* [Qemu-devel] [PATCH V6 01/11] NUMA: add NumaOptions, NumaNodeOptions and NumaMemOptions
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 10:52 ` Laszlo Ersek
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 02/11] NUMA: split -numa option Wanlong Gao
` (9 subsequent siblings)
10 siblings, 1 reply; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
qapi-schema.json | 58 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 58 insertions(+)
diff --git a/qapi-schema.json b/qapi-schema.json
index a51f7d2..b31fd08 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -3773,3 +3773,61 @@
##
{ 'command': 'query-rx-filter', 'data': { '*name': 'str' },
'returns': ['RxFilterInfo'] }
+
+##
+# @UInt16
+#
+# Let "uint16" be entry-wise
+#
+# @u16: a "uint16" type value
+#
+# Since 1.7
+##
+{ 'type': 'UInt16', 'data': { 'u16': 'uint16' }}
+
+##
+# @NumaOptions
+#
+# A discriminated record of NUMA options.
+#
+# Since 1.7
+##
+{ 'union': 'NumaOptions',
+ 'data': {
+ 'node': 'NumaNodeOptions',
+ 'mem': 'NumaMemOptions' }}
+
+##
+# @NumaNodeOptions
+#
+# Create a guest NUMA node.
+#
+# @nodeid: #optional NUMA node ID
+#
+# @cpus: #optional VCPUs belong to this node
+#
+# @mem: #optional memory size of this node (remain as legacy)
+#
+# Since: 1.7
+##
+{ 'type': 'NumaNodeOptions',
+ 'data': {
+ '*nodeid': 'uint16',
+ '*cpus': ['UInt16'],
+ '*mem': 'str' }}
+
+##
+# @NumaMemOptions
+#
+# Set memory information of guest NUMA node.
+#
+# @nodeid: #optional NUMA node ID
+#
+# @size: #optional memory size of this node
+#
+# Since 1.7
+##
+{ 'type': 'NumaMemOptions',
+ 'data': {
+ '*nodeid': 'uint16',
+ '*size': 'size' }}
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* Re: [Qemu-devel] [PATCH V6 01/11] NUMA: add NumaOptions, NumaNodeOptions and NumaMemOptions
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 01/11] NUMA: add NumaOptions, NumaNodeOptions and NumaMemOptions Wanlong Gao
@ 2013-07-30 10:52 ` Laszlo Ersek
0 siblings, 0 replies; 15+ messages in thread
From: Laszlo Ersek @ 2013-07-30 10:52 UTC (permalink / raw)
To: Wanlong Gao
Cc: aliguori, ehabkost, hutao, qemu-devel, lcapitulino, bsd, y-goto,
pbonzini, peter.huangpeng, afaerber
On 07/30/13 09:32, Wanlong Gao wrote:
> +##
> +# @UInt16
> +#
> +# Let "uint16" be entry-wise
> +#
> +# @u16: a "uint16" type value
> +#
> +# Since 1.7
> +##
> +{ 'type': 'UInt16', 'data': { 'u16': 'uint16' }}
Ah sorry, I should have spelled this out more clearly -- I only
mentioned it in my OptsVisitor series' cover letter (not the RFC one,
just the PATCH one):
On 07/22/13 23:07, Laszlo Ersek wrote:
> rfc->v1:
> ...
> - updated the cover letter to take native lists into account, plus
> cleaned it up.
> ...
> (Commit eb7ee2cb ("qapi: introduce OptsVisitor") had originally
> documented OptsVisitor's general schema requirements for parsing
> repeated options such that the list element type had to be a struct
> with one mandatory scalar field. Accordingly, the RFC version of this
> series required for interval flattening that this underlying scalar
> type be an integer type. However, since commit a678e26c ("qapi: pad
> GenericList value fields to 64 bits") we've had reliable native lists;
> OptsVisitor turns out to support them automatically.)
IOW, you can drop the UInt16 wrapper struct, and below:
> +
> +##
> +# @NumaOptions
> +#
> +# A discriminated record of NUMA options.
> +#
> +# Since 1.7
> +##
> +{ 'union': 'NumaOptions',
> + 'data': {
> + 'node': 'NumaNodeOptions',
> + 'mem': 'NumaMemOptions' }}
> +
> +##
> +# @NumaNodeOptions
> +#
> +# Create a guest NUMA node.
> +#
> +# @nodeid: #optional NUMA node ID
> +#
> +# @cpus: #optional VCPUs belong to this node
> +#
> +# @mem: #optional memory size of this node (remain as legacy)
> +#
> +# Since: 1.7
> +##
> +{ 'type': 'NumaNodeOptions',
> + 'data': {
> + '*nodeid': 'uint16',
> + '*cpus': ['UInt16'],
> + '*mem': 'str' }}
you can use plain ['uint16'] as type. (You can also glance at 8/8 (unit
test schema) in my series.)
> +
> +##
> +# @NumaMemOptions
> +#
> +# Set memory information of guest NUMA node.
> +#
> +# @nodeid: #optional NUMA node ID
> +#
> +# @size: #optional memory size of this node
> +#
> +# Since 1.7
> +##
> +{ 'type': 'NumaMemOptions',
> + 'data': {
> + '*nodeid': 'uint16',
> + '*size': 'size' }}
>
Rest looks good for this patch I think.
Laszlo
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Qemu-devel] [PATCH V6 02/11] NUMA: split -numa option
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 01/11] NUMA: add NumaOptions, NumaNodeOptions and NumaMemOptions Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 11:14 ` Laszlo Ersek
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 03/11] NUMA: move numa related code to numa.c Wanlong Gao
` (8 subsequent siblings)
10 siblings, 1 reply; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Change -numa option like following as Paolo suggested:
-numa node,nodeid=0,cpus=0-1 \
-numa mem,nodeid=0,size=1G
This new option will make later coming memory hotplug better.
And this new option is implemented using OptsVisitor.
And just remain "-numa node,mem=xx" as legacy.
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
Makefile.target | 2 +-
include/sysemu/sysemu.h | 3 +
numa.c | 145 ++++++++++++++++++++++++++++++++++++++++++++++++
qemu-options.hx | 6 +-
vl.c | 113 ++++++-------------------------------
5 files changed, 169 insertions(+), 100 deletions(-)
create mode 100644 numa.c
diff --git a/Makefile.target b/Makefile.target
index 9a49852..7e1fddf 100644
--- a/Makefile.target
+++ b/Makefile.target
@@ -113,7 +113,7 @@ endif #CONFIG_BSD_USER
#########################################################
# System emulator target
ifdef CONFIG_SOFTMMU
-obj-y += arch_init.o cpus.o monitor.o gdbstub.o balloon.o ioport.o
+obj-y += arch_init.o cpus.o monitor.o gdbstub.o balloon.o ioport.o numa.o
obj-y += qtest.o
obj-y += hw/
obj-$(CONFIG_FDT) += device_tree.o
diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h
index 3caeb66..380acfe 100644
--- a/include/sysemu/sysemu.h
+++ b/include/sysemu/sysemu.h
@@ -130,8 +130,11 @@ extern QEMUClock *rtc_clock;
#define MAX_NODES 64
#define MAX_CPUMASK_BITS 255
extern int nb_numa_nodes;
+extern int nb_numa_mem_nodes;
extern uint64_t node_mem[MAX_NODES];
extern unsigned long *node_cpumask[MAX_NODES];
+extern QemuOptsList qemu_numa_opts;
+int numa_init_func(QemuOpts *opts, void *opaque);
#define MAX_OPTION_ROMS 16
typedef struct QEMUOptionRom {
diff --git a/numa.c b/numa.c
new file mode 100644
index 0000000..fe0269d
--- /dev/null
+++ b/numa.c
@@ -0,0 +1,145 @@
+/*
+ * QEMU System Emulator
+ *
+ * Copyright (c) 2013 Fujitsu Ltd.
+ * Author: Wanlong Gao <gaowanlong@cn.fujitsu.com>
+ *
+ * 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.
+ */
+
+#include "sysemu/sysemu.h"
+#include "qemu/bitmap.h"
+#include "qapi-visit.h"
+#include "qapi/opts-visitor.h"
+#include "qapi/dealloc-visitor.h"
+
+QemuOptsList qemu_numa_opts = {
+ .name = "numa",
+ .implied_opt_name = "type",
+ .head = QTAILQ_HEAD_INITIALIZER(qemu_numa_opts.head),
+ .desc = { { 0 } } /* validated with OptsVisitor */
+};
+
+static int numa_node_parse(NumaNodeOptions *opts)
+{
+ uint16_t nodenr;
+ UInt16List *cpus = NULL;
+
+ if (opts->has_nodeid) {
+ nodenr = opts->nodeid;
+ if (nodenr >= MAX_NODES) {
+ fprintf(stderr, "qemu: Max number of NUMA nodes reached: %"
+ PRIu16 "\n", nodenr);
+ return -1;
+ }
+ } else {
+ nodenr = nb_numa_nodes;
+ }
+
+ for (cpus = opts->cpus; cpus; cpus = cpus->next) {
+ bitmap_set(node_cpumask[nodenr], cpus->value->u16, 1);
+ }
+
+ if (opts->has_mem) {
+ int64_t mem_size;
+ char *endptr;
+ mem_size = strtosz(opts->mem, &endptr);
+ if (mem_size < 0 || *endptr) {
+ fprintf(stderr, "qemu: invalid numa mem size: %s\n", opts->mem);
+ return -1;
+ }
+ node_mem[nodenr] = mem_size;
+ }
+
+ return 0;
+}
+
+static int numa_mem_parse(NumaMemOptions *opts)
+{
+ uint16_t nodenr;
+ uint64_t mem_size;
+
+ if (opts->has_nodeid) {
+ nodenr = opts->nodeid;
+ if (nodenr >= MAX_NODES) {
+ fprintf(stderr, "qemu: Max number of NUMA nodes reached: %"
+ PRIu16 "\n", nodenr);
+ return -1;
+ }
+ } else {
+ nodenr = nb_numa_mem_nodes;
+ }
+
+ if (opts->has_size) {
+ mem_size = opts->size;
+ node_mem[nodenr] = mem_size;
+ }
+
+ return 0;
+}
+
+int numa_init_func(QemuOpts *opts, void *opaque)
+{
+ NumaOptions *object = NULL;
+ Error *err = NULL;
+ int ret = 0;
+
+ {
+ OptsVisitor *ov = opts_visitor_new(opts);
+ visit_type_NumaOptions(opts_get_visitor(ov), &object, NULL, &err);
+ opts_visitor_cleanup(ov);
+ }
+
+ if (error_is_set(&err)) {
+ fprintf(stderr, "qemu: %s\n", error_get_pretty(err));
+ error_free(err);
+ ret = -1;
+ goto error;
+ }
+
+ switch (object->kind) {
+ case NUMA_OPTIONS_KIND_NODE:
+ if (nb_numa_nodes >= MAX_NODES) {
+ fprintf(stderr, "qemu: too many NUMA nodes\n");
+ ret = -1;
+ goto error;
+ }
+ ret = numa_node_parse(object->node);
+ nb_numa_nodes++;
+ break;
+ case NUMA_OPTIONS_KIND_MEM:
+ ret = numa_mem_parse(object->mem);
+ nb_numa_mem_nodes++;
+ break;
+ default:
+ fprintf(stderr, "qemu: Invalid NUMA options type.\n");
+ ret = -1;
+ }
+
+error:
+ if (object) {
+ QapiDeallocVisitor *dv = qapi_dealloc_visitor_new();
+ visit_type_NumaOptions(qapi_dealloc_get_visitor(dv),
+ &object, NULL, NULL);
+ qapi_dealloc_visitor_cleanup(dv);
+ }
+
+ return ret;
+}
+
diff --git a/qemu-options.hx b/qemu-options.hx
index d15338e..e9123b8 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -95,11 +95,13 @@ specifies the maximum number of hotpluggable CPUs.
ETEXI
DEF("numa", HAS_ARG, QEMU_OPTION_numa,
- "-numa node[,mem=size][,cpus=cpu[-cpu]][,nodeid=node]\n", QEMU_ARCH_ALL)
+ "-numa node[,nodeid=node][,cpus=cpu[-cpu]]\n"
+ "-numa mem[,nodeid=node][,size=size]\n"
+ , QEMU_ARCH_ALL)
STEXI
@item -numa @var{opts}
@findex -numa
-Simulate a multi node NUMA system. If mem and cpus are omitted, resources
+Simulate a multi node NUMA system. If @var{size} and @var{cpus} are omitted, resources
are split equally.
ETEXI
diff --git a/vl.c b/vl.c
index 25b8f2f..129b7c6 100644
--- a/vl.c
+++ b/vl.c
@@ -250,6 +250,7 @@ static QTAILQ_HEAD(, FWBootEntry) fw_boot_order =
QTAILQ_HEAD_INITIALIZER(fw_boot_order);
int nb_numa_nodes;
+int nb_numa_mem_nodes;
uint64_t node_mem[MAX_NODES];
unsigned long *node_cpumask[MAX_NODES];
@@ -1330,102 +1331,6 @@ char *get_boot_devices_list(size_t *size)
return list;
}
-static void numa_node_parse_cpus(int nodenr, const char *cpus)
-{
- char *endptr;
- unsigned long long value, endvalue;
-
- /* Empty CPU range strings will be considered valid, they will simply
- * not set any bit in the CPU bitmap.
- */
- if (!*cpus) {
- return;
- }
-
- if (parse_uint(cpus, &value, &endptr, 10) < 0) {
- goto error;
- }
- if (*endptr == '-') {
- if (parse_uint_full(endptr + 1, &endvalue, 10) < 0) {
- goto error;
- }
- } else if (*endptr == '\0') {
- endvalue = value;
- } else {
- goto error;
- }
-
- if (endvalue >= MAX_CPUMASK_BITS) {
- endvalue = MAX_CPUMASK_BITS - 1;
- fprintf(stderr,
- "qemu: NUMA: A max of %d VCPUs are supported\n",
- MAX_CPUMASK_BITS);
- }
-
- if (endvalue < value) {
- goto error;
- }
-
- bitmap_set(node_cpumask[nodenr], value, endvalue-value+1);
- return;
-
-error:
- fprintf(stderr, "qemu: Invalid NUMA CPU range: %s\n", cpus);
- exit(1);
-}
-
-static void numa_add(const char *optarg)
-{
- char option[128];
- char *endptr;
- unsigned long long nodenr;
-
- optarg = get_opt_name(option, 128, optarg, ',');
- if (*optarg == ',') {
- optarg++;
- }
- if (!strcmp(option, "node")) {
-
- if (nb_numa_nodes >= MAX_NODES) {
- fprintf(stderr, "qemu: too many NUMA nodes\n");
- exit(1);
- }
-
- if (get_param_value(option, 128, "nodeid", optarg) == 0) {
- nodenr = nb_numa_nodes;
- } else {
- if (parse_uint_full(option, &nodenr, 10) < 0) {
- fprintf(stderr, "qemu: Invalid NUMA nodeid: %s\n", option);
- exit(1);
- }
- }
-
- if (nodenr >= MAX_NODES) {
- fprintf(stderr, "qemu: invalid NUMA nodeid: %llu\n", nodenr);
- exit(1);
- }
-
- if (get_param_value(option, 128, "mem", optarg) == 0) {
- node_mem[nodenr] = 0;
- } else {
- int64_t sval;
- sval = strtosz(option, &endptr);
- if (sval < 0 || *endptr) {
- fprintf(stderr, "qemu: invalid numa mem size: %s\n", optarg);
- exit(1);
- }
- node_mem[nodenr] = sval;
- }
- if (get_param_value(option, 128, "cpus", optarg) != 0) {
- numa_node_parse_cpus(nodenr, option);
- }
- nb_numa_nodes++;
- } else {
- fprintf(stderr, "Invalid -numa option: %s\n", option);
- exit(1);
- }
-}
-
static QemuOptsList qemu_smp_opts = {
.name = "smp-opts",
.implied_opt_name = "cpus",
@@ -2961,6 +2866,7 @@ int main(int argc, char **argv, char **envp)
qemu_add_opts(&qemu_tpmdev_opts);
qemu_add_opts(&qemu_realtime_opts);
qemu_add_opts(&qemu_msg_opts);
+ qemu_add_opts(&qemu_numa_opts);
runstate_init();
@@ -2986,6 +2892,7 @@ int main(int argc, char **argv, char **envp)
}
nb_numa_nodes = 0;
+ nb_numa_mem_nodes = 0;
nb_nics = 0;
bdrv_init_with_whitelist();
@@ -3147,7 +3054,10 @@ int main(int argc, char **argv, char **envp)
}
break;
case QEMU_OPTION_numa:
- numa_add(optarg);
+ opts = qemu_opts_parse(qemu_find_opts("numa"), optarg, 1);
+ if (!opts) {
+ exit(1);
+ }
break;
case QEMU_OPTION_display:
display_type = select_display(optarg);
@@ -4226,6 +4136,15 @@ int main(int argc, char **argv, char **envp)
register_savevm_live(NULL, "ram", 0, 4, &savevm_ram_handlers, NULL);
+ if (qemu_opts_foreach(qemu_find_opts("numa"), numa_init_func,
+ NULL, 1) != 0) {
+ exit(1);
+ }
+
+ if (nb_numa_mem_nodes > nb_numa_nodes) {
+ nb_numa_nodes = nb_numa_mem_nodes;
+ }
+
if (nb_numa_nodes > 0) {
int i;
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* Re: [Qemu-devel] [PATCH V6 02/11] NUMA: split -numa option
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 02/11] NUMA: split -numa option Wanlong Gao
@ 2013-07-30 11:14 ` Laszlo Ersek
2013-07-31 2:10 ` Wanlong Gao
0 siblings, 1 reply; 15+ messages in thread
From: Laszlo Ersek @ 2013-07-30 11:14 UTC (permalink / raw)
To: Wanlong Gao
Cc: aliguori, ehabkost, hutao, qemu-devel, lcapitulino, bsd, y-goto,
pbonzini, peter.huangpeng, afaerber
On 07/30/13 09:32, Wanlong Gao wrote:
> +static int numa_node_parse(NumaNodeOptions *opts)
> +{
> + uint16_t nodenr;
> + UInt16List *cpus = NULL;
> +
> + if (opts->has_nodeid) {
> + nodenr = opts->nodeid;
> + if (nodenr >= MAX_NODES) {
> + fprintf(stderr, "qemu: Max number of NUMA nodes reached: %"
> + PRIu16 "\n", nodenr);
> + return -1;
> + }
> + } else {
> + nodenr = nb_numa_nodes;
> + }
> +
> + for (cpus = opts->cpus; cpus; cpus = cpus->next) {
> + bitmap_set(node_cpumask[nodenr], cpus->value->u16, 1);
> + }
> +
> + if (opts->has_mem) {
> + int64_t mem_size;
> + char *endptr;
> + mem_size = strtosz(opts->mem, &endptr);
> + if (mem_size < 0 || *endptr) {
> + fprintf(stderr, "qemu: invalid numa mem size: %s\n", opts->mem);
> + return -1;
> + }
> + node_mem[nodenr] = mem_size;
> + }
> +
> + return 0;
> +}
> +
> +static int numa_mem_parse(NumaMemOptions *opts)
> +{
> + uint16_t nodenr;
> + uint64_t mem_size;
> +
> + if (opts->has_nodeid) {
> + nodenr = opts->nodeid;
> + if (nodenr >= MAX_NODES) {
> + fprintf(stderr, "qemu: Max number of NUMA nodes reached: %"
> + PRIu16 "\n", nodenr);
> + return -1;
> + }
> + } else {
> + nodenr = nb_numa_mem_nodes;
> + }
> +
> + if (opts->has_size) {
> + mem_size = opts->size;
> + node_mem[nodenr] = mem_size;
> + }
> +
> + return 0;
> +}
Unless I'm missing something:
Just like "NumaMemOptions.size" (which has type 'size' in the JSON),
OptsVisitor could parse "NumaNodeOptions.mem" for you as well, if you
switched the latter's type to 'size' too.
