Re: [PATCH v10 3/8] dt, numa: adding numa dt binding implementation.

[Ganapatrao Kulkarni]

Hi Rob,

On Thu, Feb 11, 2016 at 9:02 PM, Robert Richter
<robert.rich...@caviumnetworks.com> wrote:
> On 11.02.16 08:50:41, Rob Herring wrote:
>> On Tue, Feb 2, 2016 at 4:09 AM, Ganapatrao Kulkarni
>> <gkulka...@caviumnetworks.com> wrote:
>> > dt node parsing for numa topology is done using device property
>> > numa-node-id and device node distance-map.
>>
>> How is it that powerpc doesn't need flat DT parsing for NUMA? Both
>> arches are memblock based and the binding is similar IIRC, so you
>> should be able to find a way to do this with the unflattened tree.
>> Ideally, there should be some common code shared as well.

powerpc has implemented numa binding using property named
"ibm,associativity", we too started adding using the same binding[
refer version V6]
however had discussion/review with Mark R about mapping and we could not map to
some of the topology like mesh and ring topology.

So we have defined and implemented this new binding which is simple
and scalable.
This binding is very much aligned to ACPI NUMA specification.

>>
>> Do you have a git tree with this series?
>
> I have pushed it here for you:
>
>  
> https://git.kernel.org/cgit/linux/kernel/git/rric/linux.git/log/?h=thunder/numa-v10

thanks Robert.
>
> -Robert

thanks
Ganapat
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Re: [PATCH v10 3/8] dt, numa: adding numa dt binding implementation.

[Ganapatrao Kulkarni]

Hi Rob,

On Wed, Feb 3, 2016 at 5:04 AM, Rob Herring <r...@kernel.org> wrote:
> On Tue, Feb 02, 2016 at 03:39:18PM +0530, Ganapatrao Kulkarni wrote:
>> dt node parsing for numa topology is done using device property
>> numa-node-id and device node distance-map.
>>
>> Reviewed-by: Robert Richter <rrich...@cavium.com>
>> Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
>> ---
>>  drivers/of/Kconfig   |  11 +++
>>  drivers/of/Makefile  |   1 +
>>  drivers/of/of_numa.c | 207 
>> +++
>>  include/linux/of.h   |   4 +
>>  4 files changed, 223 insertions(+)
>>  create mode 100644 drivers/of/of_numa.c
>>
>> diff --git a/drivers/of/Kconfig b/drivers/of/Kconfig
>> index e2a4841..8f9cc3a 100644
>> --- a/drivers/of/Kconfig
>> +++ b/drivers/of/Kconfig
>> @@ -112,4 +112,15 @@ config OF_OVERLAY
>> While this option is selected automatically when needed, you can
>> enable it manually to improve device tree unit test coverage.
>>
>> +config OF_NUMA
>> + bool "Device Tree NUMA support"
>
> Does this need to be user visible?

thanks, not required, arch can select
>
>> + depends on NUMA
>> + depends on OF
>> + depends on ARM64
>
> drop this (and make sure it compiles on other arches). It will fail
> because you also have a dependency on FDT.

thanks will add OF_EARLY_FLATTREE
>
>> + default y
>> + help
>> +   Enable Device Tree NUMA support.
>> +   This enables the numa mapping of cpu, memory, io and
>> +   inter node distances using dt bindings.
>> +
>>  endif # OF
>> diff --git a/drivers/of/Makefile b/drivers/of/Makefile
>> index 156c072..bee3fa9 100644
>> --- a/drivers/of/Makefile
>> +++ b/drivers/of/Makefile
>> @@ -14,5 +14,6 @@ obj-$(CONFIG_OF_MTD)+= of_mtd.o
>>  obj-$(CONFIG_OF_RESERVED_MEM) += of_reserved_mem.o
>>  obj-$(CONFIG_OF_RESOLVE)  += resolver.o
>>  obj-$(CONFIG_OF_OVERLAY) += overlay.o
>> +obj-$(CONFIG_OF_NUMA) += of_numa.o
>>
>>  obj-$(CONFIG_OF_UNITTEST) += unittest-data/
>> diff --git a/drivers/of/of_numa.c b/drivers/of/of_numa.c
>> new file mode 100644
>> index 000..1142cdb
>> --- /dev/null
>> +++ b/drivers/of/of_numa.c
>> @@ -0,0 +1,207 @@
>> +/*
>> + * OF NUMA Parsing support.
>> + *
>> + * Copyright (C) 2015 Cavium Inc.
>> + * Author: Ganapatrao Kulkarni <gkulka...@cavium.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
>> + * published by the Free Software Foundation.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> + * GNU General Public License for more details.
>> + *
>> + * 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 
>> +#include 
>
> Surely you need some numa related includes.
ok
>
>> +
>> +/* define default numa node to 0 */
>> +#define DEFAULT_NODE 0
>> +
>> +/* Returns nid in the range [0..MAX_NUMNODES-1],
>> + * or NUMA_NO_NODE if no valid numa-node-id entry found
>> + * or DEFAULT_NODE if no numa-node-id entry exists
>> + */
>> +static int of_numa_prop_to_nid(const __be32 *of_numa_prop, int length)
>> +{
>> + int nid;
>> +
>> + if (!of_numa_prop)
>> + return DEFAULT_NODE;
>> +
>> + if (length != sizeof(*of_numa_prop)) {
>> + pr_warn("NUMA: Invalid of_numa_prop length %d found.\n",
>> + length);
>> + return NUMA_NO_NODE;
>> + }
>> +
>> + nid = of_read_number(of_numa_prop, 1);
>> + if (nid >= MAX_NUMNODES) {
>> + pr_warn("NUMA: Invalid numa node %d found.\n", nid);
>> + return NUMA_NO_NODE;
>> + }
>> +
>> + return nid;
>> +}
>> +
>> +static int __init early_init_of_node_to_nid(unsigned long node)
>> +{
>> + int length;
>> + const __be32 *of_numa_prop;
>> +
>> + of_numa_prop = of_get_flat_dt_prop(node, "numa-node-id", );
>> +
>> + return of_numa_prop_to_nid(of_numa_prop

[PATCH v10 7/8] topology, cleanup: Avoid redefinition of cpumask_of_pcibus in asm header files.

[Ganapatrao Kulkarni]

At present cpumask_of_pcibus is defined for !CONFIG_NUMA and moving out
to common will allow to use for numa too. This also avoids
redefinition of this macro in respective architecture header files.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/include/asm/topology.h   | 3 ---
 arch/ia64/include/asm/topology.h| 4 
 arch/metag/include/asm/topology.h   | 3 ---
 arch/powerpc/include/asm/topology.h | 4 
 arch/s390/include/asm/pci.h | 2 +-
 arch/s390/include/asm/topology.h| 1 +
 arch/sh/include/asm/topology.h  | 3 ---
 arch/tile/include/asm/pci.h | 2 --
 arch/tile/include/asm/topology.h| 3 +++
 arch/x86/include/asm/pci.h  | 2 +-
 arch/x86/include/asm/topology.h | 1 +
 include/asm-generic/topology.h  | 4 ++--
 12 files changed, 9 insertions(+), 23 deletions(-)

diff --git a/arch/arm64/include/asm/topology.h 
b/arch/arm64/include/asm/topology.h
index 8b57339..6e1f62c 100644
--- a/arch/arm64/include/asm/topology.h
+++ b/arch/arm64/include/asm/topology.h
@@ -26,9 +26,6 @@ const struct cpumask *cpu_coregroup_mask(int cpu);
 
 struct pci_bus;
 int pcibus_to_node(struct pci_bus *bus);
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?\
-cpu_all_mask : \
-cpumask_of_node(pcibus_to_node(bus)))
 
 #endif /* CONFIG_NUMA */
 
diff --git a/arch/ia64/include/asm/topology.h b/arch/ia64/include/asm/topology.h
index 3ad8f69..2778eb6 100644
--- a/arch/ia64/include/asm/topology.h
+++ b/arch/ia64/include/asm/topology.h
@@ -58,10 +58,6 @@ void build_cpu_to_node_map(void);
 
 extern void arch_fix_phys_package_id(int num, u32 slot);
 
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?\
-cpu_all_mask : \
-cpumask_of_node(pcibus_to_node(bus)))
-
 #include 
 
 #endif /* _ASM_IA64_TOPOLOGY_H */
diff --git a/arch/metag/include/asm/topology.h 
b/arch/metag/include/asm/topology.h
index e95f874..b285196 100644
--- a/arch/metag/include/asm/topology.h
+++ b/arch/metag/include/asm/topology.h
@@ -9,9 +9,6 @@
 #define cpumask_of_node(node)  ((void)node, cpu_online_mask)
 
 #define pcibus_to_node(bus)((void)(bus), -1)
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ? \
-   cpu_all_mask : \
-   cpumask_of_node(pcibus_to_node(bus)))
 
 #endif
 
diff --git a/arch/powerpc/include/asm/topology.h 
b/arch/powerpc/include/asm/topology.h
index 8b3b46b..eee025d 100644
--- a/arch/powerpc/include/asm/topology.h
+++ b/arch/powerpc/include/asm/topology.h
@@ -32,10 +32,6 @@ static inline int pcibus_to_node(struct pci_bus *bus)
 }
 #endif
 
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?\
-cpu_all_mask : \
-cpumask_of_node(pcibus_to_node(bus)))
-
 extern int __node_distance(int, int);
 #define node_distance(a, b) __node_distance(a, b)
 
diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index c873e68..539fb2d 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -205,7 +205,7 @@ static inline int __pcibus_to_node(const struct pci_bus 
*bus)
 }
 
 static inline const struct cpumask *
-cpumask_of_pcibus(const struct pci_bus *bus)
+__cpumask_of_pcibus(const struct pci_bus *bus)
 {
return cpu_online_mask;
 }
diff --git a/arch/s390/include/asm/topology.h b/arch/s390/include/asm/topology.h
index 6b53962..ac2f88f 100644
--- a/arch/s390/include/asm/topology.h
+++ b/arch/s390/include/asm/topology.h
@@ -78,6 +78,7 @@ static inline const struct cpumask *cpumask_of_node(int node)
 #define parent_node(node) (node)
 
 #define pcibus_to_node(bus) __pcibus_to_node(bus)
+#define cpumask_of_pcibus(bus) __cpumask_of_pcibus(bus)
 
 #define node_distance(a, b) __node_distance(a, b)
 
diff --git a/arch/sh/include/asm/topology.h b/arch/sh/include/asm/topology.h
index b0a282d..357983d 100644
--- a/arch/sh/include/asm/topology.h
+++ b/arch/sh/include/asm/topology.h
@@ -9,9 +9,6 @@
 #define cpumask_of_node(node)  ((void)node, cpu_online_mask)
 
 #define pcibus_to_node(bus)((void)(bus), -1)
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ? \
-   cpu_all_mask : \
-   cpumask_of_node(pcibus_to_node(bus)))
 
 #endif
 
diff --git a/arch/tile/include/asm/pci.h b/arch/tile/include/asm/pci.h
index dfedd7a..473ed46 100644
--- a/arch/tile/include/asm/pci.h
+++ b/arch/tile/include/asm/pci.h
@@ -223,8 +223,6 @@ static inline int pcibios_assign_all_busses(void)
 /* Minimum PCI I/O address, starting at the page boundary. */
 #define PCIBIOS_MIN_IO PAGE_SIZE
 
-/* Use any cpu for PCI.

[PATCH v10 1/8] arm64, numa: adding numa support for arm64 platforms.

[Ganapatrao Kulkarni]

Adding numa support for arm64 based platforms.
This patch adds by default the dummy numa node and
maps all memory and cpus to node 0.
using this patch, numa can be simulated on single node arm64 platforms.

Tested-by: Shannon Zhao <shannon.z...@linaro.org>
Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/Kconfig|  25 +++
 arch/arm64/include/asm/mmzone.h   |  12 ++
 arch/arm64/include/asm/numa.h |  43 +
 arch/arm64/include/asm/topology.h |  10 +
 arch/arm64/kernel/pci.c   |  10 +
 arch/arm64/kernel/setup.c |   4 +
 arch/arm64/kernel/smp.c   |   2 +
 arch/arm64/mm/Makefile|   1 +
 arch/arm64/mm/init.c  |  34 +++-
 arch/arm64/mm/mmu.c   |   1 +
 arch/arm64/mm/numa.c  | 387 ++
 11 files changed, 524 insertions(+), 5 deletions(-)
 create mode 100644 arch/arm64/include/asm/mmzone.h
 create mode 100644 arch/arm64/include/asm/numa.h
 create mode 100644 arch/arm64/mm/numa.c

diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 8cc6228..fcf3950 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -74,6 +74,7 @@ config ARM64
select HAVE_HW_BREAKPOINT if PERF_EVENTS
select HAVE_IRQ_TIME_ACCOUNTING
select HAVE_MEMBLOCK
+   select HAVE_MEMBLOCK_NODE_MAP if NUMA
select HAVE_PATA_PLATFORM
select HAVE_PERF_EVENTS
select HAVE_PERF_REGS
@@ -534,6 +535,30 @@ config HOTPLUG_CPU
  Say Y here to experiment with turning CPUs off and on.  CPUs
  can be controlled through /sys/devices/system/cpu.
 
+# Common NUMA Features
+config NUMA
+   bool "Numa Memory Allocation and Scheduler Support"
+   depends on SMP
+   help
+ Enable NUMA (Non Uniform Memory Access) support.
+
+ The kernel will try to allocate memory used by a CPU on the
+ local memory of the CPU and add some more
+ NUMA awareness to the kernel.
+
+config NODES_SHIFT
+   int "Maximum NUMA Nodes (as a power of 2)"
+   range 1 10
+   default "2"
+   depends on NEED_MULTIPLE_NODES
+   help
+ Specify the maximum number of NUMA Nodes available on the target
+ system.  Increases memory reserved to accommodate various tables.
+
+config USE_PERCPU_NUMA_NODE_ID
+   def_bool y
+   depends on NUMA
+
 source kernel/Kconfig.preempt
 source kernel/Kconfig.hz
 
diff --git a/arch/arm64/include/asm/mmzone.h b/arch/arm64/include/asm/mmzone.h
new file mode 100644
index 000..a0de9e6
--- /dev/null
+++ b/arch/arm64/include/asm/mmzone.h
@@ -0,0 +1,12 @@
+#ifndef __ASM_MMZONE_H
+#define __ASM_MMZONE_H
+
+#ifdef CONFIG_NUMA
+
+#include 
+
+extern struct pglist_data *node_data[];
+#define NODE_DATA(nid) (node_data[(nid)])
+
+#endif /* CONFIG_NUMA */
+#endif /* __ASM_MMZONE_H */
diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h
new file mode 100644
index 000..574267f
--- /dev/null
+++ b/arch/arm64/include/asm/numa.h
@@ -0,0 +1,43 @@
+#ifndef __ASM_NUMA_H
+#define __ASM_NUMA_H
+
+#include 
+
+#ifdef CONFIG_NUMA
+
+/* currently, arm64 implements flat NUMA topology */
+#define parent_node(node)  (node)
+
+int __node_distance(int from, int to);
+#define node_distance(a, b) __node_distance(a, b)
+
+extern nodemask_t numa_nodes_parsed __initdata;
+
+/* Mappings between node number and cpus on that node. */
+extern cpumask_var_t node_to_cpumask_map[MAX_NUMNODES];
+void numa_clear_node(unsigned int cpu);
+
+#ifdef CONFIG_DEBUG_PER_CPU_MAPS
+const struct cpumask *cpumask_of_node(int node);
+#else
+/* Returns a pointer to the cpumask of CPUs on Node 'node'. */
+static inline const struct cpumask *cpumask_of_node(int node)
+{
+   return node_to_cpumask_map[node];
+}
+#endif
+
+void __init arm64_numa_init(void);
+int __init numa_add_memblk(int nodeid, u64 start, u64 end);
+void __init numa_set_distance(int from, int to, int distance);
+void __init numa_free_distance(void);
+void numa_store_cpu_info(unsigned int cpu);
+
+#else  /* CONFIG_NUMA */
+
+static inline void numa_store_cpu_info(unsigned int cpu) { }
+static inline void arm64_numa_init(void) { }
+
+#endif /* CONFIG_NUMA */
+
+#endif /* __ASM_NUMA_H */
diff --git a/arch/arm64/include/asm/topology.h 
b/arch/arm64/include/asm/topology.h
index a3e9d6f..8b57339 100644
--- a/arch/arm64/include/asm/topology.h
+++ b/arch/arm64/include/asm/topology.h
@@ -22,6 +22,16 @@ void init_cpu_topology(void);
 void store_cpu_topology(unsigned int cpuid);
 const struct cpumask *cpu_coregroup_mask(int cpu);
 
+#ifdef CONFIG_NUMA
+
+struct pci_bus;
+int pcibus_to_node(struct pci_bus *bus);
+#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?\
+cpu_all_mask : \
+cpumask_of_node(pcibus_to_nod

[PATCH v10 3/8] dt, numa: adding numa dt binding implementation.

[Ganapatrao Kulkarni]

dt node parsing for numa topology is done using device property
numa-node-id and device node distance-map.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 drivers/of/Kconfig   |  11 +++
 drivers/of/Makefile  |   1 +
 drivers/of/of_numa.c | 207 +++
 include/linux/of.h   |   4 +
 4 files changed, 223 insertions(+)
 create mode 100644 drivers/of/of_numa.c

diff --git a/drivers/of/Kconfig b/drivers/of/Kconfig
index e2a4841..8f9cc3a 100644
--- a/drivers/of/Kconfig
+++ b/drivers/of/Kconfig
@@ -112,4 +112,15 @@ config OF_OVERLAY
  While this option is selected automatically when needed, you can
  enable it manually to improve device tree unit test coverage.
 
+config OF_NUMA
+   bool "Device Tree NUMA support"
+   depends on NUMA
+   depends on OF
+   depends on ARM64
+   default y
+   help
+ Enable Device Tree NUMA support.
+ This enables the numa mapping of cpu, memory, io and
+ inter node distances using dt bindings.
+
 endif # OF
diff --git a/drivers/of/Makefile b/drivers/of/Makefile
index 156c072..bee3fa9 100644
--- a/drivers/of/Makefile
+++ b/drivers/of/Makefile
@@ -14,5 +14,6 @@ obj-$(CONFIG_OF_MTD)  += of_mtd.o
 obj-$(CONFIG_OF_RESERVED_MEM) += of_reserved_mem.o
 obj-$(CONFIG_OF_RESOLVE)  += resolver.o
 obj-$(CONFIG_OF_OVERLAY) += overlay.o
+obj-$(CONFIG_OF_NUMA) += of_numa.o
 
 obj-$(CONFIG_OF_UNITTEST) += unittest-data/
diff --git a/drivers/of/of_numa.c b/drivers/of/of_numa.c
new file mode 100644
index 000..1142cdb
--- /dev/null
+++ b/drivers/of/of_numa.c
@@ -0,0 +1,207 @@
+/*
+ * OF NUMA Parsing support.
+ *
+ * Copyright (C) 2015 Cavium Inc.
+ * Author: Ganapatrao Kulkarni <gkulka...@cavium.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
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * 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 
+#include 
+
+/* define default numa node to 0 */
+#define DEFAULT_NODE 0
+
+/* Returns nid in the range [0..MAX_NUMNODES-1],
+ * or NUMA_NO_NODE if no valid numa-node-id entry found
+ * or DEFAULT_NODE if no numa-node-id entry exists
+ */
+static int of_numa_prop_to_nid(const __be32 *of_numa_prop, int length)
+{
+   int nid;
+
+   if (!of_numa_prop)
+   return DEFAULT_NODE;
+
+   if (length != sizeof(*of_numa_prop)) {
+   pr_warn("NUMA: Invalid of_numa_prop length %d found.\n",
+   length);
+   return NUMA_NO_NODE;
+   }
+
+   nid = of_read_number(of_numa_prop, 1);
+   if (nid >= MAX_NUMNODES) {
+   pr_warn("NUMA: Invalid numa node %d found.\n", nid);
+   return NUMA_NO_NODE;
+   }
+
+   return nid;
+}
+
+static int __init early_init_of_node_to_nid(unsigned long node)
+{
+   int length;
+   const __be32 *of_numa_prop;
+
+   of_numa_prop = of_get_flat_dt_prop(node, "numa-node-id", );
+
+   return of_numa_prop_to_nid(of_numa_prop, length);
+}
+
+/*
+ * Even though we connect cpus to numa domains later in SMP
+ * init, we need to know the node ids now for all cpus.
+*/
+static int __init early_init_parse_cpu_node(unsigned long node)
+{
+   int nid;
+   const char *type = of_get_flat_dt_prop(node, "device_type", NULL);
+
+   if (type == NULL)
+   return 0;
+
+   if (strcmp(type, "cpu") != 0)
+   return 0;
+
+   nid = early_init_of_node_to_nid(node);
+   if (nid == NUMA_NO_NODE)
+   return -EINVAL;
+
+   node_set(nid, numa_nodes_parsed);
+   return 0;
+}
+
+static int __init early_init_parse_memory_node(unsigned long node)
+{
+   const __be32 *reg, *endp;
+   int length;
+   int nid;
+   const char *type = of_get_flat_dt_prop(node, "device_type", NULL);
+
+   if (type == NULL)
+   return 0;
+
+   if (strcmp(type, "memory") != 0)
+   return 0;
+
+   nid = early_init_of_node_to_nid(node);
+   if (nid == NUMA_NO_NODE)
+   return -EINVAL;
+
+   reg = of_get_flat_dt_prop(node, "reg", );
+   endp = reg + (length / sizeof(__be32));
+
+   while ((endp - reg) >= (dt_root_addr_cells + dt_root_size_cells)) {
+   u64 base, size;
+
+   base = dt_mem_next_cell(dt_root_addr_cells, );
+

[PATCH v10 4/8] arm64, numa : Enable numa dt for arm64 platforms.

[Ganapatrao Kulkarni]

Adding numa dt binding support for arm64 based platforms.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/include/asm/numa.h |  2 ++
 arch/arm64/kernel/smp.c   |  2 ++
 arch/arm64/mm/numa.c  | 17 +
 3 files changed, 21 insertions(+)

diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h
index 574267f..e9b4f29 100644
--- a/arch/arm64/include/asm/numa.h
+++ b/arch/arm64/include/asm/numa.h
@@ -31,12 +31,14 @@ void __init arm64_numa_init(void);
 int __init numa_add_memblk(int nodeid, u64 start, u64 end);
 void __init numa_set_distance(int from, int to, int distance);
 void __init numa_free_distance(void);
+void __init early_map_cpu_to_node(unsigned int cpu, int nid);
 void numa_store_cpu_info(unsigned int cpu);
 
 #else  /* CONFIG_NUMA */
 
 static inline void numa_store_cpu_info(unsigned int cpu) { }
 static inline void arm64_numa_init(void) { }
+static inline void early_map_cpu_to_node(unsigned int cpu, int nid) { }
 
 #endif /* CONFIG_NUMA */
 
diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c
index d6e7d6a..46c45c8 100644
--- a/arch/arm64/kernel/smp.c
+++ b/arch/arm64/kernel/smp.c
@@ -520,6 +520,8 @@ static void __init of_parse_and_init_cpus(void)
 
pr_debug("cpu logical map 0x%llx\n", hwid);
cpu_logical_map(cpu_count) = hwid;
+
+   early_map_cpu_to_node(cpu_count, of_node_to_nid(dn));
 next:
cpu_count++;
}
diff --git a/arch/arm64/mm/numa.c b/arch/arm64/mm/numa.c
index 60f19a2..7aad5d4 100644
--- a/arch/arm64/mm/numa.c
+++ b/arch/arm64/mm/numa.c
@@ -20,6 +20,7 @@
 #include 
 #include 
 #include 
+#include 
 
 struct pglist_data *node_data[MAX_NUMNODES] __read_mostly;
 EXPORT_SYMBOL(node_data);
@@ -122,6 +123,15 @@ void numa_store_cpu_info(unsigned int cpu)
map_cpu_to_node(cpu, numa_off ? 0 : cpu_to_node_map[cpu]);
 }
 
+void __init early_map_cpu_to_node(unsigned int cpu, int nid)
+{
+   /* fallback to node 0 */
+   if (nid < 0 || nid >= MAX_NUMNODES)
+   nid = 0;
+
+   cpu_to_node_map[cpu] = nid;
+}
+
 /**
  * numa_add_memblk - Set node id to memblk
  * @nid: NUMA node ID of the new memblk
@@ -383,5 +393,12 @@ static int __init dummy_numa_init(void)
  */
 void __init arm64_numa_init(void)
 {
+   if (!numa_off) {
+#ifdef CONFIG_OF_NUMA
+   if (!numa_init(of_numa_init))
+   return;
+#endif
+   }
+
numa_init(dummy_numa_init);
 }
-- 
1.8.1.4

___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

[PATCH v10 5/8] arm64, dt, thunderx: Add initial dts for Cavium Thunderx in 2 node topology.

[Ganapatrao Kulkarni]

Adding dt file for Cavium's Thunderx dual socket platform.

Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/boot/dts/cavium/Makefile |   2 +-
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts  |  83 +++
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi | 806 
 3 files changed, 890 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts
 create mode 100644 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi

diff --git a/arch/arm64/boot/dts/cavium/Makefile 
b/arch/arm64/boot/dts/cavium/Makefile
index e34f89d..7fe7067 100644
--- a/arch/arm64/boot/dts/cavium/Makefile
+++ b/arch/arm64/boot/dts/cavium/Makefile
@@ -1,4 +1,4 @@
-dtb-$(CONFIG_ARCH_THUNDER) += thunder-88xx.dtb
+dtb-$(CONFIG_ARCH_THUNDER) += thunder-88xx.dtb thunder-88xx-2n.dtb
 
 always := $(dtb-y)
 subdir-y   := $(dts-dirs)
diff --git a/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts 
b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts
new file mode 100644
index 000..5601e87
--- /dev/null
+++ b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts
@@ -0,0 +1,83 @@
+/*
+ * Cavium Thunder DTS file - Thunder board description
+ *
+ * Copyright (C) 2014, Cavium Inc.
+ *
+ * This file is dual-licensed: you can use it either under the terms
+ * of the GPL or the X11 license, at your option. Note that this dual
+ * licensing only applies to this file, and not this project as a
+ * whole.
+ *
+ *  a) This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this library; if not, write to the Free
+ * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston,
+ * MA 02110-1301 USA
+ *
+ * Or, alternatively,
+ *
+ *  b) Permission is hereby granted, free of charge, to any person
+ * obtaining a copy of this software and associated documentation
+ * files (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use,
+ * copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following
+ * conditions:
+ *
+ * The above copyright notice and this permission notice shall be
+ * included in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
+ * OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
+ * HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
+ * WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+/dts-v1/;
+
+/include/ "thunder-88xx-2n.dtsi"
+
+/ {
+   model = "Cavium ThunderX CN88XX board";
+   compatible = "cavium,thunder-88xx";
+
+   aliases {
+   serial0 = 
+   serial1 = 
+   };
+
+   memory@ {
+   device_type = "memory";
+   reg = <0x0 0x0140 0x3 0xFEC0>;
+   /* socket 0 */
+   numa-node-id = <0>;
+   };
+
+   memory@100 {
+   device_type = "memory";
+   reg = <0x100 0x0040 0x3 0xFFC0>;
+/* socket 1 */
+   numa-node-id = <1>;
+   };
+
+   distance-map {
+   compatible = "numa-distance-map-v1";
+   distance-matrix = <0 0  10>,
+ <0 1  20>,
+ <1 1  10>;
+   };
+};
diff --git a/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi 
b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi
new file mode 100644
index 000..b58e5c7
--- /dev/null
+++ b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi
@@ -0,0 +1,806 @@
+/*
+ * Cavium Thunder DTS file - Thunder SoC description
+ *
+ * Copyright (C) 2014, Cavium Inc.
+ *
+ * This file is dual-licensed: you can use it either under the terms
+ * of the GPL or the X11 license, at your option. Note that this dual
+ * licensing only applies to this file, and no

[PATCH v10 2/8] Documentation, dt, numa: dt bindings for numa.

[Ganapatrao Kulkarni]

DT bindings for numa mapping of memory, cores and IOs.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 Documentation/devicetree/bindings/numa.txt | 272 +
 1 file changed, 272 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/numa.txt

diff --git a/Documentation/devicetree/bindings/numa.txt 
b/Documentation/devicetree/bindings/numa.txt
new file mode 100644
index 000..ec5ed7c
--- /dev/null
+++ b/Documentation/devicetree/bindings/numa.txt
@@ -0,0 +1,272 @@
+==
+NUMA binding description.
+==
+
+==
+1 - Introduction
+==
+
+Systems employing a Non Uniform Memory Access (NUMA) architecture contain
+collections of hardware resources including processors, memory, and I/O buses,
+that comprise what is commonly known as a NUMA node.
+Processor accesses to memory within the local NUMA node is generally faster
+than processor accesses to memory outside of the local NUMA node.
+DT defines interfaces that allow the platform to convey NUMA node
+topology information to OS.
+
+==
+2 - numa-node-id
+==
+
+For the purpose of identification, each NUMA node is associated with a unique
+token known as a node id. For the purpose of this binding
+a node id is a 32-bit integer.
+
+A device node is associated with a NUMA node by the presence of a
+numa-node-id property which contains the node id of the device.
+
+Example:
+   /* numa node 0 */
+   numa-node-id = <0>;
+
+   /* numa node 1 */
+   numa-node-id = <1>;
+
+==
+3 - distance-map
+==
+
+The device tree node distance-map describes the relative
+distance (memory latency) between all numa nodes.
+
+- compatible : Should at least contain "numa-distance-map-v1".
+
+- distance-matrix
+  This property defines a matrix to describe the relative distances
+  between all numa nodes.
+  It is represented as a list of node pairs and their relative distance.
+
+  Note:
+   1. Each entry represents distance from first node to second node.
+   The distances are equal in either direction.
+   2. The distance from a node to self (local distance) is represented
+   with value 10 and all internode distance should be represented with
+   a value greater than 10.
+   3. distance-matrix should have entries in lexicographical ascending
+   order of nodes.
+   4. There must be only one device node distance-map which must reside in 
the root node.
+
+Example:
+   4 nodes connected in mesh/ring topology as below,
+
+   0___20__1
+   |   |
+   |   |
+   20 20
+   |   |
+   |   |
+   |___|
+   3   20  2
+
+   if relative distance for each hop is 20,
+   then internode distance would be,
+ 0 -> 1 = 20
+ 1 -> 2 = 20
+ 2 -> 3 = 20
+ 3 -> 0 = 20
+ 0 -> 2 = 40
+ 1 -> 3 = 40
+
+ and dt presentation for this distance matrix is,
+
+   distance-map {
+compatible = "numa-distance-map-v1";
+distance-matrix = <0 0  10>,
+  <0 1  20>,
+  <0 2  40>,
+  <0 3  20>,
+  <1 0  20>,
+  <1 1  10>,
+  <1 2  20>,
+  <1 3  40>,
+  <2 0  40>,
+  <2 1  20>,
+  <2 2  10>,
+  <2 3  20>,
+  <3 0  20>,
+  <3 1  40>,
+  <3 2  20>,
+  <3 3  10>;
+   };
+
+==
+4 - Example dts
+=

[PATCH v10 0/8] arm64, numa: Add numa support for arm64 platforms

[Ganapatrao Kulkarni]

v10:
- Incorporated review comments from Rob Herring.
- Moved numa binding and implementation to devicetree core.
- Added cleanup patch to remove redundant NODE_DATA macro from asm 
header files
- Include numa balancing support for arm64 patch in this series.
- Fix tile build issue reported by the kbuild robot(patch 7)

v9: - Added cleanup patch to reuse and avoid redefinition of 
cpumask_of_pcibus
  as suggested from Will Deacon and Bjorn Helgaas.
- Including patch to Make pci-host-generic driver numa aware.
- Incorporated comment from Shannon Zhao.

v8:
- Incorporated review comments of Mark Rutland and Will Deacon.
- Added pci helper function and macro for numa.

v7:
- managing numa memory mapping using memblock.
- Incorporated review comments of Mark Rutland.

v6:
- defined and implemented the numa dt binding using
node property proximity and device node distance-map.
- renamed dt_numa to of_numa

v5:
- created base verion of numa.c which creates dummy numa without using 
dt
  on single socket platforms. Then added patches for dt support.
- Incorporated review comments from Hanjun Guo.

v4:
done changes as per Arnd review comments.

v3:
Added changes to support numa on arm64 based platforms.
Tested these patches on cavium's multinode(2 node topology) platform.
In this patchset, defined and implemented dt bindings for numa mapping
for core and memory using device node property arm,associativity.

v2:
Defined and implemented numa map for memory, cores to node and
proximity distance matrix of nodes.

v1:
Initial patchset to support numa on arm64 platforms.

Note:
1. This patchset is tested for numa with dt on
   thunderx single socket and dual socket boards.
2. Numa DT booting needs the dt memory nodes, which are deleted in 
current efi-stub,
hence to try numa with dt, you need to rebase with ard's patchset.

http://git.linaro.org/people/ard.biesheuvel/linux-arm.git/shortlog/refs/heads/arm64-uefi-early-fdt-handling
3. PATCH[7,8] are not tested for other architectures.

