernel/devtree.c
@@ -100,18 +100,3 @@ struct machine_desc * __init setup_machine_fdt(void *dt)
 
 	return mdesc_best;
 }
-
-/*
- * Copy the flattened DT out of .init since unflattening doesn't copy strings
- * and the normal DT APIs refs them from orig flat DT
- */
-void __init copy_devtree(void)
-{
-	void *alloc = early_init_dt_alloc_memory_arch(
-			be32_to_cpu(initial_boot_params->totalsize), 64);
-	if (alloc) {
-		memcpy(alloc, initial_boot_params,
-				be32_to_cpu(initial_boot_params->totalsize));
-		initial_boot_params = alloc;
-	}
-}
diff --git a/arch/arc/kernel/setup.c b/arch/arc/kernel/setup.c
index 2c68bc7e6a78..710bf895dc2c 100644
--- a/arch/arc/kernel/setup.c
+++ b/arch/arc/kernel/setup.c
@@ -346,8 +346,7 @@ void __init setup_arch(char **cmdline_p)
 	setup_arch_memory();
 
 	/* copy flat DT out of .init and then unflatten it */
-	copy_devtree();
-	unflatten_device_tree();
+	unflatten_and_copy_device_tree();
 
 	/* Can be issue if someone passes cmd line arg "ro"
 	 * But that is unlikely so keeping it as it is
-- 
cgit v1.3-14-g43fede


From ad81fcb53033be98ab29c6cfb6ef0249a8e2ffaa Mon Sep 17 00:00:00 2001
From: Rob Herring <rob.herring@calxeda.com>
Date: Mon, 26 Aug 2013 10:12:22 -0500
Subject: arc: use early_init_dt_scan

Convert arc to use new early_init_dt_scan function.

Signed-off-by: Rob Herring <rob.herring@calxeda.com>
Acked-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/devtree.c | 14 +-------------
 1 file changed, 1 insertion(+), 13 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/devtree.c b/arch/arc/kernel/devtree.c
index eeb613aa44ff..1ab6f3569f60 100644
--- a/arch/arc/kernel/devtree.c
+++ b/arch/arc/kernel/devtree.c
@@ -27,7 +27,6 @@
  */
 struct machine_desc * __init setup_machine_fdt(void *dt)
 {
-	struct boot_param_header *devtree = dt;
 	struct machine_desc *mdesc = NULL, *mdesc_best = NULL;
 	unsigned int score, mdesc_score = ~1;
 	unsigned long dt_root;
@@ -36,11 +35,9 @@ struct machine_desc * __init setup_machine_fdt(void *dt)
 	char manufacturer[16];
 	unsigned long len;
 
-	/* check device tree validity */
-	if (be32_to_cpu(devtree->magic) != OF_DT_HEADER)
+	if (!early_init_dt_scan(dt))
 		return NULL;
 
-	initial_boot_params = devtree;
 	dt_root = of_get_flat_dt_root();
 
 	/*
@@ -85,15 +82,6 @@ struct machine_desc * __init setup_machine_fdt(void *dt)
 
 	pr_info("Board \"%s\" from %s (Manufacturer)\n", model, manufacturer);
 
-	/* Retrieve various information from the /chosen node */
-	of_scan_flat_dt(early_init_dt_scan_chosen, boot_command_line);
-
-	/* Initialize {size,address}-cells info */
-	of_scan_flat_dt(early_init_dt_scan_root, NULL);
-
-	/* Setup memory, calling early_init_dt_add_memory_arch */
-	of_scan_flat_dt(early_init_dt_scan_memory, NULL);
-
 	clk = of_get_flat_dt_prop(dt_root, "clock-frequency", &len);
 	if (clk)
 		arc_set_core_freq(of_read_ulong(clk, len/4));
-- 
cgit v1.3-14-g43fede


From 880beb8807814c7785680bcbdc81dcf076f585db Mon Sep 17 00:00:00 2001
From: Rob Herring <rob.herring@calxeda.com>
Date: Tue, 27 Aug 2013 21:43:12 -0500
Subject: arc: use common of_flat_dt_match_machine

Convert arc to use the common of_flat_dt_match_machine function.

Signed-off-by: Rob Herring <rob.herring@calxeda.com>
Acked-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/include/asm/mach_desc.h | 16 ++--------
 arch/arc/kernel/devtree.c        | 67 ++++++++++++----------------------------
 arch/arc/kernel/setup.c          |  2 +-
 3 files changed, 23 insertions(+), 62 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/include/asm/mach_desc.h b/arch/arc/include/asm/mach_desc.h
index d3e9c0a05be8..e8993a2be6c2 100644
--- a/arch/arc/include/asm/mach_desc.h
+++ b/arch/arc/include/asm/mach_desc.h
@@ -51,22 +51,12 @@ struct machine_desc {
 /*
  * Current machine - only accessible during boot.
  */
-extern struct machine_desc *machine_desc;
+extern const struct machine_desc *machine_desc;
 
 /*
  * Machine type table - also only accessible during boot
  */
-extern struct machine_desc __arch_info_begin[], __arch_info_end[];
-#define for_each_machine_desc(p)			\
-	for (p = __arch_info_begin; p < __arch_info_end; p++)
-
-static inline struct machine_desc *default_machine_desc(void)
-{
-	/* the default machine is the last one linked in */
-	if (__arch_info_end - 1 < __arch_info_begin)
-		return NULL;
-	return __arch_info_end - 1;
-}
+extern const struct machine_desc __arch_info_begin[], __arch_info_end[];
 
 /*
  * Set of macros to define architecture features.
@@ -81,6 +71,6 @@ __attribute__((__section__(".arch.info.init"))) = {	\
 #define MACHINE_END				\
 };
 
-extern struct machine_desc *setup_machine_fdt(void *dt);
+extern const struct machine_desc *setup_machine_fdt(void *dt);
 
 #endif
diff --git a/arch/arc/kernel/devtree.c b/arch/arc/kernel/devtree.c
index 1ab6f3569f60..d7b5322ffed5 100644
--- a/arch/arc/kernel/devtree.c
+++ b/arch/arc/kernel/devtree.c
@@ -18,6 +18,19 @@
 #include <asm/clk.h>
 #include <asm/mach_desc.h>
 
+static const void * __init arch_get_next_mach(const char *const **match)
+{
+	static const struct machine_desc *mdesc = __arch_info_begin;
+	const struct machine_desc *m = mdesc;
+
+	if (m >= __arch_info_end)
+		return NULL;
+
+	mdesc++;
+	*match = m->dt_compat;
+	return m;
+}
+
 /**
  * setup_machine_fdt - Machine setup when an dtb was passed to the kernel
  * @dt:		virtual address pointer to dt blob
@@ -25,66 +38,24 @@
  * If a dtb was passed to the kernel, then use it to choose the correct
  * machine_desc and to setup the system.
  */
-struct machine_desc * __init setup_machine_fdt(void *dt)
+const struct machine_desc * __init setup_machine_fdt(void *dt)
 {
-	struct machine_desc *mdesc = NULL, *mdesc_best = NULL;
-	unsigned int score, mdesc_score = ~1;
+	const struct machine_desc *mdesc;
 	unsigned long dt_root;
-	const char *model, *compat;
 	void *clk;
-	char manufacturer[16];
 	unsigned long len;
 
 	if (!early_init_dt_scan(dt))
 		return NULL;
 
-	dt_root = of_get_flat_dt_root();
-
-	/*
-	 * The kernel could be multi-platform enabled, thus could have many
-	 * "baked-in" machine descriptors. Search thru all for the best
-	 * "compatible" string match.
-	 */
-	for_each_machine_desc(mdesc) {
-		score = of_flat_dt_match(dt_root, mdesc->dt_compat);
-		if (score > 0 && score < mdesc_score) {
-			mdesc_best = mdesc;
-			mdesc_score = score;
-		}
-	}
-	if (!mdesc_best) {
-		const char *prop;
-		long size;
-
-		pr_err("\n unrecognized device tree list:\n[ ");
-
-		prop = of_get_flat_dt_prop(dt_root, "compatible", &size);
-		if (prop) {
-			while (size > 0) {
-				printk("'%s' ", prop);
-				size -= strlen(prop) + 1;
-				prop += strlen(prop) + 1;
-			}
-		}
-		printk("]\n\n");
-
+	mdesc = of_flat_dt_match_machine(NULL, arch_get_next_mach);
+	if (!mdesc)
 		machine_halt();
-	}
-
-	/* compat = "<manufacturer>,<model>" */
-	compat =  mdesc_best->dt_compat[0];
-
-	model = strchr(compat, ',');
-	if (model)
-		model++;
-
-	strlcpy(manufacturer, compat, model ? model - compat : strlen(compat));
-
-	pr_info("Board \"%s\" from %s (Manufacturer)\n", model, manufacturer);
 
+	dt_root = of_get_flat_dt_root();
 	clk = of_get_flat_dt_prop(dt_root, "clock-frequency", &len);
 	if (clk)
 		arc_set_core_freq(of_read_ulong(clk, len/4));
 
-	return mdesc_best;
+	return mdesc;
 }
diff --git a/arch/arc/kernel/setup.c b/arch/arc/kernel/setup.c
index 710bf895dc2c..8a9cf4e7403b 100644
--- a/arch/arc/kernel/setup.c
+++ b/arch/arc/kernel/setup.c
@@ -31,7 +31,7 @@
 int running_on_hw = 1;	/* vs. on ISS */
 
 char __initdata command_line[COMMAND_LINE_SIZE];
-struct machine_desc *machine_desc;
+const struct machine_desc *machine_desc;
 
 struct task_struct *_current_task[NR_CPUS];	/* For stack switching */
 
-- 
cgit v1.3-14-g43fede


From be5973117bd48624d9e7d7728d754f4c9a322193 Mon Sep 17 00:00:00 2001
From: Rob Herring <rob.herring@calxeda.com>
Date: Sun, 8 Sep 2013 14:26:28 -0500
Subject: arc: remove unnecessary prom.h includes

Remove unnecessary prom.h includes in preparation to remove prom.h.

Signed-off-by: Rob Herring <rob.herring@calxeda.com>
Acked-by: Vineet Gupta <vgupta@synopsys.com>
Acked-by: Grant Likely <grant.likely@linaro.org>
---
 arch/arc/kernel/devtree.c | 1 -
 arch/arc/kernel/setup.c   | 1 -
 2 files changed, 2 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/devtree.c b/arch/arc/kernel/devtree.c
index d7b5322ffed5..b6dc4e21fd32 100644
--- a/arch/arc/kernel/devtree.c
+++ b/arch/arc/kernel/devtree.c
@@ -14,7 +14,6 @@
 #include <linux/memblock.h>
 #include <linux/of.h>
 #include <linux/of_fdt.h>
-#include <asm/prom.h>
 #include <asm/clk.h>
 #include <asm/mach_desc.h>
 
diff --git a/arch/arc/kernel/setup.c b/arch/arc/kernel/setup.c
index 8a9cf4e7403b..f7d4a41d0629 100644
--- a/arch/arc/kernel/setup.c
+++ b/arch/arc/kernel/setup.c
@@ -21,7 +21,6 @@
 #include <asm/setup.h>
 #include <asm/page.h>
 #include <asm/irq.h>
-#include <asm/prom.h>
 #include <asm/unwind.h>
 #include <asm/clk.h>
 #include <asm/mach_desc.h>
-- 
cgit v1.3-14-g43fede


From 23d6d3db66151d9f8e7f2ea6b2c4fc6cc0097a66 Mon Sep 17 00:00:00 2001
From: Anoop Thomas Mathew <atm@profoundis.com>
Date: Fri, 20 Sep 2013 09:25:41 +0530
Subject: doc: typo on word accounting in kprobes.c in mutliple architectures

Signed-off-by: Anoop Thomas Mathew <atm@profoundis.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 arch/arc/kernel/kprobes.c     | 2 +-
 arch/ia64/kernel/kprobes.c    | 2 +-
 arch/powerpc/kernel/kprobes.c | 2 +-
 arch/s390/kernel/kprobes.c    | 2 +-
 arch/sparc/kernel/kprobes.c   | 2 +-
 5 files changed, 5 insertions(+), 5 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/kprobes.c b/arch/arc/kernel/kprobes.c
index 72f97822784a..7446c8dcb915 100644
--- a/arch/arc/kernel/kprobes.c
+++ b/arch/arc/kernel/kprobes.c
@@ -327,7 +327,7 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, unsigned long trapnr)
 		 */
 
 		/* We increment the nmissed count for accounting,
-		 * we can also use npre/npostfault count for accouting
+		 * we can also use npre/npostfault count for accounting
 		 * these specific fault cases.
 		 */
 		kprobes_inc_nmissed_count(cur);
diff --git a/arch/ia64/kernel/kprobes.c b/arch/ia64/kernel/kprobes.c
index f8280a766a78..074fde49c9e6 100644
--- a/arch/ia64/kernel/kprobes.c
+++ b/arch/ia64/kernel/kprobes.c
@@ -947,7 +947,7 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
 	case KPROBE_HIT_SSDONE:
 		/*
 		 * We increment the nmissed count for accounting,
-		 * we can also use npre/npostfault count for accouting
+		 * we can also use npre/npostfault count for accounting
 		 * these specific fault cases.
 		 */
 		kprobes_inc_nmissed_count(cur);
diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c
index 2156ea90eb54..90fab64d911d 100644
--- a/arch/powerpc/kernel/kprobes.c
+++ b/arch/powerpc/kernel/kprobes.c
@@ -429,7 +429,7 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
 	case KPROBE_HIT_SSDONE:
 		/*
 		 * We increment the nmissed count for accounting,
-		 * we can also use npre/npostfault count for accouting
+		 * we can also use npre/npostfault count for accounting
 		 * these specific fault cases.
 		 */
 		kprobes_inc_nmissed_count(cur);
diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c
index adbbe7f1cb0d..4f869b2e63f4 100644
--- a/arch/s390/kernel/kprobes.c
+++ b/arch/s390/kernel/kprobes.c
@@ -543,7 +543,7 @@ static int __kprobes kprobe_trap_handler(struct pt_regs *regs, int trapnr)
 	case KPROBE_HIT_SSDONE:
 		/*
 		 * We increment the nmissed count for accounting,
-		 * we can also use npre/npostfault count for accouting
+		 * we can also use npre/npostfault count for accounting
 		 * these specific fault cases.
 		 */
 		kprobes_inc_nmissed_count(p);
diff --git a/arch/sparc/kernel/kprobes.c b/arch/sparc/kernel/kprobes.c
index e72212148d2a..d3d5d388c557 100644
--- a/arch/sparc/kernel/kprobes.c
+++ b/arch/sparc/kernel/kprobes.c
@@ -349,7 +349,7 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
 	case KPROBE_HIT_SSDONE:
 		/*
 		 * We increment the nmissed count for accounting,
-		 * we can also use npre/npostfault count for accouting
+		 * we can also use npre/npostfault count for accounting
 		 * these specific fault cases.
 		 */
 		kprobes_inc_nmissed_count(cur);
-- 
cgit v1.3-14-g43fede


From 6855e95ce3256cdb5f4fbc988fe8ee925b051ef7 Mon Sep 17 00:00:00 2001
From: Christoph Lameter <cl@linux.com>
Date: Wed, 28 Aug 2013 19:48:15 +0000
Subject: arc: Replace __get_cpu_var uses

__get_cpu_var() is used for multiple purposes in the kernel source. One of them is
address calculation via the form &__get_cpu_var(x). This calculates the address for
the instance of the percpu variable of the current processor based on an offset.

Other use cases are for storing and retrieving data from the current processors percpu area.
__get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment.

__get_cpu_var() is defined as :

#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))

__get_cpu_var() always only does an address determination. However, store and retrieve operations
could use a segment prefix (or global register on other platforms) to avoid the address calculation.

this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use
optimized assembly code to read and write per cpu variables.

This patch converts __get_cpu_var into either an explicit address calculation using this_cpu_ptr()
or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided
and less registers are used when code is generated.

At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too.

The patchset includes passes over all arches as well. Once these operations are used throughout then
specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by
f.e. using a global register that may be set to the per cpu base.

Transformations done to __get_cpu_var()

1. Determine the address of the percpu instance of the current processor.

	DEFINE_PER_CPU(int, y);
	int *x = &__get_cpu_var(y);

    Converts to

	int *x = this_cpu_ptr(&y);

2. Same as #1 but this time an array structure is involved.

	DEFINE_PER_CPU(int, y[20]);
	int *x = __get_cpu_var(y);

    Converts to

	int *x = this_cpu_ptr(y);

3. Retrieve the content of the current processors instance of a per cpu variable.

	DEFINE_PER_CPU(int, u);
	int x = __get_cpu_var(y)

   Converts to

	int x = __this_cpu_read(y);

4. Retrieve the content of a percpu struct

	DEFINE_PER_CPU(struct mystruct, y);
	struct mystruct x = __get_cpu_var(y);

   Converts to

	memcpy(this_cpu_ptr(&x), y, sizeof(x));

5. Assignment to a per cpu variable

	DEFINE_PER_CPU(int, y)
	__get_cpu_var(y) = x;

   Converts to

	this_cpu_write(y, x);

6. Increment/Decrement etc of a per cpu variable

	DEFINE_PER_CPU(int, y);
	__get_cpu_var(y)++

   Converts to

	this_cpu_inc(y)

Acked-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Christoph Lameter <cl@linux.com>
---
 arch/arc/kernel/kprobes.c | 6 +++---
 arch/arc/kernel/time.c    | 2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/kprobes.c b/arch/arc/kernel/kprobes.c
index 72f97822784a..eb1c2ee5eaf0 100644
--- a/arch/arc/kernel/kprobes.c
+++ b/arch/arc/kernel/kprobes.c
@@ -87,13 +87,13 @@ static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
 
 static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
 {
-	__get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
+	__this_cpu_write(current_kprobe, kcb->prev_kprobe.kp);
 	kcb->kprobe_status = kcb->prev_kprobe.status;
 }
 
 static inline void __kprobes set_current_kprobe(struct kprobe *p)
 {
-	__get_cpu_var(current_kprobe) = p;
+	__this_cpu_write(current_kprobe, p);
 }
 
 static void __kprobes resume_execution(struct kprobe *p, unsigned long addr,
@@ -237,7 +237,7 @@ int __kprobes arc_kprobe_handler(unsigned long addr, struct pt_regs *regs)
 
 		return 1;
 	} else if (kprobe_running()) {
-		p = __get_cpu_var(current_kprobe);
+		p = __this_cpu_read(current_kprobe);
 		if (p->break_handler && p->break_handler(p, regs)) {
 			setup_singlestep(p, regs);
 			kcb->kprobe_status = KPROBE_HIT_SS;
diff --git a/arch/arc/kernel/time.c b/arch/arc/kernel/time.c
index 3fde7de3ea67..4c21dde2f6a9 100644
--- a/arch/arc/kernel/time.c
+++ b/arch/arc/kernel/time.c
@@ -206,7 +206,7 @@ static DEFINE_PER_CPU(struct clock_event_device, arc_clockevent_device) = {
 
 static irqreturn_t timer_irq_handler(int irq, void *dev_id)
 {
-	struct clock_event_device *clk = &__get_cpu_var(arc_clockevent_device);
+	struct clock_event_device *clk = this_cpu_ptr(&arc_clockevent_device);
 
 	arc_timer_event_ack(clk->mode == CLOCK_EVT_MODE_PERIODIC);
 	clk->event_handler(clk);
-- 
cgit v1.3-14-g43fede


From 8e457d6a75421372c7721388c46d21381fd5b451 Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Wed, 4 Sep 2013 16:13:35 +0530
Subject: ARC: Annotate some functions as static

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/setup.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/setup.c b/arch/arc/kernel/setup.c
index 2c68bc7e6a78..d9e15f16633e 100644
--- a/arch/arc/kernel/setup.c
+++ b/arch/arc/kernel/setup.c
@@ -37,8 +37,7 @@ struct task_struct *_current_task[NR_CPUS];	/* For stack switching */
 
 struct cpuinfo_arc cpuinfo_arc700[NR_CPUS];
 
-
-void read_arc_build_cfg_regs(void)
+static void read_arc_build_cfg_regs(void)
 {
 	struct bcr_perip uncached_space;
 	struct cpuinfo_arc *cpu = &cpuinfo_arc700[smp_processor_id()];
@@ -106,7 +105,7 @@ static const struct cpuinfo_data arc_cpu_tbl[] = {
 	{ {0x00, NULL		} }
 };
 