... Hm, not really. This is probably a compatibility thing. Without any
specific suffix, strtosz() passes STRTOSZ_DEFSUFFIX_MB to
strtosz_suffix() (ie. the legacy optarg is expressed in megabytes, if
the user doesn't specify a suffix him/herself), while opts_type_size()
passes STRTOSZ_DEFSUFFIX_B.
I think this patch is good; if you adapt it to the ['UInt16'] ->
['uint16'] change I proposed for 01/11, you can add
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Thanks!
Laszlo
^ permalink raw reply [flat|nested] 15+ messages in thread* Re: [Qemu-devel] [PATCH V6 02/11] NUMA: split -numa option
2013-07-30 11:14 ` Laszlo Ersek
@ 2013-07-31 2:10 ` Wanlong Gao
0 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-31 2:10 UTC (permalink / raw)
To: Laszlo Ersek
Cc: aliguori, ehabkost, hutao, qemu-devel, lcapitulino, bsd, y-goto,
pbonzini, peter.huangpeng, afaerber, Wanlong Gao
On 07/30/2013 07:14 PM, Laszlo Ersek wrote:
> On 07/30/13 09:32, Wanlong Gao wrote:
>
>> +static int numa_node_parse(NumaNodeOptions *opts)
>> +{
>> + uint16_t nodenr;
>> + UInt16List *cpus = NULL;
>> +
>> + if (opts->has_nodeid) {
>> + nodenr = opts->nodeid;
>> + if (nodenr >= MAX_NODES) {
>> + fprintf(stderr, "qemu: Max number of NUMA nodes reached: %"
>> + PRIu16 "\n", nodenr);
>> + return -1;
>> + }
>> + } else {
>> + nodenr = nb_numa_nodes;
>> + }
>> +
>> + for (cpus = opts->cpus; cpus; cpus = cpus->next) {
>> + bitmap_set(node_cpumask[nodenr], cpus->value->u16, 1);
>> + }
>> +
>> + if (opts->has_mem) {
>> + int64_t mem_size;
>> + char *endptr;
>> + mem_size = strtosz(opts->mem, &endptr);
>> + if (mem_size < 0 || *endptr) {
>> + fprintf(stderr, "qemu: invalid numa mem size: %s\n", opts->mem);
>> + return -1;
>> + }
>> + node_mem[nodenr] = mem_size;
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static int numa_mem_parse(NumaMemOptions *opts)
>> +{
>> + uint16_t nodenr;
>> + uint64_t mem_size;
>> +
>> + if (opts->has_nodeid) {
>> + nodenr = opts->nodeid;
>> + if (nodenr >= MAX_NODES) {
>> + fprintf(stderr, "qemu: Max number of NUMA nodes reached: %"
>> + PRIu16 "\n", nodenr);
>> + return -1;
>> + }
>> + } else {
>> + nodenr = nb_numa_mem_nodes;
>> + }
>> +
>> + if (opts->has_size) {
>> + mem_size = opts->size;
>> + node_mem[nodenr] = mem_size;
>> + }
>> +
>> + return 0;
>> +}
>
> Unless I'm missing something:
>
> Just like "NumaMemOptions.size" (which has type 'size' in the JSON),
> OptsVisitor could parse "NumaNodeOptions.mem" for you as well, if you
> switched the latter's type to 'size' too.
>
> ... Hm, not really. This is probably a compatibility thing. Without any
> specific suffix, strtosz() passes STRTOSZ_DEFSUFFIX_MB to
> strtosz_suffix() (ie. the legacy optarg is expressed in megabytes, if
> the user doesn't specify a suffix him/herself), while opts_type_size()
> passes STRTOSZ_DEFSUFFIX_B.
Yeah, right.
>
> I think this patch is good; if you adapt it to the ['UInt16'] ->
> ['uint16'] change I proposed for 01/11, you can add
Done, thank you.
Wanlong Gao
>
> Reviewed-by: Laszlo Ersek <lersek@redhat.com>
>
> Thanks!
> Laszlo
>
^ permalink raw reply [flat|nested] 15+ messages in thread
* [Qemu-devel] [PATCH V6 03/11] NUMA: move numa related code to numa.c
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 01/11] NUMA: add NumaOptions, NumaNodeOptions and NumaMemOptions Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 02/11] NUMA: split -numa option Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 04/11] NUMA: Add numa_info structure to contain numa nodes info Wanlong Gao
` (7 subsequent siblings)
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
cpus.c | 14 -----------
include/sysemu/cpus.h | 1 -
include/sysemu/sysemu.h | 2 ++
numa.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++
vl.c | 47 +----------------------------------
5 files changed, 68 insertions(+), 61 deletions(-)
diff --git a/cpus.c b/cpus.c
index 0f65e76..08b8ef1 100644
--- a/cpus.c
+++ b/cpus.c
@@ -1201,20 +1201,6 @@ static void tcg_exec_all(void)
exit_request = 0;
}
-void set_numa_modes(void)
-{
- CPUState *cpu;
- int i;
-
- for (cpu = first_cpu; cpu != NULL; cpu = cpu->next_cpu) {
- for (i = 0; i < nb_numa_nodes; i++) {
- if (test_bit(cpu->cpu_index, node_cpumask[i])) {
- cpu->numa_node = i;
- }
- }
- }
-}
-
void list_cpus(FILE *f, fprintf_function cpu_fprintf, const char *optarg)
{
/* XXX: implement xxx_cpu_list for targets that still miss it */
diff --git a/include/sysemu/cpus.h b/include/sysemu/cpus.h
index 6502488..4f79081 100644
--- a/include/sysemu/cpus.h
+++ b/include/sysemu/cpus.h
@@ -23,7 +23,6 @@ extern int smp_threads;
#define smp_threads 1
#endif
-void set_numa_modes(void);
void list_cpus(FILE *f, fprintf_function cpu_fprintf, const char *optarg);
#endif
diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h
index 380acfe..5277fd9 100644
--- a/include/sysemu/sysemu.h
+++ b/include/sysemu/sysemu.h
@@ -135,6 +135,8 @@ extern uint64_t node_mem[MAX_NODES];
extern unsigned long *node_cpumask[MAX_NODES];
extern QemuOptsList qemu_numa_opts;
int numa_init_func(QemuOpts *opts, void *opaque);
+void set_numa_nodes(void);
+void set_numa_modes(void);
#define MAX_OPTION_ROMS 16
typedef struct QEMUOptionRom {
diff --git a/numa.c b/numa.c
index fe0269d..e684777 100644
--- a/numa.c
+++ b/numa.c
@@ -143,3 +143,68 @@ error:
return ret;
}
+void set_numa_nodes(void)
+{
+ if (nb_numa_mem_nodes > nb_numa_nodes) {
+ nb_numa_nodes = nb_numa_mem_nodes;
+ }
+
+ if (nb_numa_nodes > 0) {
+ int i;
+
+ if (nb_numa_nodes > MAX_NODES) {
+ nb_numa_nodes = MAX_NODES;
+ }
+
+ /* If no memory size if given for any node, assume the default case
+ * and distribute the available memory equally across all nodes
+ */
+ for (i = 0; i < nb_numa_nodes; i++) {
+ if (node_mem[i] != 0) {
+ break;
+ }
+ }
+
+ if (i == nb_numa_nodes) {
+ uint64_t usedmem = 0;
+
+ /* On Linux, the each node's border has to be 8MB aligned,
+ * the final node gets the rest.