Ganapatrao Kulkarni (8):
  arm64, numa: adding numa support for arm64 platforms.
  Documentation, dt, numa: dt bindings for numa.
  dt, numa: adding numa dt binding implementation.
  arm64, numa : Enable numa dt for arm64 platforms.
  arm64, dt, thunderx: Add initial dts for Cavium Thunderx in 2 node
topology.
  arm64, mm, numa: Adding numa balancing support for arm64.
  topology, cleanup: Avoid redefinition of cpumask_of_pcibus in asm
header files.
  numa, mm, cleanup: remove redundant NODE_DATA macro from asm header
files.

 Documentation/devicetree/bindings/numa.txt  | 272 
 arch/arm64/Kconfig  |  26 +
 arch/arm64/boot/dts/cavium/Makefile |   2 +-
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts  |  83 +++
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi | 806 
 arch/arm64/include/asm/mmzone.h |  10 +
 arch/arm64/include/asm/numa.h   |  45 ++
 arch/arm64/include/asm/pgtable.h|  18 +
 arch/arm64/include/asm/topology.h   |   7 +
 arch/arm64/kernel/pci.c |  10 +
 arch/arm64/kernel/setup.c   |   4 +
 arch/arm64/kernel/smp.c |   4 +
 arch/arm64/mm/Makefile  |   1 +
 arch/arm64/mm/init.c|  34 +-
 arch/arm64/mm/mmu.c |   1 +
 arch/arm64/mm/numa.c| 404 
 arch/ia64/include/asm/topology.h|   4 -
 arch/m32r/include/asm/mmzone.h  |   4 +-
 arch/metag/include/asm/mmzone.h |   4 +-
 arch/metag/include/asm/topology.h   |   3 -
 arch/powerpc/include/asm/mmzone.h   |   8 +-
 arch/powerpc/include/asm/topology.h |   4 -
 arch/s390/include/asm/mmzone.h  |   6 +-
 arch/s390/include/asm/pci.h |   2 +-
 arch/s390/include/asm/topology.h|   1 +
 arch/sh/include/asm/mmzone.h|   4 +-
 arch/sh/include/asm/topology.h  |   3 -
 arch/sparc/include/asm/mmzone.h |   6 +-
 arch/tile/include/asm/pci.h |   2 -
 arch/tile/include/asm/topology.h|   3 +
 arch/x86/include/asm/mmzone.h   |   3 +-
 arch/x86/include/asm/mmzone_32.h|   5 -
 arch/x86/include/asm/mmzone_64.h|  17 -
 arch/x86/include/asm/pci.h  |   2 +-
 arch/x86/include/asm/topology.h |   1 +
 drivers/of/Kconfig  |  11 +
 drivers/of/Makefile |   1 +
 drivers/of/of_numa.c| 207 ++
 include/asm-generic

[PATCH v10 6/8] arm64, mm, numa: Adding numa balancing support for arm64.

[Ganapatrao Kulkarni]

enabled numa balancing for arm64 platforms.
added pte, pmd protnone helpers for use by automatic NUMA balancing.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/Kconfig   |  1 +
 arch/arm64/include/asm/pgtable.h | 18 ++
 2 files changed, 19 insertions(+)

diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index fcf3950..8a7c02a 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -11,6 +11,7 @@ config ARM64
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_USE_CMPXCHG_LOCKREF
select ARCH_SUPPORTS_ATOMIC_RMW
+   select ARCH_SUPPORTS_NUMA_BALANCING if NUMA
select ARCH_WANT_OPTIONAL_GPIOLIB
select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
select ARCH_WANT_FRAME_POINTERS
diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
index 2d545d7a..24ce546 100644
--- a/arch/arm64/include/asm/pgtable.h
+++ b/arch/arm64/include/asm/pgtable.h
@@ -347,6 +347,24 @@ static inline pgprot_t mk_sect_prot(pgprot_t prot)
return __pgprot(pgprot_val(prot) & ~PTE_TABLE_BIT);
 }
 
+#ifdef CONFIG_NUMA_BALANCING
+
+/*
+ * These work without NUMA balancing but the kernel does not care. See the
+ * comment in include/asm-generic/pgtable.h
+ */
+static inline int pte_protnone(pte_t pte)
+{
+   return ((pte_val(pte) & (PTE_VALID | PTE_PROT_NONE)) == PTE_PROT_NONE);
+}
+
+static inline int pmd_protnone(pmd_t pmd)
+{
+   return pte_protnone(pmd_pte(pmd));
+}
+
+#endif /* CONFIG_NUMA_BALANCING */
+
 /*
  * THP definitions.
  */
-- 
1.8.1.4

___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

[PATCH v10 8/8] numa, mm, cleanup: remove redundant NODE_DATA macro from asm header files.

[Ganapatrao Kulkarni]

NODE_DATA is defined across multiple asm header files.
Moving generic definition to asm-generic/mmzone.h to
remove redundant definitions.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/include/asm/mmzone.h   |  4 +---
 arch/m32r/include/asm/mmzone.h|  4 +---
 arch/metag/include/asm/mmzone.h   |  4 +---
 arch/powerpc/include/asm/mmzone.h |  8 ++--
 arch/s390/include/asm/mmzone.h|  6 +-
 arch/sh/include/asm/mmzone.h  |  4 +---
 arch/sparc/include/asm/mmzone.h   |  6 ++
 arch/x86/include/asm/mmzone.h |  3 +--
 arch/x86/include/asm/mmzone_32.h  |  5 -
 arch/x86/include/asm/mmzone_64.h  | 17 -
 include/asm-generic/mmzone.h  | 24 
 11 files changed, 34 insertions(+), 51 deletions(-)
 delete mode 100644 arch/x86/include/asm/mmzone_64.h
 create mode 100644 include/asm-generic/mmzone.h

diff --git a/arch/arm64/include/asm/mmzone.h b/arch/arm64/include/asm/mmzone.h
index a0de9e6..611a1cf 100644
--- a/arch/arm64/include/asm/mmzone.h
+++ b/arch/arm64/include/asm/mmzone.h
@@ -4,9 +4,7 @@
 #ifdef CONFIG_NUMA
 
 #include 
-
-extern struct pglist_data *node_data[];
-#define NODE_DATA(nid) (node_data[(nid)])
+#include  
 
 #endif /* CONFIG_NUMA */
 #endif /* __ASM_MMZONE_H */
diff --git a/arch/m32r/include/asm/mmzone.h b/arch/m32r/include/asm/mmzone.h
index 115ced3..e3d66a0 100644
--- a/arch/m32r/include/asm/mmzone.h
+++ b/arch/m32r/include/asm/mmzone.h
@@ -7,12 +7,10 @@
 #define _ASM_MMZONE_H_
 
 #include 
+#include  
 
 #ifdef CONFIG_DISCONTIGMEM
 
-extern struct pglist_data *node_data[];
-#define NODE_DATA(nid) (node_data[nid])
-
 #define node_localnr(pfn, nid) ((pfn) - NODE_DATA(nid)->node_start_pfn)
 
 #define pmd_page(pmd)  (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT))
diff --git a/arch/metag/include/asm/mmzone.h b/arch/metag/include/asm/mmzone.h
index 9c88a9c..b1e95b3 100644
--- a/arch/metag/include/asm/mmzone.h
+++ b/arch/metag/include/asm/mmzone.h
@@ -3,9 +3,7 @@
 
 #ifdef CONFIG_NEED_MULTIPLE_NODES
 #include 
-
-extern struct pglist_data *node_data[];
-#define NODE_DATA(nid) (node_data[nid])
+#include  
 
 static inline int pfn_to_nid(unsigned long pfn)
 {
diff --git a/arch/powerpc/include/asm/mmzone.h 
b/arch/powerpc/include/asm/mmzone.h
index 7b58917..da0c5ba 100644
--- a/arch/powerpc/include/asm/mmzone.h
+++ b/arch/powerpc/include/asm/mmzone.h
@@ -19,12 +19,6 @@
 
 #ifdef CONFIG_NEED_MULTIPLE_NODES
 
-extern struct pglist_data *node_data[];
-/*
- * Return a pointer to the node data for node n.
- */
-#define NODE_DATA(nid) (node_data[nid])
-
 /*
  * Following are specific to this numa platform.
  */
@@ -42,5 +36,7 @@ u64 memory_hotplug_max(void);
 #define memory_hotplug_max() memblock_end_of_DRAM()
 #endif /* CONFIG_NEED_MULTIPLE_NODES */
 
+#include  
+
 #endif /* __KERNEL__ */
 #endif /* _ASM_MMZONE_H_ */
diff --git a/arch/s390/include/asm/mmzone.h b/arch/s390/include/asm/mmzone.h
index a9e834e..91f1fcc 100644
--- a/arch/s390/include/asm/mmzone.h
+++ b/arch/s390/include/asm/mmzone.h
@@ -7,10 +7,6 @@
 #ifndef _ASM_S390_MMZONE_H
 #define _ASM_S390_MMZONE_H
 
-#ifdef CONFIG_NUMA
+#include  
 
-extern struct pglist_data *node_data[];
-#define NODE_DATA(nid) (node_data[nid])
-
-#endif /* CONFIG_NUMA */
 #endif /* _ASM_S390_MMZONE_H */
diff --git a/arch/sh/include/asm/mmzone.h b/arch/sh/include/asm/mmzone.h
index 15a8496..c070d00 100644
--- a/arch/sh/include/asm/mmzone.h
+++ b/arch/sh/include/asm/mmzone.h
@@ -5,9 +5,7 @@
 
 #ifdef CONFIG_NEED_MULTIPLE_NODES
 #include 
-
-extern struct pglist_data *node_data[];
-#define NODE_DATA(nid) (node_data[nid])
+#include  
 
 static inline int pfn_to_nid(unsigned long pfn)
 {
diff --git a/arch/sparc/include/asm/mmzone.h b/arch/sparc/include/asm/mmzone.h
index 99d9b9f..ef1365b 100644
--- a/arch/sparc/include/asm/mmzone.h
+++ b/arch/sparc/include/asm/mmzone.h
@@ -5,13 +5,11 @@
 
 #include 
 
-extern struct pglist_data *node_data[];
-
-#define NODE_DATA(nid) (node_data[nid])
-
 extern int numa_cpu_lookup_table[];
 extern cpumask_t numa_cpumask_lookup_table[];
 
 #endif /* CONFIG_NEED_MULTIPLE_NODES */
 
+#include  
+
 #endif /* _SPARC64_MMZONE_H */
diff --git a/arch/x86/include/asm/mmzone.h b/arch/x86/include/asm/mmzone.h
index d497bc4..5a52815 100644
--- a/arch/x86/include/asm/mmzone.h
+++ b/arch/x86/include/asm/mmzone.h
@@ -1,5 +1,4 @@
 #ifdef CONFIG_X86_32
 # include 
-#else
-# include 
 #endif
+#include  
diff --git a/arch/x86/include/asm/mmzone_32.h b/arch/x86/include/asm/mmzone_32.h
index 1ec990b..09f7cfb 100644
--- a/arch/x86/include/asm/mmzone_32.h
+++ b/arch/x86/include/asm/mmzone_32.h
@@ -8,11 +8,6 @@
 
 #include 
 
-#ifdef CONFIG_NUMA
-extern struct pglist_data *node_data[];
-#define NODE_DATA(nid) (node_data[nid])
-#endif /* CONFIG_NUMA */
-
 #ifdef CONFIG_DISCONTIGMEM
 
 /*
diff --git a/arch/x86/include/asm/mmzone_64.h b/arch/x86/in

Re: [PATCH v9 3/6] arm64/arm, numa, dt: adding numa dt binding implementation for arm64 platforms.