-char *arc_cpu_mumbojumbo(int cpu_id, char *buf, int len)
+static char *arc_cpu_mumbojumbo(int cpu_id, char *buf, int len)
 {
 	int n = 0;
 	struct cpuinfo_arc *cpu = &cpuinfo_arc700[cpu_id];
@@ -171,7 +170,7 @@ static const struct id_to_str mac_mul_nm[] = {
 	{0x6, "Dual 16x16 and 32x16"}
 };
 
-char *arc_extn_mumbojumbo(int cpu_id, char *buf, int len)
+static char *arc_extn_mumbojumbo(int cpu_id, char *buf, int len)
 {
 	int n = 0;
 	struct cpuinfo_arc *cpu = &cpuinfo_arc700[cpu_id];
@@ -234,7 +233,7 @@ char *arc_extn_mumbojumbo(int cpu_id, char *buf, int len)
 	return buf;
 }
 
-void arc_chk_ccms(void)
+static void arc_chk_ccms(void)
 {
 #if defined(CONFIG_ARC_HAS_DCCM) || defined(CONFIG_ARC_HAS_ICCM)
 	struct cpuinfo_arc *cpu = &cpuinfo_arc700[smp_processor_id()];
@@ -269,7 +268,7 @@ void arc_chk_ccms(void)
  * hardware has dedicated regs which need to be saved/restored on ctx-sw
  * (Single Precision uses core regs), thus kernel is kind of oblivious to it
  */
-void arc_chk_fpu(void)
+static void arc_chk_fpu(void)
 {
 	struct cpuinfo_arc *cpu = &cpuinfo_arc700[smp_processor_id()];
 
-- 
cgit v1.3-14-g43fede


From 064a6269245655b15054ad9783bca9bfd8dc1f9e Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Thu, 31 Oct 2013 13:53:54 +0530
Subject: ARC: use __weak instead of __attribute__((weak))

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/smp.c  | 2 +-
 arch/arc/kernel/time.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/smp.c b/arch/arc/kernel/smp.c
index bca3052c956d..7f0ab1ecd640 100644
--- a/arch/arc/kernel/smp.c
+++ b/arch/arc/kernel/smp.c
@@ -95,7 +95,7 @@ void __init smp_cpus_done(unsigned int max_cpus)
  *        If it turns out to be elaborate, it's better to code it in assembly
  *
  */
-void __attribute__((weak)) arc_platform_smp_wait_to_boot(int cpu)
+void __weak arc_platform_smp_wait_to_boot(int cpu)
 {
 	/*
 	 * As a hack for debugging - since debugger will single-step over the
diff --git a/arch/arc/kernel/time.c b/arch/arc/kernel/time.c
index 4c21dde2f6a9..0a9b6b289c4f 100644
--- a/arch/arc/kernel/time.c
+++ b/arch/arc/kernel/time.c
@@ -223,7 +223,7 @@ static struct irqaction arc_timer_irq = {
  * Setup the local event timer for @cpu
  * N.B. weak so that some exotic ARC SoCs can completely override it
  */
-void __attribute__((weak)) arc_local_timer_setup(unsigned int cpu)
+void __weak arc_local_timer_setup(unsigned int cpu)
 {
 	struct clock_event_device *clk = &per_cpu(arc_clockevent_device, cpu);
 
-- 
cgit v1.3-14-g43fede


From 21a63b56044706aa37637315dd27d9d465bbd5c4 Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Wed, 18 Sep 2013 16:25:40 +0530
Subject: ARC: Change calling convention of do_page_fault()

switch the args (address, pt_regs) to match with all the other "C"
exception handlers.

This removes the awkwardness in EV_ProtV for page fault vs. unaligned
access.

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/entry.S | 9 ++++-----
 arch/arc/mm/fault.c     | 2 +-
 arch/arc/mm/tlbex.S     | 4 ++--
 3 files changed, 7 insertions(+), 8 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/entry.S b/arch/arc/kernel/entry.S
index b908dde8a331..eb3f3c5a700a 100644
--- a/arch/arc/kernel/entry.S
+++ b/arch/arc/kernel/entry.S
@@ -337,9 +337,9 @@ ARC_ENTRY EV_TLBProtV
 	;  vineetg: Mar 6th: Random Seg Fault issue #1
 	;  ecr and efa were not saved in case an Intr sneaks in
 	;  after fake rtie
-	;
+
 	lr  r2, [ecr]
-	lr  r1, [efa]	; Faulting Data address
+	lr  r0, [efa]	; Faulting Data address
 
 	; --------(4) Return from CPU Exception Mode ---------
 	;  Fake a rtie, but rtie to next label
@@ -348,6 +348,8 @@ ARC_ENTRY EV_TLBProtV
 
 	FAKE_RET_FROM_EXCPN r9
 
+	mov   r1, sp
+
 	;------ (5) Type of Protection Violation? ----------
 	;
 	; ProtV Hardware Exception is triggered for Access Faults of 2 types
@@ -358,14 +360,11 @@ ARC_ENTRY EV_TLBProtV
 	bbit1 r2, ECR_C_BIT_PROTV_MISALIG_DATA, 4f
 
 	;========= (6a) Access Violation Processing ========
-	mov r0, sp              ; pt_regs
 	bl  do_page_fault
 	b   ret_from_exception
 
 	;========== (6b) Non aligned access ============
 4:
-	mov r0, r1
-	mov r1, sp              ; pt_regs
 
 #ifdef  CONFIG_ARC_MISALIGN_ACCESS
 	SAVE_CALLEE_SAVED_USER
diff --git a/arch/arc/mm/fault.c b/arch/arc/mm/fault.c
index 0c14d8a52683..9c69552350c4 100644
--- a/arch/arc/mm/fault.c
+++ b/arch/arc/mm/fault.c
@@ -52,7 +52,7 @@ bad_area:
 	return 1;
 }
 
-void do_page_fault(struct pt_regs *regs, unsigned long address)
+void do_page_fault(unsigned long address, struct pt_regs *regs)
 {
 	struct vm_area_struct *vma = NULL;
 	struct task_struct *tsk = current;
diff --git a/arch/arc/mm/tlbex.S b/arch/arc/mm/tlbex.S
index cf7d7d9ad695..3fcfdb38d242 100644
--- a/arch/arc/mm/tlbex.S
+++ b/arch/arc/mm/tlbex.S
@@ -369,8 +369,8 @@ do_slow_path_pf:
 	EXCEPTION_PROLOGUE
 
 	; ------- setup args for Linux Page fault Hanlder ---------
-	mov_s r0, sp
-	lr  r1, [efa]
+	mov_s r1, sp
+	lr    r0, [efa]
 
 	; We don't want exceptions to be disabled while the fault is handled.
 	; Now that we have saved the context we return from exception hence
-- 
cgit v1.3-14-g43fede


From 07ba69a46c3394fe6675483275337e5a79a959e0 Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Wed, 18 Sep 2013 18:08:01 +0530
Subject: ARC: Reduce #ifdef'ery for unaligned access emulation

Emulation not enabled is treated as if the fixup failed, so no need for
special #ifdef checks.

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/include/asm/unaligned.h | 3 ++-
 arch/arc/kernel/entry.S          | 4 ----
 arch/arc/kernel/traps.c          | 3 +--
 3 files changed, 3 insertions(+), 7 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/include/asm/unaligned.h b/arch/arc/include/asm/unaligned.h
index 60702f3751d2..3e5f071bc00c 100644
--- a/arch/arc/include/asm/unaligned.h
+++ b/arch/arc/include/asm/unaligned.h
@@ -22,7 +22,8 @@ static inline int
 misaligned_fixup(unsigned long address, struct pt_regs *regs,
 		 struct callee_regs *cregs)
 {
-	return 0;
+	/* Not fixed */
+	return 1;
 }
 #endif
 
diff --git a/arch/arc/kernel/entry.S b/arch/arc/kernel/entry.S
index eb3f3c5a700a..daa0e7990270 100644
--- a/arch/arc/kernel/entry.S
+++ b/arch/arc/kernel/entry.S
@@ -366,7 +366,6 @@ ARC_ENTRY EV_TLBProtV
 	;========== (6b) Non aligned access ============
 4:
 
-#ifdef  CONFIG_ARC_MISALIGN_ACCESS
 	SAVE_CALLEE_SAVED_USER
 	mov r2, sp              ; callee_regs
 
@@ -375,9 +374,6 @@ ARC_ENTRY EV_TLBProtV
 	; TBD: optimize - do this only if a callee reg was involved
 	; either a dst of emulated LD/ST or src with address-writeback
 	RESTORE_CALLEE_SAVED_USER
-#else
-	bl  do_misaligned_error
-#endif
 
 	b   ret_from_exception
 
diff --git a/arch/arc/kernel/traps.c b/arch/arc/kernel/traps.c
index e21692d2fdab..3eadfdabc322 100644
--- a/arch/arc/kernel/traps.c
+++ b/arch/arc/kernel/traps.c
@@ -84,19 +84,18 @@ DO_ERROR_INFO(SIGBUS, "Invalid Mem Access", do_memory_error, BUS_ADRERR)
 DO_ERROR_INFO(SIGTRAP, "Breakpoint Set", trap_is_brkpt, TRAP_BRKPT)
 DO_ERROR_INFO(SIGBUS, "Misaligned Access", do_misaligned_error, BUS_ADRALN)
 
-#ifdef CONFIG_ARC_MISALIGN_ACCESS
 /*
  * Entry Point for Misaligned Data access Exception, for emulating in software
  */
 int do_misaligned_access(unsigned long address, struct pt_regs *regs,
 			 struct callee_regs *cregs)
 {
+	/* If emulation not enabled, or failed, kill the task */
 	if (misaligned_fixup(address, regs, cregs) != 0)
 		return do_misaligned_error(address, regs);
 
 	return 0;
 }
-#endif
 
 /*
  * Entry point for miscll errors such as Nested Exceptions
-- 
cgit v1.3-14-g43fede


From 54c8bff14d604de23d0718eee59c5436a4703fe5 Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Thu, 12 Sep 2013 13:53:06 +0530
Subject: ARC: Reset the value of Interrupt Priority Register

In case bootloader has changed the priority of one/more IRQ lines

Reported-by: Noam Camus <noamc@ezchip.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/irq.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/irq.c b/arch/arc/kernel/irq.c
index 5fc92455da36..02684e1a0cf0 100644
--- a/arch/arc/kernel/irq.c
+++ b/arch/arc/kernel/irq.c
@@ -39,10 +39,14 @@ void arc_init_IRQ(void)
 	level_mask |= IS_ENABLED(CONFIG_ARC_IRQ5_LV2) << 5;
 	level_mask |= IS_ENABLED(CONFIG_ARC_IRQ6_LV2) << 6;
 
-	if (level_mask) {
+	/*
+	 * Write to register, even if no LV2 IRQs configured to reset it
+	 * in case bootloader had mucked with it
+	 */
+	write_aux_reg(AUX_IRQ_LEV, level_mask);
+
+	if (level_mask)
 		pr_info("Level-2 interrupts bitset %x\n", level_mask);
-		write_aux_reg(AUX_IRQ_LEV, level_mask);
-	}
 }
 
 /*
-- 
cgit v1.3-14-g43fede


From 0dafafc3ef42bad34fd446725cb9778c3bdd83a8 Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Fri, 6 Sep 2013 14:18:17 +0530
Subject: ARC: Add support for irqflags tracing and lockdep

Lockdep required a small fix to stacktrace API which was incorrectly
unwindign out of __switch_to for the current call frame.