+ */
+ for (i = 0; i < nb_numa_nodes - 1; i++) {
+ node_mem[i] = (ram_size / nb_numa_nodes) & ~((1 << 23UL) - 1);
+ usedmem += node_mem[i];
+ }
+ node_mem[i] = ram_size - usedmem;
+ }
+
+ for (i = 0; i < nb_numa_nodes; i++) {
+ if (!bitmap_empty(node_cpumask[i], MAX_CPUMASK_BITS)) {
+ break;
+ }
+ }
+ /* assigning the VCPUs round-robin is easier to implement, guest OSes
+ * must cope with this anyway, because there are BIOSes out there in
+ * real machines which also use this scheme.
+ */
+ if (i == nb_numa_nodes) {
+ for (i = 0; i < max_cpus; i++) {
+ set_bit(i, node_cpumask[i % nb_numa_nodes]);
+ }
+ }
+ }
+}
+
+void set_numa_modes(void)
+{
+ CPUState *cpu;
+ int i;
+
+ for (cpu = first_cpu; cpu != NULL; cpu = cpu->next_cpu) {
+ for (i = 0; i < nb_numa_nodes; i++) {
+ if (test_bit(cpu->cpu_index, node_cpumask[i])) {
+ cpu->numa_node = i;
+ }
+ }
+ }
+}
diff --git a/vl.c b/vl.c
index 129b7c6..0304af4 100644
--- a/vl.c
+++ b/vl.c
@@ -4141,52 +4141,7 @@ int main(int argc, char **argv, char **envp)
exit(1);
}
- if (nb_numa_mem_nodes > nb_numa_nodes) {
- nb_numa_nodes = nb_numa_mem_nodes;
- }
-
- if (nb_numa_nodes > 0) {
- int i;
-
- if (nb_numa_nodes > MAX_NODES) {
- nb_numa_nodes = MAX_NODES;
- }
-
- /* If no memory size if given for any node, assume the default case
- * and distribute the available memory equally across all nodes
- */
- for (i = 0; i < nb_numa_nodes; i++) {
- if (node_mem[i] != 0)
- break;
- }
- if (i == nb_numa_nodes) {
- uint64_t usedmem = 0;
-
- /* On Linux, the each node's border has to be 8MB aligned,
- * the final node gets the rest.
- */
- for (i = 0; i < nb_numa_nodes - 1; i++) {
- node_mem[i] = (ram_size / nb_numa_nodes) & ~((1 << 23UL) - 1);
- usedmem += node_mem[i];
- }
- node_mem[i] = ram_size - usedmem;
- }
-
- for (i = 0; i < nb_numa_nodes; i++) {
- if (!bitmap_empty(node_cpumask[i], MAX_CPUMASK_BITS)) {
- break;
- }
- }
- /* assigning the VCPUs round-robin is easier to implement, guest OSes
- * must cope with this anyway, because there are BIOSes out there in
- * real machines which also use this scheme.
- */
- if (i == nb_numa_nodes) {
- for (i = 0; i < max_cpus; i++) {
- set_bit(i, node_cpumask[i % nb_numa_nodes]);
- }
- }
- }
+ set_numa_nodes();
if (qemu_opts_foreach(qemu_find_opts("mon"), mon_init_func, NULL, 1) != 0) {
exit(1);
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 04/11] NUMA: Add numa_info structure to contain numa nodes info
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (2 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 03/11] NUMA: move numa related code to numa.c Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 05/11] NUMA: Add Linux libnuma detection Wanlong Gao
` (6 subsequent siblings)
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Add the numa_info structure to contain the numa nodes memory,
VCPUs information and the future added numa nodes host memory
policies.
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
hw/i386/pc.c | 4 ++--
include/sysemu/sysemu.h | 8 ++++++--
monitor.c | 2 +-
numa.c | 21 +++++++++++----------
vl.c | 7 +++----
5 files changed, 23 insertions(+), 19 deletions(-)
diff --git a/hw/i386/pc.c b/hw/i386/pc.c
index 2a87563..7240c07 100644
--- a/hw/i386/pc.c
+++ b/hw/i386/pc.c
@@ -654,14 +654,14 @@ static FWCfgState *bochs_bios_init(void)
unsigned int apic_id = x86_cpu_apic_id_from_index(i);
assert(apic_id < apic_id_limit);
for (j = 0; j < nb_numa_nodes; j++) {
- if (test_bit(i, node_cpumask[j])) {
+ if (test_bit(i, numa_info[j].node_cpu)) {
numa_fw_cfg[apic_id + 1] = cpu_to_le64(j);
break;
}
}
}
for (i = 0; i < nb_numa_nodes; i++) {
- numa_fw_cfg[apic_id_limit + 1 + i] = cpu_to_le64(node_mem[i]);
+ numa_fw_cfg[apic_id_limit + 1 + i] = cpu_to_le64(numa_info[i].node_mem);
}
fw_cfg_add_bytes(fw_cfg, FW_CFG_NUMA, numa_fw_cfg,
(1 + apic_id_limit + nb_numa_nodes) *
diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h
index 5277fd9..0f6d1bc 100644
--- a/include/sysemu/sysemu.h
+++ b/include/sysemu/sysemu.h
@@ -9,6 +9,7 @@
#include "qapi-types.h"
#include "qemu/notify.h"
#include "qemu/main-loop.h"
+#include "qemu/bitmap.h"
/* vl.c */
@@ -131,8 +132,11 @@ extern QEMUClock *rtc_clock;
#define MAX_CPUMASK_BITS 255
extern int nb_numa_nodes;
extern int nb_numa_mem_nodes;
-extern uint64_t node_mem[MAX_NODES];
-extern unsigned long *node_cpumask[MAX_NODES];
+typedef struct node_info {
+ uint64_t node_mem;
+ DECLARE_BITMAP(node_cpu, MAX_CPUMASK_BITS);
+} NodeInfo;
+extern NodeInfo numa_info[MAX_NODES];
extern QemuOptsList qemu_numa_opts;
int numa_init_func(QemuOpts *opts, void *opaque);
void set_numa_nodes(void);
diff --git a/monitor.c b/monitor.c
index 5dc0aa9..513af25 100644
--- a/monitor.c
+++ b/monitor.c
@@ -1826,7 +1826,7 @@ static void do_info_numa(Monitor *mon, const QDict *qdict)
}
monitor_printf(mon, "\n");
monitor_printf(mon, "node %d size: %" PRId64 " MB\n", i,
- node_mem[i] >> 20);
+ numa_info[i].node_mem >> 20);
}
}
diff --git a/numa.c b/numa.c
index e684777..7aaa663 100644
--- a/numa.c
+++ b/numa.c
@@ -53,7 +53,7 @@ static int numa_node_parse(NumaNodeOptions *opts)
}
for (cpus = opts->cpus; cpus; cpus = cpus->next) {
- bitmap_set(node_cpumask[nodenr], cpus->value->u16, 1);
+ bitmap_set(numa_info[nodenr].node_cpu, cpus->value->u16, 1);
}
if (opts->has_mem) {
@@ -64,7 +64,7 @@ static int numa_node_parse(NumaNodeOptions *opts)
fprintf(stderr, "qemu: invalid numa mem size: %s\n", opts->mem);
return -1;
}
- node_mem[nodenr] = mem_size;
+ numa_info[nodenr].node_mem = mem_size;
}
return 0;
@@ -88,7 +88,7 @@ static int numa_mem_parse(NumaMemOptions *opts)
if (opts->has_size) {
mem_size = opts->size;
- node_mem[nodenr] = mem_size;
+ numa_info[nodenr].node_mem = mem_size;
}
return 0;
@@ -160,7 +160,7 @@ void set_numa_nodes(void)
* and distribute the available memory equally across all nodes
*/
for (i = 0; i < nb_numa_nodes; i++) {
- if (node_mem[i] != 0) {
+ if (numa_info[i].node_mem != 0) {
break;
}
}
@@ -172,14 +172,15 @@ void set_numa_nodes(void)
* the final node gets the rest.
*/
for (i = 0; i < nb_numa_nodes - 1; i++) {
- node_mem[i] = (ram_size / nb_numa_nodes) & ~((1 << 23UL) - 1);
- usedmem += node_mem[i];
+ numa_info[i].node_mem = (ram_size / nb_numa_nodes) &
+ ~((1 << 23UL) - 1);
+ usedmem += numa_info[i].node_mem;
}
- node_mem[i] = ram_size - usedmem;
+ numa_info[i].node_mem = ram_size - usedmem;
}
for (i = 0; i < nb_numa_nodes; i++) {
- if (!bitmap_empty(node_cpumask[i], MAX_CPUMASK_BITS)) {
+ if (!bitmap_empty(numa_info[i].node_cpu, MAX_CPUMASK_BITS)) {
break;
}
}
@@ -189,7 +190,7 @@ void set_numa_nodes(void)
*/
if (i == nb_numa_nodes) {
for (i = 0; i < max_cpus; i++) {
- set_bit(i, node_cpumask[i % nb_numa_nodes]);
+ set_bit(i, numa_info[i % nb_numa_nodes].node_cpu);
}
}
}
@@ -202,7 +203,7 @@ void set_numa_modes(void)
for (cpu = first_cpu; cpu != NULL; cpu = cpu->next_cpu) {
for (i = 0; i < nb_numa_nodes; i++) {
- if (test_bit(cpu->cpu_index, node_cpumask[i])) {
+ if (test_bit(cpu->cpu_index, numa_info[i].node_cpu)) {
cpu->numa_node = i;
}
}
diff --git a/vl.c b/vl.c
index 0304af4..d9976c7 100644
--- a/vl.c
+++ b/vl.c
@@ -251,8 +251,7 @@ static QTAILQ_HEAD(, FWBootEntry) fw_boot_order =
int nb_numa_nodes;
int nb_numa_mem_nodes;
-uint64_t node_mem[MAX_NODES];
-unsigned long *node_cpumask[MAX_NODES];
+NodeInfo numa_info[MAX_NODES];
uint8_t qemu_uuid[16];
@@ -2887,8 +2886,8 @@ int main(int argc, char **argv, char **envp)
translation = BIOS_ATA_TRANSLATION_AUTO;
for (i = 0; i < MAX_NODES; i++) {
- node_mem[i] = 0;
- node_cpumask[i] = bitmap_new(MAX_CPUMASK_BITS);
+ numa_info[i].node_mem = 0;
+ bitmap_zero(numa_info[i].node_cpu, MAX_CPUMASK_BITS);
}
nb_numa_nodes = 0;
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 05/11] NUMA: Add Linux libnuma detection
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (3 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 04/11] NUMA: Add numa_info structure to contain numa nodes info Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 06/11] NUMA: parse guest numa nodes memory policy Wanlong Gao
` (5 subsequent siblings)
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Add detection of libnuma (mostly contained in the numactl package)
to the configure script. Can be enabled or disabled on the command line,
default is use if available.
Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
configure | 32 ++++++++++++++++++++++++++++++++
1 file changed, 32 insertions(+)
diff --git a/configure b/configure
index f0761ea..d6736e0 100755
--- a/configure
+++ b/configure
@@ -243,6 +243,7 @@ gtk=""
gtkabi="2.0"
tpm="no"
libssh2=""
+numa=""
# parse CC options first
for opt do
@@ -945,6 +946,10 @@ for opt do
;;
--enable-libssh2) libssh2="yes"
;;
+ --disable-numa) numa="no"
+ ;;
+ --enable-numa) numa="yes"
+ ;;
*) echo "ERROR: unknown option $opt"; show_help="yes"
;;
esac
@@ -1159,6 +1164,8 @@ echo " --gcov=GCOV use specified gcov [$gcov_tool]"
echo " --enable-tpm enable TPM support"
echo " --disable-libssh2 disable ssh block device support"
echo " --enable-libssh2 enable ssh block device support"
+echo " --disable-numa disable libnuma support"
+echo " --enable-numa enable libnuma support"
echo ""
echo "NOTE: The object files are built at the place where configure is launched"
exit 1
@@ -2412,6 +2419,27 @@ EOF
fi
##########################################
+# libnuma probe
+
+if test "$numa" != "no" ; then
+ numa=no
+ cat > $TMPC << EOF
+#include <numa.h>
+int main(void) { return numa_available(); }
+EOF
+
+ if compile_prog "" "-lnuma" ; then
+ numa=yes
+ libs_softmmu="-lnuma $libs_softmmu"
+ else
+ if test "$numa" = "yes" ; then
+ feature_not_found "linux NUMA (install numactl?)"
+ fi
+ numa=no
+ fi
+fi
+
+##########################################
# linux-aio probe
if test "$linux_aio" != "no" ; then
@@ -3608,6 +3636,7 @@ echo "TPM support $tpm"
echo "libssh2 support $libssh2"
echo "TPM passthrough $tpm_passthrough"
echo "QOM debugging $qom_cast_debug"
+echo "NUMA host support $numa"
if test "$sdl_too_old" = "yes"; then
echo "-> Your SDL version is too old - please upgrade to have SDL support"
@@ -3641,6 +3670,9 @@ echo "extra_cflags=$EXTRA_CFLAGS" >> $config_host_mak
echo "extra_ldflags=$EXTRA_LDFLAGS" >> $config_host_mak
echo "qemu_localedir=$qemu_localedir" >> $config_host_mak
echo "libs_softmmu=$libs_softmmu" >> $config_host_mak
+if test "$numa" = "yes"; then
+ echo "CONFIG_NUMA=y" >> $config_host_mak
+fi
echo "ARCH=$ARCH" >> $config_host_mak
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 06/11] NUMA: parse guest numa nodes memory policy
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (4 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 05/11] NUMA: Add Linux libnuma detection Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 07/11] NUMA: set " Wanlong Gao
` (4 subsequent siblings)
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
The memory policy setting format is like:
policy={default|membind|interleave|preferred}[,relative=true],host-nodes=N-N
And we are adding this setting as a suboption of "-numa mem,",
the memory policy then can be set like following:
-numa node,nodeid=0,cpus=0 \
-numa node,nodeid=1,cpus=1 \
-numa mem,nodeid=0,size=1G,policy=membind,host-nodes=0-1 \
-numa mem,nodeid=1,size=1G,policy=interleave,relative=true,host-nodes=1
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
include/sysemu/sysemu.h | 3 +++
numa.c | 13 +++++++++++++
qapi-schema.json | 31 +++++++++++++++++++++++++++++--
vl.c | 3 +++
4 files changed, 48 insertions(+), 2 deletions(-)
diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h
index 0f6d1bc..09d5c10 100644
--- a/include/sysemu/sysemu.h
+++ b/include/sysemu/sysemu.h
@@ -135,6 +135,9 @@ extern int nb_numa_mem_nodes;
typedef struct node_info {
uint64_t node_mem;
DECLARE_BITMAP(node_cpu, MAX_CPUMASK_BITS);
+ DECLARE_BITMAP(host_mem, MAX_CPUMASK_BITS);
+ NumaNodePolicy policy;
+ bool relative;
} NodeInfo;
extern NodeInfo numa_info[MAX_NODES];
extern QemuOptsList qemu_numa_opts;
diff --git a/numa.c b/numa.c
index 7aaa663..45325a1 100644
--- a/numa.c
+++ b/numa.c
@@ -74,6 +74,7 @@ static int numa_mem_parse(NumaMemOptions *opts)
{
uint16_t nodenr;
uint64_t mem_size;
+ UInt16List *nodes;
if (opts->has_nodeid) {
nodenr = opts->nodeid;
@@ -91,6 +92,18 @@ static int numa_mem_parse(NumaMemOptions *opts)
numa_info[nodenr].node_mem = mem_size;
}
+ if (opts->has_policy) {
+ numa_info[nodenr].policy = opts->policy;
+ }
+
+ if (opts->has_relative) {
+ numa_info[nodenr].relative = opts->relative;
+ }
+
+ for (nodes = opts->host_nodes; nodes; nodes = nodes->next) {
+ bitmap_set(numa_info[nodenr].host_mem, nodes->value->u16, 1);
+ }
+
return 0;
}
diff --git a/qapi-schema.json b/qapi-schema.json
index b31fd08..8ec83dc 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -3817,6 +3817,24 @@
'*mem': 'str' }}
##
+# @NumaNodePolicy
+#
+# NUMA node policy types
+#
+# @default: restore default policy, remove any nondefault policy
+#
+# @membind: a strict policy that restricts memory allocation to the
+# nodes specified
+#
+# @interleave: the page allocations is interleaved across the set
+# of nodes specified
+#
+# @preferred: set the preferred node for allocation
+##
+{ 'enum': 'NumaNodePolicy',
+ 'data': [ 'default', 'membind', 'interleave', 'preferred' ] }
+
+##
# @NumaMemOptions
#
# Set memory information of guest NUMA node.
@@ -3825,9 +3843,18 @@
#
# @size: #optional memory size of this node
#
+# @policy: #optional memory policy of this node
+#
+# @relative: #optional if the nodes specified are relative
+#
+# @host-nodes: #optional host nodes for its memory policy
+#
# Since 1.7
##
{ 'type': 'NumaMemOptions',
'data': {
- '*nodeid': 'uint16',
- '*size': 'size' }}
+ '*nodeid': 'uint16',
+ '*size': 'size',
+ '*policy': 'NumaNodePolicy',
+ '*relative': 'bool',
+ '*host-nodes': ['UInt16'] }}
diff --git a/vl.c b/vl.c
index d9976c7..3982e46 100644
--- a/vl.c
+++ b/vl.c
@@ -2888,6 +2888,9 @@ int main(int argc, char **argv, char **envp)
for (i = 0; i < MAX_NODES; i++) {
numa_info[i].node_mem = 0;
bitmap_zero(numa_info[i].node_cpu, MAX_CPUMASK_BITS);
+ bitmap_zero(numa_info[i].host_mem, MAX_CPUMASK_BITS);
+ numa_info[i].policy = NUMA_NODE_POLICY_DEFAULT;
+ numa_info[i].relative = false;
}
nb_numa_nodes = 0;
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 07/11] NUMA: set guest numa nodes memory policy
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (5 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 06/11] NUMA: parse guest numa nodes memory policy Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 08/11] NUMA: add qmp command set-mem-policy to set memory policy for NUMA node Wanlong Gao
` (3 subsequent siblings)
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Set the guest numa nodes memory policies using the mbind(2)
system call node by node.