[Ganapatrao Kulkarni]

On Thu, Jan 28, 2016 at 11:38 PM, Will Deacon <will.dea...@arm.com> wrote:
> On Thu, Jan 28, 2016 at 10:42:17PM +0530, Ganapatrao Kulkarni wrote:
>> On Thu, Jan 28, 2016 at 8:09 PM, Will Deacon <will.dea...@arm.com> wrote:
>> > On Tue, Jan 26, 2016 at 02:36:04PM -0600, Bjorn Helgaas wrote:
>> >> Subject is "arm64/arm, numa, dt: adding ..."  What is the significance
>> >> of the "arm" part?  The other patches only mention "arm64".
>> >>
>> >> General comment: the code below has little, if anything, that is
>> >> actually arm64-specific.  Maybe this is the first DT-based NUMA
>> >> platform?  I don't see other similar code for other arches, so maybe
>> >> it's too early to try to generalize it, but we should try to avoid
>> >> adding duplicates of this code if/when other arches do show up.
>> >
>> > Having it in the core code would allow us to share it with arch/arm/
>> > fairly straightforwardly.
>> This binding can be used for arm too.
>> however at this moment it is the need of arm64 platforms.
>> can we please keep this to arm64 as it's too early to try to
>> generalize it(as Bjorn suggested)
>> I prefer to keep it as it is, otherwise ok.
>> Please suggest.
>
> My suggestions time and time again on the NUMA patches from you have
> consistently been around consolidation of existing code, or moving things
> that aren't architecture-specific out of the architecture code.

thanks, i shall move this out to drivers/of
>
> Will

thanks
Ganapat
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Re: [PATCH v9 3/6] arm64/arm, numa, dt: adding numa dt binding implementation for arm64 platforms.

[Ganapatrao Kulkarni]

Hi Will,


On Thu, Jan 28, 2016 at 8:09 PM, Will Deacon  wrote:
> On Tue, Jan 26, 2016 at 02:36:04PM -0600, Bjorn Helgaas wrote:
>> Subject is "arm64/arm, numa, dt: adding ..."  What is the significance
>> of the "arm" part?  The other patches only mention "arm64".
>>
>> General comment: the code below has little, if anything, that is
>> actually arm64-specific.  Maybe this is the first DT-based NUMA
>> platform?  I don't see other similar code for other arches, so maybe
>> it's too early to try to generalize it, but we should try to avoid
>> adding duplicates of this code if/when other arches do show up.
>
> Having it in the core code would allow us to share it with arch/arm/
> fairly straightforwardly.
This binding can be used for arm too.
however at this moment it is the need of arm64 platforms.
can we please keep this to arm64 as it's too early to try to
generalize it(as Bjorn suggested)
I prefer to keep it as it is, otherwise ok.
Please suggest.
>
> Will
thanks
Ganapat
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Re: [PATCH v9 2/6] Documentation, dt, arm64/arm: dt bindings for numa.

[Ganapatrao Kulkarni]

Hi Rob, Mark,


On Wed, Jan 20, 2016 at 7:48 PM, Rob Herring <r...@kernel.org> wrote:
> On Mon, Jan 18, 2016 at 10:06:01PM +0530, Ganapatrao Kulkarni wrote:
>> DT bindings for numa mapping of memory, cores and IOs.
>>
>> Reviewed-by: Robert Richter <rrich...@cavium.com>
>> Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
>> ---
>>  Documentation/devicetree/bindings/arm/numa.txt | 272 
>> +
>>  1 file changed, 272 insertions(+)
>>  create mode 100644 Documentation/devicetree/bindings/arm/numa.txt
>
> This is looks okay to me, but some cosmetic things on the example.
can i have your Ack please?
>
>> +==
>> +4 - Example dts
>> +==
>> +
>> +2 sockets system consists of 2 boards connected through ccn bus and
>> +each board having one socket/soc of 8 cpus, memory and pci bus.
>> +
>> + memory@00c0 {
>
> Drop the leading 0s on unit addresses.
i will correct these in next version.
>
>> + device_type = "memory";
>> + reg = <0x0 0x00c0 0x0 0x8000>;
>> + /* node 0 */
>> + numa-node-id = <0>;
>> + };
>> +
>> + memory@100 {
>> + device_type = "memory";
>> + reg = <0x100 0x 0x0 0x8000>;
>> + /* node 1 */
>> + numa-node-id = <1>;
>> + };
>> +
>> + cpus {
>> + #address-cells = <2>;
>> + #size-cells = <0>;
>> +
>> + cpu@000 {
>
> Same here (leaving one of course).
>
>> + device_type = "cpu";
>> + compatible =  "arm,armv8";
>> + reg = <0x0 0x000>;
>> + enable-method = "psci";
>> + /* node 0 */
>> + numa-node-id = <0>;
>> + };
>> + cpu@001 {
>
> and so on...
>
>> + device_type = "cpu";
>> + compatible =  "arm,armv8";
>> + reg = <0x0 0x001>;
>
> Either all leading 0s or none.
>
>> + reg = <0x0 0x008>;
>> + enable-method = "psci";
>> + /* node 1 */
>
> Kind of a pointless comment.
>
> Wouldn't each cluster of cpus for a given numa node be in a different
> cpu affinity? Certainly not required by the architecture, but the common
> case at least.
>
>> + numa-node-id = <1>;
>> + };
>
> [...]
>
>> + pcie0: pcie0@0x8480, {
>
> Drop the 0x and the comma.
>
>> + compatible = "arm,armv8";
>> + device_type = "pci";
>> + bus-range = <0 255>;
>> + #size-cells = <2>;
>> + #address-cells = <3>;
>> + reg = <0x8480 0x 0 0x1000>;  /* Configuration 
>> space */
>> + ranges = <0x0300 0x8010 0x 0x8010 0x 0x70 
>> 0x>;
>> + /* node 0 */
>> + numa-node-id = <0>;
>> +};
>> +
>> + pcie1: pcie1@0x9480, {
>
> ditto
>
>> + compatible = "arm,armv8";
>> + device_type = "pci";
>> + bus-range = <0 255>;
>> + #size-cells = <2>;
>> + #address-cells = <3>;
>> + reg = <0x9480 0x 0 0x1000>;  /* Configuration 
>> space */
>> + ranges = <0x0300 0x9010 0x 0x9010 0x 0x70 
>> 0x>;
>> + /* node 1 */
>> + numa-node-id = <1>;
>> +};
>> +
>> + distance-map {
>> + compatible = "numa-distance-map-v1";
>> + distance-matrix = <0 0 10>,
>> +   <0 1 20>,
>> +   <1 1 10>;
>> + };
>> --
>> 1.8.1.4
thanks
Ganapat
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe devicetree" in
>> the body of a message to majord...@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Re: [PATCH v9 5/6] PCI: generic: Make pci-host-generic driver numa aware

[Ganapatrao Kulkarni]

On Tue, Jan 19, 2016 at 3:11 PM, Lorenzo Pieralisi
<lorenzo.pieral...@arm.com> wrote:
> On Tue, Jan 19, 2016 at 11:28:56AM +0530, Ganapatrao Kulkarni wrote:
>> On Mon, Jan 18, 2016 at 11:11 PM, David Daney <ddaney.c...@gmail.com> wrote:
>> > On 01/18/2016 08:36 AM, Ganapatrao Kulkarni wrote:
>> >>
>> >> update numa_node of device associated with pci bus.
>> >> moved down devm_kzalloc to allocate from node memory.
>> >>
>> >> Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
>> >> ---
>> >>   drivers/pci/host/pci-host-generic.c | 9 ++---
>> >>   1 file changed, 6 insertions(+), 3 deletions(-)
>> >>
>> >> diff --git a/drivers/pci/host/pci-host-generic.c
>> >> b/drivers/pci/host/pci-host-generic.c
>> >> index 5434c90..0e1ce06 100644
>> >> --- a/drivers/pci/host/pci-host-generic.c
>> >> +++ b/drivers/pci/host/pci-host-generic.c
>> >> @@ -215,11 +215,9 @@ static int gen_pci_probe(struct platform_device
>> >> *pdev)
>> >> const struct of_device_id *of_id;
>> >> struct device *dev = >dev;
>> >> struct device_node *np = dev->of_node;
>> >> -   struct gen_pci *pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
>> >> +   struct gen_pci *pci;
>> >> struct pci_bus *bus, *child;
>> >>
>> >> -   if (!pci)
>> >> -   return -ENOMEM;
>> >>
>> >> type = of_get_property(np, "device_type", NULL);
>> >> if (!type || strcmp(type, "pci")) {
>> >> @@ -230,6 +228,11 @@ static int gen_pci_probe(struct platform_device
>> >> *pdev)
>> >> of_pci_check_probe_only();
>> >>
>> >> of_id = of_match_node(gen_pci_of_match, np);
>> >> +   set_dev_node(dev, of_node_to_nid(np));
>> >
>> >
>> > This shouldn't be done in individual platform_drivers, but instead in the
>> > device probing code.
>> >
>> > There is code that does this in drivers/of/platform.c and
>> > drivers/of/device.c  Is that not being called for the pci-host-gweneric
>> > devices?  If not, we should figure out why, and perhaps attempt to fix it
>> > rather than doing it here.
>> is it more appropriate to call of_platform_device_create ?
>
> That's already done to create the platform device by OF core when
> populating devices from DT, what David suggested is that you should
> add set_dev_node() to core OF code instead of adding it specifically
> to the PCI host generic code.
This patch was worked on 4.2(inherited to 4.4) to have correct
numa_node populated on sysfs.
however, on 4.4, OF  is fixed to set numa_node(fixed by
56f2de81e020c537f7e35550d13840143cb765cd)

This patch (patch #5) is not required and will be dropped.
thanks for your comments.


>
> Lorenzo
thanks
Ganapat
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Re: [PATCH v9 5/6] PCI: generic: Make pci-host-generic driver numa aware

[Ganapatrao Kulkarni]

On Mon, Jan 18, 2016 at 11:11 PM, David Daney <ddaney.c...@gmail.com> wrote:
> On 01/18/2016 08:36 AM, Ganapatrao Kulkarni wrote:
>>
>> update numa_node of device associated with pci bus.
>> moved down devm_kzalloc to allocate from node memory.
>>
>> Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
>> ---
>>   drivers/pci/host/pci-host-generic.c | 9 ++---
>>   1 file changed, 6 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/pci/host/pci-host-generic.c
>> b/drivers/pci/host/pci-host-generic.c
>> index 5434c90..0e1ce06 100644
>> --- a/drivers/pci/host/pci-host-generic.c
>> +++ b/drivers/pci/host/pci-host-generic.c
>> @@ -215,11 +215,9 @@ static int gen_pci_probe(struct platform_device
>> *pdev)
>> const struct of_device_id *of_id;
>> struct device *dev = >dev;
>> struct device_node *np = dev->of_node;
>> -   struct gen_pci *pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
>> +   struct gen_pci *pci;
>> struct pci_bus *bus, *child;
>>
>> -   if (!pci)
>> -   return -ENOMEM;
>>
>> type = of_get_property(np, "device_type", NULL);
>> if (!type || strcmp(type, "pci")) {
>> @@ -230,6 +228,11 @@ static int gen_pci_probe(struct platform_device
>> *pdev)
>> of_pci_check_probe_only();
>>
>> of_id = of_match_node(gen_pci_of_match, np);
>> +   set_dev_node(dev, of_node_to_nid(np));
>
>
> This shouldn't be done in individual platform_drivers, but instead in the
> device probing code.
>
> There is code that does this in drivers/of/platform.c and
> drivers/of/device.c  Is that not being called for the pci-host-gweneric
> devices?  If not, we should figure out why, and perhaps attempt to fix it
> rather than doing it here.
is it more appropriate to call of_platform_device_create ?

