[PATCH 0/4] drivers: base: cacheinfo: fixes/updates

[PATCH 0/4] drivers: base: cacheinfo: fixes/updates

[Sudeep Holla]

Hi Greg,

Since the couple of fixes here are not too severe, I am considering as
updates only.

Now the x86 allow CONFIG_OF to be enabled and ACPI on arm64, we have
couple of minor bugs in those configurations. The first 2 patches fixes
them. The 3rd patch is cosmetic update to help identify the logs easily.

The last patch adds the basic support for overriding cache properties
using the device tree.

Regards,
Sudeep

Sudeep Holla (4):
  drivers: base: cacheinfo: fix x86 with CONFIG_OF enabled
  drivers: base: cacheinfo: fix boot error message when acpi is enabled
  drivers: base: cacheinfo: add pr_fmt logging
  drivers: base: cacheinfo: support DT overrides for cache properties

 arch/x86/kernel/cpu/intel_cacheinfo.c |   2 +
 drivers/base/cacheinfo.c              | 138 +++++++++++++++++++++++++++++++++-
 include/linux/cacheinfo.h             |   1 +
 3 files changed, 137 insertions(+), 4 deletions(-)

-- 
2.7.4

[PATCH 1/4] drivers: base: cacheinfo: fix x86 with CONFIG_OF enabled

[Sudeep Holla]

With CONFIG_OF enabled on x86, we get the following error on boot:
"
	Failed to find cpu0 device node
 	Unable to detect cache hierarchy from DT for CPU 0
"
and the cacheinfo fails to get populated in the corresponding sysfs
entries. This is because cache_setup_of_node looks for of_node for
setting up the shared cpu_map without checking that it's already
populated in the architecture specific callback.

In order to indicate that the shared cpu_map is already populated, this
patch introduces a boolean `cpu_map_populated` in struct cpu_cacheinfo
that can be used by the generic code to skip cache_shared_cpu_map_setup.

This patch also sets that boolean for x86.

Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 arch/x86/kernel/cpu/intel_cacheinfo.c | 2 ++
 drivers/base/cacheinfo.c              | 3 +++
 include/linux/cacheinfo.h             | 1 +
 3 files changed, 6 insertions(+)

diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c
index de6626c18e42..be6337156502 100644
--- a/arch/x86/kernel/cpu/intel_cacheinfo.c
+++ b/arch/x86/kernel/cpu/intel_cacheinfo.c
@@ -934,6 +934,8 @@ static int __populate_cache_leaves(unsigned int cpu)
 		ci_leaf_init(this_leaf++, &id4_regs);
 		__cache_cpumap_setup(cpu, idx, &id4_regs);
 	}
+	this_cpu_ci->cpu_map_populated = true;
+
 	return 0;
 }
 
diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c
index e9fd32e91668..ecde8957835a 100644
--- a/drivers/base/cacheinfo.c
+++ b/drivers/base/cacheinfo.c
@@ -106,6 +106,9 @@ static int cache_shared_cpu_map_setup(unsigned int cpu)
 	unsigned int index;
 	int ret;
 
+	if (this_cpu_ci->cpu_map_populated)
+		return 0;
+
 	ret = cache_setup_of_node(cpu);
 	if (ret)
 		return ret;
diff --git a/include/linux/cacheinfo.h b/include/linux/cacheinfo.h
index 2189935075b4..a951fd10aaaa 100644
--- a/include/linux/cacheinfo.h
+++ b/include/linux/cacheinfo.h
@@ -71,6 +71,7 @@ struct cpu_cacheinfo {
 	struct cacheinfo *info_list;
 	unsigned int num_levels;
 	unsigned int num_leaves;
+	bool cpu_map_populated;
 };
 
 /*
-- 
2.7.4

[PATCH 2/4] drivers: base: cacheinfo: fix boot error message when acpi is enabled

[Sudeep Holla]

ARM64 enables both CONFIG_OF and CONFIG_ACPI and the firmware can pass
both ACPI tables and the device tree. Based on the kernel parameter, one
of the two will be chosen. If acpi is enabled, then device tree is not
unflattened.

Currently ARM64 platforms report:
"
	Failed to find cpu0 device node
	Unable to detect cache hierarchy from DT for CPU 0
"
which is incorrect when booting with ACPI. Also latest ACPI v6.1 has no
support for cache properties/hierarchy.

This patch adds check for unflattened device tree and also returns as
"not supported" if ACPI is runtime enabled.

It also removes the reference to DT from the error message as the cache
hierarchy can be detected from the firmware(OF/DT/ACPI)

Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/base/cacheinfo.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c
index ecde8957835a..70e13cf06ed0 100644
--- a/drivers/base/cacheinfo.c
+++ b/drivers/base/cacheinfo.c
@@ -16,6 +16,7 @@
  * You should have received a copy of the GNU General Public License
  * along with this program.  If not, see <http://www.gnu.org/licenses/>.
  */
+#include <linux/acpi.h>
 #include <linux/bitops.h>
 #include <linux/cacheinfo.h>
 #include <linux/compiler.h>
@@ -104,12 +105,16 @@ static int cache_shared_cpu_map_setup(unsigned int cpu)
 	struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu);
 	struct cacheinfo *this_leaf, *sib_leaf;
 	unsigned int index;
-	int ret;
+	int ret = 0;
 
 	if (this_cpu_ci->cpu_map_populated)
 		return 0;
 
-	ret = cache_setup_of_node(cpu);
+	if (of_have_populated_dt())
+		ret = cache_setup_of_node(cpu);
+	else if (!acpi_disabled)
+		/* No cache property/hierarchy support yet in ACPI */
+		ret = -ENOTSUPP;
 	if (ret)
 		return ret;
 
@@ -206,8 +211,7 @@ static int detect_cache_attributes(unsigned int cpu)
 	 */
 	ret = cache_shared_cpu_map_setup(cpu);
 	if (ret) {
-		pr_warn("Unable to detect cache hierarchy from DT for CPU %d\n",
-			cpu);
+		pr_warn("Unable to detect cache hierarchy for CPU %d\n", cpu);
 		goto free_ci;
 	}
 	return 0;
-- 
2.7.4

[PATCH 3/4] drivers: base: cacheinfo: add pr_fmt logging

[Sudeep Holla]

This cleanup patch just adds pr_fmt style logging for cacheinfo.

Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/base/cacheinfo.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c
index 70e13cf06ed0..f19d50bd8925 100644
--- a/drivers/base/cacheinfo.c
+++ b/drivers/base/cacheinfo.c
@@ -16,6 +16,8 @@
  * You should have received a copy of the GNU General Public License
  * along with this program.  If not, see <http://www.gnu.org/licenses/>.
  */
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/acpi.h>
 #include <linux/bitops.h>
 #include <linux/cacheinfo.h>
-- 
2.7.4

[PATCH 4/4] drivers: base: cacheinfo: support DT overrides for cache properties

[Sudeep Holla]

Few architectures like x86, ia64 and s390 derive the cache topology and
all the properties using a specific architected mechanism while some
other architectures like powerpc all those information id derived from
the device tree.

On ARM, both the mechanism is used. While all the cache properties can
be derived in a architected way, it needs to rely on device tree to get
the cache topology information.

However there are few platforms where this architected mechanism is
broken and the device tree properties can be used to override these
incorrect values.

This patch adds support for overriding the cache properties values to
the values specified in the device tree.

Cc: Alex Van Brunt <avanbrunt@nvidia.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/base/cacheinfo.c | 121 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 121 insertions(+)

diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c
index f19d50bd8925..2376628c599c 100644
--- a/drivers/base/cacheinfo.c
+++ b/drivers/base/cacheinfo.c
@@ -88,7 +88,120 @@ static inline bool cache_leaves_are_shared(struct cacheinfo *this_leaf,
 {
 	return sib_leaf->of_node == this_leaf->of_node;
 }
+
+/* OF properties to query for a given cache type */
+struct cache_type_info {
+	const char *size_prop;
+	const char *line_size_props[2];
+	const char *nr_sets_prop;
+};
+
+static const struct cache_type_info cache_type_info[] = {
+	{
+		.size_prop       = "cache-size",
+		.line_size_props = { "cache-line-size",
+				     "cache-block-size", },
+		.nr_sets_prop    = "cache-sets",
+	}, {
+		.size_prop       = "i-cache-size",
+		.line_size_props = { "i-cache-line-size",
+				     "i-cache-block-size", },
+		.nr_sets_prop    = "i-cache-sets",
+	}, {
+		.size_prop       = "d-cache-size",
+		.line_size_props = { "d-cache-line-size",
+				     "d-cache-block-size", },
+		.nr_sets_prop    = "d-cache-sets",
+	},
+};
+
+static inline int get_cacheinfo_idx(enum cache_type type)
+{
+	if (type == CACHE_TYPE_UNIFIED)
+		return 0;
+	return type;
+}
+
+static void cache_size(struct cacheinfo *this_leaf)
+{
+	const char *propname;
+	const __be32 *cache_size;
+	int ct_idx;
+
+	ct_idx = get_cacheinfo_idx(this_leaf->type);
+	propname = cache_type_info[ct_idx].size_prop;
+
+	cache_size = of_get_property(this_leaf->of_node, propname, NULL);
+	if (cache_size)
+		this_leaf->size = of_read_number(cache_size, 1);
+}
+
+/* not cache_line_size() because that's a macro in include/linux/cache.h */
+static void cache_get_line_size(struct cacheinfo *this_leaf)
+{
+	const __be32 *line_size;
+	int i, lim, ct_idx;
+
+	ct_idx = get_cacheinfo_idx(this_leaf->type);
+	lim = ARRAY_SIZE(cache_type_info[ct_idx].line_size_props);
+
+	for (i = 0; i < lim; i++) {
+		const char *propname;
+
+		propname = cache_type_info[ct_idx].line_size_props[i];
+		line_size = of_get_property(this_leaf->of_node, propname, NULL);
+		if (line_size)
+			break;
+	}
+
+	if (line_size)
+		this_leaf->coherency_line_size = of_read_number(line_size, 1);
+}
+
+static void cache_nr_sets(struct cacheinfo *this_leaf)
+{
+	const char *propname;
+	const __be32 *nr_sets;
+	int ct_idx;
+
+	ct_idx = get_cacheinfo_idx(this_leaf->type);
+	propname = cache_type_info[ct_idx].nr_sets_prop;
+
+	nr_sets = of_get_property(this_leaf->of_node, propname, NULL);
+	if (nr_sets)
+		this_leaf->number_of_sets = of_read_number(nr_sets, 1);
+}
+
+static void cache_associativity(struct cacheinfo *this_leaf)
+{
+	unsigned int line_size = this_leaf->coherency_line_size;
+	unsigned int nr_sets = this_leaf->number_of_sets;
+	unsigned int size = this_leaf->size;
+
+	/*
+	 * If the cache is fully associative, there is no need to
+	 * check the other properties.
+	 */
+	if (!(nr_sets == 1) && (nr_sets > 0 && size > 0 && line_size > 0))
+		this_leaf->ways_of_associativity = (size / nr_sets) / line_size;
+}
+
+static void cache_of_override_properties(unsigned int cpu)
+{
+	int index;
+	struct cacheinfo *this_leaf;
+	struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu);
+
+	for (index = 0; index < cache_leaves(cpu); index++) {
+		this_leaf = this_cpu_ci->info_list + index;
+		cache_size(this_leaf);
+		cache_get_line_size(this_leaf);
+		cache_nr_sets(this_leaf);
+		cache_associativity(this_leaf);
+	}
+}
 #else
+static void cache_of_override_properties(unsigned int cpu) { }
 static inline int cache_setup_of_node(unsigned int cpu) { return 0; }
 static inline bool cache_leaves_are_shared(struct cacheinfo *this_leaf,
 					   struct cacheinfo *sib_leaf)
@@ -171,6 +284,12 @@ static void cache_shared_cpu_map_remove(unsigned int cpu)
 	}
 }
 
+static void cache_override_properties(unsigned int cpu)
+{
+	if (of_have_populated_dt())
+		return cache_of_override_properties(cpu);
+}
+
 static void free_cache_attributes(unsigned int cpu)
 {
 	if (!per_cpu_cacheinfo(cpu))
@@ -216,6 +335,8 @@ static int detect_cache_attributes(unsigned int cpu)
 		pr_warn("Unable to detect cache hierarchy for CPU %d\n", cpu);
 		goto free_ci;
 	}
+
+	cache_override_properties(cpu);
 	return 0;
 
 free_ci:
-- 
2.7.4