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/Kconfig                |  6 ++++++
 arch/arc/include/asm/irqflags.h | 22 ++++++++++++++++++++++
 arch/arc/kernel/entry.S         | 11 +++++++++++
 arch/arc/kernel/stacktrace.c    |  4 +++-
 4 files changed, 42 insertions(+), 1 deletion(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig
index 91dbb2757afd..524e6fce2c89 100644
--- a/arch/arc/Kconfig
+++ b/arch/arc/Kconfig
@@ -35,6 +35,12 @@ config ARC
 	select PERF_USE_VMALLOC
 	select HAVE_DEBUG_STACKOVERFLOW
 
+config TRACE_IRQFLAGS_SUPPORT
+	def_bool y
+
+config LOCKDEP_SUPPORT
+	def_bool y
+
 config SCHED_OMIT_FRAME_POINTER
 	def_bool y
 
diff --git a/arch/arc/include/asm/irqflags.h b/arch/arc/include/asm/irqflags.h
index b68b53f458d1..cb7efc29f16f 100644
--- a/arch/arc/include/asm/irqflags.h
+++ b/arch/arc/include/asm/irqflags.h
@@ -151,16 +151,38 @@ static inline void arch_unmask_irq(unsigned int irq)
 
 #else
 
+#ifdef CONFIG_TRACE_IRQFLAGS
+
+.macro TRACE_ASM_IRQ_DISABLE
+	bl	trace_hardirqs_off
+.endm
+
+.macro TRACE_ASM_IRQ_ENABLE
+	bl	trace_hardirqs_on
+.endm
+
+#else
+
+.macro TRACE_ASM_IRQ_DISABLE
+.endm
+
+.macro TRACE_ASM_IRQ_ENABLE
+.endm
+
+#endif
+
 .macro IRQ_DISABLE  scratch
 	lr	\scratch, [status32]
 	bic	\scratch, \scratch, (STATUS_E1_MASK | STATUS_E2_MASK)
 	flag	\scratch
+	TRACE_ASM_IRQ_DISABLE
 .endm
 
 .macro IRQ_ENABLE  scratch
 	lr	\scratch, [status32]
 	or	\scratch, \scratch, (STATUS_E1_MASK | STATUS_E2_MASK)
 	flag	\scratch
+	TRACE_ASM_IRQ_ENABLE
 .endm
 
 #endif	/* __ASSEMBLY__ */
diff --git a/arch/arc/kernel/entry.S b/arch/arc/kernel/entry.S
index daa0e7990270..47d09d07f093 100644
--- a/arch/arc/kernel/entry.S
+++ b/arch/arc/kernel/entry.S
@@ -250,6 +250,14 @@ ARC_ENTRY handle_interrupt_level1
 	lr  r0, [icause1]
 	and r0, r0, 0x1f
 
+#ifdef CONFIG_TRACE_IRQFLAGS
+	; icause1 needs to be read early, before calling tracing, which
+	; can clobber scratch regs, hence use of stack to stash it
+	push r0
+	TRACE_ASM_IRQ_DISABLE
+	pop  r0
+#endif
+
 	bl.d  @arch_do_IRQ
 	mov r1, sp
 
@@ -570,6 +578,7 @@ resume_user_mode_begin:
 	; --- (Slow Path #2) pending signal  ---
 	mov r0, sp	; pt_regs for arg to do_signal()/do_notify_resume()
 
+	GET_CURR_THR_INFO_FLAGS   r9
 	bbit0  r9, TIF_SIGPENDING, .Lchk_notify_resume
 
 	; Normal Trap/IRQ entry only saves Scratch (caller-saved) regs
@@ -635,6 +644,8 @@ resume_kernel_mode:
 
 restore_regs :
 
+	TRACE_ASM_IRQ_ENABLE
+
 	lr	r10, [status32]
 
 	; Restore REG File. In case multiple Events outstanding,
diff --git a/arch/arc/kernel/stacktrace.c b/arch/arc/kernel/stacktrace.c
index f8b7d880304d..ab97b034922f 100644
--- a/arch/arc/kernel/stacktrace.c
+++ b/arch/arc/kernel/stacktrace.c
@@ -237,11 +237,13 @@ unsigned int get_wchan(struct task_struct *tsk)
  */
 void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace)
 {
+	/* Assumes @tsk is sleeping so unwinds from __switch_to */
 	arc_unwind_core(tsk, NULL, __collect_all_but_sched, trace);
 }
 
 void save_stack_trace(struct stack_trace *trace)
 {
-	arc_unwind_core(current, NULL, __collect_all, trace);
+	/* Pass NULL for task so it unwinds the current call frame */
+	arc_unwind_core(NULL, NULL, __collect_all, trace);
 }
 #endif
-- 
cgit v1.3-14-g43fede


From 3d01c1ce41fae0fb1d8e50589f9fb3b1fa376c1c Mon Sep 17 00:00:00 2001
From: Chen Gang <gang.chen@asianux.com>
Date: Thu, 24 Oct 2013 11:50:09 +0800
Subject: arc: kgdb: add default implementation for kgdb_roundup_cpus()

arc supports kgdb, but need update -- add function kgdb_roundup_cpus(),
or can not pass compiling. At present, add the simple generic one just
like other architectures(e.g. tile, mips ...).

The related error (with allmodconfig):

  kernel/built-in.o: In function `kgdb_cpu_enter':
  kernel/debug/debug_core.c:580: undefined reference to `kgdb_roundup_cpus'

Signed-off-by: Chen Gang <gang.chen@asianux.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/kgdb.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/kgdb.c b/arch/arc/kernel/kgdb.c
index a7698fb14818..a2ff5c5d1450 100644
--- a/arch/arc/kernel/kgdb.c
+++ b/arch/arc/kernel/kgdb.c
@@ -196,6 +196,18 @@ void kgdb_arch_set_pc(struct pt_regs *regs, unsigned long ip)
 	instruction_pointer(regs) = ip;
 }
 
+static void kgdb_call_nmi_hook(void *ignored)
+{
+	kgdb_nmicallback(raw_smp_processor_id(), NULL);
+}
+
+void kgdb_roundup_cpus(unsigned long flags)
+{
+	local_irq_enable();
+	smp_call_function(kgdb_call_nmi_hook, NULL, 0);
+	local_irq_disable();
+}
+
 struct kgdb_arch arch_kgdb_ops = {
 	/* breakpoint instruction: TRAP_S 0x3 */
 #ifdef CONFIG_CPU_BIG_ENDIAN
-- 
cgit v1.3-14-g43fede


From 8f5d221b0641a9a16311a6e5c056a5b583ef0cbb Mon Sep 17 00:00:00 2001
From: Chen Gang <gang.chen@asianux.com>
Date: Wed, 23 Oct 2013 10:16:38 +0800
Subject: arc: remove '__init' for first_lines_of_secondary()

first_lines_of_secondary() is a '__init' function, but it may be called
by __cpu_up() by _cpu_up() by cpu_up() which is a normal export symbol
function. So recommend to remove '__init'.

The related warning (with allmodconfig):

    MODPOST vmlinux.o
  WARNING: vmlinux.o(.text+0x315c): Section mismatch in reference from the function __cpu_up() to the function .init.text:first_lines_of_secondary()
  The function __cpu_up() references
  the function __init first_lines_of_secondary().
  This is often because __cpu_up lacks a __init
  annotation or the annotation of first_lines_of_secondary is wrong.

Signed-off-by: Chen Gang <gang.chen@asianux.com>
---
 arch/arc/include/asm/smp.h | 2 +-
 arch/arc/kernel/head.S     | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/include/asm/smp.h b/arch/arc/include/asm/smp.h
index c4fb211dcd25..eefc29f08cdb 100644
--- a/arch/arc/include/asm/smp.h
+++ b/arch/arc/include/asm/smp.h
@@ -30,7 +30,7 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask);
  * APIs provided by arch SMP code to rest of arch code
  */
 extern void __init smp_init_cpus(void);
-extern void __init first_lines_of_secondary(void);
+extern void first_lines_of_secondary(void);
 extern const char *arc_platform_smp_cpuinfo(void);
 
 /*
diff --git a/arch/arc/kernel/head.S b/arch/arc/kernel/head.S
index 0f944f024513..2c878e964a64 100644
--- a/arch/arc/kernel/head.S
+++ b/arch/arc/kernel/head.S
@@ -95,7 +95,7 @@ stext:
 ;----------------------------------------------------------------
 ;     First lines of code run by secondary before jumping to 'C'
 ;----------------------------------------------------------------
-	.section .init.text, "ax",@progbits
+	.section .text, "ax",@progbits
 	.type first_lines_of_secondary, @function
 	.globl first_lines_of_secondary
 
-- 
cgit v1.3-14-g43fede


From 4782f7f9aeba4dfd383c532e96bebf83ce0b9748 Mon Sep 17 00:00:00 2001
From: Chen Gang <gang.chen@asianux.com>
Date: Wed, 23 Oct 2013 11:02:51 +0800
Subject: arc: remove '__init' for get_hw_config_num_irq()

get_hw_config_num_irq() may be called by normal iss_model_init_smp()
which is a function pointer for 'init_smp' which may be called by
first_lines_of_secondary() which also need be normal too.

The related warning (with allmodconfig):

    MODPOST vmlinux.o
  WARNING: vmlinux.o(.text+0x5814): Section mismatch in reference from the function iss_model_init_smp() to the function .init.text:get_hw_config_num_irq()
  The function iss_model_init_smp() references
  the function __init get_hw_config_num_irq().
  This is often because iss_model_init_smp lacks a __init
  annotation or the annotation of get_hw_config_num_irq is wrong.

Signed-off-by: Chen Gang <gang.chen@asianux.com>
---
 arch/arc/include/asm/irq.h | 2 +-
 arch/arc/kernel/irq.c      | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/include/asm/irq.h b/arch/arc/include/asm/irq.h
index 548207f073da..291a70db68b8 100644
--- a/arch/arc/include/asm/irq.h
+++ b/arch/arc/include/asm/irq.h
@@ -19,7 +19,7 @@
 #include <asm-generic/irq.h>
 
 extern void arc_init_IRQ(void);
-extern int __init get_hw_config_num_irq(void);
+extern int get_hw_config_num_irq(void);
 
 void arc_local_timer_setup(unsigned int cpu);
 
diff --git a/arch/arc/kernel/irq.c b/arch/arc/kernel/irq.c
index 02684e1a0cf0..a4b141ee9a6a 100644
--- a/arch/arc/kernel/irq.c
+++ b/arch/arc/kernel/irq.c
@@ -150,7 +150,7 @@ void arch_do_IRQ(unsigned int irq, struct pt_regs *regs)
 	set_irq_regs(old_regs);
 }
 
-int __init get_hw_config_num_irq(void)
+int get_hw_config_num_irq(void)
 {
 	uint32_t val = read_aux_reg(ARC_REG_VECBASE_BCR);
 
-- 
cgit v1.3-14-g43fede


From 8f146d0204a874b6b1f5344ed578796e2fcd7e9c Mon Sep 17 00:00:00 2001
From: Chen Gang <gang.chen@asianux.com>
Date: Mon, 28 Oct 2013 11:00:38 +0800
Subject: arc: export symbol for save_stack_trace() in stacktrace.c

Need export its symbol just like other architectures done, or can not
pass compiling with allmodconfig, the related error:

    MODPOST 2994 modules
  ERROR: "save_stack_trace" [kernel/backtracetest.ko] undefined!
  ERROR: "save_stack_trace" [drivers/md/persistent-data/dm-persistent-data.ko] undefined!

Signed-off-by: Chen Gang <gang.chen@asianux.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/stacktrace.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/stacktrace.c b/arch/arc/kernel/stacktrace.c
index ab97b034922f..9ce47cfe2303 100644
--- a/arch/arc/kernel/stacktrace.c
+++ b/arch/arc/kernel/stacktrace.c
@@ -246,4 +246,5 @@ void save_stack_trace(struct stack_trace *trace)
 	/* Pass NULL for task so it unwinds the current call frame */
 	arc_unwind_core(NULL, NULL, __collect_all, trace);
 }
+EXPORT_SYMBOL_GPL(save_stack_trace);
 #endif
-- 
cgit v1.3-14-g43fede


From b6fe8e7c0189d017fdba90d1cd134337098c19c3 Mon Sep 17 00:00:00 2001
From: Chen Gang <gang.chen@asianux.com>
Date: Mon, 28 Oct 2013 11:49:47 +0800
Subject: arc: export symbol for pm_power_off in reset.c

Need export symbol for it, or can not pass compiling, the related error
with allmodconfig:

    MODPOST 2994 modules
  ERROR: "pm_power_off" [drivers/mfd/retu-mfd.ko] undefined!
  ERROR: "pm_power_off" [drivers/char/ipmi/ipmi_poweroff.ko] undefined!

Signed-off-by: Chen Gang <gang.chen@asianux.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/reset.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/reset.c b/arch/arc/kernel/reset.c
index e227a2b1c943..2768fa1e39b9 100644
--- a/arch/arc/kernel/reset.c
+++ b/arch/arc/kernel/reset.c
@@ -31,3 +31,4 @@ void machine_power_off(void)
 }
 
 void (*pm_power_off) (void) = NULL;
+EXPORT_SYMBOL(pm_power_off);
-- 
cgit v1.3-14-g43fede


From 5ea72a90261552ed5fdca35239feb6cba498301e Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Sun, 27 Oct 2013 14:49:02 +0530
Subject: ARC: [SMP] TLB flush

- Add mm_cpumask setting (aggregating only, unlike some other arches)
  used to restrict the TLB flush cross-calling

- cross-calling versions of TLB flush routines (thanks to Noam)

Signed-off-by: Noam Camus <noamc@ezchip.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/include/asm/mmu_context.h | 17 ++++++++-
 arch/arc/include/asm/tlbflush.h    | 11 ++++--
 arch/arc/kernel/smp.c              |  1 +
 arch/arc/mm/tlb.c                  | 73 ++++++++++++++++++++++++++++++++++++++
 4 files changed, 99 insertions(+), 3 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/include/asm/mmu_context.h b/arch/arc/include/asm/mmu_context.h
index 45f06f566b02..1fd467ef658f 100644
--- a/arch/arc/include/asm/mmu_context.h
+++ b/arch/arc/include/asm/mmu_context.h
@@ -80,7 +80,7 @@ static inline void get_new_mmu_context(struct mm_struct *mm)
 	/* move to new ASID and handle rollover */
 	if (unlikely(!(++asid_cpu(cpu) & MM_CTXT_ASID_MASK))) {
 
-		flush_tlb_all();
+		local_flush_tlb_all();
 
 		/*
 		 * Above checke for rollover of 8 bit ASID in 32 bit container.
@@ -131,6 +131,21 @@ static inline void destroy_context(struct mm_struct *mm)
 static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 			     struct task_struct *tsk)
 {
+	const int cpu = smp_processor_id();
+
+	/*
+	 * Note that the mm_cpumask is "aggregating" only, we don't clear it
+	 * for the switched-out task, unlike some other arches.
+	 * It is used to enlist cpus for sending TLB flush IPIs and not sending
+	 * it to CPUs where a task once ran-on, could cause stale TLB entry
+	 * re-use, specially for a multi-threaded task.
+	 * e.g. T1 runs on C1, migrates to C3. T2 running on C2 munmaps.
+	 *      For a non-aggregating mm_cpumask, IPI not sent C1, and if T1
+	 *      were to re-migrate to C1, it could access the unmapped region
+	 *      via any existing stale TLB entries.
+	 */
+	cpumask_set_cpu(cpu, mm_cpumask(next));
+
 #ifndef CONFIG_SMP
 	/* PGD cached in MMU reg to avoid 3 mem lookups: task->mm->pgd */
 	write_aux_reg(ARC_REG_SCRATCH_DATA0, next->pgd);
diff --git a/arch/arc/include/asm/tlbflush.h b/arch/arc/include/asm/tlbflush.h
index b2f9bc7f68c8..71c7b2e4b874 100644
--- a/arch/arc/include/asm/tlbflush.h
+++ b/arch/arc/include/asm/tlbflush.h
@@ -18,11 +18,18 @@ void local_flush_tlb_kernel_range(unsigned long start, unsigned long end);
 void local_flush_tlb_range(struct vm_area_struct *vma,
 			   unsigned long start, unsigned long end);
 
-/* XXX: Revisit for SMP */
+#ifndef CONFIG_SMP
 #define flush_tlb_range(vma, s, e)	local_flush_tlb_range(vma, s, e)
 #define flush_tlb_page(vma, page)	local_flush_tlb_page(vma, page)
 #define flush_tlb_kernel_range(s, e)	local_flush_tlb_kernel_range(s, e)
 #define flush_tlb_all()			local_flush_tlb_all()
 #define flush_tlb_mm(mm)		local_flush_tlb_mm(mm)
-
+#else
+extern void flush_tlb_range(struct vm_area_struct *vma, unsigned long start,
+							 unsigned long end);
+extern void flush_tlb_page(struct vm_area_struct *vma, unsigned long page);
+extern void flush_tlb_kernel_range(unsigned long start, unsigned long end);
+extern void flush_tlb_all(void);
+extern void flush_tlb_mm(struct mm_struct *mm);
+#endif /* CONFIG_SMP */
 #endif
diff --git a/arch/arc/kernel/smp.c b/arch/arc/kernel/smp.c
index 7f0ab1ecd640..41bc4c703f42 100644
--- a/arch/arc/kernel/smp.c
+++ b/arch/arc/kernel/smp.c
@@ -128,6 +128,7 @@ void start_kernel_secondary(void)
 	atomic_inc(&mm->mm_users);
 	atomic_inc(&mm->mm_count);
 	current->active_mm = mm;
+	cpumask_set_cpu(cpu, mm_cpumask(mm));
 
 	notify_cpu_starting(cpu);
 	set_cpu_online(cpu, true);
diff --git a/arch/arc/mm/tlb.c b/arch/arc/mm/tlb.c
index db0f0f823980..e1acf0ce5647 100644
--- a/arch/arc/mm/tlb.c
+++ b/arch/arc/mm/tlb.c
@@ -363,6 +363,79 @@ void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page)
 	local_irq_restore(flags);
 }
 
+#ifdef CONFIG_SMP
+
+struct tlb_args {
+	struct vm_area_struct *ta_vma;
+	unsigned long ta_start;
+	unsigned long ta_end;
+};
+
+static inline void ipi_flush_tlb_page(void *arg)
+{
+	struct tlb_args *ta = arg;
+
+	local_flush_tlb_page(ta->ta_vma, ta->ta_start);
+}
+
+static inline void ipi_flush_tlb_range(void *arg)
+{
+	struct tlb_args *ta = arg;
+
+	local_flush_tlb_range(ta->ta_vma, ta->ta_start, ta->ta_end);
+}
+
+static inline void ipi_flush_tlb_kernel_range(void *arg)
+{
+	struct tlb_args *ta = (struct tlb_args *)arg;
+
+	local_flush_tlb_kernel_range(ta->ta_start, ta->ta_end);
+}
+
+void flush_tlb_all(void)
+{
+	on_each_cpu((smp_call_func_t)local_flush_tlb_all, NULL, 1);
+}
+
+void flush_tlb_mm(struct mm_struct *mm)
+{
+	on_each_cpu_mask(mm_cpumask(mm), (smp_call_func_t)local_flush_tlb_mm,
+			 mm, 1);
+}
+
+void flush_tlb_page(struct vm_area_struct *vma, unsigned long uaddr)
+{
+	struct tlb_args ta = {
+		.ta_vma = vma,
+		.ta_start = uaddr
+	};
+
+	on_each_cpu_mask(mm_cpumask(vma->vm_mm), ipi_flush_tlb_page, &ta, 1);
+}
+
+void flush_tlb_range(struct vm_area_struct *vma, unsigned long start,
+		     unsigned long end)
+{
+	struct tlb_args ta = {
+		.ta_vma = vma,
+		.ta_start = start,
+		.ta_end = end
+	};
+
+	on_each_cpu_mask(mm_cpumask(vma->vm_mm), ipi_flush_tlb_range, &ta, 1);
+}
+
+void flush_tlb_kernel_range(unsigned long start, unsigned long end)
+{
+	struct tlb_args ta = {
+		.ta_start = start,
+		.ta_end = end
+	};
+
+	on_each_cpu(ipi_flush_tlb_kernel_range, &ta, 1);
+}
+#endif
+
 /*
  * Routine to create a TLB entry
  */
-- 
cgit v1.3-14-g43fede


From 57e26e57454fae4f1d15c2e9fa965b7a8046ab34 Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Fri, 1 Nov 2013 10:46:40 +0530
Subject: ARC: [SMP] Fix build failures for large NR_CPUS

ST.as only takes S9 (255) for offset. This was going out of range when
accessing a task_struct field with 4k NR_CPUS (due to 128b of coumaks
itself in there).

Workaround by using an intermediate register to do the address scaling.

There is some duplication of fix for ctx_sw.c and ctx_sw_asm.S however
given that C version will go away soon I'm not bothering to factor out
the common code.

Reported-by: Noam Camus <noamc@ezchip.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/ctx_sw.c     | 13 ++++++++++++-
 arch/arc/kernel/ctx_sw_asm.S | 11 +++++++++--
 2 files changed, 21 insertions(+), 3 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/ctx_sw.c b/arch/arc/kernel/ctx_sw.c
index 34410eb1a308..c14a5bea0c76 100644
--- a/arch/arc/kernel/ctx_sw.c
+++ b/arch/arc/kernel/ctx_sw.c
@@ -17,6 +17,8 @@
 #include <asm/asm-offsets.h>
 #include <linux/sched.h>
 
+#define KSP_WORD_OFF 	((TASK_THREAD + THREAD_KSP) / 4)
+
 struct task_struct *__sched
 __switch_to(struct task_struct *prev_task, struct task_struct *next_task)
 {
@@ -45,7 +47,16 @@ __switch_to(struct task_struct *prev_task, struct task_struct *next_task)
 #endif
 
 		/* set ksp of outgoing task in tsk->thread.ksp */
+#if KSP_WORD_OFF <= 255
 		"st.as   sp, [%3, %1]    \n\t"
+#else
+		/*
+		 * Workaround for NR_CPUS=4k
+		 * %1 is bigger than 255 (S9 offset for st.as)
+		 */
+		"add2    r24, %3, %1     \n\t"
+		"st      sp, [r24]       \n\t"
+#endif
 
 		"sync   \n\t"
 
@@ -97,7 +108,7 @@ __switch_to(struct task_struct *prev_task, struct task_struct *next_task)
 		/* FP/BLINK restore generated by gcc (standard func epilogue */
 
 		: "=r"(tmp)
-		: "n"((TASK_THREAD + THREAD_KSP) / 4), "r"(next), "r"(prev)
+		: "n"(KSP_WORD_OFF), "r"(next), "r"(prev)
 		: "blink"
 	);
 
diff --git a/arch/arc/kernel/ctx_sw_asm.S b/arch/arc/kernel/ctx_sw_asm.S
index d8972345e4c2..65690e7fcc8c 100644
--- a/arch/arc/kernel/ctx_sw_asm.S
+++ b/arch/arc/kernel/ctx_sw_asm.S
@@ -14,6 +14,8 @@
 #include <asm/asm-offsets.h>
 #include <asm/linkage.h>
 
+#define KSP_WORD_OFF 	((TASK_THREAD + THREAD_KSP) / 4)
+
 ;################### Low Level Context Switch ##########################
 
 	.section .sched.text,"ax",@progbits
@@ -28,8 +30,13 @@ __switch_to:
 	SAVE_CALLEE_SAVED_KERNEL
 
 	/* Save the now KSP in task->thread.ksp */
-	st.as  sp, [r0, (TASK_THREAD + THREAD_KSP)/4]
-
+#if KSP_WORD_OFF  <= 255
+	st.as  sp, [r0, KSP_WORD_OFF]
+#else
+	/* Workaround for NR_CPUS=4k as ST.as can only take s9 offset */
+	add2	r24, r0, KSP_WORD_OFF
+	st	sp, [r24]
+#endif
 	/*
 	* Return last task in r0 (return reg)
 	* On ARC, Return reg = First Arg reg = r0.
-- 
cgit v1.3-14-g43fede


From 7d0857a54aedbd47b3de503933d65ce462970bd6 Mon Sep 17 00:00:00 2001
From: Vineet Gupta <vgupta@synopsys.com>
Date: Mon, 9 Sep 2013 16:04:15 +0530
Subject: ARC: [SMP] Disallow RTSC

RTSC is strictly incore and must not be allowed in SMP configs

Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/Kconfig       | 6 +-----
 arch/arc/kernel/time.c | 7 ++++---
 2 files changed, 5 insertions(+), 8 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig
index fb4177e48260..5ede5460c806 100644
--- a/arch/arc/Kconfig
+++ b/arch/arc/Kconfig
@@ -136,9 +136,6 @@ if SMP
 config ARC_HAS_COH_CACHES
 	def_bool n
 
-config ARC_HAS_COH_RTSC
-	def_bool n
-
 config ARC_HAS_REENTRANT_IRQ_LV2
 	def_bool n
 
@@ -332,8 +329,7 @@ config ARC_HAS_RTSC
 	bool "Insn: RTSC (64-bit r/o cycle counter)"
 	default y
 	depends on ARC_CPU_REL_4_10
-	# if SMP, enable RTSC only if counter is coherent across cores
-	depends on !SMP || ARC_HAS_COH_RTSC
+	depends on !SMP
 
 endmenu   # "ARC CPU Configuration"
 
diff --git a/arch/arc/kernel/time.c b/arch/arc/kernel/time.c
index 0a9b6b289c4f..e5f3a837fb35 100644
--- a/arch/arc/kernel/time.c
+++ b/arch/arc/kernel/time.c
@@ -63,9 +63,10 @@
 
 int arc_counter_setup(void)
 {
-	/* RTSC insn taps into cpu clk, needs no setup */
-
-	/* For SMP, only allowed if cross-core-sync, hence usable as cs */
+	/*
+	 * For SMP this needs to be 0. However Kconfig glue doesn't
+	 * enable this option for SMP configs
+	 */
 	return 1;
 }
 
-- 
cgit v1.3-14-g43fede


From 06b93c347c93d53f5cd76c3efa277ea114cb2000 Mon Sep 17 00:00:00 2001
From: Jiang Liu <jiang.liu@huawei.com>
Date: Thu, 12 Sep 2013 00:07:07 +0800
Subject: smp, ARC: kill SMP single function call interrupt

Commit 9a46ad6d6df3b54 "smp: make smp_call_function_many() use logic
similar to smp_call_function_single()" has unified the way to handle
single and multiple cross-CPU function calls. Now only one interrupt
is needed for architecture specific code to support generic SMP function
call interfaces, so kill the redundant single function call interrupt.

Signed-off-by: Jiang Liu <jiang.liu@huawei.com>
Cc: Jiang Liu <liuj97@gmail.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/smp.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/smp.c b/arch/arc/kernel/smp.c
index 41bc4c703f42..c2f9ebbc38f6 100644
--- a/arch/arc/kernel/smp.c
+++ b/arch/arc/kernel/smp.c
@@ -211,7 +211,6 @@ enum ipi_msg_type {
 	IPI_NOP = 0,
 	IPI_RESCHEDULE = 1,
 	IPI_CALL_FUNC,
-	IPI_CALL_FUNC_SINGLE,
 	IPI_CPU_STOP
 };
 
@@ -255,7 +254,7 @@ void smp_send_stop(void)
 
 void arch_send_call_function_single_ipi(int cpu)
 {
-	ipi_send_msg(cpumask_of(cpu), IPI_CALL_FUNC_SINGLE);
+	ipi_send_msg(cpumask_of(cpu), IPI_CALL_FUNC);
 }
 
 void arch_send_call_function_ipi_mask(const struct cpumask *mask)
@@ -287,10 +286,6 @@ static inline void __do_IPI(unsigned long *ops, struct ipi_data *ipi, int cpu)
 			generic_smp_call_function_interrupt();
 			break;
 
-		case IPI_CALL_FUNC_SINGLE:
-			generic_smp_call_function_single_interrupt();
-			break;
-
 		case IPI_CPU_STOP:
 			ipi_cpu_stop(cpu);
 			break;
-- 
cgit v1.3-14-g43fede


From 0dd450fe13da4aeacc69916ecfe39d3d0b531900 Mon Sep 17 00:00:00 2001
From: Mischa Jonker <mjonker@synopsys.com>
Date: Thu, 7 Nov 2013 14:55:11 +0100
Subject: ARC: Add perf support for ARC700 cores

This adds basic perf support for ARC700 cores. Most PERF_COUNT_HW* events
are supported now.

Signed-off-by: Mischa Jonker <mjonker@synopsys.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/boot/dts/angel4.dts      |   4 +
 arch/arc/include/asm/perf_event.h | 204 +++++++++++++++++++++++-
 arch/arc/kernel/Makefile          |   1 +
 arch/arc/kernel/perf_event.c      | 322 ++++++++++++++++++++++++++++++++++++++
 4 files changed, 530 insertions(+), 1 deletion(-)
 create mode 100644 arch/arc/kernel/perf_event.c

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/boot/dts/angel4.dts b/arch/arc/boot/dts/angel4.dts
index 4fb2d6f655bd..bcf662d21a57 100644
--- a/arch/arc/boot/dts/angel4.dts
+++ b/arch/arc/boot/dts/angel4.dts
@@ -67,5 +67,9 @@
 				reg = <1>;
 			};
 		};
+
+		arcpmu0: pmu {
+			compatible = "snps,arc700-pmu";
+		};
 	};
 };
diff --git a/arch/arc/include/asm/perf_event.h b/arch/arc/include/asm/perf_event.h
index 115ad96480e6..cbf755e32a03 100644
--- a/arch/arc/include/asm/perf_event.h
+++ b/arch/arc/include/asm/perf_event.h
@@ -1,5 +1,7 @@
 /*
- * Copyright (C) 2011-2012 Synopsys, Inc. (www.synopsys.com)
+ * Linux performance counter support for ARC
+ *
+ * Copyright (C) 2011-2013 Synopsys, Inc. (www.synopsys.com)
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -10,4 +12,204 @@
 #ifndef __ASM_PERF_EVENT_H
 #define __ASM_PERF_EVENT_H
 
+/* real maximum varies per CPU, this is the maximum supported by the driver */
+#define ARC_PMU_MAX_HWEVENTS	64
+
+#define ARC_REG_CC_BUILD	0xF6
+#define ARC_REG_CC_INDEX	0x240
+#define ARC_REG_CC_NAME0	0x241
+#define ARC_REG_CC_NAME1	0x242
+
+#define ARC_REG_PCT_BUILD	0xF5
+#define ARC_REG_PCT_COUNTL	0x250
+#define ARC_REG_PCT_COUNTH	0x251
+#define ARC_REG_PCT_SNAPL	0x252
+#define ARC_REG_PCT_SNAPH	0x253
+#define ARC_REG_PCT_CONFIG	0x254
+#define ARC_REG_PCT_CONTROL	0x255
+#define ARC_REG_PCT_INDEX	0x256
+
+#define ARC_REG_PCT_CONTROL_CC	(1 << 16)	/* clear counts */
+#define ARC_REG_PCT_CONTROL_SN	(1 << 17)	/* snapshot */
+
+struct arc_reg_pct_build {
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	unsigned int m:8, c:8, r:6, s:2, v:8;
+#else
+	unsigned int v:8, s:2, r:6, c:8, m:8;
+#endif
+};
+
+struct arc_reg_cc_build {
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	unsigned int c:16, r:8, v:8;
+#else
+	unsigned int v:8, r:8, c:16;
+#endif
+};
+
+#define PERF_COUNT_ARC_DCLM	(PERF_COUNT_HW_MAX + 0)
+#define PERF_COUNT_ARC_DCSM	(PERF_COUNT_HW_MAX + 1)
+#define PERF_COUNT_ARC_ICM	(PERF_COUNT_HW_MAX + 2)
+#define PERF_COUNT_ARC_BPOK	(PERF_COUNT_HW_MAX + 3)
+#define PERF_COUNT_ARC_EDTLB	(PERF_COUNT_HW_MAX + 4)
+#define PERF_COUNT_ARC_EITLB	(PERF_COUNT_HW_MAX + 5)
+#define PERF_COUNT_ARC_HW_MAX	(PERF_COUNT_HW_MAX + 6)
+
+/*
+ * The "generalized" performance events seem to really be a copy
+ * of the available events on x86 processors; the mapping to ARC
+ * events is not always possible 1-to-1. Fortunately, there doesn't
+ * seem to be an exact definition for these events, so we can cheat
+ * a bit where necessary.
+ *
+ * In particular, the following PERF events may behave a bit differently
+ * compared to other architectures:
+ *
+ * PERF_COUNT_HW_CPU_CYCLES
+ *	Cycles not in halted state
+ *
+ * PERF_COUNT_HW_REF_CPU_CYCLES
+ *	Reference cycles not in halted state, same as PERF_COUNT_HW_CPU_CYCLES
+ *	for now as we don't do Dynamic Voltage/Frequency Scaling (yet)
+ *
+ * PERF_COUNT_HW_BUS_CYCLES
+ *	Unclear what this means, Intel uses 0x013c, which according to
+ *	their datasheet means "unhalted reference cycles". It sounds similar
+ *	to PERF_COUNT_HW_REF_CPU_CYCLES, and we use the same counter for it.
+ *
+ * PERF_COUNT_HW_STALLED_CYCLES_BACKEND
+ * PERF_COUNT_HW_STALLED_CYCLES_FRONTEND
+ *	The ARC 700 can either measure stalls per pipeline stage, or all stalls
+ *	combined; for now we assign all stalls to STALLED_CYCLES_BACKEND
+ *	and all pipeline flushes (e.g. caused by mispredicts, etc.) to
+ *	STALLED_CYCLES_FRONTEND.
+ *
+ *	We could start multiple performance counters and combine everything
+ *	afterwards, but that makes it complicated.
+ *
+ *	Note that I$ cache misses aren't counted by either of the two!
+ */
+
+static const char * const arc_pmu_ev_hw_map[] = {
+	[PERF_COUNT_HW_CPU_CYCLES] = "crun",
+	[PERF_COUNT_HW_REF_CPU_CYCLES] = "crun",
+	[PERF_COUNT_HW_BUS_CYCLES] = "crun",
+	[PERF_COUNT_HW_INSTRUCTIONS] = "iall",
+	[PERF_COUNT_HW_BRANCH_MISSES] = "bpfail",
+	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = "ijmp",
+	[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = "bflush",
+	[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = "bstall",
+	[PERF_COUNT_ARC_DCLM] = "dclm",
+	[PERF_COUNT_ARC_DCSM] = "dcsm",
+	[PERF_COUNT_ARC_ICM] = "icm",
+	[PERF_COUNT_ARC_BPOK] = "bpok",
+	[PERF_COUNT_ARC_EDTLB] = "edtlb",
+	[PERF_COUNT_ARC_EITLB] = "eitlb",
+};
+
+#define C(_x)			PERF_COUNT_HW_CACHE_##_x
+#define CACHE_OP_UNSUPPORTED	0xffff
+
+static const unsigned arc_pmu_cache_map[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = {
+	[C(L1D)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= PERF_COUNT_ARC_DCLM,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= PERF_COUNT_ARC_DCSM,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(L1I)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= PERF_COUNT_ARC_ICM,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(LL)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(DTLB)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= PERF_COUNT_ARC_EDTLB,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(ITLB)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= PERF_COUNT_ARC_EITLB,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(BPU)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)] = PERF_COUNT_HW_BRANCH_INSTRUCTIONS,
+			[C(RESULT_MISS)]	= PERF_COUNT_HW_BRANCH_MISSES,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+	[C(NODE)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+		[C(OP_PREFETCH)] = {
+			[C(RESULT_ACCESS)]	= CACHE_OP_UNSUPPORTED,
+			[C(RESULT_MISS)]	= CACHE_OP_UNSUPPORTED,
+		},
+	},
+};
+
 #endif /* __ASM_PERF_EVENT_H */
diff --git a/arch/arc/kernel/Makefile b/arch/arc/kernel/Makefile
index c242ef07ba70..8004b4fa6461 100644
--- a/arch/arc/kernel/Makefile
+++ b/arch/arc/kernel/Makefile
@@ -19,6 +19,7 @@ obj-$(CONFIG_KPROBES)      		+= kprobes.o
 obj-$(CONFIG_ARC_MISALIGN_ACCESS) 	+= unaligned.o
 obj-$(CONFIG_KGDB)			+= kgdb.o
 obj-$(CONFIG_ARC_METAWARE_HLINK)	+= arc_hostlink.o
+obj-$(CONFIG_PERF_EVENTS)		+= perf_event.o
 
 obj-$(CONFIG_ARC_FPU_SAVE_RESTORE)	+= fpu.o
 CFLAGS_fpu.o   += -mdpfp
diff --git a/arch/arc/kernel/perf_event.c b/arch/arc/kernel/perf_event.c
new file mode 100644
index 000000000000..759e5f3e9029
--- /dev/null
+++ b/arch/arc/kernel/perf_event.c
@@ -0,0 +1,322 @@
+/*
+ * Linux performance counter support for ARC700 series
+ *
+ * Copyright (C) 2013 Synopsys, Inc. (www.synopsys.com)
+ *
+ * This code is inspired by the perf support of various other architectures.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/perf_event.h>
+#include <linux/platform_device.h>
+#include <asm/arcregs.h>
+
+struct arc_pmu {
+	struct pmu	pmu;
+	int		counter_size;	/* in bits */
+	int		n_counters;
+	unsigned long	used_mask[BITS_TO_LONGS(ARC_PMU_MAX_HWEVENTS)];
+	int		ev_hw_idx[PERF_COUNT_ARC_HW_MAX];
+};
+
+/* read counter #idx; note that counter# != event# on ARC! */
+static uint64_t arc_pmu_read_counter(int idx)
+{
+	uint32_t tmp;
+	uint64_t result;
+
+	/*
+	 * ARC supports making 'snapshots' of the counters, so we don't
+	 * need to care about counters wrapping to 0 underneath our feet
+	 */
+	write_aux_reg(ARC_REG_PCT_INDEX, idx);
+	tmp = read_aux_reg(ARC_REG_PCT_CONTROL);
+	write_aux_reg(ARC_REG_PCT_CONTROL, tmp | ARC_REG_PCT_CONTROL_SN);
+	result = (uint64_t) (read_aux_reg(ARC_REG_PCT_SNAPH)) << 32;
+	result |= read_aux_reg(ARC_REG_PCT_SNAPL);
+
+	return result;
+}
+
+static void arc_perf_event_update(struct perf_event *event,
+				  struct hw_perf_event *hwc, int idx)
+{
+	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
+	uint64_t prev_raw_count, new_raw_count;
+	int64_t delta;
+
+	do {
+		prev_raw_count = local64_read(&hwc->prev_count);
+		new_raw_count = arc_pmu_read_counter(idx);
+	} while (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
+				 new_raw_count) != prev_raw_count);
+
+	delta = (new_raw_count - prev_raw_count) &
+		((1ULL << arc_pmu->counter_size) - 1ULL);
+
+	local64_add(delta, &event->count);
+	local64_sub(delta, &hwc->period_left);
+}
+
+static void arc_pmu_read(struct perf_event *event)
+{
+	arc_perf_event_update(event, &event->hw, event->hw.idx);
+}
+
+static int arc_pmu_cache_event(u64 config)
+{
+	unsigned int cache_type, cache_op, cache_result;
+	int ret;
+
+	cache_type	= (config >>  0) & 0xff;
+	cache_op	= (config >>  8) & 0xff;
+	cache_result	= (config >> 16) & 0xff;
+	if (cache_type >= PERF_COUNT_HW_CACHE_MAX)
+		return -EINVAL;
+	if (cache_type >= PERF_COUNT_HW_CACHE_OP_MAX)
+		return -EINVAL;
+	if (cache_type >= PERF_COUNT_HW_CACHE_RESULT_MAX)
+		return -EINVAL;
+
+	ret = arc_pmu_cache_map[cache_type][cache_op][cache_result];
+
+	if (ret == CACHE_OP_UNSUPPORTED)
+		return -ENOENT;
+
+	return ret;
+}
+
+/* initializes hw_perf_event structure if event is supported */
+static int arc_pmu_event_init(struct perf_event *event)
+{
+	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
+	struct hw_perf_event *hwc = &event->hw;
+	int ret;
+
+	switch (event->attr.type) {
+	case PERF_TYPE_HARDWARE:
+		if (event->attr.config >= PERF_COUNT_HW_MAX)
+			return -ENOENT;
+		if (arc_pmu->ev_hw_idx[event->attr.config] < 0)
+			return -ENOENT;
+		hwc->config = arc_pmu->ev_hw_idx[event->attr.config];
+		pr_debug("initializing event %d with cfg %d\n",
+			 (int) event->attr.config, (int) hwc->config);
+		return 0;
+	case PERF_TYPE_HW_CACHE:
+		ret = arc_pmu_cache_event(event->attr.config);
+		if (ret < 0)
+			return ret;
+		hwc->config = arc_pmu->ev_hw_idx[ret];
+		return 0;
+	default:
+		return -ENOENT;
+	}
+}
+
+/* starts all counters */
+static void arc_pmu_enable(struct pmu *pmu)
+{
+	uint32_t tmp;
+	tmp = read_aux_reg(ARC_REG_PCT_CONTROL);
+	write_aux_reg(ARC_REG_PCT_CONTROL, (tmp & 0xffff0000) | 0x1);
+}
+
+/* stops all counters */
+static void arc_pmu_disable(struct pmu *pmu)
+{
+	uint32_t tmp;
+	tmp = read_aux_reg(ARC_REG_PCT_CONTROL);
+	write_aux_reg(ARC_REG_PCT_CONTROL, (tmp & 0xffff0000) | 0x0);
+}
+
+/*
+ * Assigns hardware counter to hardware condition.
+ * Note that there is no separate start/stop mechanism;
+ * stopping is achieved by assigning the 'never' condition
+ */
+static void arc_pmu_start(struct perf_event *event, int flags)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (WARN_ON_ONCE(idx == -1))
+		return;
+
+	if (flags & PERF_EF_RELOAD)
+		WARN_ON_ONCE(!(event->hw.state & PERF_HES_UPTODATE));
+
+	event->hw.state = 0;
+
+	/* enable ARC pmu here */
+	write_aux_reg(ARC_REG_PCT_INDEX, idx);
+	write_aux_reg(ARC_REG_PCT_CONFIG, hwc->config);
+}
+
+static void arc_pmu_stop(struct perf_event *event, int flags)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (!(event->hw.state & PERF_HES_STOPPED)) {
+		/* stop ARC pmu here */
+		write_aux_reg(ARC_REG_PCT_INDEX, idx);
+
+		/* condition code #0 is always "never" */
+		write_aux_reg(ARC_REG_PCT_CONFIG, 0);
+
+		event->hw.state |= PERF_HES_STOPPED;
+	}
+
+	if ((flags & PERF_EF_UPDATE) &&
+	    !(event->hw.state & PERF_HES_UPTODATE)) {
+		arc_perf_event_update(event, &event->hw, idx);
+		event->hw.state |= PERF_HES_UPTODATE;
+	}
+}
+
+static void arc_pmu_del(struct perf_event *event, int flags)
+{
+	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
+
+	arc_pmu_stop(event, PERF_EF_UPDATE);
+	__clear_bit(event->hw.idx, arc_pmu->used_mask);
+
+	perf_event_update_userpage(event);
+}
+
+/* allocate hardware counter and optionally start counting */
+static int arc_pmu_add(struct perf_event *event, int flags)
+{
+	struct arc_pmu *arc_pmu = container_of(event->pmu, struct arc_pmu, pmu);
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (__test_and_set_bit(idx, arc_pmu->used_mask)) {
+		idx = find_first_zero_bit(arc_pmu->used_mask,
+					  arc_pmu->n_counters);
+		if (idx == arc_pmu->n_counters)
+			return -EAGAIN;
+
+		__set_bit(idx, arc_pmu->used_mask);
+		hwc->idx = idx;
+	}
+
+	write_aux_reg(ARC_REG_PCT_INDEX, idx);
+	write_aux_reg(ARC_REG_PCT_CONFIG, 0);
+	write_aux_reg(ARC_REG_PCT_COUNTL, 0);
+	write_aux_reg(ARC_REG_PCT_COUNTH, 0);
+	local64_set(&hwc->prev_count, 0);
+
+	hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
+	if (flags & PERF_EF_START)
+		arc_pmu_start(event, PERF_EF_RELOAD);
+
+	perf_event_update_userpage(event);
+
+	return 0;
+}
+
+static int arc_pmu_device_probe(struct platform_device *pdev)
+{
+	struct arc_pmu *arc_pmu;
+	struct arc_reg_pct_build pct_bcr;
+	struct arc_reg_cc_build cc_bcr;
+	int i, j, ret;
+
+	union cc_name {
+		struct {
+			uint32_t word0, word1;
+			char sentinel;
+		} indiv;
+		char str[9];
+	} cc_name;
+
+
+	READ_BCR(ARC_REG_PCT_BUILD, pct_bcr);
+	if (!pct_bcr.v) {
+		pr_err("This core does not have performance counters!\n");
+		return -ENODEV;
+	}
+
+	arc_pmu = devm_kzalloc(&pdev->dev, sizeof(struct arc_pmu),
+			       GFP_KERNEL);
+	if (!arc_pmu)
+		return -ENOMEM;
+
+	arc_pmu->n_counters = pct_bcr.c;
+	BUG_ON(arc_pmu->n_counters > ARC_PMU_MAX_HWEVENTS);
+
+	arc_pmu->counter_size = 32 + (pct_bcr.s << 4);
+	pr_info("ARC PMU found with %d counters of size %d bits\n",
+		arc_pmu->n_counters, arc_pmu->counter_size);
+
+	READ_BCR(ARC_REG_CC_BUILD, cc_bcr);
+
+	if (!cc_bcr.v)
+		pr_err("Strange! Performance counters exist, but no countable conditions?\n");
+
+	pr_info("ARC PMU has %d countable conditions\n", cc_bcr.c);
+
+	cc_name.str[8] = 0;
+	for (i = 0; i < PERF_COUNT_HW_MAX; i++)
+		arc_pmu->ev_hw_idx[i] = -1;
+
+	for (j = 0; j < cc_bcr.c; j++) {
+		write_aux_reg(ARC_REG_CC_INDEX, j);
+		cc_name.indiv.word0 = read_aux_reg(ARC_REG_CC_NAME0);
+		cc_name.indiv.word1 = read_aux_reg(ARC_REG_CC_NAME1);
+		for (i = 0; i < ARRAY_SIZE(arc_pmu_ev_hw_map); i++) {
+			if (arc_pmu_ev_hw_map[i] &&
+			    !strcmp(arc_pmu_ev_hw_map[i], cc_name.str) &&
+			    strlen(arc_pmu_ev_hw_map[i])) {
+				pr_debug("mapping %d to idx %d with name %s\n",
+					 i, j, cc_name.str);
+				arc_pmu->ev_hw_idx[i] = j;
+			}
+		}
+	}
+
+	arc_pmu->pmu = (struct pmu) {
+		.pmu_enable	= arc_pmu_enable,
+		.pmu_disable	= arc_pmu_disable,
+		.event_init	= arc_pmu_event_init,
+		.add		= arc_pmu_add,
+		.del		= arc_pmu_del,
+		.start		= arc_pmu_start,
+		.stop		= arc_pmu_stop,
+		.read		= arc_pmu_read,
+	};
+
+	ret = perf_pmu_register(&arc_pmu->pmu, pdev->name, PERF_TYPE_RAW);
+
+	return ret;
+}
+
+#ifdef CONFIG_OF
+static const struct of_device_id arc_pmu_match[] = {
+	{ .compatible = "snps,arc700-pmu" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, arc_pmu_match);
+#endif
+
+static struct platform_driver arc_pmu_driver = {
+	.driver	= {
+		.name		= "arc700-pmu",
+		.of_match_table = of_match_ptr(arc_pmu_match),
+	},
+	.probe		= arc_pmu_device_probe,
+};
+
+module_platform_driver(arc_pmu_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Mischa Jonker <mjonker@synopsys.com>");
+MODULE_DESCRIPTION("ARC PMU driver");
-- 
cgit v1.3-14-g43fede


From 230c4aadcc394fb48d865a1cd96c8725dabad56d Mon Sep 17 00:00:00 2001
From: Mischa Jonker <mjonker@synopsys.com>
Date: Wed, 13 Nov 2013 16:06:06 +0100
Subject: ARC: perf: ARC 700 PMU doesn't support sampling events

The ARC 700 does not have an interrupt associated with it, and as
such it cannot trigger when a counter overflows. As the counters are
48 bit, it will usually take at least 100 days before a counter
overflows, so for mere counting of events, there is no problem.
Sampling is not supported though.

Signed-off-by: Mischa Jonker <mjonker@synopsys.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
---
 arch/arc/kernel/perf_event.c | 4 ++++
 1 file changed, 4 insertions(+)

(limited to 'arch/arc/kernel')

diff --git a/arch/arc/kernel/perf_event.c b/arch/arc/kernel/perf_event.c
index 759e5f3e9029..e46d81f70979 100644
--- a/arch/arc/kernel/perf_event.c
+++ b/arch/arc/kernel/perf_event.c
@@ -99,6 +99,10 @@ static int arc_pmu_event_init(struct perf_event *event)
 	struct hw_perf_event *hwc = &event->hw;
 	int ret;
 
+	/* ARC 700 PMU does not support sampling events */
+	if (is_sampling_event(event))
+		return -ENOENT;
+
 	switch (event->attr.type) {
 	case PERF_TYPE_HARDWARE:
 		if (event->attr.config >= PERF_COUNT_HW_MAX)
-- 
cgit v1.3-14-g43fede