After this patch, we are able to set guest nodes memory policies
through the QEMU options, this arms to solve the guest cross
nodes memory access performance issue.
And as you all know, if PCI-passthrough is used,
direct-attached-device uses DMA transfer between device and qemu process.
All pages of the guest will be pinned by get_user_pages().
KVM_ASSIGN_PCI_DEVICE ioctl
kvm_vm_ioctl_assign_device()
=>kvm_assign_device()
=> kvm_iommu_map_memslots()
=> kvm_iommu_map_pages()
=> kvm_pin_pages()
So, with direct-attached-device, all guest page's page count will be +1 and
any page migration will not work. AutoNUMA won't too.
So, we should set the guest nodes memory allocation policies before
the pages are really mapped.
Signed-off-by: Andre Przywara <andre.przywara@amd.com>
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
numa.c | 89 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 89 insertions(+)
diff --git a/numa.c b/numa.c
index 45325a1..c5a2b90 100644
--- a/numa.c
+++ b/numa.c
@@ -28,6 +28,16 @@
#include "qapi-visit.h"
#include "qapi/opts-visitor.h"
#include "qapi/dealloc-visitor.h"
+#include "exec/memory.h"
+
+#ifdef CONFIG_NUMA
+#include <numa.h>
+#include <numaif.h>
+#ifndef MPOL_F_RELATIVE_NODES
+#define MPOL_F_RELATIVE_NODES (1 << 14)
+#define MPOL_F_STATIC_NODES (1 << 15)
+#endif
+#endif
QemuOptsList qemu_numa_opts = {
.name = "numa",
@@ -209,6 +219,78 @@ void set_numa_nodes(void)
}
}
+#ifdef CONFIG_NUMA
+static int node_parse_bind_mode(unsigned int nodeid)
+{
+ int bind_mode;
+
+ switch (numa_info[nodeid].policy) {
+ case NUMA_NODE_POLICY_MEMBIND:
+ bind_mode = MPOL_BIND;
+ break;
+ case NUMA_NODE_POLICY_INTERLEAVE:
+ bind_mode = MPOL_INTERLEAVE;
+ break;
+ case NUMA_NODE_POLICY_PREFERRED:
+ bind_mode = MPOL_PREFERRED;
+ break;
+ case NUMA_NODE_POLICY_DEFAULT:
+ default:
+ bind_mode = MPOL_DEFAULT;
+ return bind_mode;
+ }
+
+ bind_mode |= numa_info[nodeid].relative ?
+ MPOL_F_RELATIVE_NODES : MPOL_F_STATIC_NODES;
+
+ return bind_mode;
+}
+#endif
+
+static int set_node_mem_policy(int nodeid)
+{
+#ifdef CONFIG_NUMA
+ void *ram_ptr;
+ RAMBlock *block;
+ ram_addr_t len, ram_offset = 0;
+ int bind_mode;
+ int i;
+
+ QTAILQ_FOREACH(block, &ram_list.blocks, next) {
+ if (!strcmp(block->mr->name, "pc.ram")) {
+ break;
+ }
+ }
+
+ if (block->host == NULL) {
+ return -1;
+ }
+
+ ram_ptr = block->host;
+ for (i = 0; i < nodeid; i++) {
+ len = numa_info[i].node_mem;
+ ram_offset += len;
+ }
+
+ len = numa_info[i].node_mem;
+ bind_mode = node_parse_bind_mode(i);
+
+ /* This is a workaround for a long standing bug in Linux'
+ * mbind implementation, which cuts off the last specified
+ * node. To stay compatible should this bug be fixed, we
+ * specify one more node and zero this one out.
+ */
+ clear_bit(numa_num_configured_nodes() + 1, numa_info[i].host_mem);
+ if (mbind(ram_ptr + ram_offset, len, bind_mode,
+ numa_info[i].host_mem, numa_num_configured_nodes() + 1, 0)) {
+ perror("mbind");
+ return -1;
+ }
+#endif
+
+ return 0;
+}
+
void set_numa_modes(void)
{
CPUState *cpu;
@@ -221,4 +303,11 @@ void set_numa_modes(void)
}
}
}
+
+ for (i = 0; i < nb_numa_nodes; i++) {
+ if (set_node_mem_policy(i) == -1) {
+ fprintf(stderr,
+ "qemu: can not set host memory policy for node%d\n", i);
+ }
+ }
}
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 08/11] NUMA: add qmp command set-mem-policy to set memory policy for NUMA node
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (6 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 07/11] NUMA: set " Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 09/11] NUMA: add hmp command set-mem-policy Wanlong Gao
` (2 subsequent siblings)
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
This QMP command allows user set guest node's memory policy
through the QMP protocol. The qmp-shell command is like:
set-mem-policy nodeid=0 policy=membind relative=true host-nodes=0-1
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
numa.c | 62 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
qapi-schema.json | 21 +++++++++++++++++++
qmp-commands.hx | 41 +++++++++++++++++++++++++++++++++++++
3 files changed, 124 insertions(+)
diff --git a/numa.c b/numa.c
index c5a2b90..edcda25 100644
--- a/numa.c
+++ b/numa.c
@@ -29,6 +29,7 @@
#include "qapi/opts-visitor.h"
#include "qapi/dealloc-visitor.h"
#include "exec/memory.h"
+#include "qmp-commands.h"
#ifdef CONFIG_NUMA
#include <numa.h>
@@ -311,3 +312,64 @@ void set_numa_modes(void)
}
}
}
+
+void qmp_set_mem_policy(uint16_t nodeid, bool has_policy, NumaNodePolicy policy,
+ bool has_relative, bool relative,
+ bool has_host_nodes, UInt16List *host_nodes,
+ Error **errp)
+{
+ NumaNodePolicy old_policy;
+ bool old_relative;
+ DECLARE_BITMAP(host_mem, MAX_CPUMASK_BITS);
+ UInt16List *nodes;
+
+ if (nodeid >= nb_numa_nodes) {
+ error_setg(errp, "Only has '%d' NUMA nodes", nb_numa_nodes);
+ return;
+ }
+
+ bitmap_copy(host_mem, numa_info[nodeid].host_mem, MAX_CPUMASK_BITS);
+ old_policy = numa_info[nodeid].policy;
+ old_relative = numa_info[nodeid].relative;
+
+ numa_info[nodeid].policy = NUMA_NODE_POLICY_DEFAULT;
+ numa_info[nodeid].relative = false;
+ bitmap_zero(numa_info[nodeid].host_mem, MAX_CPUMASK_BITS);
+
+ if (!has_policy) {
+ if (set_node_mem_policy(nodeid) == -1) {
+ error_setg(errp, "Failed to set memory policy for node%" PRIu16,
+ nodeid);
+ goto error;
+ }
+ return;
+ }
+
+ numa_info[nodeid].policy = policy;
+
+ if (has_relative) {
+ numa_info[nodeid].relative = relative;
+ }
+
+ if (!has_host_nodes) {
+ bitmap_fill(numa_info[nodeid].host_mem, MAX_CPUMASK_BITS);
+ }
+
+ for (nodes = host_nodes; nodes; nodes = nodes->next) {
+ bitmap_set(numa_info[nodeid].host_mem, nodes->value->u16, 1);
+ }
+
+ if (set_node_mem_policy(nodeid) == -1) {
+ error_setg(errp, "Failed to set memory policy for node%" PRIu16,
+ nodeid);
+ goto error;
+ }
+
+ return;
+
+error:
+ bitmap_copy(numa_info[nodeid].host_mem, host_mem, MAX_CPUMASK_BITS);
+ numa_info[nodeid].policy = old_policy;
+ numa_info[nodeid].relative = old_relative;
+ return;
+}
diff --git a/qapi-schema.json b/qapi-schema.json
index 8ec83dc..16cd2b5 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -3858,3 +3858,24 @@
'*policy': 'NumaNodePolicy',
'*relative': 'bool',
'*host-nodes': ['UInt16'] }}
+
+##
+# @set-mem-policy:
+#
+# Set the host memory binding policy for guest NUMA node.
+#
+# @nodeid: The node ID of guest NUMA node to set memory policy to.
+#
+# @policy: #optional The memory policy to be set (default 'default').
+#
+# @relative: #optional If the specified nodes are relative (default 'false')
+#
+# @host-nodes: #optional The host nodes range for memory policy.
+#
+# Returns: Nothing on success
+#
+# Since: 1.7
+##
+{ 'command': 'set-mem-policy',
+ 'data': {'nodeid': 'uint16', '*policy': 'NumaNodePolicy',
+ '*relative': 'bool', '*host-nodes': ['UInt16'] } }
diff --git a/qmp-commands.hx b/qmp-commands.hx
index 2e59b0d..51561ae 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -3051,6 +3051,7 @@ Example:
<- { "return": {} }
EQMP
+
{
.name = "query-rx-filter",
.args_type = "name:s?",
@@ -3114,3 +3115,43 @@ Example:
}
EQMP
+
+ {
+ .name = "set-mem-policy",
+ .args_type = "nodeid:i,policy:s?,relative:b?,host-nodes:s?",
+ .help = "Set the host memory binding policy for guest NUMA node",
+ .mhandler.cmd_new = qmp_marshal_input_set_mem_policy,
+ },
+
+SQMP
+set-mem-policy
+------
+
+Set the host memory binding policy for guest NUMA node
+
+Arguments:
+
+- "nodeid": The nodeid of guest NUMA node to set memory policy to.
+ (json-int)
+- "policy": The memory policy to set.
+ (json-string, optional)
+- "relative": If the specified nodes are relative.
+ (json-bool, optional)
+- "host-nodes": The host nodes contained to this memory policy.
+ (json-string, optional)
+
+Example:
+
+-> { "execute": "set-mem-policy", "arguments": { "nodeid": 0,
+ "policy": "membind",
+ "relative": true,
+ "host-nodes": "0-1" } }
+<- { "return": {} }
+
+Notes:
+ 1. If "policy" is not set, the memory policy of this "nodeid" will be set
+ to "default".
+ 2. If "host-nodes" is not set, the node mask of this "policy" will be set
+ to all available host nodes.
+
+EQMP
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 09/11] NUMA: add hmp command set-mem-policy
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (7 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 08/11] NUMA: add qmp command set-mem-policy to set memory policy for NUMA node Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 10/11] NUMA: add qmp command query-numa Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 11/11] NUMA: convert hmp command info_numa to use qmp command query_numa Wanlong Gao
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Add hmp command set-mem-policy to set host memory policy for a guest
NUMA node. Then we can also set node's memory policy using
the monitor command like:
(qemu) set-mem-policy 0 policy=membind,relative=false,host-nodes=0-1
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
hmp-commands.hx | 16 ++++++++++++++
hmp.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
hmp.h | 1 +
3 files changed, 82 insertions(+)
diff --git a/hmp-commands.hx b/hmp-commands.hx
index 8c6b91a..fe3a26f 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -1587,6 +1587,22 @@ Executes a qemu-io command on the given block device.
ETEXI
{
+ .name = "set-mem-policy",
+ .args_type = "nodeid:i,args:s?",
+ .params = "nodeid [args]",
+ .help = "set host memory policy for a guest NUMA node",
+ .mhandler.cmd = hmp_set_mem_policy,
+ },
+
+STEXI
+@item set-mem-policy @var{nodeid} @var{args}
+@findex set-mem-policy
+
+Set host memory policy for a guest NUMA node
+
+ETEXI
+
+ {
.name = "info",
.args_type = "item:s?",
.params = "[subcommand]",
diff --git a/hmp.c b/hmp.c
index c45514b..b199409 100644
--- a/hmp.c
+++ b/hmp.c
@@ -24,6 +24,9 @@
#include "ui/console.h"
#include "block/qapi.h"
#include "qemu-io.h"
+#include "qapi-visit.h"
+#include "qapi/opts-visitor.h"
+#include "qapi/dealloc-visitor.h"
static void hmp_handle_error(Monitor *mon, Error **errp)
{
@@ -1514,3 +1517,65 @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
hmp_handle_error(mon, &err);
}
+
+void hmp_set_mem_policy(Monitor *mon, const QDict *qdict)
+{
+ Error *local_err = NULL;
+ bool has_policy = true;
+ bool has_relative = true;
+ bool has_host_nodes = true;
+ QemuOpts *opts;
+ NumaMemOptions *object = NULL;
+ NumaNodePolicy policy = NUMA_NODE_POLICY_DEFAULT;
+ bool relative = false;
+ UInt16List *host_nodes = NULL;
+
+ uint64_t nodeid = qdict_get_int(qdict, "nodeid");
+ const char *args = qdict_get_try_str(qdict, "args");
+
+ if (args == NULL) {
+ has_policy = false;
+ has_relative = false;
+ has_host_nodes = false;
+ } else {
+ opts = qemu_opts_parse(qemu_find_opts("numa"), args, 1);
+ if (opts == NULL) {
+ monitor_printf(mon, "Parsing memory policy args failed\n");
+ return;
+ } else {
+ OptsVisitor *ov = opts_visitor_new(opts);
+ visit_type_NumaMemOptions(opts_get_visitor(ov), &object, NULL,
+ &local_err);
+ opts_visitor_cleanup(ov);
+
+ if (error_is_set(&local_err)) {
+ goto error;
+ }
+
+ has_policy = object->has_policy;
+ if (has_policy) {
+ policy = object->policy;
+ }
+ has_relative = object->has_relative;
+ if (has_relative) {
+ relative = object->relative;
+ }
+ has_host_nodes = object->has_host_nodes;
+ if (has_host_nodes) {
+ host_nodes = object->host_nodes;
+ }
+ }
+ }
+
+ qmp_set_mem_policy(nodeid, has_policy, policy, has_relative, relative,
+ has_host_nodes, host_nodes, &local_err);
+error:
+ if (object) {
+ QapiDeallocVisitor *dv = qapi_dealloc_visitor_new();
+ visit_type_NumaMemOptions(qapi_dealloc_get_visitor(dv),
+ &object, NULL, NULL);
+ qapi_dealloc_visitor_cleanup(dv);
+ }
+
+ hmp_handle_error(mon, &local_err);
+}
diff --git a/hmp.h b/hmp.h
index 6c3bdcd..ae09525 100644
--- a/hmp.h
+++ b/hmp.h
@@ -87,5 +87,6 @@ void hmp_nbd_server_stop(Monitor *mon, const QDict *qdict);
void hmp_chardev_add(Monitor *mon, const QDict *qdict);
void hmp_chardev_remove(Monitor *mon, const QDict *qdict);
void hmp_qemu_io(Monitor *mon, const QDict *qdict);
+void hmp_set_mem_policy(Monitor *mon, const QDict *qdict);
#endif
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 10/11] NUMA: add qmp command query-numa
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (8 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 09/11] NUMA: add hmp command set-mem-policy Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 11/11] NUMA: convert hmp command info_numa to use qmp command query_numa Wanlong Gao
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Add qmp command query-numa to show guest NUMA information.
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
numa.c | 69 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
qapi-schema.json | 36 +++++++++++++++++++++++++++++
qmp-commands.hx | 49 ++++++++++++++++++++++++++++++++++++++++
3 files changed, 154 insertions(+)
diff --git a/numa.c b/numa.c
index edcda25..f5b1bec 100644
--- a/numa.c
+++ b/numa.c
@@ -373,3 +373,72 @@ error:
numa_info[nodeid].relative = old_relative;
return;
}
+
+NUMAInfoList *qmp_query_numa(Error **errp)
+{
+ NUMAInfoList *head = NULL, *cur_item = NULL;
+ CPUState *cpu;
+ int i;
+
+ for (i = 0; i < nb_numa_nodes; i++) {
+ NUMAInfoList *info;
+ uint16List *cur_cpu_item = NULL;
+ info = g_malloc0(sizeof(*info));
+ info->value = g_malloc0(sizeof(*info->value));
+ info->value->nodeid = i;
+ for (cpu = first_cpu; cpu != NULL; cpu = cpu->next_cpu) {
+ if (cpu->numa_node == i) {
+ uint16List *node_cpu = g_malloc0(sizeof(*node_cpu));
+ node_cpu->value = cpu->cpu_index;
+
+ if (!cur_cpu_item) {
+ info->value->cpus = cur_cpu_item = node_cpu;
+ } else {
+ cur_cpu_item->next = node_cpu;
+ cur_cpu_item = node_cpu;
+ }
+ }
+ }
+ info->value->memory = numa_info[i].node_mem;
+
+#ifdef CONFIG_NUMA
+ info->value->policy = numa_info[i].policy;
+ info->value->relative = numa_info[i].relative;
+
+ unsigned long first, next;
+ next = first = find_first_bit(numa_info[i].host_mem, MAX_CPUMASK_BITS);
+ if (first == MAX_CPUMASK_BITS) {
+ goto end;
+ }
+ uint16List *cur_node_item = g_malloc0(sizeof(*cur_node_item));
+ cur_node_item->value = first;
+ info->value->host_nodes = cur_node_item;
+ do {
+ if (next == numa_max_node()) {
+ break;
+ }
+
+ next = find_next_bit(numa_info[i].host_mem, MAX_CPUMASK_BITS,
+ next + 1);
+ if (next > numa_max_node() || next == MAX_CPUMASK_BITS) {
+ break;
+ }
+
+ uint16List *host_node = g_malloc0(sizeof(*host_node));
+ host_node->value = next;
+ cur_node_item->next = host_node;
+ cur_node_item = host_node;
+ } while (true);
+end:
+#endif
+
+ if (!cur_item) {
+ head = cur_item = info;
+ } else {
+ cur_item->next = info;
+ cur_item = info;
+ }
+ }
+
+ return head;
+}
diff --git a/qapi-schema.json b/qapi-schema.json
index 16cd2b5..cfb1db8 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -3879,3 +3879,39 @@
{ 'command': 'set-mem-policy',
'data': {'nodeid': 'uint16', '*policy': 'NumaNodePolicy',
'*relative': 'bool', '*host-nodes': ['UInt16'] } }
+
+##
+# @NUMAInfo:
+#
+# Information about guest NUMA nodes
+#
+# @nodeid: NUMA node ID
+#
+# @cpus: VCPUs contained in this node
+#
+# @memory: memory size of this node
+#
+# @policy: memory policy of this node
+#
+# @relative: if host nodes are relative for memory policy
+#
+# @host-nodes: host nodes for its memory policy
+#
+# Since: 1.7
+#
+##
+{ 'type': 'NUMAInfo',
+ 'data': {'nodeid': 'uint16', 'cpus': ['uint16'], 'memory': 'uint64',
+ 'policy': 'NumaNodePolicy', 'relative': 'bool',
+ 'host-nodes': ['uint16'] }}
+
+##
+# @query-numa:
+#
+# Returns a list of information about each guest node.
+#
+# Returns: a list of @NUMAInfo for each guest node
+#
+# Since: 1.7
+##
+{ 'command': 'query-numa', 'returns': ['NUMAInfo'] }
diff --git a/qmp-commands.hx b/qmp-commands.hx
index 51561ae..de3799c 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -3155,3 +3155,52 @@ Notes:
to all available host nodes.
EQMP
+
+ {
+ .name = "query-numa",
+ .args_type = "",
+ .mhandler.cmd_new = qmp_marshal_input_query_numa,
+ },
+
+SQMP
+query-numa
+---------
+
+Show NUMA information.
+
+Return a json-array. Each NUMA node is represented by a json-object,
+which contains:
+
+- "nodeid": NUMA node ID (json-int)
+- "cpus": a json-arry of contained VCPUs
+- "memory": amount of memory in each node in Byte (json-int)
+- "policy": memory policy of this node (json-string)
+- "relative": if host nodes is relative for its memory policy (json-bool)
+- "host-nodes": a json-array of host nodes for its memory policy
+
+Arguments:
+
+Example:
+
+-> { "excute": "query-numa" }
+<- { "return":[
+ {
+ "nodeid": 0,
+ "cpus": [0, 1],
+ "memory": 536870912,
+ "policy": "membind",
+ "relative": false,
+ "host-nodes": [0, 1]
+ },
+ {
+ "nodeid": 1,
+ "cpus": [2, 3],
+ "memory": 536870912,
+ "policy": "interleave",
+ "relative": false,
+ "host-nodes": [1]
+ }
+ ]
+ }
+
+EQMP
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread* [Qemu-devel] [PATCH V6 11/11] NUMA: convert hmp command info_numa to use qmp command query_numa
2013-07-30 7:32 [Qemu-devel] [PATCH V6 00/11] Add support for binding guest numa nodes to host numa nodes Wanlong Gao
` (9 preceding siblings ...)
2013-07-30 7:32 ` [Qemu-devel] [PATCH V6 10/11] NUMA: add qmp command query-numa Wanlong Gao
@ 2013-07-30 7:32 ` Wanlong Gao
10 siblings, 0 replies; 15+ messages in thread
From: Wanlong Gao @ 2013-07-30 7:32 UTC (permalink / raw)
To: qemu-devel
Cc: aliguori, ehabkost, lersek, peter.huangpeng, lcapitulino, bsd,
hutao, y-goto, pbonzini, afaerber, gaowanlong
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>
---
hmp.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
hmp.h | 1 +
monitor.c | 21 +--------------------
3 files changed, 56 insertions(+), 20 deletions(-)
diff --git a/hmp.c b/hmp.c
index b199409..3152f2e 100644
--- a/hmp.c
+++ b/hmp.c
@@ -27,6 +27,7 @@
#include "qapi-visit.h"
#include "qapi/opts-visitor.h"
#include "qapi/dealloc-visitor.h"
+#include "sysemu/sysemu.h"
static void hmp_handle_error(Monitor *mon, Error **errp)
{
@@ -1579,3 +1580,56 @@ error:
hmp_handle_error(mon, &local_err);
}
+
+void hmp_info_numa(Monitor *mon, const QDict *qdict)
+{
+ NUMAInfoList *node_list, *node;
+ uint16List *head;
+ int nodeid;
+ char *policy_str = NULL;
+
+ node_list = qmp_query_numa(NULL);
+
+ monitor_printf(mon, "%d nodes\n", nb_numa_nodes);
+ for (node = node_list; node; node = node->next) {
+ nodeid = node->value->nodeid;
+ monitor_printf(mon, "node %d cpus:", nodeid);
+ head = node->value->cpus;
+ for (head = node->value->cpus; head != NULL; head = head->next) {
+ monitor_printf(mon, " %d", (int)head->value);
+ }
+ monitor_printf(mon, "\n");
+ monitor_printf(mon, "node %d size: %" PRId64 " MB\n",
+ nodeid, node->value->memory >> 20);
+ switch (node->value->policy) {
+ case NUMA_NODE_POLICY_DEFAULT:
+ policy_str = g_strdup("default");
+ break;
+ case NUMA_NODE_POLICY_MEMBIND:
+ policy_str = g_strdup("membind");
+ break;
+ case NUMA_NODE_POLICY_INTERLEAVE:
+ policy_str = g_strdup("interleave");
+ break;
+ case NUMA_NODE_POLICY_PREFERRED:
+ policy_str = g_strdup("preferred");
+ break;
+ default:
+ break;
+ }
+ monitor_printf(mon, "node %d policy: %s\n",
+ nodeid, policy_str ? : " ");
+ if (policy_str) {
+ free(policy_str);
+ }
+ monitor_printf(mon, "node %d relative: %s\n", nodeid,
+ node->value->relative ? "true" : "false");
+ monitor_printf(mon, "node %d host-nodes:", nodeid);
+ for (head = node->value->host_nodes; head != NULL; head = head->next) {
+ monitor_printf(mon, " %d", (int)head->value);
+ }
+ monitor_printf(mon, "\n");
+ }
+
+ qapi_free_NUMAInfoList(node_list);
+}
diff --git a/hmp.h b/hmp.h
index ae09525..56a5efd 100644
--- a/hmp.h
+++ b/hmp.h
@@ -37,6 +37,7 @@ void hmp_info_balloon(Monitor *mon, const QDict *qdict);
void hmp_info_pci(Monitor *mon, const QDict *qdict);
void hmp_info_block_jobs(Monitor *mon, const QDict *qdict);
void hmp_info_tpm(Monitor *mon, const QDict *qdict);
+void hmp_info_numa(Monitor *mon, const QDict *qdict);
void hmp_quit(Monitor *mon, const QDict *qdict);
void hmp_stop(Monitor *mon, const QDict *qdict);
void hmp_system_reset(Monitor *mon, const QDict *qdict);
diff --git a/monitor.c b/monitor.c
index 513af25..13fbe2b 100644
--- a/monitor.c
+++ b/monitor.c
@@ -1811,25 +1811,6 @@ static void do_info_mtree(Monitor *mon, const QDict *qdict)
mtree_info((fprintf_function)monitor_printf, mon);
}
-static void do_info_numa(Monitor *mon, const QDict *qdict)
-{
- int i;
- CPUState *cpu;
-
- monitor_printf(mon, "%d nodes\n", nb_numa_nodes);
- for (i = 0; i < nb_numa_nodes; i++) {
- monitor_printf(mon, "node %d cpus:", i);
- for (cpu = first_cpu; cpu != NULL; cpu = cpu->next_cpu) {
- if (cpu->numa_node == i) {
- monitor_printf(mon, " %d", cpu->cpu_index);
- }
- }
- monitor_printf(mon, "\n");
- monitor_printf(mon, "node %d size: %" PRId64 " MB\n", i,
- numa_info[i].node_mem >> 20);
- }
-}
-
#ifdef CONFIG_PROFILER
int64_t qemu_time;
@@ -2597,7 +2578,7 @@ static mon_cmd_t info_cmds[] = {
.args_type = "",
.params = "",
.help = "show NUMA information",
- .mhandler.cmd = do_info_numa,
+ .mhandler.cmd = hmp_info_numa,
},
{
.name = "usb",
--
1.8.3.3.754.g9c3c367
^ permalink raw reply related [flat|nested] 15+ messages in thread