--- a/drivers/pci/host/pci-host-generic.c
+++ b/drivers/pci/host/pci-host-generic.c
@@ -24,6 +24,7 @@
 #include 
 #include 
 #include 
+#include 

 struct gen_pci;

@@ -466,7 +467,7 @@ static int gen_pci_probe(struct platform_device *pdev)
of_pci_check_probe_only();

of_id = of_match_node(gen_pci_of_match, np);
-   set_dev_node(dev, of_node_to_nid(np));
+   of_platform_device_create(np, NULL, NULL);
pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
if (!pci)
return -ENOMEM;

>
>
>
>> +   pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
>> +   if (!pci)
>> +   return -ENOMEM;
>> +
>> pci->cfg.ops = (struct gen_pci_cfg_bus_ops *)of_id->data;
>> pci->host.dev.parent = dev;
>> INIT_LIST_HEAD(>host.windows);
>>
>
thanks
Ganapat
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

[PATCH v9 1/6] arm64, numa: adding numa support for arm64 platforms.

[Ganapatrao Kulkarni]

Adding numa support for arm64 based platforms.
This patch adds by default the dummy numa node and
maps all memory and cpus to node 0.
using this patch, numa can be simulated on single node arm64 platforms.

Tested-by: Shannon Zhao <shannon.z...@linaro.org>
Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/Kconfig|  25 +++
 arch/arm64/include/asm/mmzone.h   |  18 ++
 arch/arm64/include/asm/numa.h |  40 
 arch/arm64/include/asm/topology.h |  10 +
 arch/arm64/kernel/pci.c   |   8 +
 arch/arm64/kernel/setup.c |   4 +
 arch/arm64/kernel/smp.c   |   2 +
 arch/arm64/mm/Makefile|   1 +
 arch/arm64/mm/init.c  |  31 ++-
 arch/arm64/mm/mmu.c   |   1 +
 arch/arm64/mm/numa.c  | 387 ++
 11 files changed, 522 insertions(+), 5 deletions(-)
 create mode 100644 arch/arm64/include/asm/mmzone.h
 create mode 100644 arch/arm64/include/asm/numa.h
 create mode 100644 arch/arm64/mm/numa.c

diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 871f217..74f5d73 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -71,6 +71,7 @@ config ARM64
select HAVE_GENERIC_DMA_COHERENT
select HAVE_HW_BREAKPOINT if PERF_EVENTS
select HAVE_MEMBLOCK
+   select HAVE_MEMBLOCK_NODE_MAP if NUMA
select HAVE_PATA_PLATFORM
select HAVE_PERF_EVENTS
select HAVE_PERF_REGS
@@ -503,6 +504,30 @@ config HOTPLUG_CPU
  Say Y here to experiment with turning CPUs off and on.  CPUs
  can be controlled through /sys/devices/system/cpu.
 
+# Common NUMA Features
+config NUMA
+   bool "Numa Memory Allocation and Scheduler Support"
+   depends on SMP
+   help
+ Enable NUMA (Non Uniform Memory Access) support.
+
+ The kernel will try to allocate memory used by a CPU on the
+ local memory of the CPU and add some more
+ NUMA awareness to the kernel.
+
+config NODES_SHIFT
+   int "Maximum NUMA Nodes (as a power of 2)"
+   range 1 10
+   default "2"
+   depends on NEED_MULTIPLE_NODES
+   help
+ Specify the maximum number of NUMA Nodes available on the target
+ system.  Increases memory reserved to accommodate various tables.
+
+config USE_PERCPU_NUMA_NODE_ID
+   def_bool y
+   depends on NUMA
+
 source kernel/Kconfig.preempt
 source kernel/Kconfig.hz
 
diff --git a/arch/arm64/include/asm/mmzone.h b/arch/arm64/include/asm/mmzone.h
new file mode 100644
index 000..2cd804d
--- /dev/null
+++ b/arch/arm64/include/asm/mmzone.h
@@ -0,0 +1,18 @@
+#ifndef __ASM_MMZONE_H
+#define __ASM_MMZONE_H
+
+#ifdef CONFIG_NUMA
+
+#include 
+#include 
+
+#include 
+#include 
+
+extern struct pglist_data *node_data[];
+
+#define NODE_DATA(nid) (node_data[(nid)])
+
+#endif /* CONFIG_NUMA */
+
+#endif /* __ASM_MMZONE_H */
diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h
new file mode 100644
index 000..f28f15b0
--- /dev/null
+++ b/arch/arm64/include/asm/numa.h
@@ -0,0 +1,40 @@
+#ifndef __ASM_NUMA_H
+#define __ASM_NUMA_H
+
+#include 
+#include 
+
+#ifdef CONFIG_NUMA
+
+/* currently, arm64 implements flat NUMA topology */
+#define parent_node(node)  (node)
+
+extern int __node_distance(int from, int to);
+#define node_distance(a, b) __node_distance(a, b)
+
+extern int cpu_to_node_map[NR_CPUS];
+extern nodemask_t numa_nodes_parsed __initdata;
+
+/* Mappings between node number and cpus on that node. */
+extern cpumask_var_t node_to_cpumask_map[MAX_NUMNODES];
+extern void numa_clear_node(unsigned int cpu);
+#ifdef CONFIG_DEBUG_PER_CPU_MAPS
+extern const struct cpumask *cpumask_of_node(int node);
+#else
+/* Returns a pointer to the cpumask of CPUs on Node 'node'. */
+static inline const struct cpumask *cpumask_of_node(int node)
+{
+   return node_to_cpumask_map[node];
+}
+#endif
+
+void __init arm64_numa_init(void);
+int __init numa_add_memblk(int nodeid, u64 start, u64 end);
+void __init numa_set_distance(int from, int to, int distance);
+void __init numa_reset_distance(void);
+void numa_store_cpu_info(unsigned int cpu);
+#else  /* CONFIG_NUMA */
+static inline void numa_store_cpu_info(unsigned int cpu)   { }
+static inline void arm64_numa_init(void)   { }
+#endif /* CONFIG_NUMA */
+#endif /* __ASM_NUMA_H */
diff --git a/arch/arm64/include/asm/topology.h 
b/arch/arm64/include/asm/topology.h
index a3e9d6f..8b57339 100644
--- a/arch/arm64/include/asm/topology.h
+++ b/arch/arm64/include/asm/topology.h
@@ -22,6 +22,16 @@ void init_cpu_topology(void);
 void store_cpu_topology(unsigned int cpuid);
 const struct cpumask *cpu_coregroup_mask(int cpu);
 
+#ifdef CONFIG_NUMA
+
+struct pci_bus;
+int pcibus_to_node(struct pci_bus *bus);
+#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?   

[PATCH v9 2/6] Documentation, dt, arm64/arm: dt bindings for numa.

[Ganapatrao Kulkarni]

DT bindings for numa mapping of memory, cores and IOs.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 Documentation/devicetree/bindings/arm/numa.txt | 272 +
 1 file changed, 272 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/arm/numa.txt

diff --git a/Documentation/devicetree/bindings/arm/numa.txt 
b/Documentation/devicetree/bindings/arm/numa.txt
new file mode 100644
index 000..734e5b4
--- /dev/null
+++ b/Documentation/devicetree/bindings/arm/numa.txt
@@ -0,0 +1,272 @@
+==
+NUMA binding description.
+==
+
+==
+1 - Introduction
+==
+
+Systems employing a Non Uniform Memory Access (NUMA) architecture contain
+collections of hardware resources including processors, memory, and I/O buses,
+that comprise what is commonly known as a NUMA node.
+Processor accesses to memory within the local NUMA node is generally faster
+than processor accesses to memory outside of the local NUMA node.
+DT defines interfaces that allow the platform to convey NUMA node
+topology information to OS.
+
+==
+2 - numa-node-id
+==
+
+For the purpose of identification, each NUMA node is associated with a unique
+token known as a node id. For the purpose of this binding
+a node id is a 32-bit integer.
+
+A device node is associated with a NUMA node by the presence of a
+numa-node-id property which contains the node id of the device.
+
+Example:
+   /* numa node 0 */
+   numa-node-id = <0>;
+
+   /* numa node 1 */
+   numa-node-id = <1>;
+
+==
+3 - distance-map
+==
+
+The device tree node distance-map describes the relative
+distance (memory latency) between all numa nodes.
+
+- compatible : Should at least contain "numa-distance-map-v1".
+
+- distance-matrix
+  This property defines a matrix to describe the relative distances
+  between all numa nodes.
+  It is represented as a list of node pairs and their relative distance.
+
+  Note:
+   1. Each entry represents distance from first node to second node.
+   The distances are equal in either direction.
+   2. The distance from a node to self (local distance) is represented
+   with value 10 and all internode distance should be represented with
+   a value greater than 10.
+   3. distance-matrix should have entries in lexicographical ascending
+   order of nodes.
+   4. There must be only one device node distance-map which must reside in 
the root node.
+
+Example:
+   4 nodes connected in mesh/ring topology as below,
+
+   0___20__1
+   |   |
+   |   |
+   20 20
+   |   |
+   |   |
+   |___|
+   3   20  2
+
+   if relative distance for each hop is 20,
+   then internode distance would be,
+ 0 -> 1 = 20
+ 1 -> 2 = 20
+ 2 -> 3 = 20
+ 3 -> 0 = 20
+ 0 -> 2 = 40
+ 1 -> 3 = 40
+
+ and dt presentation for this distance matrix is,
+
+   distance-map {
+compatible = "numa-distance-map-v1";
+distance-matrix = <0 0  10>,
+  <0 1  20>,
+  <0 2  40>,
+  <0 3  20>,
+  <1 0  20>,
+  <1 1  10>,
+  <1 2  20>,
+  <1 3  40>,
+  <2 0  40>,
+  <2 1  20>,
+  <2 2  10>,
+  <2 3  20>,
+  <3 0  20>,
+  <3 1  40>,
+  <3 2  20>,
+  <3 3  10>;
+   };
+
+==
+4 - Example dts
+=

[PATCH v9 0/6] arm64, numa: Add numa support for arm64 platforms

[Ganapatrao Kulkarni]

v9: - Added cleanup patch to reuse and avoid redefinition of 
cpumask_of_pcibus
  as suggested from Will Deacon and Bjorn Helgaas.
- Including patch to Make pci-host-generic driver numa aware.
- Incorporated comment from Shannon Zhao.

v8:
- Incorporated review comments of Mark Rutland and Will Deacon.
- Added pci helper function and macro for numa.

v7:
- managing numa memory mapping using memblock.
- Incorporated review comments of Mark Rutland.

v6:
- defined and implemented the numa dt binding using
node property proximity and device node distance-map.
- renamed dt_numa to of_numa

v5:
- created base verion of numa.c which creates dummy numa without using 
dt
  on single socket platforms. Then added patches for dt support.
- Incorporated review comments from Hanjun Guo.

v4:
done changes as per Arnd review comments.

v3:
Added changes to support numa on arm64 based platforms.
Tested these patches on cavium's multinode(2 node topology) platform.
In this patchset, defined and implemented dt bindings for numa mapping
for core and memory using device node property arm,associativity.

v2:
Defined and implemented numa map for memory, cores to node and
proximity distance matrix of nodes.

v1:
Initial patchset to support numa on arm64 platforms.

Note:
1. This patchset is tested for numa with dt on
   thunderx single socket and dual socket boards.
2. Numa DT booting needs the dt memory nodes, which are deleted in 
current efi-stub,
hence to try numa with dt, you need to rebase with ard's patchset.

http://git.linaro.org/people/ard.biesheuvel/linux-arm.git/shortlog/refs/heads/arm64-uefi-early-fdt-handling
3. PATCH 6 is not tested for other architectures.

Ganapatrao Kulkarni (6):
  arm64, numa: adding numa support for arm64 platforms.
  Documentation, dt, arm64/arm: dt bindings for numa.
  arm64/arm, numa, dt: adding numa dt binding implementation for arm64
platforms.
  arm64, dt, thunderx: Add initial dts for Cavium Thunderx in 2 node
topology.
  PCI: generic: Make pci-host-generic driver numa aware
  topology, cleanup: Avoid redefinition of cpumask_of_pcibus in asm
header files.

 Documentation/devicetree/bindings/arm/numa.txt  | 272 
 arch/arm64/Kconfig  |  35 +
 arch/arm64/boot/dts/cavium/Makefile |   2 +-
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts  |  83 +++
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi | 806 
 arch/arm64/include/asm/mmzone.h |  18 +
 arch/arm64/include/asm/numa.h   |  50 ++
 arch/arm64/include/asm/topology.h   |   7 +
 arch/arm64/kernel/Makefile  |   1 +
 arch/arm64/kernel/of_numa.c | 257 
 arch/arm64/kernel/pci.c |   8 +
 arch/arm64/kernel/setup.c   |   4 +
 arch/arm64/kernel/smp.c |   4 +
 arch/arm64/mm/Makefile  |   1 +
 arch/arm64/mm/init.c|  31 +-
 arch/arm64/mm/mmu.c |   1 +
 arch/arm64/mm/numa.c| 395 
 arch/ia64/include/asm/topology.h|   4 -
 arch/metag/include/asm/topology.h   |   3 -
 arch/powerpc/include/asm/topology.h |   4 -
 arch/s390/include/asm/pci.h |   2 +-
 arch/s390/include/asm/topology.h|   1 +
 arch/sh/include/asm/topology.h  |   3 -
 arch/x86/include/asm/pci.h  |   2 +-
 arch/x86/include/asm/topology.h |   1 +
 drivers/pci/host/pci-host-generic.c |   9 +-
 include/asm-generic/topology.h  |   4 +-
 27 files changed, 1981 insertions(+), 27 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/arm/numa.txt
 create mode 100644 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts
 create mode 100644 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi
 create mode 100644 arch/arm64/include/asm/mmzone.h
 create mode 100644 arch/arm64/include/asm/numa.h
 create mode 100644 arch/arm64/kernel/of_numa.c
 create mode 100644 arch/arm64/mm/numa.c

-- 
1.8.1.4

___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

[PATCH v9 5/6] PCI: generic: Make pci-host-generic driver numa aware

[Ganapatrao Kulkarni]

update numa_node of device associated with pci bus.
moved down devm_kzalloc to allocate from node memory.

Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 drivers/pci/host/pci-host-generic.c | 9 ++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/pci/host/pci-host-generic.c 
b/drivers/pci/host/pci-host-generic.c
index 5434c90..0e1ce06 100644
--- a/drivers/pci/host/pci-host-generic.c
+++ b/drivers/pci/host/pci-host-generic.c
@@ -215,11 +215,9 @@ static int gen_pci_probe(struct platform_device *pdev)
const struct of_device_id *of_id;
struct device *dev = >dev;
struct device_node *np = dev->of_node;
-   struct gen_pci *pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
+   struct gen_pci *pci;
struct pci_bus *bus, *child;
 
-   if (!pci)
-   return -ENOMEM;
 
type = of_get_property(np, "device_type", NULL);
if (!type || strcmp(type, "pci")) {
@@ -230,6 +228,11 @@ static int gen_pci_probe(struct platform_device *pdev)
of_pci_check_probe_only();
 
of_id = of_match_node(gen_pci_of_match, np);
+   set_dev_node(dev, of_node_to_nid(np));
+   pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
+   if (!pci)
+   return -ENOMEM;
+
pci->cfg.ops = (struct gen_pci_cfg_bus_ops *)of_id->data;
pci->host.dev.parent = dev;
INIT_LIST_HEAD(>host.windows);
-- 
1.8.1.4

___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

[RFC PATCH v9 6/6] topology, cleanup: Avoid redefinition of cpumask_of_pcibus in asm header files.

[Ganapatrao Kulkarni]

At present cpumask_of_pcibus is defined for !CONFIG_NUMA and moving out
to common will allow to use for numa too. This also avoids
redefinition of this macro in respective architecture header files.

Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/include/asm/topology.h   | 3 ---
 arch/ia64/include/asm/topology.h| 4 
 arch/metag/include/asm/topology.h   | 3 ---
 arch/powerpc/include/asm/topology.h | 4 
 arch/s390/include/asm/pci.h | 2 +-
 arch/s390/include/asm/topology.h| 1 +
 arch/sh/include/asm/topology.h  | 3 ---
 arch/x86/include/asm/pci.h  | 2 +-
 arch/x86/include/asm/topology.h | 1 +
 include/asm-generic/topology.h  | 4 ++--
 10 files changed, 6 insertions(+), 21 deletions(-)

diff --git a/arch/arm64/include/asm/topology.h 
b/arch/arm64/include/asm/topology.h
index 8b57339..6e1f62c 100644
--- a/arch/arm64/include/asm/topology.h
+++ b/arch/arm64/include/asm/topology.h
@@ -26,9 +26,6 @@ const struct cpumask *cpu_coregroup_mask(int cpu);
 
 struct pci_bus;
 int pcibus_to_node(struct pci_bus *bus);
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?\
-cpu_all_mask : \
-cpumask_of_node(pcibus_to_node(bus)))
 
 #endif /* CONFIG_NUMA */
 
diff --git a/arch/ia64/include/asm/topology.h b/arch/ia64/include/asm/topology.h
index 3ad8f69..2778eb6 100644
--- a/arch/ia64/include/asm/topology.h
+++ b/arch/ia64/include/asm/topology.h
@@ -58,10 +58,6 @@ void build_cpu_to_node_map(void);
 
 extern void arch_fix_phys_package_id(int num, u32 slot);
 
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?\
-cpu_all_mask : \
-cpumask_of_node(pcibus_to_node(bus)))
-
 #include 
 
 #endif /* _ASM_IA64_TOPOLOGY_H */
diff --git a/arch/metag/include/asm/topology.h 
b/arch/metag/include/asm/topology.h
index e95f874..b285196 100644
--- a/arch/metag/include/asm/topology.h
+++ b/arch/metag/include/asm/topology.h
@@ -9,9 +9,6 @@
 #define cpumask_of_node(node)  ((void)node, cpu_online_mask)
 
 #define pcibus_to_node(bus)((void)(bus), -1)
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ? \
-   cpu_all_mask : \
-   cpumask_of_node(pcibus_to_node(bus)))
 
 #endif
 
diff --git a/arch/powerpc/include/asm/topology.h 
b/arch/powerpc/include/asm/topology.h
index 8b3b46b..eee025d 100644
--- a/arch/powerpc/include/asm/topology.h
+++ b/arch/powerpc/include/asm/topology.h
@@ -32,10 +32,6 @@ static inline int pcibus_to_node(struct pci_bus *bus)
 }
 #endif
 
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ?\
-cpu_all_mask : \
-cpumask_of_node(pcibus_to_node(bus)))
-
 extern int __node_distance(int, int);
 #define node_distance(a, b) __node_distance(a, b)
 
diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index c873e68..539fb2d 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -205,7 +205,7 @@ static inline int __pcibus_to_node(const struct pci_bus 
*bus)
 }
 
 static inline const struct cpumask *
-cpumask_of_pcibus(const struct pci_bus *bus)
+__cpumask_of_pcibus(const struct pci_bus *bus)
 {
return cpu_online_mask;
 }
diff --git a/arch/s390/include/asm/topology.h b/arch/s390/include/asm/topology.h
index 94fc55f..280eafc 100644
--- a/arch/s390/include/asm/topology.h
+++ b/arch/s390/include/asm/topology.h
@@ -78,6 +78,7 @@ static inline const struct cpumask *cpumask_of_node(int node)
 #define parent_node(node) (node)
 
 #define pcibus_to_node(bus) __pcibus_to_node(bus)
+#define cpumask_of_pcibus(bus) __cpumask_of_pcibus(bus)
 
 #define node_distance(a, b) __node_distance(a, b)
 
diff --git a/arch/sh/include/asm/topology.h b/arch/sh/include/asm/topology.h
index b0a282d..357983d 100644
--- a/arch/sh/include/asm/topology.h
+++ b/arch/sh/include/asm/topology.h
@@ -9,9 +9,6 @@
 #define cpumask_of_node(node)  ((void)node, cpu_online_mask)
 
 #define pcibus_to_node(bus)((void)(bus), -1)
-#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ? \
-   cpu_all_mask : \
-   cpumask_of_node(pcibus_to_node(bus)))
 
 #endif
 
diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h
index 4625943..4a6f358 100644
--- a/arch/x86/include/asm/pci.h
+++ b/arch/x86/include/asm/pci.h
@@ -121,7 +121,7 @@ static inline int __pcibus_to_node(const struct pci_bus 
*bus)
 }
 
 static inline const struct cpumask *
-cpumask_of_pcibus(const struct pci_bus *bus)
+__cpumask_of_pcibus(const struct pci_bus *bus)
 {
int node;
 
diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h
index 0fb4648..980f6ee 100644
--- a/ar

[PATCH v9 4/6] arm64, dt, thunderx: Add initial dts for Cavium Thunderx in 2 node topology.

[Ganapatrao Kulkarni]

Adding dt file for Cavium's Thunderx dual socket platform.

Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/boot/dts/cavium/Makefile |   2 +-
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts  |  83 +++
 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi | 806 
 3 files changed, 890 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts
 create mode 100644 arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi

diff --git a/arch/arm64/boot/dts/cavium/Makefile 
b/arch/arm64/boot/dts/cavium/Makefile
index e34f89d..7fe7067 100644
--- a/arch/arm64/boot/dts/cavium/Makefile
+++ b/arch/arm64/boot/dts/cavium/Makefile
@@ -1,4 +1,4 @@
-dtb-$(CONFIG_ARCH_THUNDER) += thunder-88xx.dtb
+dtb-$(CONFIG_ARCH_THUNDER) += thunder-88xx.dtb thunder-88xx-2n.dtb
 
 always := $(dtb-y)
 subdir-y   := $(dts-dirs)
diff --git a/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts 
b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts
new file mode 100644
index 000..5601e87
--- /dev/null
+++ b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dts
@@ -0,0 +1,83 @@
+/*
+ * Cavium Thunder DTS file - Thunder board description
+ *
+ * Copyright (C) 2014, Cavium Inc.
+ *
+ * This file is dual-licensed: you can use it either under the terms
+ * of the GPL or the X11 license, at your option. Note that this dual
+ * licensing only applies to this file, and not this project as a
+ * whole.
+ *
+ *  a) This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this library; if not, write to the Free
+ * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston,
+ * MA 02110-1301 USA
+ *
+ * Or, alternatively,
+ *
+ *  b) Permission is hereby granted, free of charge, to any person
+ * obtaining a copy of this software and associated documentation
+ * files (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use,
+ * copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following
+ * conditions:
+ *
+ * The above copyright notice and this permission notice shall be
+ * included in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
+ * OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
+ * HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
+ * WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+/dts-v1/;
+
+/include/ "thunder-88xx-2n.dtsi"
+
+/ {
+   model = "Cavium ThunderX CN88XX board";
+   compatible = "cavium,thunder-88xx";
+
+   aliases {
+   serial0 = 
+   serial1 = 
+   };
+
+   memory@ {
+   device_type = "memory";
+   reg = <0x0 0x0140 0x3 0xFEC0>;
+   /* socket 0 */
+   numa-node-id = <0>;
+   };
+
+   memory@100 {
+   device_type = "memory";
+   reg = <0x100 0x0040 0x3 0xFFC0>;
+/* socket 1 */
+   numa-node-id = <1>;
+   };
+
+   distance-map {
+   compatible = "numa-distance-map-v1";
+   distance-matrix = <0 0  10>,
+ <0 1  20>,
+ <1 1  10>;
+   };
+};
diff --git a/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi 
b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi
new file mode 100644
index 000..b58e5c7
--- /dev/null
+++ b/arch/arm64/boot/dts/cavium/thunder-88xx-2n.dtsi
@@ -0,0 +1,806 @@
+/*
+ * Cavium Thunder DTS file - Thunder SoC description
+ *
+ * Copyright (C) 2014, Cavium Inc.
+ *
+ * This file is dual-licensed: you can use it either under the terms
+ * of the GPL or the X11 license, at your option. Note that this dual
+ * licensing only applies to this file, and no

[PATCH v9 3/6] arm64/arm, numa, dt: adding numa dt binding implementation for arm64 platforms.

[Ganapatrao Kulkarni]

Adding numa dt binding support for arm64 based platforms.
dt node parsing for numa topology is done using device property
numa-node-id and device node distance-map.

Reviewed-by: Robert Richter <rrich...@cavium.com>
Signed-off-by: Ganapatrao Kulkarni <gkulka...@caviumnetworks.com>
---
 arch/arm64/Kconfig|  10 ++
 arch/arm64/include/asm/numa.h |  10 ++
 arch/arm64/kernel/Makefile|   1 +
 arch/arm64/kernel/of_numa.c   | 257 ++
 arch/arm64/kernel/smp.c   |   2 +
 arch/arm64/mm/numa.c  |  10 +-
 6 files changed, 289 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm64/kernel/of_numa.c

diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 74f5d73..775cf4a 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -515,6 +515,16 @@ config NUMA
  local memory of the CPU and add some more
  NUMA awareness to the kernel.
 
+config OF_NUMA
+   bool "Device Tree NUMA support"
+   depends on NUMA
+   depends on OF
+   default y
+   help
+ Enable Device Tree NUMA support.
+ This enables the numa mapping of cpu, memory, io and
+ inter node distances using dt bindings.
+
 config NODES_SHIFT
int "Maximum NUMA Nodes (as a power of 2)"
range 1 10
diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h
index f28f15b0..54deb38 100644
--- a/arch/arm64/include/asm/numa.h
+++ b/arch/arm64/include/asm/numa.h
@@ -37,4 +37,14 @@ void numa_store_cpu_info(unsigned int cpu);
 static inline void numa_store_cpu_info(unsigned int cpu)   { }
 static inline void arm64_numa_init(void)   { }
 #endif /* CONFIG_NUMA */
+
+struct device_node;
+#ifdef CONFIG_OF_NUMA
+int __init arm64_of_numa_init(void);
+void __init of_numa_set_node_info(unsigned int cpu, struct device_node *dn);
+#else
+static inline void of_numa_set_node_info(unsigned int cpu,
+   struct device_node *dn) { }
+#endif
+
 #endif /* __ASM_NUMA_H */
diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile
index 474691f..7987763 100644
--- a/arch/arm64/kernel/Makefile
+++ b/arch/arm64/kernel/Makefile
@@ -41,6 +41,7 @@ arm64-obj-$(CONFIG_EFI)   += efi.o 
efi-entry.stub.o
 arm64-obj-$(CONFIG_PCI)+= pci.o
 arm64-obj-$(CONFIG_ARMV8_DEPRECATED)   += armv8_deprecated.o
 arm64-obj-$(CONFIG_ACPI)   += acpi.o
+arm64-obj-$(CONFIG_OF_NUMA)+= of_numa.o
 
 obj-y  += $(arm64-obj-y) vdso/
 obj-m  += $(arm64-obj-m)
diff --git a/arch/arm64/kernel/of_numa.c b/arch/arm64/kernel/of_numa.c
new file mode 100644
index 000..2f9e34b
--- /dev/null
+++ b/arch/arm64/kernel/of_numa.c
@@ -0,0 +1,257 @@
+/*
+ * OF NUMA Parsing support.
+ *
+ * Copyright (C) 2015 Cavium Inc.
+ * Author: Ganapatrao Kulkarni <gkulka...@cavium.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
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * 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 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include 
+
+/* define default numa node to 0 */
+#define DEFAULT_NODE 0
+#define OF_NUMA_PROP "numa-node-id"
+
+/* Returns nid in the range [0..MAX_NUMNODES-1],
+ * or NUMA_NO_NODE if no valid numa-node-id entry found
+ * or DEFAULT_NODE if no numa-node-id entry exists
+ */
+static int of_numa_prop_to_nid(const __be32 *of_numa_prop, int length)
+{
+   int nid;
+
+   if (!of_numa_prop)
+   return DEFAULT_NODE;
+
+   if (length != sizeof(*of_numa_prop)) {
+   pr_warn("NUMA: Invalid of_numa_prop length %d found.\n",
+   length);
+   return NUMA_NO_NODE;
+   }
+
+   nid = of_read_number(of_numa_prop, 1);
+   if (nid >= MAX_NUMNODES) {
+   pr_warn("NUMA: Invalid numa node %d found.\n", nid);
+   return NUMA_NO_NODE;
+   }
+
+   return nid;
+}
+
+/* Must hold reference to node during call */
+static int of_get_numa_nid(struct device_node *device)
+{
+   int length;
+   const __be32 *of_numa_prop;
+
+   of_numa_prop = of_get_property(device, OF_NUMA_PROP, );
+
+   return of_numa_prop_to_nid(of_numa_prop, length);
+}
+
+static int __init early_init_of_get_numa_nid(unsigned long node)
+{
+   int length;
+   const __be32 *of_numa_prop;
+
+   of_numa_prop = of

Re: [RFC PATCH] dt:numa: adding numa node mapping for memory nodes.

[Ganapatrao Kulkarni]

looks like previous email not had the email ids added by kumar gala.
adding the missing ids and sending again.

On Thu, Sep 18, 2014 at 9:32 AM, Ganapatrao Kulkarni
gpkulka...@gmail.com wrote:
 On Thu, Sep 18, 2014 at 4:42 AM, Zi Shen Lim zlim@gmail.com wrote:
 On Wed, Sep 17, 2014 at 2:48 PM, Nathan Lynch nathan_ly...@mentor.com 
 wrote:
 On 09/17/2014 03:56 AM, Ganapatrao Kulkarni wrote:
 From: Ganapatrao Kulkarni ganapatrao.kulka...@cavium.com

 This patch adds property nid to memory node to provide the memory range 
 to
 numa node id mapping.

 Signed-off-by: Ganapatrao Kulkarni ganapatrao.kulka...@cavium.com

 ---
  Documentation/devicetree/bindings/numa.txt | 58 
 ++
  1 file changed, 58 insertions(+)
  create mode 100644 Documentation/devicetree/bindings/numa.txt

 diff --git a/Documentation/devicetree/bindings/numa.txt 
 b/Documentation/devicetree/bindings/numa.txt
 new file mode 100644
 index 000..c4a94f2
 --- /dev/null
 +++ b/Documentation/devicetree/bindings/numa.txt
 @@ -0,0 +1,58 @@
 +==
 +numa id binding description
 +==
 +
 +==
 +1 - Introduction
 +==
 +The device node  property nid(numa node id) can be added to memory
 +device node to map the range of memory addresses as defined in property 
 reg.
 +The property nid maps the memory range to the numa node id, which is 
 used to
 +find the local and remory pages on numa aware systems.

 Local and remote memory are notions that relate to some other
 resource -- typically a CPU, but also I/O resources on some systems.  It
 seems to me that a useful NUMA binding would at least specify a nid
 property, or something like it, for both cpu and memory nodes.  But this
 document speaks only of memory nodes.
 IMO, nid can be extended to cpu node also if the arch dont have any
 other way to do mapping, in fact this can be scaled to any node.
 for ARM arch, cpu mapping can be fetched from topology parsing and
 more over cpu mapping is required much later when secondary cpus comes
 online.
 i have added for memory, since this is parsed very early in the boot
 sequence and there is no support for dt based map at this moment(at
 least i am not aware off).
 numa distance matrix can also be defined, if the architecture has
 variable node distance other than just local/remote.

 As Kumar said, the device tree on powerpc server systems already has
 properties that express NUMA information.  If you can get hold of a copy
 of the PAPR (not ePAPR) from power.org, refer to the description of
 ibm,associativity and related properties.  I recall that it's a bit
 more complex than this proposal, though.

 I'm not able to find a link to the actual PAPR (not ePAPR)
 specification. Anyone has a linky?

 I did find some code in tree, but couldn't find the bindings in
 Documentation/devicetree/.
 yes, there is nothing in Documentation, if PPC are using since several
 years, then i humbly request to publish the documentation for numa
 bindings.
 No point in reinventing the wheel. Also i did not find any spec named
 PAPR, is it open spec, available to all like ePAPR?


 Seems like we'd care about form1?
 -8-
 commit 41eab6f88f24124df89e38067b3766b7bef06ddb
 Author: Anton Blanchard an...@samba.org
 Date:   Sun May 16 20:22:31 2010 +

 powerpc/numa: Use form 1 affinity to setup node distance

 Form 1 affinity allows multiple entries in
 ibm,associativity-reference-points
 which represent affinity domains in decreasing order of importance. The
 Linux concept of a node is always the first entry, but using the other
 values as an input to node_distance() allows the memory allocator to make
 better decisions on which node to go first when local memory has been
 exhausted.

 We keep things simple and create an array indexed by NUMA node, capped at
 4 entries. Each time we lookup an associativity property we initialise
 the array which is overkill, but since we should only hit this path 
 during
 boot it didn't seem worth adding a per node valid bit.

 Signed-off-by: Anton Blanchard an...@samba.org
 Signed-off-by: Benjamin Herrenschmidt b...@kernel.crashing.org
 -8-



 ___
 linux-arm-kernel mailing list
 linux-arm-ker...@lists.infradead.org
 http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev