* [PATCH v5 0/6]: Make 24x7 and GPCI events available in sysfs
From: Sukadev Bhattiprolu @ 2014-12-03 2:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
The current support for the 24x7 and GPCI counters in the kernel requires
users to specify the domain and offset of the event numerically, which is
obviously hard to use:
perf stat -C 0 -e \
'hv_24x7/domain=2,offset=0xd58,starting_index=0,lpar=0xffffffff/' \
sleep 1
This patchset exports the 24x7 and GPCI counters info in sysfs so users can
specify the events by name:
$ cd /sys/bus/event_source/devices/hv_24x7/events
$ cat HPM_CS_FROM_L4_LDATA__PHYS_CORE
domain=0x2,offset=0xd58,starting_index=$core,lpar=0x0
$ cat HPM_TLBIE__VCPU_HOME_CHIP
domain=0x4,offset=0x358,starting_index=$vcpu,lpar=$sibling_guest_id
perf stat -C 0 -e \
'hv_24x7/HPM_CS_FROM_L4_LDATA__PHYS_CORE,starting_index=0'
sleep 1
This patchset adds the kernel support to export events in sysfs. A follow-on
patchset will add support to the perf tool to parse the event parameters like
'lpar=$sibling_guest_id' and display them via 'perf list'.
Changelog[v5]
- [Jiri Olsa, Peter Zijlstra] Use '$arg' notation rather than ?
to indicate event parameters.
- [Michael Ellerman] Separate the kernel and tool patches in the
patchset into different patchsets.
Changelog[v4]
- [Jiri Olsa Rebase to perf/core tree to fix small merge conflict.
Changelog[v3]
- [Jiri Olsa] Changed the event parameters are specified. If
event file specifes 'param=val' make the usage 'param=123'
rather than 'val=123'. (patch 1,2/10)
- Shortened event names using "PHYS" and "VCPU" (patch 4/10)
- Print help message if invalid parameter is specified or required
parameter is missing.
- Moved 3 patches that are unrelated to parametrized events into
a separate patchset.
- Reordered patches so code changes come first.
Changelog[v2]
- [Joe Perches, David Laight] Use beNN_to_cpu() instead of guessing
the size from type.
- Use kmem_cache_free() to free page allocated with kmem_cache_alloc().
- Rebase to recent kernel
Cody P Schafer (6):
perf: provide sysfs_show for struct perf_pmu_events_attr
perf: add PMU_EVENT_ATTR_STRING() helper
powerpc/perf/hv-24x7: parse catalog and populate sysfs with events
powerpc/perf/{hv-gpci, hv-common}: generate requests with counters
annotated
powerpc/perf/hv-gpci: add the remaining gpci requests
powerpc/perf/hv-24x7: Document sysfs event description entries
.../testing/sysfs-bus-event_source-devices-hv_24x7 | 22 +
arch/powerpc/perf/hv-24x7-catalog.h | 25 +
arch/powerpc/perf/hv-24x7-domains.h | 28 +
arch/powerpc/perf/hv-24x7.c | 787 ++++++++++++++++++++-
arch/powerpc/perf/hv-24x7.h | 12 +-
arch/powerpc/perf/hv-common.c | 10 +-
arch/powerpc/perf/hv-gpci-requests.h | 262 +++++++
arch/powerpc/perf/hv-gpci.c | 8 +
arch/powerpc/perf/hv-gpci.h | 37 +-
arch/powerpc/perf/req-gen/_begin.h | 13 +
arch/powerpc/perf/req-gen/_clear.h | 5 +
arch/powerpc/perf/req-gen/_end.h | 4 +
arch/powerpc/perf/req-gen/_request-begin.h | 15 +
arch/powerpc/perf/req-gen/_request-end.h | 8 +
arch/powerpc/perf/req-gen/perf.h | 155 ++++
include/linux/perf_event.h | 10 +
kernel/events/core.c | 8 +
17 files changed, 1365 insertions(+), 44 deletions(-)
create mode 100644 arch/powerpc/perf/hv-24x7-domains.h
create mode 100644 arch/powerpc/perf/hv-gpci-requests.h
create mode 100644 arch/powerpc/perf/req-gen/_begin.h
create mode 100644 arch/powerpc/perf/req-gen/_clear.h
create mode 100644 arch/powerpc/perf/req-gen/_end.h
create mode 100644 arch/powerpc/perf/req-gen/_request-begin.h
create mode 100644 arch/powerpc/perf/req-gen/_request-end.h
create mode 100644 arch/powerpc/perf/req-gen/perf.h
--
1.8.3.1
^ permalink raw reply
* [PATCH v5 1/6] perf: provide sysfs_show for struct perf_pmu_events_attr
From: Sukadev Bhattiprolu @ 2014-12-03 2:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572235-8197-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
(struct perf_pmu_events_attr) is defined in include/linux/perf_event.h,
but the only "show" for it is in x86 and contains x86 specific stuff.
Make a generic one for those of us who are just using the event_str.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
include/linux/perf_event.h | 3 +++
kernel/events/core.c | 8 ++++++++
2 files changed, 11 insertions(+)
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 486e84c..58f59bd 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -897,6 +897,9 @@ struct perf_pmu_events_attr {
const char *event_str;
};
+ssize_t perf_event_sysfs_show(struct device *dev, struct device_attribute *attr,
+ char *page);
+
#define PMU_EVENT_ATTR(_name, _var, _id, _show) \
static struct perf_pmu_events_attr _var = { \
.attr = __ATTR(_name, 0444, _show, NULL), \
diff --git a/kernel/events/core.c b/kernel/events/core.c
index 3e19d3e..1e92387 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -8276,6 +8276,14 @@ void __init perf_event_init(void)
!= 1024);
}
+ssize_t perf_event_sysfs_show(struct device *dev, struct device_attribute *attr,
+ char *page)
+{
+ struct perf_pmu_events_attr *pmu_attr =
+ container_of(attr, struct perf_pmu_events_attr, attr);
+ return sprintf(page, "%s\n", pmu_attr->event_str);
+}
+
static int __init perf_event_sysfs_init(void)
{
struct pmu *pmu;
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 3/6] powerpc/perf/hv-24x7: parse catalog and populate sysfs with events
From: Sukadev Bhattiprolu @ 2014-12-03 2:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572235-8197-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Retrieves and parses the 24x7 catalog on POWER systems that supply it
(right now, only POWER 8). Events are exposed via sysfs in the standard
fashion, and are all parameterized.
$ cd /sys/bus/event_source/devices/hv_24x7/events
$ cat HPM_CS_FROM_L4_LDATA__PHYS_CORE
domain=0x2,offset=0xd58,starting_index=$core,lpar=0x0
$ cat HPM_TLBIE__VCPU_HOME_CHIP
domain=0x4,offset=0x358,starting_index=$vcpu,lpar=$sibling_guest_id
where user has to specify values for $core when specifying the first event
and $vcpu and $sibling_guest_id when specifying the second event with the
perf tool.
Catalog is (at the moment) only parsed on boot. It needs re-parsing
when a some hypervisor events occur. At that point we'll also need to
prevent old events from continuing to function (counter that is passed
in via spare space in the config values?).
Changelog[v5]
[Jiri Olsa, Peter Zijlstra] Prefix required parameters with '$'
to make it easy for user to recognize.
Changelog[v4]
[Sukadev Bhattiprolu] Use PHYS and VCPU in place of PHYSICAL and
VIRTUAL_PROCESSOR to shorten the names of the domains and hence,
events;
Changelog[v2]
[Joe Perches, David Laight] Use beNN_to_cpu() instead of guessing
the size from type.
Use kmem_cache_free() to free page allocated with kmem_cache_alloc().
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
arch/powerpc/perf/hv-24x7-catalog.h | 25 ++
arch/powerpc/perf/hv-24x7-domains.h | 28 ++
arch/powerpc/perf/hv-24x7.c | 787 +++++++++++++++++++++++++++++++++++-
arch/powerpc/perf/hv-24x7.h | 12 +-
4 files changed, 838 insertions(+), 14 deletions(-)
create mode 100644 arch/powerpc/perf/hv-24x7-domains.h
diff --git a/arch/powerpc/perf/hv-24x7-catalog.h b/arch/powerpc/perf/hv-24x7-catalog.h
index 21b19dd..69e2e1f 100644
--- a/arch/powerpc/perf/hv-24x7-catalog.h
+++ b/arch/powerpc/perf/hv-24x7-catalog.h
@@ -30,4 +30,29 @@ struct hv_24x7_catalog_page_0 {
__u8 reserved6[2];
} __packed;
+struct hv_24x7_event_data {
+ __be16 length; /* in bytes, must be a multiple of 16 */
+ __u8 reserved1[2];
+ __u8 domain; /* Chip = 1, Core = 2 */
+ __u8 reserved2[1];
+ __be16 event_group_record_offs; /* in bytes, must be 8 byte aligned */
+ __be16 event_group_record_len; /* in bytes */
+
+ /* in bytes, offset from event_group_record */
+ __be16 event_counter_offs;
+
+ /* verified_state, unverified_state, caveat_state, broken_state, ... */
+ __be32 flags;
+
+ __be16 primary_group_ix;
+ __be16 group_count;
+ __be16 event_name_len;
+ __u8 remainder[];
+ /* __u8 event_name[event_name_len - 2]; */
+ /* __be16 event_description_len; */
+ /* __u8 event_desc[event_description_len - 2]; */
+ /* __be16 detailed_desc_len; */
+ /* __u8 detailed_desc[detailed_desc_len - 2]; */
+} __packed;
+
#endif
diff --git a/arch/powerpc/perf/hv-24x7-domains.h b/arch/powerpc/perf/hv-24x7-domains.h
new file mode 100644
index 0000000..49c1efd
--- /dev/null
+++ b/arch/powerpc/perf/hv-24x7-domains.h
@@ -0,0 +1,28 @@
+
+/*
+ * DOMAIN(name, num, index_kind, is_physical)
+ *
+ * @name: An all caps token, suitable for use in generating an enum
+ * member and appending to an event name in sysfs.
+ *
+ * @num: The number corresponding to the domain as given in
+ * documentation. We assume the catalog domain and the hcall
+ * domain have the same numbering (so far they do), but this
+ * may need to be changed in the future.
+ *
+ * @index_kind: A stringifiable token describing the meaning of the index
+ * within the given domain. Must fit the parsing rules of the
+ * perf sysfs api.
+ *
+ * @is_physical: True if the domain is physical, false otherwise (if virtual).
+ *
+ * Note: The terms PHYS_CHIP, PHYS_CORE, VCPU correspond to physical chip,
+ * physical core and virtual processor in 24x7 Counters specifications.
+ */
+
+DOMAIN(PHYS_CHIP, 0x01, chip, true)
+DOMAIN(PHYS_CORE, 0x02, core, true)
+DOMAIN(VCPU_HOME_CORE, 0x03, vcpu, false)
+DOMAIN(VCPU_HOME_CHIP, 0x04, vcpu, false)
+DOMAIN(VCPU_HOME_NODE, 0x05, vcpu, false)
+DOMAIN(VCPU_REMOTE_NODE, 0x06, vcpu, false)
diff --git a/arch/powerpc/perf/hv-24x7.c b/arch/powerpc/perf/hv-24x7.c
index dba3408..73d5bfc 100644
--- a/arch/powerpc/perf/hv-24x7.c
+++ b/arch/powerpc/perf/hv-24x7.c
@@ -13,16 +13,80 @@
#define pr_fmt(fmt) "hv-24x7: " fmt
#include <linux/perf_event.h>
+#include <linux/rbtree.h>
#include <linux/module.h>
#include <linux/slab.h>
+#include <linux/vmalloc.h>
+
#include <asm/firmware.h>
#include <asm/hvcall.h>
#include <asm/io.h>
+#include <linux/byteorder/generic.h>
#include "hv-24x7.h"
#include "hv-24x7-catalog.h"
#include "hv-common.h"
+static const char *domain_to_index_string(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ return "$"#x;
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ default:
+ WARN(1, "unknown domain %d\n", domain);
+ return "UNKNOWN_DOMAIN_INDEX_STRING";
+ }
+}
+
+static const char *event_domain_suffix(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ return "__" #n;
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ default:
+ WARN(1, "unknown domain %d\n", domain);
+ return "__UNKNOWN_DOMAIN_SUFFIX";
+ }
+}
+
+static bool domain_is_valid(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ /* fall through */
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ return true;
+ default:
+ return false;
+ }
+}
+
+static bool is_physical_domain(unsigned domain)
+{
+ switch (domain) {
+#define DOMAIN(n, v, x, c) \
+ case HV_PERF_DOMAIN_##n: \
+ return c;
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+ default:
+ return false;
+ }
+}
+
+static bool catalog_entry_domain_is_valid(unsigned domain)
+{
+ return is_physical_domain(domain);
+}
+
/*
* TODO: Merging events:
* - Think of the hcall as an interface to a 4d array of counters:
@@ -73,8 +137,115 @@ static struct attribute_group format_group = {
.attrs = format_attrs,
};
+static struct attribute_group event_group = {
+ .name = "events",
+ /* .attrs is set in init */
+};
+
+static struct attribute_group event_desc_group = {
+ .name = "event_descs",
+ /* .attrs is set in init */
+};
+
+static struct attribute_group event_long_desc_group = {
+ .name = "event_long_descs",
+ /* .attrs is set in init */
+};
+
static struct kmem_cache *hv_page_cache;
+static char *event_name(struct hv_24x7_event_data *ev, int *len)
+{
+ *len = be16_to_cpu(ev->event_name_len) - 2;
+ return (char *)ev->remainder;
+}
+
+static char *event_desc(struct hv_24x7_event_data *ev, int *len)
+{
+ unsigned nl = be16_to_cpu(ev->event_name_len);
+ __be16 *desc_len = (__be16 *)(ev->remainder + nl - 2);
+ *len = be16_to_cpu(*desc_len) - 2;
+ return (char *)ev->remainder + nl;
+}
+
+static char *event_long_desc(struct hv_24x7_event_data *ev, int *len)
+{
+ unsigned nl = be16_to_cpu(ev->event_name_len);
+ __be16 *desc_len_ = (__be16 *)(ev->remainder + nl - 2);
+ unsigned desc_len = be16_to_cpu(*desc_len_);
+ __be16 *long_desc_len = (__be16 *)(ev->remainder + nl + desc_len - 2);
+ *len = be16_to_cpu(*long_desc_len) - 2;
+ return (char *)ev->remainder + nl + desc_len;
+}
+
+static bool event_fixed_portion_is_within(struct hv_24x7_event_data *ev,
+ void *end)
+{
+ void *start = ev;
+
+ return (start + offsetof(struct hv_24x7_event_data, remainder)) < end;
+}
+
+/*
+ * Things we don't check:
+ * - padding for desc, name, and long/detailed desc is required to be '\0'
+ * bytes.
+ *
+ * Return NULL if we pass end,
+ * Otherwise return the address of the byte just following the event.
+ */
+static void *event_end(struct hv_24x7_event_data *ev, void *end)
+{
+ void *start = ev;
+ __be16 *dl_, *ldl_;
+ unsigned dl, ldl;
+ unsigned nl = be16_to_cpu(ev->event_name_len);
+
+ if (nl < 2) {
+ pr_debug("%s: name length too short: %d", __func__, nl);
+ return NULL;
+ }
+
+ if (start + nl > end) {
+ pr_debug("%s: start=%p + nl=%u > end=%p",
+ __func__, start, nl, end);
+ return NULL;
+ }
+
+ dl_ = (__be16 *)(ev->remainder + nl - 2);
+ if (!IS_ALIGNED((uintptr_t)dl_, 2))
+ pr_warn("desc len not aligned %p", dl_);
+ dl = be16_to_cpu(*dl_);
+ if (dl < 2) {
+ pr_debug("%s: desc len too short: %d", __func__, dl);
+ return NULL;
+ }
+
+ if (start + nl + dl > end) {
+ pr_debug("%s: (start=%p + nl=%u + dl=%u)=%p > end=%p",
+ __func__, start, nl, dl, start + nl + dl, end);
+ return NULL;
+ }
+
+ ldl_ = (__be16 *)(ev->remainder + nl + dl - 2);
+ if (!IS_ALIGNED((uintptr_t)ldl_, 2))
+ pr_warn("long desc len not aligned %p", ldl_);
+ ldl = be16_to_cpu(*ldl_);
+ if (ldl < 2) {
+ pr_debug("%s: long desc len too short (ldl=%u)",
+ __func__, ldl);
+ return NULL;
+ }
+
+ if (start + nl + dl + ldl > end) {
+ pr_debug("%s: start=%p + nl=%u + dl=%u + ldl=%u > end=%p",
+ __func__, start, nl, dl, ldl, end);
+ return NULL;
+ }
+
+ return start + nl + dl + ldl;
+}
+
static unsigned long h_get_24x7_catalog_page_(unsigned long phys_4096,
unsigned long version,
unsigned long index)
@@ -97,6 +268,605 @@ static unsigned long h_get_24x7_catalog_page(char page[],
version, index);
}
+static unsigned core_domains[] = {
+ HV_PERF_DOMAIN_PHYS_CORE,
+ HV_PERF_DOMAIN_VCPU_HOME_CORE,
+ HV_PERF_DOMAIN_VCPU_HOME_CHIP,
+ HV_PERF_DOMAIN_VCPU_HOME_NODE,
+ HV_PERF_DOMAIN_VCPU_REMOTE_NODE,
+};
+/* chip event data always yeilds a single event, core yeilds multiple */
+#define MAX_EVENTS_PER_EVENT_DATA ARRAY_SIZE(core_domains)
+
+static char *event_fmt(struct hv_24x7_event_data *event, unsigned domain)
+{
+ const char *lpar;
+
+ if (is_physical_domain(domain))
+ lpar = "0x0";
+ else
+ lpar = "$sibling_guest_id";
+
+ return kasprintf(GFP_KERNEL,
+ "domain=0x%x,offset=0x%x,starting_index=%s,lpar=%s",
+ domain,
+ be16_to_cpu(event->event_counter_offs) +
+ be16_to_cpu(event->event_group_record_offs),
+ domain_to_index_string(domain),
+ lpar);
+}
+
+/* Avoid trusting fw to NUL terminate strings */
+static char *memdup_to_str(char *maybe_str, int max_len, gfp_t gfp)
+{
+ return kasprintf(gfp, "%.*s", max_len, maybe_str);
+}
+
+static ssize_t device_show_string(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct dev_ext_attribute *d;
+
+ d = container_of(attr, struct dev_ext_attribute, attr);
+ return sprintf(buf, "%s\n", (char *)d->var);
+}
+
+static struct attribute *device_str_attr_create_(char *name, char *str)
+{
+ struct dev_ext_attribute *attr = kzalloc(sizeof(*attr), GFP_KERNEL);
+
+ if (!attr)
+ return NULL;
+
+ attr->var = str;
+ attr->attr.attr.name = name;
+ attr->attr.attr.mode = 0444;
+ attr->attr.show = device_show_string;
+ return &attr->attr.attr;
+}
+
+static struct attribute *device_str_attr_create(char *name, int name_max,
+ int name_nonce,
+ char *str, size_t str_max)
+{
+ char *n;
+ char *s = memdup_to_str(str, str_max, GFP_KERNEL);
+ struct attribute *a;
+
+ if (!s)
+ return NULL;
+
+ if (!name_nonce)
+ n = kasprintf(GFP_KERNEL, "%.*s", name_max, name);
+ else
+ n = kasprintf(GFP_KERNEL, "%.*s__%d", name_max, name,
+ name_nonce);
+ if (!n)
+ goto out_s;
+
+ a = device_str_attr_create_(n, s);
+ if (!a)
+ goto out_n;
+
+ return a;
+out_n:
+ kfree(n);
+out_s:
+ kfree(s);
+ return NULL;
+}
+
+static void device_str_attr_destroy(struct attribute *attr)
+{
+ struct dev_ext_attribute *d;
+
+ d = container_of(attr, struct dev_ext_attribute, attr.attr);
+ kfree(d->var);
+ kfree(d->attr.attr.name);
+ kfree(d);
+}
+
+static struct attribute *event_to_attr(unsigned ix,
+ struct hv_24x7_event_data *event,
+ unsigned domain,
+ int nonce)
+{
+ int event_name_len;
+ char *ev_name, *a_ev_name, *val;
+ const char *ev_suffix;
+ struct attribute *attr;
+
+ if (!domain_is_valid(domain)) {
+ pr_warn("catalog event %u has invalid domain %u\n",
+ ix, domain);
+ return NULL;
+ }
+
+ val = event_fmt(event, domain);
+ if (!val)
+ return NULL;
+
+ ev_suffix = event_domain_suffix(domain);
+ ev_name = event_name(event, &event_name_len);
+ if (!nonce)
+ a_ev_name = kasprintf(GFP_KERNEL, "%.*s%s",
+ (int)event_name_len, ev_name, ev_suffix);
+ else
+ a_ev_name = kasprintf(GFP_KERNEL, "%.*s%s__%d",
+ (int)event_name_len, ev_name, ev_suffix, nonce);
+
+
+ if (!a_ev_name)
+ goto out_val;
+
+ attr = device_str_attr_create_(a_ev_name, val);
+ if (!attr)
+ goto out_name;
+
+ return attr;
+out_name:
+ kfree(a_ev_name);
+out_val:
+ kfree(val);
+ return NULL;
+}
+
+static struct attribute *event_to_desc_attr(struct hv_24x7_event_data *event,
+ int nonce)
+{
+ int nl, dl;
+ char *name = event_name(event, &nl);
+ char *desc = event_desc(event, &dl);
+
+ /* If there isn't a description, don't create the sysfs file */
+ if (!dl)
+ return NULL;
+
+ return device_str_attr_create(name, nl, nonce, desc, dl);
+}
+
+static struct attribute *
+event_to_long_desc_attr(struct hv_24x7_event_data *event, int nonce)
+{
+ int nl, dl;
+ char *name = event_name(event, &nl);
+ char *desc = event_long_desc(event, &dl);
+
+ /* If there isn't a description, don't create the sysfs file */
+ if (!dl)
+ return NULL;
+
+ return device_str_attr_create(name, nl, nonce, desc, dl);
+}
+
+static ssize_t event_data_to_attrs(unsigned ix, struct attribute **attrs,
+ struct hv_24x7_event_data *event, int nonce)
+{
+ unsigned i;
+
+ switch (event->domain) {
+ case HV_PERF_DOMAIN_PHYS_CHIP:
+ *attrs = event_to_attr(ix, event, event->domain, nonce);
+ return 1;
+ case HV_PERF_DOMAIN_PHYS_CORE:
+ for (i = 0; i < ARRAY_SIZE(core_domains); i++) {
+ attrs[i] = event_to_attr(ix, event, core_domains[i],
+ nonce);
+ if (!attrs[i]) {
+ pr_warn("catalog event %u: individual attr %u "
+ "creation failure\n", ix, i);
+ for (; i; i--)
+ device_str_attr_destroy(attrs[i - 1]);
+ return -1;
+ }
+ }
+ return i;
+ default:
+ pr_warn("catalog event %u: domain %u is not allowed in the "
+ "catalog\n", ix, event->domain);
+ return -1;
+ }
+}
+
+static size_t event_to_attr_ct(struct hv_24x7_event_data *event)
+{
+ switch (event->domain) {
+ case HV_PERF_DOMAIN_PHYS_CHIP:
+ return 1;
+ case HV_PERF_DOMAIN_PHYS_CORE:
+ return ARRAY_SIZE(core_domains);
+ default:
+ return 0;
+ }
+}
+
+static unsigned long vmalloc_to_phys(void *v)
+{
+ struct page *p = vmalloc_to_page(v);
+
+ BUG_ON(!p);
+ return page_to_phys(p) + offset_in_page(v);
+}
+
+/* */
+struct event_uniq {
+ struct rb_node node;
+ const char *name;
+ int nl;
+ unsigned ct;
+ unsigned domain;
+};
+
+static int memord(const void *d1, size_t s1, const void *d2, size_t s2)
+{
+ if (s1 < s2)
+ return 1;
+ if (s2 > s1)
+ return -1;
+
+ return memcmp(d1, d2, s1);
+}
+
+static int ev_uniq_ord(const void *v1, size_t s1, unsigned d1, const void *v2,
+ size_t s2, unsigned d2)
+{
+ int r = memord(v1, s1, v2, s2);
+
+ if (r)
+ return r;
+ if (d1 > d2)
+ return 1;
+ if (d2 > d1)
+ return -1;
+ return 0;
+}
+
+static int event_uniq_add(struct rb_root *root, const char *name, int nl,
+ unsigned domain)
+{
+ struct rb_node **new = &(root->rb_node), *parent = NULL;
+ struct event_uniq *data;
+
+ /* Figure out where to put new node */
+ while (*new) {
+ struct event_uniq *it;
+ int result;
+
+ it = container_of(*new, struct event_uniq, node);
+ result = ev_uniq_ord(name, nl, domain, it->name, it->nl,
+ it->domain);
+
+ parent = *new;
+ if (result < 0)
+ new = &((*new)->rb_left);
+ else if (result > 0)
+ new = &((*new)->rb_right);
+ else {
+ it->ct++;
+ pr_info("found a duplicate event %.*s, ct=%u\n", nl,
+ name, it->ct);
+ return it->ct;
+ }
+ }
+
+ data = kmalloc(sizeof(*data), GFP_KERNEL);
+ if (!data)
+ return -ENOMEM;
+
+ *data = (struct event_uniq) {
+ .name = name,
+ .nl = nl,
+ .ct = 0,
+ .domain = domain,
+ };
+
+ /* Add new node and rebalance tree. */
+ rb_link_node(&data->node, parent, new);
+ rb_insert_color(&data->node, root);
+
+ /* data->ct */
+ return 0;
+}
+
+static void event_uniq_destroy(struct rb_root *root)
+{
+ /*
+ * the strings we point to are in the giant block of memory filled by
+ * the catalog, and are freed separately.
+ */
+ struct event_uniq *pos, *n;
+
+ rbtree_postorder_for_each_entry_safe(pos, n, root, node)
+ kfree(pos);
+}
+
+
+/*
+ * ensure the event structure's sizes are self consistent and don't cause us to
+ * read outside of the event
+ *
+ * On success, return the event length in bytes.
+ * Otherwise, return -1 (and print as appropriate).
+ */
+static ssize_t catalog_event_len_validate(struct hv_24x7_event_data *event,
+ size_t event_idx,
+ size_t event_data_bytes,
+ size_t event_entry_count,
+ size_t offset, void *end)
+{
+ ssize_t ev_len;
+ void *ev_end, *calc_ev_end;
+
+ if (offset >= event_data_bytes)
+ return -1;
+
+ if (event_idx >= event_entry_count) {
+ pr_devel("catalog event data has %zu bytes of padding after last event\n",
+ event_data_bytes - offset);
+ return -1;
+ }
+
+ if (!event_fixed_portion_is_within(event, end)) {
+ pr_warn("event %zu fixed portion is not within range\n",
+ event_idx);
+ return -1;
+ }
+
+ ev_len = be16_to_cpu(event->length);
+
+ if (ev_len % 16)
+ pr_info("event %zu has length %zu not divisible by 16: event=%pK\n",
+ event_idx, ev_len, event);
+
+ ev_end = (__u8 *)event + ev_len;
+ if (ev_end > end) {
+ pr_warn("event %zu has .length=%zu, ends after buffer end: ev_end=%pK > end=%pK, offset=%zu\n",
+ event_idx, ev_len, ev_end, end,
+ offset);
+ return -1;
+ }
+
+ calc_ev_end = event_end(event, end);
+ if (!calc_ev_end) {
+ pr_warn("event %zu has a calculated length which exceeds buffer length %zu: event=%pK end=%pK, offset=%zu\n",
+ event_idx, event_data_bytes, event, end,
+ offset);
+ return -1;
+ }
+
+ if (calc_ev_end > ev_end) {
+ pr_warn("event %zu exceeds it's own length: event=%pK, end=%pK, offset=%zu, calc_ev_end=%pK\n",
+ event_idx, event, ev_end, offset, calc_ev_end);
+ return -1;
+ }
+
+ return ev_len;
+}
+
+#define MAX_4K (SIZE_MAX / 4096)
+
+static void create_events_from_catalog(struct attribute ***events_,
+ struct attribute ***event_descs_,
+ struct attribute ***event_long_descs_)
+{
+ unsigned long hret;
+ size_t catalog_len, catalog_page_len, event_entry_count,
+ event_data_len, event_data_offs,
+ event_data_bytes, junk_events, event_idx, event_attr_ct, i,
+ attr_max, event_idx_last, desc_ct, long_desc_ct;
+ ssize_t ct, ev_len;
+ uint32_t catalog_version_num;
+ struct attribute **events, **event_descs, **event_long_descs;
+ struct hv_24x7_catalog_page_0 *page_0 =
+ kmem_cache_alloc(hv_page_cache, GFP_KERNEL);
+ void *page = page_0;
+ void *event_data, *end;
+ struct hv_24x7_event_data *event;
+ struct rb_root ev_uniq = RB_ROOT;
+
+ if (!page)
+ goto e_out;
+
+ hret = h_get_24x7_catalog_page(page, 0, 0);
+ if (hret)
+ goto e_free;
+
+ catalog_version_num = be64_to_cpu(page_0->version);
+ catalog_page_len = be32_to_cpu(page_0->length);
+
+ if (MAX_4K < catalog_page_len) {
+ pr_err("invalid page count: %zu\n", catalog_page_len);
+ goto e_free;
+ }
+
+ catalog_len = catalog_page_len * 4096;
+
+ event_entry_count = be16_to_cpu(page_0->event_entry_count);
+ event_data_offs = be16_to_cpu(page_0->event_data_offs);
+ event_data_len = be16_to_cpu(page_0->event_data_len);
+
+ pr_devel("cv %zu cl %zu eec %zu edo %zu edl %zu\n",
+ (size_t)catalog_version_num, catalog_len,
+ event_entry_count, event_data_offs, event_data_len);
+
+ if ((MAX_4K < event_data_len)
+ || (MAX_4K < event_data_offs)
+ || (MAX_4K - event_data_offs < event_data_len)) {
+ pr_err("invalid event data offs %zu and/or len %zu\n",
+ event_data_offs, event_data_len);
+ goto e_free;
+ }
+
+ if ((event_data_offs + event_data_len) > catalog_page_len) {
+ pr_err("event data %zu-%zu does not fit inside catalog 0-%zu\n",
+ event_data_offs,
+ event_data_offs + event_data_len,
+ catalog_page_len);
+ goto e_free;
+ }
+
+ if (SIZE_MAX / MAX_EVENTS_PER_EVENT_DATA - 1 < event_entry_count) {
+ pr_err("event_entry_count %zu is invalid\n",
+ event_entry_count);
+ goto e_free;
+ }
+
+ event_data_bytes = event_data_len * 4096;
+
+ /*
+ * event data can span several pages, events can cross between these
+ * pages. Use vmalloc to make this easier.
+ */
+ event_data = vmalloc(event_data_bytes);
+ if (!event_data) {
+ pr_err("could not allocate event data\n");
+ goto e_free;
+ }
+
+ end = event_data + event_data_bytes;
+
+ /*
+ * using vmalloc_to_phys() like this only works if PAGE_SIZE is
+ * divisible by 4096
+ */
+ BUILD_BUG_ON(PAGE_SIZE % 4096);
+
+ for (i = 0; i < event_data_len; i++) {
+ hret = h_get_24x7_catalog_page_(
+ vmalloc_to_phys(event_data + i * 4096),
+ catalog_version_num,
+ i + event_data_offs);
+ if (hret) {
+ pr_err("failed to get event data in page %zu\n",
+ i + event_data_offs);
+ goto e_event_data;
+ }
+ }
+
+ /*
+ * scan the catalog to determine the number of attributes we need, and
+ * verify it at the same time.
+ */
+ for (junk_events = 0, event = event_data, event_idx = 0, attr_max = 0;
+ ;
+ event_idx++, event = (void *)event + ev_len) {
+ size_t offset = (void *)event - (void *)event_data;
+ char *name;
+ int nl;
+
+ ev_len = catalog_event_len_validate(event, event_idx,
+ event_data_bytes,
+ event_entry_count,
+ offset, end);
+ if (ev_len < 0)
+ break;
+
+ name = event_name(event, &nl);
+
+ if (event->event_group_record_len == 0) {
+ pr_devel("invalid event %zu (%.*s): group_record_len == 0, skipping\n",
+ event_idx, nl, name);
+ junk_events++;
+ continue;
+ }
+
+ if (!catalog_entry_domain_is_valid(event->domain)) {
+ pr_info("event %zu (%.*s) has invalid domain %d\n",
+ event_idx, nl, name, event->domain);
+ junk_events++;
+ continue;
+ }
+
+ attr_max += event_to_attr_ct(event);
+ }
+
+ event_idx_last = event_idx;
+ if (event_idx_last != event_entry_count)
+ pr_warn("event buffer ended before listed # of events were parsed (got %zu, wanted %zu, junk %zu)\n",
+ event_idx_last, event_entry_count, junk_events);
+
+ events = kmalloc_array(attr_max + 1, sizeof(*events), GFP_KERNEL);
+ if (!events)
+ goto e_event_data;
+
+ event_descs = kmalloc_array(event_idx + 1, sizeof(*event_descs),
+ GFP_KERNEL);
+ if (!event_descs)
+ goto e_event_attrs;
+
+ event_long_descs = kmalloc_array(event_idx + 1,
+ sizeof(*event_long_descs), GFP_KERNEL);
+ if (!event_long_descs)
+ goto e_event_descs;
+
+ /* Iterate over the catalog filling in the attribute vector */
+ for (junk_events = 0, event_attr_ct = 0, desc_ct = 0, long_desc_ct = 0,
+ event = event_data, event_idx = 0;
+ event_idx < event_idx_last;
+ event_idx++, ev_len = be16_to_cpu(event->length),
+ event = (void *)event + ev_len) {
+ char *name;
+ int nl;
+ int nonce;
+ /*
+ * these are the only "bad" events that are intermixed and that
+ * we can ignore without issue. make sure to skip them here
+ */
+ if (event->event_group_record_len == 0)
+ continue;
+ if (!catalog_entry_domain_is_valid(event->domain))
+ continue;
+
+ name = event_name(event, &nl);
+ nonce = event_uniq_add(&ev_uniq, name, nl, event->domain);
+ ct = event_data_to_attrs(event_idx, events + event_attr_ct,
+ event, nonce);
+ if (ct <= 0) {
+ pr_warn("event %zu (%.*s) creation failure, skipping\n",
+ event_idx, nl, name);
+ junk_events++;
+ } else {
+ event_attr_ct += ct;
+ event_descs[desc_ct] = event_to_desc_attr(event, nonce);
+ if (event_descs[desc_ct])
+ desc_ct++;
+ event_long_descs[long_desc_ct] =
+ event_to_long_desc_attr(event, nonce);
+ if (event_long_descs[long_desc_ct])
+ long_desc_ct++;
+ }
+ }
+
+ pr_info("read %zu catalog entries, created %zu event attrs (%zu failures), %zu descs\n",
+ event_idx, event_attr_ct, junk_events, desc_ct);
+
+ events[event_attr_ct] = NULL;
+ event_descs[desc_ct] = NULL;
+ event_long_descs[long_desc_ct] = NULL;
+
+ event_uniq_destroy(&ev_uniq);
+ vfree(event_data);
+ kmem_cache_free(hv_page_cache, page);
+
+ *events_ = events;
+ *event_descs_ = event_descs;
+ *event_long_descs_ = event_long_descs;
+ return;
+
+e_event_descs:
+ kfree(event_descs);
+e_event_attrs:
+ kfree(events);
+e_event_data:
+ vfree(event_data);
+e_free:
+ kmem_cache_free(hv_page_cache, page);
+e_out:
+ *events_ = NULL;
+ *event_descs_ = NULL;
+ *event_long_descs_ = NULL;
+}
+
static ssize_t catalog_read(struct file *filp, struct kobject *kobj,
struct bin_attribute *bin_attr, char *buf,
loff_t offset, size_t count)
@@ -177,7 +947,7 @@ static ssize_t _name##_show(struct device *dev, \
} \
ret = sprintf(buf, _fmt, _expr); \
e_free: \
- kfree(page); \
+ kmem_cache_free(hv_page_cache, page); \
return ret; \
} \
static DEVICE_ATTR_RO(_name)
@@ -207,16 +977,13 @@ static struct attribute_group if_group = {
static const struct attribute_group *attr_groups[] = {
&format_group,
+ &event_group,
+ &event_desc_group,
+ &event_long_desc_group,
&if_group,
NULL,
};
-static bool is_physical_domain(int domain)
-{
- return domain == HV_24X7_PERF_DOMAIN_PHYSICAL_CHIP ||
- domain == HV_24X7_PERF_DOMAIN_PHYSICAL_CORE;
-}
-
static unsigned long single_24x7_request(u8 domain, u32 offset, u16 ix,
u16 lpar, u64 *res,
bool success_expected)
@@ -359,7 +1126,7 @@ static int h_24x7_event_init(struct perf_event *event)
return -EIO;
}
- /* PHYSICAL domains & other lpars require extra capabilities */
+ /* Physical domains & other lpars require extra capabilities */
if (!caps.collect_privileged && (is_physical_domain(domain) ||
(event_get_lpar(event) != event_get_lpar_max()))) {
pr_devel("hv permisions disallow: is_physical_domain:%d, lpar=0x%llx\n",
@@ -455,6 +1222,10 @@ static int hv_24x7_init(void)
/* sampling not supported */
h_24x7_pmu.capabilities |= PERF_PMU_CAP_NO_INTERRUPT;
+ create_events_from_catalog(&event_group.attrs,
+ &event_desc_group.attrs,
+ &event_long_desc_group.attrs);
+
r = perf_pmu_register(&h_24x7_pmu, h_24x7_pmu.name, -1);
if (r)
return r;
diff --git a/arch/powerpc/perf/hv-24x7.h b/arch/powerpc/perf/hv-24x7.h
index 720ebce..69cd4e6 100644
--- a/arch/powerpc/perf/hv-24x7.h
+++ b/arch/powerpc/perf/hv-24x7.h
@@ -3,14 +3,14 @@
#include <linux/types.h>
+enum hv_perf_domains {
+#define DOMAIN(n, v, x, c) HV_PERF_DOMAIN_##n = v,
+#include "hv-24x7-domains.h"
+#undef DOMAIN
+};
+
struct hv_24x7_request {
/* PHYSICAL domains require enabling via phyp/hmc. */
-#define HV_24X7_PERF_DOMAIN_PHYSICAL_CHIP 0x01
-#define HV_24X7_PERF_DOMAIN_PHYSICAL_CORE 0x02
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_HOME_CORE 0x03
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_HOME_CHIP 0x04
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_HOME_NODE 0x05
-#define HV_24X7_PERF_DOMAIN_VIRTUAL_PROCESSOR_REMOTE_NODE 0x06
__u8 performance_domain;
__u8 reserved[0x1];
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 4/6] powerpc/perf/{hv-gpci, hv-common}: generate requests with counters annotated
From: Sukadev Bhattiprolu @ 2014-12-03 2:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572235-8197-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
This adds (in req-gen/) a framework for defining gpci counter requests.
It uses macro magic similar to ftrace.
Also convert the existing hv-gpci request structures and enum values to
use the new framework (and adjust old users of the structs and enum
values to cope with changes in naming).
In exchange for this macro disaster, we get autogenerated event listing
for GPCI in sysfs, build time field offset checking, and zero
duplication of information about GPCI requests.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
arch/powerpc/perf/hv-common.c | 10 +-
arch/powerpc/perf/hv-gpci-requests.h | 79 +++++++++++++++
arch/powerpc/perf/hv-gpci.c | 8 ++
arch/powerpc/perf/hv-gpci.h | 37 +++----
arch/powerpc/perf/req-gen/_begin.h | 13 +++
arch/powerpc/perf/req-gen/_clear.h | 5 +
arch/powerpc/perf/req-gen/_end.h | 4 +
arch/powerpc/perf/req-gen/_request-begin.h | 15 +++
arch/powerpc/perf/req-gen/_request-end.h | 8 ++
arch/powerpc/perf/req-gen/perf.h | 155 +++++++++++++++++++++++++++++
10 files changed, 304 insertions(+), 30 deletions(-)
create mode 100644 arch/powerpc/perf/hv-gpci-requests.h
create mode 100644 arch/powerpc/perf/req-gen/_begin.h
create mode 100644 arch/powerpc/perf/req-gen/_clear.h
create mode 100644 arch/powerpc/perf/req-gen/_end.h
create mode 100644 arch/powerpc/perf/req-gen/_request-begin.h
create mode 100644 arch/powerpc/perf/req-gen/_request-end.h
create mode 100644 arch/powerpc/perf/req-gen/perf.h
diff --git a/arch/powerpc/perf/hv-common.c b/arch/powerpc/perf/hv-common.c
index 47e02b3..7dce8f10 100644
--- a/arch/powerpc/perf/hv-common.c
+++ b/arch/powerpc/perf/hv-common.c
@@ -9,13 +9,13 @@ unsigned long hv_perf_caps_get(struct hv_perf_caps *caps)
unsigned long r;
struct p {
struct hv_get_perf_counter_info_params params;
- struct cv_system_performance_capabilities caps;
+ struct hv_gpci_system_performance_capabilities caps;
} __packed __aligned(sizeof(uint64_t));
struct p arg = {
.params = {
.counter_request = cpu_to_be32(
- CIR_SYSTEM_PERFORMANCE_CAPABILITIES),
+ HV_GPCI_system_performance_capabilities),
.starting_index = cpu_to_be32(-1),
.counter_info_version_in = 0,
}
@@ -31,9 +31,9 @@ unsigned long hv_perf_caps_get(struct hv_perf_caps *caps)
caps->version = arg.params.counter_info_version_out;
caps->collect_privileged = !!arg.caps.perf_collect_privileged;
- caps->ga = !!(arg.caps.capability_mask & CV_CM_GA);
- caps->expanded = !!(arg.caps.capability_mask & CV_CM_EXPANDED);
- caps->lab = !!(arg.caps.capability_mask & CV_CM_LAB);
+ caps->ga = !!(arg.caps.capability_mask & HV_GPCI_CM_GA);
+ caps->expanded = !!(arg.caps.capability_mask & HV_GPCI_CM_EXPANDED);
+ caps->lab = !!(arg.caps.capability_mask & HV_GPCI_CM_LAB);
return r;
}
diff --git a/arch/powerpc/perf/hv-gpci-requests.h b/arch/powerpc/perf/hv-gpci-requests.h
new file mode 100644
index 0000000..0dfc4d9
--- /dev/null
+++ b/arch/powerpc/perf/hv-gpci-requests.h
@@ -0,0 +1,79 @@
+
+#include "req-gen/_begin.h"
+
+/*
+ * Based on the document "getPerfCountInfo v1.07"
+ */
+
+/* this needs to be -1 encoded in hex suitable for parsing by tools/perf. */
+#define M1 0xffffffffffffffff
+
+/*
+ * #define REQUEST_NAME counter_request_name
+ * #define REQUEST_NUM r_num
+ * #define REQUEST_IDX_KIND starting_index_kind
+ * #include I(REQUEST_BEGIN)
+ * REQUEST(
+ * __field(...)
+ * __field(...)
+ * __array(...)
+ * __count(...)
+ * )
+ * #include I(REQUEST_END)
+ *
+ * - starting_index_kind is one of:
+ * M1: must be -1
+ * chip_id: hardware chip id or -1 for current hw chip
+ * phys_processor_idx:
+ *
+ * __count(offset, bytes, name):
+ * a counter that should be exposed via perf
+ * __field(offset, bytes, name)
+ * a normal field
+ * __array(offset, bytes, name)
+ * an array of bytes
+ *
+ *
+ * @bytes for __count, and __field _must_ be a numeral token
+ * in decimal, not an expression and not in hex.
+ *
+ *
+ * TODO:
+ * - expose secondary index (if any counter ever uses it, only 0xA0
+ * appears to use it right now, and it doesn't have any counters)
+ * - embed versioning info
+ * - include counter descriptions
+ */
+#define REQUEST_NAME dispatch_timebase_by_processor
+#define REQUEST_NUM 0x10
+#define REQUEST_IDX_KIND phys_processor_idx
+#include I(REQUEST_BEGIN)
+REQUEST(__count(0, 8, processor_time_in_timebase_cycles)
+ __field(0x8, 4, hw_processor_id)
+ __field(0xC, 2, owning_part_id)
+ __field(0xE, 1, processor_state)
+ __field(0xF, 1, version)
+ __field(0x10, 4, hw_chip_id)
+ __field(0x14, 4, phys_module_id)
+ __field(0x18, 4, primary_affinity_domain_idx)
+ __field(0x1C, 4, secondary_affinity_domain_idx)
+ __field(0x20, 4, processor_version)
+ __field(0x24, 2, logical_processor_idx)
+ __field(0x26, 2, reserved)
+ __field(0x28, 4, processor_id_register)
+ __field(0x2C, 4, phys_processor_idx)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME system_performance_capabilities
+#define REQUEST_NUM 0x40
+#define REQUEST_IDX_KIND M1
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 1, perf_collect_privileged)
+ __field(0x1, 1 , capability_mask)
+ __array(0x2, 0xE, reserved)
+)
+#include I(REQUEST_END)
+
+#undef M1
+#include "req-gen/_end.h"
diff --git a/arch/powerpc/perf/hv-gpci.c b/arch/powerpc/perf/hv-gpci.c
index a051fe9..c1db188 100644
--- a/arch/powerpc/perf/hv-gpci.c
+++ b/arch/powerpc/perf/hv-gpci.c
@@ -57,6 +57,11 @@ static struct attribute_group format_group = {
.attrs = format_attrs,
};
+static struct attribute_group event_group = {
+ .name = "events",
+ .attrs = hv_gpci_event_attrs,
+};
+
#define HV_CAPS_ATTR(_name, _format) \
static ssize_t _name##_show(struct device *dev, \
struct device_attribute *attr, \
@@ -102,6 +107,7 @@ static struct attribute_group interface_group = {
static const struct attribute_group *attr_groups[] = {
&format_group,
+ &event_group,
&interface_group,
NULL,
};
@@ -265,6 +271,8 @@ static int hv_gpci_init(void)
unsigned long hret;
struct hv_perf_caps caps;
+ hv_gpci_assert_offsets_correct();
+
if (!firmware_has_feature(FW_FEATURE_LPAR)) {
pr_debug("not a virtualized system, not enabling\n");
return -ENODEV;
diff --git a/arch/powerpc/perf/hv-gpci.h b/arch/powerpc/perf/hv-gpci.h
index b25f460..86ede82 100644
--- a/arch/powerpc/perf/hv-gpci.h
+++ b/arch/powerpc/perf/hv-gpci.h
@@ -42,32 +42,19 @@ struct hv_get_perf_counter_info_params {
*/
#define COUNTER_INFO_VERSION_CURRENT 0x8
-/*
- * These determine the counter_value[] layout and the meaning of starting_index
- * and secondary_index.
- *
- * Unless otherwise noted, @secondary_index is unused and ignored.
- */
-enum counter_info_requests {
-
- /* GENERAL */
-
- /* @starting_index: must be -1 (to refer to the current partition)
- */
- CIR_SYSTEM_PERFORMANCE_CAPABILITIES = 0X40,
+/* capability mask masks. */
+enum {
+ HV_GPCI_CM_GA = (1 << 7),
+ HV_GPCI_CM_EXPANDED = (1 << 6),
+ HV_GPCI_CM_LAB = (1 << 5)
};
-struct cv_system_performance_capabilities {
- /* If != 0, allowed to collect data from other partitions */
- __u8 perf_collect_privileged;
-
- /* These following are only valid if counter_info_version >= 0x3 */
-#define CV_CM_GA (1 << 7)
-#define CV_CM_EXPANDED (1 << 6)
-#define CV_CM_LAB (1 << 5)
- /* remaining bits are reserved */
- __u8 capability_mask;
- __u8 reserved[0xE];
-} __packed;
+#define REQUEST_FILE "../hv-gpci-requests.h"
+#define NAME_LOWER hv_gpci
+#define NAME_UPPER HV_GPCI
+#include "req-gen/perf.h"
+#undef REQUEST_FILE
+#undef NAME_LOWER
+#undef NAME_UPPER
#endif
diff --git a/arch/powerpc/perf/req-gen/_begin.h b/arch/powerpc/perf/req-gen/_begin.h
new file mode 100644
index 0000000..acfb17a
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_begin.h
@@ -0,0 +1,13 @@
+/* Include paths to be used in interface defining headers */
+#ifndef POWERPC_PERF_REQ_GEN_H_
+#define POWERPC_PERF_REQ_GEN_H_
+
+#define CAT2_STR_(t, s) __stringify(t/s)
+#define CAT2_STR(t, s) CAT2_STR_(t, s)
+#define I(...) __VA_ARGS__
+
+#endif
+
+#define REQ_GEN_PREFIX req-gen
+#define REQUEST_BEGIN CAT2_STR(REQ_GEN_PREFIX, _request-begin.h)
+#define REQUEST_END CAT2_STR(REQ_GEN_PREFIX, _request-end.h)
diff --git a/arch/powerpc/perf/req-gen/_clear.h b/arch/powerpc/perf/req-gen/_clear.h
new file mode 100644
index 0000000..422974f
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_clear.h
@@ -0,0 +1,5 @@
+
+#undef __field_
+#undef __count_
+#undef __array_
+#undef REQUEST_
diff --git a/arch/powerpc/perf/req-gen/_end.h b/arch/powerpc/perf/req-gen/_end.h
new file mode 100644
index 0000000..8a40698
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_end.h
@@ -0,0 +1,4 @@
+
+#undef REQ_GEN_PREFIX
+#undef REQUEST_BEGIN
+#undef REQUEST_END
diff --git a/arch/powerpc/perf/req-gen/_request-begin.h b/arch/powerpc/perf/req-gen/_request-begin.h
new file mode 100644
index 0000000..f6d9864
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_request-begin.h
@@ -0,0 +1,15 @@
+
+#define REQUEST(r_contents) \
+ REQUEST_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, I(r_contents))
+
+#define __field(f_offset, f_bytes, f_name) \
+ __field_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, \
+ f_offset, f_bytes, f_name)
+
+#define __array(f_offset, f_bytes, f_name) \
+ __array_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, \
+ f_offset, f_bytes, f_name)
+
+#define __count(f_offset, f_bytes, f_name) \
+ __count_(REQUEST_NAME, REQUEST_NUM, REQUEST_IDX_KIND, \
+ f_offset, f_bytes, f_name)
diff --git a/arch/powerpc/perf/req-gen/_request-end.h b/arch/powerpc/perf/req-gen/_request-end.h
new file mode 100644
index 0000000..5573be6
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/_request-end.h
@@ -0,0 +1,8 @@
+#undef REQUEST
+#undef __field
+#undef __array
+#undef __count
+
+#undef REQUEST_NAME
+#undef REQUEST_NUM
+#undef REQUEST_IDX_KIND
diff --git a/arch/powerpc/perf/req-gen/perf.h b/arch/powerpc/perf/req-gen/perf.h
new file mode 100644
index 0000000..591d663
--- /dev/null
+++ b/arch/powerpc/perf/req-gen/perf.h
@@ -0,0 +1,155 @@
+#ifndef LINUX_POWERPC_PERF_REQ_GEN_PERF_H_
+#define LINUX_POWERPC_PERF_REQ_GEN_PERF_H_
+
+#include <linux/perf_event.h>
+
+#ifndef REQUEST_FILE
+#error "REQUEST_FILE must be defined before including"
+#endif
+
+#ifndef NAME_LOWER
+#error "NAME_LOWER must be defined before including"
+#endif
+
+#ifndef NAME_UPPER
+#error "NAME_UPPER must be defined before including"
+#endif
+
+#define BE_TYPE_b1 __u8
+#define BE_TYPE_b2 __be16
+#define BE_TYPE_b4 __be32
+#define BE_TYPE_b8 __be64
+
+#define BYTES_TO_BE_TYPE(bytes) \
+ BE_TYPE_b##bytes
+
+#define CAT2_(a, b) a ## b
+#define CAT2(a, b) CAT2_(a, b)
+#define CAT3_(a, b, c) a ## b ## c
+#define CAT3(a, b, c) CAT3_(a, b, c)
+
+/*
+ * enumerate the request values as
+ * <NAME_UPPER>_<request name> = <request value>
+ */
+#define REQUEST_VALUE__(name_upper, r_name) name_upper ## _ ## r_name
+#define REQUEST_VALUE_(name_upper, r_name) REQUEST_VALUE__(name_upper, r_name)
+#define REQUEST_VALUE(r_name) REQUEST_VALUE_(NAME_UPPER, r_name)
+
+#include "_clear.h"
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+ REQUEST_VALUE(r_name) = r_value,
+enum CAT2(NAME_LOWER, _requests) {
+#include REQUEST_FILE
+};
+
+/*
+ * For each request:
+ * struct <NAME_LOWER>_<request name> {
+ * r_fields
+ * };
+ */
+#include "_clear.h"
+#define STRUCT_NAME__(name_lower, r_name) name_lower ## _ ## r_name
+#define STRUCT_NAME_(name_lower, r_name) STRUCT_NAME__(name_lower, r_name)
+#define STRUCT_NAME(r_name) STRUCT_NAME_(NAME_LOWER, r_name)
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+struct STRUCT_NAME(r_name) { \
+ r_fields \
+};
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_bytes, f_name) \
+ BYTES_TO_BE_TYPE(f_bytes) f_name;
+#define __count_(r_name, r_value, r_idx_1, f_offset, f_bytes, f_name) \
+ __field_(r_name, r_value, r_idx_1, f_offset, f_bytes, f_name)
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_bytes, a_name) \
+ __u8 a_name[a_bytes];
+
+#include REQUEST_FILE
+
+/*
+ * Generate a check of the field offsets
+ * <NAME_LOWER>_assert_offsets_correct()
+ */
+#include "_clear.h"
+#define REQUEST_(r_name, r_value, index, r_fields) \
+r_fields
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_size, f_name) \
+ BUILD_BUG_ON(offsetof(struct STRUCT_NAME(r_name), f_name) != f_offset);
+#define __count_(r_name, r_value, r_idx_1, c_offset, c_size, c_name) \
+ __field_(r_name, r_value, r_idx_1, c_offset, c_size, c_name)
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_size, a_name) \
+ __field_(r_name, r_value, r_idx_1, a_offset, a_size, a_name)
+
+static inline void CAT2(NAME_LOWER, _assert_offsets_correct)(void)
+{
+#include REQUEST_FILE
+}
+
+/*
+ * Generate event attributes:
+ * PMU_EVENT_ATTR_STRING(<request name>_<field name>,
+ * <NAME_LOWER>_event_attr_<request name>_<field name>,
+ * "request=<request value>"
+ * "starting_index=<starting index type>"
+ * "counter_info_version=CURRENT_COUNTER_INFO_VERSION"
+ * "length=<f_size>"
+ * "offset=<f_offset>")
+ *
+ * TODO: counter_info_version may need to vary, we should interperate the
+ * value to some extent
+ */
+#define EVENT_ATTR_NAME__(name, r_name, c_name) \
+ name ## _event_attr_ ## r_name ## _ ## c_name
+#define EVENT_ATTR_NAME_(name, r_name, c_name) \
+ EVENT_ATTR_NAME__(name, r_name, c_name)
+#define EVENT_ATTR_NAME(r_name, c_name) \
+ EVENT_ATTR_NAME_(NAME_LOWER, r_name, c_name)
+
+#include "_clear.h"
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_size, f_name)
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_size, a_name)
+#define __count_(r_name, r_value, r_idx_1, c_offset, c_size, c_name) \
+PMU_EVENT_ATTR_STRING( \
+ CAT3(r_name, _, c_name), \
+ EVENT_ATTR_NAME(r_name, c_name), \
+ "request=" __stringify(r_value) "," \
+ "starting_index=" __stringify(r_idx_1) "," \
+ "counter_info_version=" \
+ __stringify(COUNTER_INFO_VERSION_CURRENT) "," \
+ "length=" #c_size "," \
+ "offset=" #c_offset)
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+ r_fields
+
+#include REQUEST_FILE
+
+/*
+ * Define event attribute array
+ * static struct attribute *hv_gpci_event_attrs[] = {
+ * &<NAME_LOWER>_event_attr_<request name>_<field name>.attr,
+ * };
+ */
+#include "_clear.h"
+#define __field_(r_name, r_value, r_idx_1, f_offset, f_size, f_name)
+#define __count_(r_name, r_value, r_idx_1, c_offset, c_size, c_name) \
+ &EVENT_ATTR_NAME(r_name, c_name).attr.attr,
+#define __array_(r_name, r_value, r_idx_1, a_offset, a_size, a_name)
+#define REQUEST_(r_name, r_value, r_idx_1, r_fields) \
+ r_fields
+
+static __maybe_unused struct attribute *hv_gpci_event_attrs[] = {
+#include REQUEST_FILE
+ NULL
+};
+
+/* cleanup */
+#include "_clear.h"
+#undef EVENT_ATTR_NAME
+#undef EVENT_ATTR_NAME_
+#undef BIT_NAME
+#undef BIT_NAME_
+#undef STRUCT_NAME
+#undef REQUEST_VALUE
+#undef REQUEST_VALUE_
+
+#endif
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 2/6] perf: add PMU_EVENT_ATTR_STRING() helper
From: Sukadev Bhattiprolu @ 2014-12-03 2:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572235-8197-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Helper for constructing static struct perf_pmu_events_attr s.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
include/linux/perf_event.h | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 58f59bd..4038299 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -906,6 +906,13 @@ static struct perf_pmu_events_attr _var = { \
.id = _id, \
};
+#define PMU_EVENT_ATTR_STRING(_name, _var, _value) \
+static struct perf_pmu_events_attr _var = { \
+ .attr = __ATTR(_name, 0444, perf_event_sysfs_show, NULL), \
+ .event_str = _value, \
+};
+
+
#define PMU_FORMAT_ATTR(_name, _format) \
static ssize_t \
_name##_show(struct device *dev, \
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 5/6] powerpc/perf/hv-gpci: add the remaining gpci requests
From: Sukadev Bhattiprolu @ 2014-12-03 2:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572235-8197-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Add the remaining gpci requests that contain counters suitable for use
by perf. Omit those that don't contain any counters (but note their
ommision).
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
arch/powerpc/perf/hv-gpci-requests.h | 183 +++++++++++++++++++++++++++++++++++
1 file changed, 183 insertions(+)
diff --git a/arch/powerpc/perf/hv-gpci-requests.h b/arch/powerpc/perf/hv-gpci-requests.h
index 0dfc4d9..bf70fdf 100644
--- a/arch/powerpc/perf/hv-gpci-requests.h
+++ b/arch/powerpc/perf/hv-gpci-requests.h
@@ -65,6 +65,33 @@ REQUEST(__count(0, 8, processor_time_in_timebase_cycles)
)
#include I(REQUEST_END)
+#define REQUEST_NAME entitled_capped_uncapped_donated_idle_timebase_by_partition
+#define REQUEST_NUM 0x20
+#define REQUEST_IDX_KIND sibling_part_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 8, partition_id)
+ __count(0x8, 8, entitled_cycles)
+ __count(0x10, 8, consumed_capped_cycles)
+ __count(0x18, 8, consumed_uncapped_cycles)
+ __count(0x20, 8, cycles_donated)
+ __count(0x28, 8, purr_idle_cycles)
+)
+#include I(REQUEST_END)
+
+/*
+ * Not available for counter_info_version >= 0x8, use
+ * run_instruction_cycles_by_partition(0x100) instead.
+ */
+#define REQUEST_NAME run_instructions_run_cycles_by_partition
+#define REQUEST_NUM 0x30
+#define REQUEST_IDX_KIND sibling_part_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 8, partition_id)
+ __count(0x8, 8, instructions_completed)
+ __count(0x10, 8, cycles)
+)
+#include I(REQUEST_END)
+
#define REQUEST_NAME system_performance_capabilities
#define REQUEST_NUM 0x40
#define REQUEST_IDX_KIND M1
@@ -75,5 +102,161 @@ REQUEST(__field(0, 1, perf_collect_privileged)
)
#include I(REQUEST_END)
+#define REQUEST_NAME processor_bus_utilization_abc_links
+#define REQUEST_NUM 0x50
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, total_link_cycles)
+ __count(0x18, 8, idle_cycles_for_a_link)
+ __count(0x20, 8, idle_cycles_for_b_link)
+ __count(0x28, 8, idle_cycles_for_c_link)
+ __array(0x30, 0x20, reserved2)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME processor_bus_utilization_wxyz_links
+#define REQUEST_NUM 0x60
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, total_link_cycles)
+ __count(0x18, 8, idle_cycles_for_w_link)
+ __count(0x20, 8, idle_cycles_for_x_link)
+ __count(0x28, 8, idle_cycles_for_y_link)
+ __count(0x30, 8, idle_cycles_for_z_link)
+ __array(0x38, 0x28, reserved2)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME processor_bus_utilization_gx_links
+#define REQUEST_NUM 0x70
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, gx0_in_address_cycles)
+ __count(0x18, 8, gx0_in_data_cycles)
+ __count(0x20, 8, gx0_in_retries)
+ __count(0x28, 8, gx0_in_bus_cycles)
+ __count(0x30, 8, gx0_in_cycles_total)
+ __count(0x38, 8, gx0_out_address_cycles)
+ __count(0x40, 8, gx0_out_data_cycles)
+ __count(0x48, 8, gx0_out_retries)
+ __count(0x50, 8, gx0_out_bus_cycles)
+ __count(0x58, 8, gx0_out_cycles_total)
+ __count(0x60, 8, gx1_in_address_cycles)
+ __count(0x68, 8, gx1_in_data_cycles)
+ __count(0x70, 8, gx1_in_retries)
+ __count(0x78, 8, gx1_in_bus_cycles)
+ __count(0x80, 8, gx1_in_cycles_total)
+ __count(0x88, 8, gx1_out_address_cycles)
+ __count(0x90, 8, gx1_out_data_cycles)
+ __count(0x98, 8, gx1_out_retries)
+ __count(0xA0, 8, gx1_out_bus_cycles)
+ __count(0xA8, 8, gx1_out_cycles_total)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME processor_bus_utilization_mc_links
+#define REQUEST_NUM 0x80
+#define REQUEST_IDX_KIND hw_chip_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, hw_chip_id)
+ __array(0x4, 0xC, reserved1)
+ __count(0x10, 8, mc0_frames)
+ __count(0x18, 8, mc0_reads)
+ __count(0x20, 8, mc0_write)
+ __count(0x28, 8, mc0_total_cycles)
+ __count(0x30, 8, mc1_frames)
+ __count(0x38, 8, mc1_reads)
+ __count(0x40, 8, mc1_writes)
+ __count(0x48, 8, mc1_total_cycles)
+)
+#include I(REQUEST_END)
+
+/* Processor_config (0x90) skipped, no counters */
+/* Current_processor_frequency (0x91) skipped, no counters */
+
+#define REQUEST_NAME processor_core_utilization
+#define REQUEST_NUM 0x94
+#define REQUEST_IDX_KIND phys_processor_idx
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 4, phys_processor_idx)
+ __field(0x4, 4, hw_processor_id)
+ __count(0x8, 8, cycles_across_any_thread)
+ __count(0x10, 8, timebase_at_collection)
+ __count(0x18, 8, purr_cycles)
+ __count(0x20, 8, sum_of_cycles_across_all_threads)
+ __count(0x28, 8, instructions_completed)
+)
+#include I(REQUEST_END)
+
+/* Processor_core_power_mode (0x95) skipped, no counters */
+/* Affinity_domain_information_by_virtual_processor (0xA0) skipped,
+ * no counters */
+/* Affinity_domain_information_by_domain (0xB0) skipped, no counters */
+/* Affinity_domain_information_by_partition (0xB1) skipped, no counters */
+/* Physical_memory_info (0xC0) skipped, no counters */
+/* Processor_bus_topology (0xD0) skipped, no counters */
+
+#define REQUEST_NAME partition_hypervisor_queuing_times
+#define REQUEST_NUM 0xE0
+#define REQUEST_IDX_KIND partition_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 2, partition_id)
+ __array(0x2, 6, reserved1)
+ __count(0x8, 8, time_waiting_for_entitlement)
+ __count(0x10, 8, times_waited_for_entitlement)
+ __count(0x18, 8, time_waiting_for_phys_processor)
+ __count(0x20, 8, times_waited_for_phys_processor)
+ __count(0x28, 8, dispatches_on_home_core)
+ __count(0x30, 8, dispatches_on_home_primary_affinity_domain)
+ __count(0x38, 8, dispatches_on_home_secondary_affinity_domain)
+ __count(0x40, 8, dispatches_off_home_secondary_affinity_domain)
+ __count(0x48, 8, dispatches_on_dedicated_processor_donating_cycles)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME system_hypervisor_times
+#define REQUEST_NUM 0xF0
+#define REQUEST_IDX_KIND M1
+#include I(REQUEST_BEGIN)
+REQUEST(__count(0, 8, time_spent_to_dispatch_virtual_processors)
+ __count(0x8, 8, time_spent_processing_virtual_processor_timers)
+ __count(0x10, 8, time_spent_managing_partitions_over_entitlement)
+ __count(0x18, 8, time_spent_on_system_management)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME system_tlbie_count_and_time
+#define REQUEST_NUM 0xF4
+#define REQUEST_IDX_KIND M1
+#include I(REQUEST_BEGIN)
+REQUEST(__count(0, 8, tlbie_instructions_issued)
+ /*
+ * FIXME: The spec says the offset here is 0x10, which I suspect
+ * is wrong.
+ */
+ __count(0x8, 8, time_spent_issuing_tlbies)
+)
+#include I(REQUEST_END)
+
+#define REQUEST_NAME partition_instruction_count_and_time
+#define REQUEST_NUM 0x100
+#define REQUEST_IDX_KIND partition_id
+#include I(REQUEST_BEGIN)
+REQUEST(__field(0, 2, partition_id)
+ __array(0x2, 0x6, reserved1)
+ __count(0x8, 8, instructions_performed)
+ __count(0x10, 8, time_collected)
+)
+#include I(REQUEST_END)
+
+/* set_mmcrh (0x80001000) skipped, no counters */
+/* retrieve_hpmcx (0x80002000) skipped, no counters */
+
#undef M1
#include "req-gen/_end.h"
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 6/6] powerpc/perf/hv-24x7: Document sysfs event description entries
From: Sukadev Bhattiprolu @ 2014-12-03 2:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572235-8197-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
.../testing/sysfs-bus-event_source-devices-hv_24x7 | 22 ++++++++++++++++++++++
1 file changed, 22 insertions(+)
diff --git a/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7 b/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7
index 32f3f5f..cf70084 100644
--- a/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7
+++ b/Documentation/ABI/testing/sysfs-bus-event_source-devices-hv_24x7
@@ -21,3 +21,25 @@ Contact: Linux on PowerPC Developer List <linuxppc-dev@lists.ozlabs.org>
Description:
Exposes the "version" field of the 24x7 catalog. This is also
extractable from the provided binary "catalog" sysfs entry.
+
+What: /sys/bus/event_source/devices/hv_24x7/event_descs/<event-name>
+Date: February 2014
+Contact: Cody P Schafer <cody@linux.vnet.ibm.com>
+Description:
+ Provides the description of a particular event as provided by
+ the firmware. If firmware does not provide a description, no
+ file will be created.
+
+ Note that the event-name lacks the domain suffix appended for
+ events in the events/ dir.
+
+What: /sys/bus/event_source/devices/hv_24x7/event_long_descs/<event-name>
+Date: February 2014
+Contact: Cody P Schafer <cody@linux.vnet.ibm.com>
+Description:
+ Provides the "long" description of a particular event as
+ provided by the firmware. If firmware does not provide a
+ description, no file will be created.
+
+ Note that the event-name lacks the domain suffix appended for
+ events in the events/ dir.
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 0/4] Add support for parametrized events
From: Sukadev Bhattiprolu @ 2014-12-03 2:09 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
Description of "event parameters" from the documentation patch:
Event parameters are a basic way for partial events to be specified in
sysfs with per-event names given to the fields that need to be filled in
when using a particular event.
It is intended for supporting cases where the single 'cpu' parameter is
insufficient. For example, POWER 8 has events for physical
sockets/cores/cpus that are accessible from with virtual machines. To
keep using the single 'cpu' parameter we'd need to perform a mapping
between Linux's cpus and the physical machine's cpus (in this case
Linux is running under a hypervisor). This isn't possible because
bindings between our cpus and physical cpus may not be fixed, and we
probably won't have a "cpu" on each physical cpu.
Description of the sysfs contents when events are parameterized (copied from an
included patch):
Examples:
domain=0x1,offset=0x8,starting_index=$core
In the case of the last example, a value replacing "$core" would need
to be provided by the user selecting the particular event. This is
refered to as "event parameterization". All non-numerical values
indicate an event parameter.
Notes on how perf-list displays parameterized events
PARAMETERIZED EVENTS
--------------------
Some pmu events listed by 'perf list' will be displayed with '$xyz' in
them. For example:
hv_24x7/HPM_THREAD_NAP_CCYC__PHYS_CORE,starting_index=$core/
This means that when provided as an event, a value for $core must also
be supplied. For example:
perf stat -e \
'hv_24x7/HPM_THREAD_NAP_CCYC__PHYS_CORE,starting_index=2'
...
Changelog[v5]
- [Jiri Olsa, Peter Zijlstra] Use '$arg' notation rather than ?
to indicate event parameters.
- [Michael Ellerman] Separate the kernel and tool patches in the
patchset into different patchsets.
Changelog[v4]
- [Jiri Olsa] Rebase to perf/core tree (fix small merge conflict)
Changelog[v3]
- [Jiri Olsa] Changed the event parameters are specified. If
event file specifes 'param=val' make the usage 'param=123'
rather than 'val=123'. (patch 1,2/10)
- Shortened event names using "PHYS" and "VCPU" (patch 4/10)
- Print help message if invalid parameter is specified or required
parameter is missing.
- Moved 3 patches that are unrelated to parametrized events into
a separate patchset.
- Reordered patches so code changes come first.
Changelog[v2]
- [Joe Perches, David Laight] Use beNN_to_cpu() instead of guessing
the size from type.
- Use kmem_cache_free() to free page allocated with kmem_cache_alloc().
- Rebase to recent kernel
Cody P Schafer (4):
tools/perf: support parsing parameterized events
tools/perf: extend format_alias() to include event parameters
perf Documentation: add event parameters
tools/perf: Document parameterized and symbolic events
.../testing/sysfs-bus-event_source-devices-events | 6 ++
tools/perf/Documentation/perf-list.txt | 13 +++
tools/perf/Documentation/perf-record.txt | 12 +++
tools/perf/Documentation/perf-stat.txt | 20 ++++-
tools/perf/util/parse-events.h | 1 +
tools/perf/util/pmu.c | 92 +++++++++++++++++++---
6 files changed, 128 insertions(+), 16 deletions(-)
--
1.8.3.1
^ permalink raw reply
* [PATCH v5 2/4] tools/perf: extend format_alias() to include event parameters
From: Sukadev Bhattiprolu @ 2014-12-03 2:09 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572578-9051-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
This causes `perf list pmu` to show parameters for parameterized events
like:
pmu/event_name,param1=$param1,param2=$param2/ [Kernel PMU event]
Example:
hv_24x7/HPM_TLBIE__PHYS_CORE,starting_index=$core/ [Kernel PMU event]
Changelog[v5]
[Jiri Olsa, Peter Zijlstra] Use '$' to prefix parameterized events.
Changelog[v4]
[Jiri Olsa] If the parameter for an event in sysfs is 'param=val',
have perf-list show the event as 'param=?' rather than 'val=?'.
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
---
tools/perf/util/pmu.c | 27 ++++++++++++++++++++++++++-
1 file changed, 26 insertions(+), 1 deletion(-)
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index cb516dd..f8674c1 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -810,10 +810,35 @@ void perf_pmu__set_format(unsigned long *bits, long from, long to)
set_bit(b, bits);
}
+static int sub_non_neg(int a, int b)
+{
+ if (b > a)
+ return 0;
+ return a - b;
+}
+
static char *format_alias(char *buf, int len, struct perf_pmu *pmu,
struct perf_pmu_alias *alias)
{
- snprintf(buf, len, "%s/%s/", pmu->name, alias->name);
+ struct parse_events_term *term;
+ int used = snprintf(buf, len, "%s/%s", pmu->name, alias->name);
+
+ list_for_each_entry(term, &alias->terms, list)
+ if (term->type_val == PARSE_EVENTS__TERM_TYPE_STR)
+ used += snprintf(buf + used, sub_non_neg(len, used),
+ ",%s=$%s", term->config,
+ term->val.str);
+
+ if (sub_non_neg(len, used) > 0) {
+ buf[used] = '/';
+ used++;
+ }
+ if (sub_non_neg(len, used) > 0) {
+ buf[used] = '\0';
+ used++;
+ } else
+ buf[len - 1] = '\0';
+
return buf;
}
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 1/4] tools/perf: support parsing parameterized events
From: Sukadev Bhattiprolu @ 2014-12-03 2:09 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572578-9051-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Enable event specification like:
pmu/event_name,param1=0x1,param2=0x4/
Assuming that
/sys/bus/event_source/devices/pmu/events/event_name
Contains something like
param2=$foo,bar=1,param1=$baz
Changelog[v4]:
[Jiri Olsa] Merge to recent perf-core and fix a small conflict.
Changelog[v3]:
[Jiri Olsa] If the sysfs event file specifies 'param=val', make the
usage 'hv_24x7/event,param=123/' rather than 'hv_24x7/event,val=123/'.
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Conflicts:
tools/perf/util/pmu.c
---
tools/perf/util/parse-events.h | 1 +
tools/perf/util/pmu.c | 65 +++++++++++++++++++++++++++++++++++-------
2 files changed, 55 insertions(+), 11 deletions(-)
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index db2cf78..ca226ce 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -71,6 +71,7 @@ struct parse_events_term {
int type_val;
int type_term;
struct list_head list;
+ bool used;
};
struct parse_events_evlist {
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index 5c9c494..cb516dd 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -551,31 +551,68 @@ static void pmu_format_value(unsigned long *format, __u64 value, __u64 *v,
}
/*
+ * Term is a string term, and might be a param-term. Try to look up it's value
+ * in the remaining terms.
+ * - We have a term like "base-or-format-term=param-term",
+ * - We need to find the value supplied for "param-term" (with param-term named
+ * in a config string) later on in the term list.
+ */
+static int pmu_resolve_param_term(struct parse_events_term *term,
+ struct list_head *head_terms,
+ __u64 *value)
+{
+ struct parse_events_term *t;
+
+ list_for_each_entry(t, head_terms, list) {
+ if (t->type_val == PARSE_EVENTS__TERM_TYPE_NUM) {
+ if (!strcmp(t->config, term->config)) {
+ t->used = true;
+ *value = t->val.num;
+ return 0;
+ }
+ }
+ }
+
+ if (verbose)
+ printf("Required parameter '%s' not specified\n", term->config);
+
+ return -1;
+}
+
+/*
* Setup one of config[12] attr members based on the
* user input data - term parameter.
*/
static int pmu_config_term(struct list_head *formats,
struct perf_event_attr *attr,
struct parse_events_term *term,
+ struct list_head *head_terms,
bool zero)
{
struct perf_pmu_format *format;
__u64 *vp;
+ __u64 val;
+
+ /*
+ * If this is a parameter we've already used for parameterized-eval,
+ * skip it in normal eval.
+ */
+ if (term->used)
+ return 0;
/*
- * Support only for hardcoded and numnerial terms.
* Hardcoded terms should be already in, so nothing
* to be done for them.
*/
if (parse_events__is_hardcoded_term(term))
return 0;
- if (term->type_val != PARSE_EVENTS__TERM_TYPE_NUM)
- return -EINVAL;
-
format = pmu_find_format(formats, term->config);
- if (!format)
+ if (!format) {
+ if (verbose)
+ printf("Invalid event/parameter '%s'\n", term->config);
return -EINVAL;
+ }
switch (format->value) {
case PERF_PMU_FORMAT_VALUE_CONFIG:
@@ -592,11 +629,16 @@ static int pmu_config_term(struct list_head *formats,
}
/*
- * XXX If we ever decide to go with string values for
- * non-hardcoded terms, here's the place to translate
- * them into value.
+ * Either directly use a numeric term, or try to translate string terms
+ * using event parameters.
*/
- pmu_format_value(format->bits, term->val.num, vp, zero);
+ if (term->type_val == PARSE_EVENTS__TERM_TYPE_NUM)
+ val = term->val.num;
+ else
+ if (pmu_resolve_param_term(term, head_terms, &val))
+ return -EINVAL;
+
+ pmu_format_value(format->bits, val, vp, zero);
return 0;
}
@@ -607,9 +649,10 @@ int perf_pmu__config_terms(struct list_head *formats,
{
struct parse_events_term *term;
- list_for_each_entry(term, head_terms, list)
- if (pmu_config_term(formats, attr, term, zero))
+ list_for_each_entry(term, head_terms, list) {
+ if (pmu_config_term(formats, attr, term, head_terms, zero))
return -EINVAL;
+ }
return 0;
}
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 3/4] perf Documentation: add event parameters
From: Sukadev Bhattiprolu @ 2014-12-03 2:09 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572578-9051-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Event parameters are a basic way for partial events to be specified in
sysfs with per-event names given to the fields that need to be filled in
when using a particular event.
It is intended for supporting cases where the single 'cpu' parameter is
insufficient. For example, POWER 8 has events for physical
sockets/cores/cpus that are accessible from with virtual machines. To
keep using the single 'cpu' parameter we'd need to perform a mapping
between Linux's cpus and the physical machine's cpus (in this case
Linux is running under a hypervisor). This isn't possible because
bindings between our cpus and physical cpus may not be fixed, and we
probably won't have a "cpu" on each physical cpu.
CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
---
Documentation/ABI/testing/sysfs-bus-event_source-devices-events | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
index 20979f8..f584b16 100644
--- a/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
+++ b/Documentation/ABI/testing/sysfs-bus-event_source-devices-events
@@ -52,12 +52,18 @@ Description: Per-pmu performance monitoring events specific to the running syste
event=0x2abc
event=0x423,inv,cmask=0x3
domain=0x1,offset=0x8,starting_index=0xffff
+ domain=0x1,offset=0x8,starting_index=$phys_cpu
Each of the assignments indicates a value to be assigned to a
particular set of bits (as defined by the format file
corresponding to the <term>) in the perf_event structure passed
to the perf_open syscall.
+ In the case of the last example, a value replacing "$phys_cpu"
+ would need to be provided by the user selecting the particular
+ event. This is referred to as "event parameterization". All
+ non-numerical values indicate an event parameter.
+
What: /sys/bus/event_source/devices/<pmu>/events/<event>.unit
Date: 2014/02/24
Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
--
1.8.3.1
^ permalink raw reply related
* [PATCH v5 4/4] tools/perf: Document parameterized and symbolic events
From: Sukadev Bhattiprolu @ 2014-12-03 2:09 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Jiri Olsa, Michael Ellerman,
Paul Mackerras
Cc: peterz, linuxppc-dev, dev, linux-kernel
In-Reply-To: <1417572578-9051-1-git-send-email-sukadev@linux.vnet.ibm.com>
From: Cody P Schafer <cody@linux.vnet.ibm.com>
Changelog[v6]:
- [Sukadev Bhattiprolu]: Update documentation of perf-list and
perf-record; Added documentation for perf-stat.
CC: Haren Myneni <hbabu@us.ibm.com>
CC: Cody P Schafer <dev@codyps.com>
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
---
tools/perf/Documentation/perf-list.txt | 13 +++++++++++++
tools/perf/Documentation/perf-record.txt | 12 ++++++++++++
tools/perf/Documentation/perf-stat.txt | 20 ++++++++++++++++----
3 files changed, 41 insertions(+), 4 deletions(-)
diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index cbb4f74..8e9a1121 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -89,6 +89,19 @@ raw encoding of 0x1A8 can be used:
You should refer to the processor specific documentation for getting these
details. Some of them are referenced in the SEE ALSO section below.
+PARAMETERIZED EVENTS
+--------------------
+
+Some pmu events listed by 'perf-list' will be displayed with '$x' in them. For
+example:
+
+ hv_gpci/dtbp_ptitc,starting_index=$core/
+
+This means that when provided as an event, a value for '$core' must
+also be supplied. For example:
+
+ perf stat -C 0 -e 'hv_gpci/dtbp_ptitc,starting_index=0x2/' ...
+
OPTIONS
-------
diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt
index af9a54e..acdcf3b 100644
--- a/tools/perf/Documentation/perf-record.txt
+++ b/tools/perf/Documentation/perf-record.txt
@@ -33,6 +33,18 @@ OPTIONS
- a raw PMU event (eventsel+umask) in the form of rNNN where NNN is a
hexadecimal event descriptor.
+ - a symbolically formed PMU event like 'pmu/param1=0x3,param2/' where
+ 'param1', 'param2', etc are defined as formats for the PMU in
+ /sys/bus/event_sources/devices/<pmu>/format/*.
+
+ - a symbolically formed event like 'pmu/config=M,config1=N,config3=K/'
+
+ where M, N, K are numbers (in decimal, hex, octal format). Acceptable
+ values for each of 'config', 'config1' and 'config2' are defined by
+ corresponding entries in /sys/bus/event_sources/devices/<pmu>/format/*
+ param1 and param2 are defined as formats for the PMU in:
+ /sys/bus/event_sources/devices/<pmu>/format/*
+
- a hardware breakpoint event in the form of '\mem:addr[:access]'
where addr is the address in memory you want to break in.
Access is the memory access type (read, write, execute) it can
diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index 29ee857..04e150d 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -25,10 +25,22 @@ OPTIONS
-e::
--event=::
- Select the PMU event. Selection can be a symbolic event name
- (use 'perf list' to list all events) or a raw PMU
- event (eventsel+umask) in the form of rNNN where NNN is a
- hexadecimal event descriptor.
+ Select the PMU event. Selection can be:
+
+ - a symbolic event name (use 'perf list' to list all events)
+
+ - a raw PMU event (eventsel+umask) in the form of rNNN where NNN is a
+ hexadecimal event descriptor.
+
+ - a symbolically formed event like 'pmu/param1=0x3,param2/' where
+ param1 and param2 are defined as formats for the PMU in
+ /sys/bus/event_sources/devices/<pmu>/format/*
+
+ - a symbolically formed event like 'pmu/config=M,config1=N,config2=K/'
+ where M, N, K are numbers (in decimal, hex, octal format).
+ Acceptable values for each of 'config', 'config1' and 'config2'
+ parameters are defined by corresponding entries in
+ /sys/bus/event_sources/devices/<pmu>/format/*
-i::
--no-inherit::
--
1.8.3.1
^ permalink raw reply related
* Re: [PATCH V2] powerpc/mm: don't do tlbie for updatepp request with NO HPTE fault
From: Michael Ellerman @ 2014-12-03 3:37 UTC (permalink / raw)
To: Aneesh Kumar K.V; +Cc: paulus, linuxppc-dev
In-Reply-To: <1417503002-22086-1-git-send-email-aneesh.kumar@linux.vnet.ibm.com>
On Tue, 2014-12-02 at 12:20 +0530, Aneesh Kumar K.V wrote:
> upatepp get called for a nohpte fault, when we find from the linux
> page table that the translation was hashed before. In that case
> we are sure that there is no existing translation, hence we could
> avoid doing tlbie.
Same comments as last time.
http://patchwork.ozlabs.org/patch/406219/
cheers
^ permalink raw reply
* [PATCH] powerpc: powernv: Return to cpu offline loop when finished in KVM guest
From: Paul Mackerras @ 2014-12-03 3:48 UTC (permalink / raw)
To: linuxppc-dev; +Cc: Alexander Graf, kvm-ppc, kvm
When a secondary hardware thread has finished running a KVM guest, we
currently put that thread into nap mode using a nap instruction in
the KVM code. This changes the code so that instead of doing a nap
instruction directly, we instead cause the call to power7_nap() that
put the thread into nap mode to return. The reason for doing this is
to avoid having the KVM code having to know what low-power mode to
put the thread into.
In the case of a secondary thread used to run a KVM guest, the thread
will be offline from the point of view of the host kernel, and the
relevant power7_nap() call is the one in pnv_smp_cpu_disable().
In this case we don't want to clear pending IPIs in the offline loop
in that function, since that might cause us to miss the wakeup for
the next time the thread needs to run a guest. To tell whether or
not to clear the interrupt, we use the SRR1 value returned from
power7_nap(), and check if it indicates an external interrupt. We
arrange that the return from power7_nap() when we have finished running
a guest returns 0, so pending interrupts don't get flushed in that
case.
Note that it is important a secondary thread that has finished
executing in the guest, or that didn't have a guest to run, should
not return to power7_nap's caller while the kvm_hstate.hwthread_req
flag in the PACA is non-zero, because the return from power7_nap
will reenable the MMU, and the MMU might still be in guest context.
In this situation we spin at low priority in real mode waiting for
hwthread_req to become zero.
Signed-off-by: Paul Mackerras <paulus@samba.org>
---
I think this would be best going through the powerpc tree. Alex,
if you can give me an acked-by for this that would be appreciated.
arch/powerpc/include/asm/processor.h | 2 +-
arch/powerpc/kernel/exceptions-64s.S | 2 ++
arch/powerpc/kernel/idle_power7.S | 12 ++++++--
arch/powerpc/kvm/book3s_hv_rmhandlers.S | 54 ++++++++++++++++++++++-----------
arch/powerpc/platforms/powernv/smp.c | 23 +++++++++++---
5 files changed, 68 insertions(+), 25 deletions(-)
diff --git a/arch/powerpc/include/asm/processor.h b/arch/powerpc/include/asm/processor.h
index dda7ac4..29c3798 100644
--- a/arch/powerpc/include/asm/processor.h
+++ b/arch/powerpc/include/asm/processor.h
@@ -451,7 +451,7 @@ extern unsigned long cpuidle_disable;
enum idle_boot_override {IDLE_NO_OVERRIDE = 0, IDLE_POWERSAVE_OFF};
extern int powersave_nap; /* set if nap mode can be used in idle loop */
-extern void power7_nap(int check_irq);
+extern unsigned long power7_nap(int check_irq);
extern void power7_sleep(void);
extern void flush_instruction_cache(void);
extern void hard_reset_now(void);
diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
index a1d45c1..7f29c5f 100644
--- a/arch/powerpc/kernel/exceptions-64s.S
+++ b/arch/powerpc/kernel/exceptions-64s.S
@@ -131,6 +131,8 @@ BEGIN_FTR_SECTION
1:
#endif
+ /* Return SRR1 from power7_nap() */
+ mfspr r3,SPRN_SRR1
beq cr1,2f
b power7_wakeup_noloss
2: b power7_wakeup_loss
diff --git a/arch/powerpc/kernel/idle_power7.S b/arch/powerpc/kernel/idle_power7.S
index c0754bb..18c0687 100644
--- a/arch/powerpc/kernel/idle_power7.S
+++ b/arch/powerpc/kernel/idle_power7.S
@@ -212,6 +212,10 @@ END_FTR_SECTION_IFSET(CPU_FTR_HVMODE)
mtspr SPRN_SRR0,r5
rfid
+/*
+ * R3 here contains the value that will be returned to the caller
+ * of power7_nap.
+ */
_GLOBAL(power7_wakeup_loss)
ld r1,PACAR1(r13)
BEGIN_FTR_SECTION
@@ -219,15 +223,19 @@ BEGIN_FTR_SECTION
END_FTR_SECTION_IFSET(CPU_FTR_HVMODE)
REST_NVGPRS(r1)
REST_GPR(2, r1)
- ld r3,_CCR(r1)
+ ld r6,_CCR(r1)
ld r4,_MSR(r1)
ld r5,_NIP(r1)
addi r1,r1,INT_FRAME_SIZE
- mtcr r3
+ mtcr r6
mtspr SPRN_SRR1,r4
mtspr SPRN_SRR0,r5
rfid
+/*
+ * R3 here contains the value that will be returned to the caller
+ * of power7_nap.
+ */
_GLOBAL(power7_wakeup_noloss)
lbz r0,PACA_NAPSTATELOST(r13)
cmpwi r0,0
diff --git a/arch/powerpc/kvm/book3s_hv_rmhandlers.S b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
index edb2ccd..65c105b 100644
--- a/arch/powerpc/kvm/book3s_hv_rmhandlers.S
+++ b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
@@ -201,8 +201,6 @@ kvmppc_primary_no_guest:
bge kvm_novcpu_exit /* another thread already exiting */
li r3, NAPPING_NOVCPU
stb r3, HSTATE_NAPPING(r13)
- li r3, 1
- stb r3, HSTATE_HWTHREAD_REQ(r13)
b kvm_do_nap
@@ -293,6 +291,8 @@ kvm_start_guest:
/* if we have no vcpu to run, go back to sleep */
beq kvm_no_guest
+kvm_secondary_got_guest:
+
/* Set HSTATE_DSCR(r13) to something sensible */
ld r6, PACA_DSCR(r13)
std r6, HSTATE_DSCR(r13)
@@ -318,27 +318,46 @@ kvm_start_guest:
stwcx. r3, 0, r4
bne 51b
+/*
+ * At this point we have finished executing in the guest.
+ * We need to wait for hwthread_req to become zero, since
+ * we may not turn on the MMU while hwthread_req is non-zero.
+ * While waiting we also need to check if we get given a vcpu to run.
+ */
kvm_no_guest:
- li r0, KVM_HWTHREAD_IN_NAP
+ lbz r3, HSTATE_HWTHREAD_REQ(r13)
+ cmpwi r3, 0
+ bne 53f
+ HMT_MEDIUM
+ li r0, KVM_HWTHREAD_IN_KERNEL
stb r0, HSTATE_HWTHREAD_STATE(r13)
-kvm_do_nap:
- /* Clear the runlatch bit before napping */
- mfspr r2, SPRN_CTRLF
- clrrdi r2, r2, 1
- mtspr SPRN_CTRLT, r2
-
+ /* need to recheck hwthread_req after a barrier, to avoid race */
+ sync
+ lbz r3, HSTATE_HWTHREAD_REQ(r13)
+ cmpwi r3, 0
+ bne 54f
+/*
+ * We jump to power7_wakeup_loss, which will return to the caller
+ * of power7_nap in the powernv cpu offline loop. The value we
+ * put in r3 becomes the return value for power7_nap.
+ */
li r3, LPCR_PECE0
mfspr r4, SPRN_LPCR
rlwimi r4, r3, 0, LPCR_PECE0 | LPCR_PECE1
mtspr SPRN_LPCR, r4
- isync
- std r0, HSTATE_SCRATCH0(r13)
- ptesync
- ld r0, HSTATE_SCRATCH0(r13)
-1: cmpd r0, r0
- bne 1b
- nap
- b .
+ li r3, 0
+ b power7_wakeup_loss
+
+53: HMT_LOW
+ ld r4, HSTATE_KVM_VCPU(r13)
+ cmpdi r4, 0
+ beq kvm_no_guest
+ HMT_MEDIUM
+ b kvm_secondary_got_guest
+
+54: li r0, KVM_HWTHREAD_IN_KVM
+ stb r0, HSTATE_HWTHREAD_STATE(r13)
+ b kvm_no_guest
/******************************************************************************
* *
@@ -2172,6 +2191,7 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_206)
* occurs, with PECE1, PECE0 and PECEDP set in LPCR. Also clear the
* runlatch bit before napping.
*/
+kvm_do_nap:
mfspr r2, SPRN_CTRLF
clrrdi r2, r2, 1
mtspr SPRN_CTRLT, r2
diff --git a/arch/powerpc/platforms/powernv/smp.c b/arch/powerpc/platforms/powernv/smp.c
index 4753958..b716f66 100644
--- a/arch/powerpc/platforms/powernv/smp.c
+++ b/arch/powerpc/platforms/powernv/smp.c
@@ -149,6 +149,7 @@ static int pnv_smp_cpu_disable(void)
static void pnv_smp_cpu_kill_self(void)
{
unsigned int cpu;
+ unsigned long srr1;
/* Standard hot unplug procedure */
local_irq_disable();
@@ -165,13 +166,25 @@ static void pnv_smp_cpu_kill_self(void)
mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1);
while (!generic_check_cpu_restart(cpu)) {
ppc64_runlatch_off();
- power7_nap(1);
+ srr1 = power7_nap(1);
ppc64_runlatch_on();
- /* Clear the IPI that woke us up */
- icp_native_flush_interrupt();
- local_paca->irq_happened &= PACA_IRQ_HARD_DIS;
- mb();
+ /*
+ * If the SRR1 value indicates that we woke up due to
+ * an external interrupt, then clear the interrupt.
+ * We clear the interrupt before checking for the
+ * reason, so as to avoid a race where we wake up for
+ * some other reason, find nothing and clear the interrupt
+ * just as some other cpu is sending us an interrupt.
+ * If we returned from power7_nap as a result of
+ * having finished executing in a KVM guest, then srr1
+ * contains 0.
+ */
+ if ((srr1 & SRR1_WAKEMASK) == SRR1_WAKEEE) {
+ icp_native_flush_interrupt();
+ local_paca->irq_happened &= PACA_IRQ_HARD_DIS;
+ smp_mb();
+ }
if (cpu_core_split_required())
continue;
--
2.1.3
^ permalink raw reply related
* [PATCH v2] powerpc/book3s: Fix flush_tlb cpu_spec hook to take a generic argument.
From: Mahesh J Salgaonkar @ 2014-12-03 4:26 UTC (permalink / raw)
To: linuxppc-dev, Michael Ellerman, Benjamin Herrenschmidt; +Cc: Paul Mackerras
From: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
The flush_tlb hook in cpu_spec was introduced as a generic function hook
to invalidate TLBs. But the current implementation of flush_tlb hook
takes IS (invalidation selector) as an argument which is architecture
dependent. Hence, It is not right to have a generic routine where caller
has to pass non-generic argument.
This patch fixes this and makes flush_tlb hook as high level API.
The existing MCE code calls flush_tlb hook with IS=0 (single page) resulting
partial invalidation of TLBs which is not right. This rework now makes
sure that whole TLB is invalidated to be able to successfully recover from
TLB and ERAT errors.
Reported-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
Changes in V2:
- Moved TLB flush action #defines to cputable.h
- Added missing ptesyncs to before and after TLB flush.
- Moved switch case to flush_tlb_206() function as suggested by
Michael Ellerman.
arch/powerpc/include/asm/cputable.h | 6 +++-
arch/powerpc/include/asm/mmu-hash64.h | 1 +
arch/powerpc/kernel/cpu_setup_power.S | 10 +-----
arch/powerpc/kernel/cputable.c | 4 +-
arch/powerpc/kernel/mce_power.c | 53 ++++++++++++++++++++++++++++++++-
arch/powerpc/kvm/book3s_hv_ras.c | 4 +-
6 files changed, 63 insertions(+), 15 deletions(-)
diff --git a/arch/powerpc/include/asm/cputable.h b/arch/powerpc/include/asm/cputable.h
index daa5af9..d053301 100644
--- a/arch/powerpc/include/asm/cputable.h
+++ b/arch/powerpc/include/asm/cputable.h
@@ -100,7 +100,7 @@ struct cpu_spec {
/*
* Processor specific routine to flush tlbs.
*/
- void (*flush_tlb)(unsigned long inval_selector);
+ void (*flush_tlb)(unsigned int action);
};
@@ -114,6 +114,10 @@ extern void do_feature_fixups(unsigned long value, void *fixup_start,
extern const char *powerpc_base_platform;
+/* TLB flush actions. Used as argument to cpu_spec.flush_tlb() hook */
+#define FLUSH_TLB_ALL 0 /* invalidate all TLBs */
+#define FLUSH_TLB_LPID 1 /* invalidate TLBs for current LPID */
+
#endif /* __ASSEMBLY__ */
/* CPU kernel features */
diff --git a/arch/powerpc/include/asm/mmu-hash64.h b/arch/powerpc/include/asm/mmu-hash64.h
index aeebc94..4f50db7 100644
--- a/arch/powerpc/include/asm/mmu-hash64.h
+++ b/arch/powerpc/include/asm/mmu-hash64.h
@@ -112,6 +112,7 @@
#define TLBIEL_INVAL_SET_SHIFT 12
#define POWER7_TLB_SETS 128 /* # sets in POWER7 TLB */
+#define POWER8_TLB_SETS 512 /* # sets in POWER8 TLB */
#ifndef __ASSEMBLY__
diff --git a/arch/powerpc/kernel/cpu_setup_power.S b/arch/powerpc/kernel/cpu_setup_power.S
index 4673353..9c9b741 100644
--- a/arch/powerpc/kernel/cpu_setup_power.S
+++ b/arch/powerpc/kernel/cpu_setup_power.S
@@ -137,15 +137,11 @@ __init_HFSCR:
/*
* Clear the TLB using the specified IS form of tlbiel instruction
* (invalidate by congruence class). P7 has 128 CCs., P8 has 512.
- *
- * r3 = IS field
*/
__init_tlb_power7:
- li r3,0xc00 /* IS field = 0b11 */
-_GLOBAL(__flush_tlb_power7)
li r6,128
mtctr r6
- mr r7,r3 /* IS field */
+ li r7,0xc00 /* IS field = 0b11 */
ptesync
2: tlbiel r7
addi r7,r7,0x1000
@@ -154,11 +150,9 @@ _GLOBAL(__flush_tlb_power7)
1: blr
__init_tlb_power8:
- li r3,0xc00 /* IS field = 0b11 */
-_GLOBAL(__flush_tlb_power8)
li r6,512
mtctr r6
- mr r7,r3 /* IS field */
+ li r7,0xc00 /* IS field = 0b11 */
ptesync
2: tlbiel r7
addi r7,r7,0x1000
diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c
index 8084059..81c1ba5 100644
--- a/arch/powerpc/kernel/cputable.c
+++ b/arch/powerpc/kernel/cputable.c
@@ -71,8 +71,8 @@ extern void __restore_cpu_power7(void);
extern void __setup_cpu_power8(unsigned long offset, struct cpu_spec* spec);
extern void __restore_cpu_power8(void);
extern void __restore_cpu_a2(void);
-extern void __flush_tlb_power7(unsigned long inval_selector);
-extern void __flush_tlb_power8(unsigned long inval_selector);
+extern void __flush_tlb_power7(unsigned int action);
+extern void __flush_tlb_power8(unsigned int action);
extern long __machine_check_early_realmode_p7(struct pt_regs *regs);
extern long __machine_check_early_realmode_p8(struct pt_regs *regs);
#endif /* CONFIG_PPC64 */
diff --git a/arch/powerpc/kernel/mce_power.c b/arch/powerpc/kernel/mce_power.c
index aa9aff3..65f6043 100644
--- a/arch/powerpc/kernel/mce_power.c
+++ b/arch/powerpc/kernel/mce_power.c
@@ -28,6 +28,55 @@
#include <asm/mce.h>
#include <asm/machdep.h>
+static void flush_tlb_206(unsigned int num_sets, unsigned int action)
+{
+ unsigned long rb;
+ unsigned int i;
+
+ switch (action) {
+ case FLUSH_TLB_ALL:
+ rb = TLBIEL_INVAL_SET;
+ break;
+ case FLUSH_TLB_LPID:
+ rb = TLBIEL_INVAL_SET_LPID;
+ break;
+ default:
+ BUG();
+ break;
+ }
+
+ asm volatile("ptesync" : : : "memory");
+ for (i = 0; i < num_sets; i++) {
+ asm volatile("tlbiel %0" : : "r" (rb));
+ rb += 1 << TLBIEL_INVAL_SET_SHIFT;
+ }
+ asm volatile("ptesync" : : : "memory");
+}
+
+/*
+ * Generic routine to flush TLB on power7. This routine is used as
+ * flush_tlb hook in cpu_spec for Power7 processor.
+ *
+ * action => FLUSH_TLB_ALL: Invalidate all TLBs.
+ * FLUSH_TLB_LPID: Invalidate TLB for current LPID.
+ */
+void __flush_tlb_power7(unsigned int action)
+{
+ flush_tlb_206(POWER7_TLB_SETS, action);
+}
+
+/*
+ * Generic routine to flush TLB on power8. This routine is used as
+ * flush_tlb hook in cpu_spec for power8 processor.
+ *
+ * action => FLUSH_TLB_ALL: Invalidate all TLBs.
+ * FLUSH_TLB_LPID: Invalidate TLB for current LPID.
+ */
+void __flush_tlb_power8(unsigned int action)
+{
+ flush_tlb_206(POWER8_TLB_SETS, action);
+}
+
/* flush SLBs and reload */
static void flush_and_reload_slb(void)
{
@@ -79,7 +128,7 @@ static long mce_handle_derror(uint64_t dsisr, uint64_t slb_error_bits)
}
if (dsisr & P7_DSISR_MC_TLB_MULTIHIT_MFTLB) {
if (cur_cpu_spec && cur_cpu_spec->flush_tlb)
- cur_cpu_spec->flush_tlb(TLBIEL_INVAL_PAGE);
+ cur_cpu_spec->flush_tlb(FLUSH_TLB_ALL);
/* reset error bits */
dsisr &= ~P7_DSISR_MC_TLB_MULTIHIT_MFTLB;
}
@@ -110,7 +159,7 @@ static long mce_handle_common_ierror(uint64_t srr1)
break;
case P7_SRR1_MC_IFETCH_TLB_MULTIHIT:
if (cur_cpu_spec && cur_cpu_spec->flush_tlb) {
- cur_cpu_spec->flush_tlb(TLBIEL_INVAL_PAGE);
+ cur_cpu_spec->flush_tlb(FLUSH_TLB_ALL);
handled = 1;
}
break;
diff --git a/arch/powerpc/kvm/book3s_hv_ras.c b/arch/powerpc/kvm/book3s_hv_ras.c
index d562c8e..8246090 100644
--- a/arch/powerpc/kvm/book3s_hv_ras.c
+++ b/arch/powerpc/kvm/book3s_hv_ras.c
@@ -84,7 +84,7 @@ static long kvmppc_realmode_mc_power7(struct kvm_vcpu *vcpu)
}
if (dsisr & DSISR_MC_TLB_MULTI) {
if (cur_cpu_spec && cur_cpu_spec->flush_tlb)
- cur_cpu_spec->flush_tlb(TLBIEL_INVAL_SET_LPID);
+ cur_cpu_spec->flush_tlb(FLUSH_TLB_LPID);
dsisr &= ~DSISR_MC_TLB_MULTI;
}
/* Any other errors we don't understand? */
@@ -102,7 +102,7 @@ static long kvmppc_realmode_mc_power7(struct kvm_vcpu *vcpu)
break;
case SRR1_MC_IFETCH_TLBMULTI:
if (cur_cpu_spec && cur_cpu_spec->flush_tlb)
- cur_cpu_spec->flush_tlb(TLBIEL_INVAL_SET_LPID);
+ cur_cpu_spec->flush_tlb(FLUSH_TLB_LPID);
break;
default:
handled = 0;
^ permalink raw reply related
* Re: [V6,1/9] elf: Add new powerpc specifc core note sections
From: Michael Ellerman @ 2014-12-03 5:22 UTC (permalink / raw)
To: Anshuman Khandual, linux-kernel, linuxppc-dev
Cc: mikey, james.hogan, avagin, Paul.Clothier, peterz, palves,
shuahkh, oleg, dhowells, kirjanov, tglx, davej, akpm, sukadev,
davem, sam.bobroff
In-Reply-To: <1417507013-11948-2-git-send-email-khandual@linux.vnet.ibm.com>
On Tue, 2014-02-12 at 07:56:45 UTC, Anshuman Khandual wrote:
> This patch adds four new ELF core note sections for powerpc
> transactional memory and one new ELF core note section for
> powerpc general miscellaneous debug registers. These addition
> of new ELF core note sections extends the existing ELF ABI
> without affecting it in any manner.
>
> Acked-by: Andrew Morton <akpm@linux-foundation.org>
> Signed-off-by: Anshuman Khandual <khandual@linux.vnet.ibm.com>
> ---
> include/uapi/linux/elf.h | 5 +++++
> 1 file changed, 5 insertions(+)
>
> diff --git a/include/uapi/linux/elf.h b/include/uapi/linux/elf.h
> index ea9bf25..2260fc0 100644
> --- a/include/uapi/linux/elf.h
> +++ b/include/uapi/linux/elf.h
> @@ -379,6 +379,11 @@ typedef struct elf64_shdr {
> #define NT_PPC_VMX 0x100 /* PowerPC Altivec/VMX registers */
> #define NT_PPC_SPE 0x101 /* PowerPC SPE/EVR registers */
> #define NT_PPC_VSX 0x102 /* PowerPC VSX registers */
> +#define NT_PPC_TM_SPR 0x103 /* PowerPC TM special registers */
> +#define NT_PPC_TM_CGPR 0x104 /* PowerpC TM checkpointed GPR */
> +#define NT_PPC_TM_CFPR 0x105 /* PowerPC TM checkpointed FPR */
> +#define NT_PPC_TM_CVMX 0x106 /* PowerPC TM checkpointed VMX */
> +#define NT_PPC_MISC 0x107 /* PowerPC miscellaneous registers */
This is a really terrible name, "MISC".
Having said that, I guess it's accurate. We have a whole bunch of regs that
have accrued over recent years that aren't accessible via ptrace.
It seems to me if we're adding a misc regset we should be adding everything we
might want to it that is currenty architected.
But currently you only include the PPR, TAR & DSCR.
Looking at Power ISA v2.07, I see the following that could be included:
MMCR2
MMCRA
PMC1
PMC2
PMC3
PMC4
PMC5
PMC6
MMCR0
EBBHR
EBBRR
BESCR
SIAR
SDAR
CFAR?
Those are all new in 2.07 except for CFAR.
There might be more I missed, that was just a quick scan.
Some are only accessible when EBB is in use, maybe those could be a separate
regset.
cheers
^ permalink raw reply
* Re: [PATCH V6 9/9] selftests: Make GIT ignore all binaries in powerpc test suite
From: Anshuman Khandual @ 2014-12-03 5:46 UTC (permalink / raw)
To: Shuah Khan, linux-kernel, linuxppc-dev
Cc: mikey, james.hogan, avagin, Paul.Clothier, peterz, palves, oleg,
davem, dhowells, kirjanov, davej, akpm, sukadev, tglx,
sam.bobroff
In-Reply-To: <547E039F.7070509@osg.samsung.com>
On 12/02/2014 11:53 PM, Shuah Khan wrote:
> On 12/02/2014 12:56 AM, Anshuman Khandual wrote:
>> > This patch includes all of the powerpc test binaries into the
>> > .gitignore file listing in their respective directories. This
>> > will make sure that GIT ignores all of these test binaries while
>> > displaying status.
>> >
>> > Signed-off-by: Anshuman Khandual <khandual@linux.vnet.ibm.com>
>> > ---
>> > .../testing/selftests/powerpc/copyloops/.gitignore | 4 ++++
>> > tools/testing/selftests/powerpc/mm/.gitignore | 1 +
>> > tools/testing/selftests/powerpc/pmu/.gitignore | 3 +++
>> > tools/testing/selftests/powerpc/pmu/ebb/.gitignore | 22 ++++++++++++++++++++++
>> > .../selftests/powerpc/primitives/.gitignore | 1 +
>> > 5 files changed, 31 insertions(+)
>> > create mode 100644 tools/testing/selftests/powerpc/copyloops/.gitignore
>> > create mode 100644 tools/testing/selftests/powerpc/mm/.gitignore
>> > create mode 100644 tools/testing/selftests/powerpc/pmu/.gitignore
>> > create mode 100644 tools/testing/selftests/powerpc/pmu/ebb/.gitignore
>> > create mode 100644 tools/testing/selftests/powerpc/primitives/.gitignore
>> >
> Creating a single .gitignore at tools/testing/selftests/powerpc will
> make this simpler without having to add one .gitignore for each
> directory underneath.
Sure, will do.
>
> Thanks for taking on the task to add .gitignore for all powerpc
> binaries.
^ permalink raw reply
* Re: [PATCH v5 1/6] perf: provide sysfs_show for struct perf_pmu_events_attr
From: Michael Ellerman @ 2014-12-03 6:18 UTC (permalink / raw)
To: Sukadev Bhattiprolu
Cc: peterz, linuxppc-dev, linux-kernel, Arnaldo Carvalho de Melo,
Paul Mackerras, dev, Jiri Olsa
In-Reply-To: <1417572235-8197-2-git-send-email-sukadev@linux.vnet.ibm.com>
On Tue, 2014-12-02 at 18:03 -0800, Sukadev Bhattiprolu wrote:
> From: Cody P Schafer <cody@linux.vnet.ibm.com>
>
> (struct perf_pmu_events_attr) is defined in include/linux/perf_event.h,
> but the only "show" for it is in x86 and contains x86 specific stuff.
>
> Make a generic one for those of us who are just using the event_str.
>
> CC: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
> CC: Haren Myneni <hbabu@us.ibm.com>
> CC: Cody P Schafer <dev@codyps.com>
> Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
> ---
> include/linux/perf_event.h | 3 +++
> kernel/events/core.c | 8 ++++++++
> 2 files changed, 11 insertions(+)
Acme/Peterz, are you happy for us to take this series through the powerpc tree?
Do you want to give us an ack for patches 1 & 2?
cheers
^ permalink raw reply
* Re: [PATCH v2 0/3] fix a kernel panic on fsl corenet board when CONFIG_CLK_PPC_CORENET is enabled
From: Kevin Hao @ 2014-12-03 6:23 UTC (permalink / raw)
To: Scott Wood
Cc: Tang Yuantian-B29983, Gerhard Sittig, Mike Turquette,
linuxppc-dev@lists.ozlabs.org, Lu Jingchang-B35083
In-Reply-To: <1417567137.15957.229.camel@freescale.com>
[-- Attachment #1: Type: text/plain, Size: 278 bytes --]
On Tue, Dec 02, 2014 at 06:38:57PM -0600, Scott Wood wrote:
> If the current code panics as of commit da788acb2838, then the revert
> (but not the other patches) should go to stable as well.
OK, I will make a respin to cc stable and add yours and Michael's ack.
Thanks,
Kevin
[-- Attachment #2: Type: application/pgp-signature, Size: 473 bytes --]
^ permalink raw reply
* Re: [V6,1/9] elf: Add new powerpc specifc core note sections
From: Anshuman Khandual @ 2014-12-03 6:48 UTC (permalink / raw)
To: Michael Ellerman, linux-kernel, linuxppc-dev
Cc: mikey, james.hogan, avagin, Paul.Clothier, peterz, palves,
shuahkh, oleg, dhowells, kirjanov, tglx, davej, akpm, sukadev,
davem, sam.bobroff
In-Reply-To: <20141203052204.9DA8F1400DD@ozlabs.org>
On 12/03/2014 10:52 AM, Michael Ellerman wrote:
> On Tue, 2014-02-12 at 07:56:45 UTC, Anshuman Khandual wrote:
>> This patch adds four new ELF core note sections for powerpc
>> transactional memory and one new ELF core note section for
>> powerpc general miscellaneous debug registers. These addition
>> of new ELF core note sections extends the existing ELF ABI
>> without affecting it in any manner.
>>
>> Acked-by: Andrew Morton <akpm@linux-foundation.org>
>> Signed-off-by: Anshuman Khandual <khandual@linux.vnet.ibm.com>
>> ---
>> include/uapi/linux/elf.h | 5 +++++
>> 1 file changed, 5 insertions(+)
>>
>> diff --git a/include/uapi/linux/elf.h b/include/uapi/linux/elf.h
>> index ea9bf25..2260fc0 100644
>> --- a/include/uapi/linux/elf.h
>> +++ b/include/uapi/linux/elf.h
>> @@ -379,6 +379,11 @@ typedef struct elf64_shdr {
>> #define NT_PPC_VMX 0x100 /* PowerPC Altivec/VMX registers */
>> #define NT_PPC_SPE 0x101 /* PowerPC SPE/EVR registers */
>> #define NT_PPC_VSX 0x102 /* PowerPC VSX registers */
>> +#define NT_PPC_TM_SPR 0x103 /* PowerPC TM special registers */
>> +#define NT_PPC_TM_CGPR 0x104 /* PowerpC TM checkpointed GPR */
>> +#define NT_PPC_TM_CFPR 0x105 /* PowerPC TM checkpointed FPR */
>> +#define NT_PPC_TM_CVMX 0x106 /* PowerPC TM checkpointed VMX */
>> +#define NT_PPC_MISC 0x107 /* PowerPC miscellaneous registers */
>
> This is a really terrible name, "MISC".
>
> Having said that, I guess it's accurate. We have a whole bunch of regs that
> have accrued over recent years that aren't accessible via ptrace.
>
> It seems to me if we're adding a misc regset we should be adding everything we
> might want to it that is currenty architected.
But I believe they also need to be part of the thread_struct structure to be
accessible from ptrace.
>
> But currently you only include the PPR, TAR & DSCR.
Yeah, thats what we started with.
>
> Looking at Power ISA v2.07, I see the following that could be included:
>
> MMCR2
> MMCRA
> PMC1
> PMC2
> PMC3
> PMC4
> PMC5
> PMC6
> MMCR0
> EBBHR
> EBBRR
> BESCR
> SIAR
> SDAR
> CFAR?
MMCRA, PMC[1..6], EBBHR, BESCR, EBBRR, CFAR are not part of the thread struct.
>
> Those are all new in 2.07 except for CFAR.
>
> There might be more I missed, that was just a quick scan.
>
> Some are only accessible when EBB is in use, maybe those could be a separate
> regset.
Yeah we can have one more regset for EBB specific registers.
^ permalink raw reply
* Re: [PATCH] ALSA: i2sbus: Deletion of unnecessary checks before the function call "release_and_free_resource"
From: Takashi Iwai @ 2014-12-03 6:59 UTC (permalink / raw)
To: SF Markus Elfring
Cc: alsa-devel, kernel-janitors, LKML, Jaroslav Kysela, Julia Lawall,
Johannes Berg, linuxppc-dev
In-Reply-To: <547E356D.9050609@users.sourceforge.net>
At Tue, 02 Dec 2014 22:55:57 +0100,
SF Markus Elfring wrote:
>
> From: Markus Elfring <elfring@users.sourceforge.net>
> Date: Tue, 2 Dec 2014 22:50:24 +0100
>
> The release_and_free_resource() function tests whether its argument is NULL
> and then returns immediately. Thus the test around the call is not needed.
>
> This issue was detected by using the Coccinelle software.
>
> Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Thanks, applied.
Takashi
> ---
> sound/aoa/soundbus/i2sbus/core.c | 6 ++----
> 1 file changed, 2 insertions(+), 4 deletions(-)
>
> diff --git a/sound/aoa/soundbus/i2sbus/core.c b/sound/aoa/soundbus/i2sbus/core.c
> index a80d5ea..4e2b4fb 100644
> --- a/sound/aoa/soundbus/i2sbus/core.c
> +++ b/sound/aoa/soundbus/i2sbus/core.c
> @@ -79,8 +79,7 @@ static void i2sbus_release_dev(struct device *dev)
> if (i2sdev->out.dbdma) iounmap(i2sdev->out.dbdma);
> if (i2sdev->in.dbdma) iounmap(i2sdev->in.dbdma);
> for (i = aoa_resource_i2smmio; i <= aoa_resource_rxdbdma; i++)
> - if (i2sdev->allocated_resource[i])
> - release_and_free_resource(i2sdev->allocated_resource[i]);
> + release_and_free_resource(i2sdev->allocated_resource[i]);
> free_dbdma_descriptor_ring(i2sdev, &i2sdev->out.dbdma_ring);
> free_dbdma_descriptor_ring(i2sdev, &i2sdev->in.dbdma_ring);
> for (i = aoa_resource_i2smmio; i <= aoa_resource_rxdbdma; i++)
> @@ -323,8 +322,7 @@ static int i2sbus_add_dev(struct macio_dev *macio,
> if (dev->out.dbdma) iounmap(dev->out.dbdma);
> if (dev->in.dbdma) iounmap(dev->in.dbdma);
> for (i=0;i<3;i++)
> - if (dev->allocated_resource[i])
> - release_and_free_resource(dev->allocated_resource[i]);
> + release_and_free_resource(dev->allocated_resource[i]);
> mutex_destroy(&dev->lock);
> kfree(dev);
> return 0;
> --
> 2.1.3
>
^ permalink raw reply
* Re: [PATCH v3 3/4] powerpc/mpc85xx: Add FSL QorIQ DPAA BMan support to device tree(s)
From: Emil Medve @ 2014-12-03 7:53 UTC (permalink / raw)
To: Scott Wood
Cc: devicetree, Poonam Aggrwal, Geoff Thorpe, linuxppc-dev,
Chunhe Lan
In-Reply-To: <1417566777.15957.227.camel@freescale.com>
Hello Scott,
On 12/02/2014 06:32 PM, Scott Wood wrote:
> On Mon, 2014-12-01 at 04:02 -0600, Emil Medve wrote:
>> diff --git a/arch/powerpc/boot/dts/t4240rdb.dts b/arch/powerpc/boot/dts/t4240rdb.dts
>> index 53761d4..431bf4e 100644
>> --- a/arch/powerpc/boot/dts/t4240rdb.dts
>> +++ b/arch/powerpc/boot/dts/t4240rdb.dts
>> @@ -69,10 +69,27 @@
>> device_type = "memory";
>> };
>>
>> + reserved-memory {
>> + #address-cells = <2>;
>> + #size-cells = <2>;
>> + ranges;
>> +
>> + bman_fbpr: bman-fbpr {
>> + compatible = "fsl,bman-fbpr";
>> + alloc-ranges = <0 0 0xffff 0xffffffff>;
>> + size = <0 0x1000000>;
>> + alignment = <0 0x1000000>;
>> + };
>> + };
>
> Can't this be done at the SoC level rather than board level?
The size of the memory is not SoC specific. Among other things is
determined by the number of MACs that are pinned-out on the board
Cheers,
^ permalink raw reply
* [PATCH v3 0/3] fix a kernel panic on fsl corenet board when CONFIG_CLK_PPC_CORENET is enabled
From: Kevin Hao @ 2014-12-03 8:53 UTC (permalink / raw)
To: linuxppc-dev
Cc: Mike Turquette, Gerhard Sittig, Yuantian Tang, Jingchang Lu,
Scott Wood
I have done a boot test on p5040ds and mpc5121eads board.
v3:
- Rebase on next.
- CC stable for patch 1.
- Add Scott and Michael's ack.
v2:
- Revert the commit da788acb2838 first.
- Invoke of_clk_init() from a common place.
v1
This tries to fix a kernel panic introduced by commit da788acb2838
("clk: ppc-corenet: Fix Section mismatch warning").
Kevin Hao (3):
Revert "clk: ppc-corenet: Fix Section mismatch warning"
powerpc: call of_clk_init() from time_init()
clk: ppc-corenet: fix section mismatch warning
arch/powerpc/kernel/time.c | 5 ++++
arch/powerpc/platforms/512x/clock-commonclk.c | 11 ++++---
drivers/clk/clk-ppc-corenet.c | 43 ++++-----------------------
3 files changed, 16 insertions(+), 43 deletions(-)
--
1.9.3
^ permalink raw reply
* [PATCH v3 1/3] Revert "clk: ppc-corenet: Fix Section mismatch warning"
From: Kevin Hao @ 2014-12-03 8:53 UTC (permalink / raw)
To: linuxppc-dev
Cc: Mike Turquette, Gerhard Sittig, Yuantian Tang, Jingchang Lu,
Scott Wood
In-Reply-To: <1417596833-31456-1-git-send-email-haokexin@gmail.com>
This reverts commit da788acb28386aa896224e784954bb73c99ff26c.
That commit tried to fix the section mismatch warning by moving the
ppc_corenet_clk_driver struct to init section. This is definitely wrong
because the kernel would free the memories occupied by this struct
after boot while this driver is still registered in the driver core.
The kernel would panic when accessing this driver struct.
Cc: stable@vger.kernel.org # 3.17
Signed-off-by: Kevin Hao <haokexin@gmail.com>
Acked-by: Scott Wood <scottwood@freescale.com>
Acked-by: Michael Turquette <mturquette@linaro.org>
---
v3: Cc stable and add ack.
v2: A new patch in v2.
drivers/clk/clk-ppc-corenet.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/clk/clk-ppc-corenet.c b/drivers/clk/clk-ppc-corenet.c
index b6e6c85507a5..0a47d6f49cd6 100644
--- a/drivers/clk/clk-ppc-corenet.c
+++ b/drivers/clk/clk-ppc-corenet.c
@@ -291,7 +291,7 @@ static const struct of_device_id ppc_clk_ids[] __initconst = {
{}
};
-static struct platform_driver ppc_corenet_clk_driver __initdata = {
+static struct platform_driver ppc_corenet_clk_driver = {
.driver = {
.name = "ppc_corenet_clock",
.of_match_table = ppc_clk_ids,
--
1.9.3
^ permalink raw reply related
* [PATCH v3 2/3] powerpc: call of_clk_init() from time_init()
From: Kevin Hao @ 2014-12-03 8:53 UTC (permalink / raw)
To: linuxppc-dev
Cc: Mike Turquette, Gerhard Sittig, Yuantian Tang, Jingchang Lu,
Scott Wood
In-Reply-To: <1417596833-31456-1-git-send-email-haokexin@gmail.com>
So the boards which has COMMON_CLK enabled don't have to
invoke this in its board specific file.
Signed-off-by: Kevin Hao <haokexin@gmail.com>
Acked-by: Scott Wood <scottwood@freescale.com>
Acked-by: Michael Turquette <mturquette@linaro.org>
---
v3: Add ack.
v2: A new patch in v2.
arch/powerpc/kernel/time.c | 5 +++++
arch/powerpc/platforms/512x/clock-commonclk.c | 11 +++++------
2 files changed, 10 insertions(+), 6 deletions(-)
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index fa7c4f12104f..df9fa05b5fd3 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -54,6 +54,7 @@
#include <linux/irq.h>
#include <linux/delay.h>
#include <linux/irq_work.h>
+#include <linux/clk-provider.h>
#include <asm/trace.h>
#include <asm/io.h>
@@ -943,6 +944,10 @@ void __init time_init(void)
init_decrementer_clockevent();
tick_setup_hrtimer_broadcast();
+
+#ifdef CONFIG_COMMON_CLK
+ of_clk_init(NULL);
+#endif
}
diff --git a/arch/powerpc/platforms/512x/clock-commonclk.c b/arch/powerpc/platforms/512x/clock-commonclk.c
index 6eb614a271fb..f691bcabd710 100644
--- a/arch/powerpc/platforms/512x/clock-commonclk.c
+++ b/arch/powerpc/platforms/512x/clock-commonclk.c
@@ -1168,6 +1168,11 @@ static void mpc5121_clk_provide_backwards_compat(void)
}
}
+/*
+ * The "fixed-clock" nodes (which includes the oscillator node if the board's
+ * DT provides one) has already been scanned by the of_clk_init() in
+ * time_init().
+ */
int __init mpc5121_clk_init(void)
{
struct device_node *clk_np;
@@ -1187,12 +1192,6 @@ int __init mpc5121_clk_init(void)
mpc512x_clk_preset_data();
/*
- * have the device tree scanned for "fixed-clock" nodes (which
- * includes the oscillator node if the board's DT provides one)
- */
- of_clk_init(NULL);
-
- /*
* add a dummy clock for those situations where a clock spec is
* required yet no real clock is involved
*/
--
1.9.3
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox