From: Suravee Suthikulpanit <Suravee.Suthikulpanit@amd.com>
To: <linux-kernel@vger.kernel.org>, <iommu@lists.linux-foundation.org>
Cc: <joro@8bytes.org>, <bp@alien8.de>, <peterz@infradead.org>,
<mingo@redhat.com>,
Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>,
Suravee Suthikulpanit <Suravee.Suthikulpanit@amd.com>
Subject: [PATCH v8 9/9] perf/amd/iommu: Enable support for multiple IOMMUs
Date: Mon, 16 Jan 2017 01:23:36 -0600 [thread overview]
Message-ID: <1484551416-5440-10-git-send-email-Suravee.Suthikulpanit@amd.com> (raw)
In-Reply-To: <1484551416-5440-1-git-send-email-Suravee.Suthikulpanit@amd.com>
From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Add multi-IOMMU support for perf by exposing an AMD IOMMU PMU
for each IOMMU found in the system via:
/bus/event_source/devices/amd_iommu_x
where x is the IOMMU index. This allows users to specify
different events to be programed onto performance counters
of each IOMMU.
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Borislav Petkov <bp@alien8.de>
Signed-off-by: Suravee Suthikulpanit <Suravee.Suthikulpanit@amd.com>
---
arch/x86/events/amd/iommu.c | 114 ++++++++++++++++++++++++++------------------
1 file changed, 67 insertions(+), 47 deletions(-)
diff --git a/arch/x86/events/amd/iommu.c b/arch/x86/events/amd/iommu.c
index 223c01d..38eafbf 100644
--- a/arch/x86/events/amd/iommu.c
+++ b/arch/x86/events/amd/iommu.c
@@ -35,16 +35,21 @@
#define _GET_PASID_MASK(ev) ((ev->hw.extra_reg.config >> 16) & 0xFFFFULL)
#define _GET_DOMID_MASK(ev) ((ev->hw.extra_reg.config >> 32) & 0xFFFFULL)
-static struct perf_amd_iommu __perf_iommu;
+#define PERF_AMD_IOMMU_NAME_SZ 16
struct perf_amd_iommu {
+ struct list_head list;
struct pmu pmu;
+ unsigned int idx;
+ char name[PERF_AMD_IOMMU_NAME_SZ];
u8 max_banks;
u8 max_counters;
u64 cntr_assign_mask;
raw_spinlock_t lock;
};
+static LIST_HEAD(perf_amd_iommu_list);
+
/*---------------------------------------------
* sysfs format attributes
*---------------------------------------------*/
@@ -202,8 +207,7 @@ static int clear_avail_iommu_bnk_cntr(struct perf_amd_iommu *perf_iommu,
static int perf_iommu_event_init(struct perf_event *event)
{
struct hw_perf_event *hwc = &event->hw;
- struct perf_amd_iommu *perf_iommu;
- u64 config, config1;
+ struct perf_amd_iommu *pi;
/* test the event attr type check for PMU enumeration */
if (event->attr.type != event->pmu->type)
@@ -225,27 +229,18 @@ static int perf_iommu_event_init(struct perf_event *event)
if (event->cpu < 0)
return -EINVAL;
- perf_iommu = &__perf_iommu;
-
- if (event->pmu != &perf_iommu->pmu)
- return -ENOENT;
-
- if (perf_iommu) {
- config = event->attr.config;
- config1 = event->attr.config1;
- } else {
- return -EINVAL;
- }
-
/* update the hw_perf_event struct with the iommu config data */
- hwc->config = config;
- hwc->extra_reg.config = config1;
+ pi = container_of(event->pmu, struct perf_amd_iommu, pmu);
+ hwc->idx = pi->idx;
+ hwc->config = event->attr.config;
+ hwc->extra_reg.config = event->attr.config1;
return 0;
}
static void perf_iommu_enable_event(struct perf_event *ev)
{
+ struct hw_perf_event *hwc = &ev->hw;
u8 csource = _GET_CSOURCE(ev);
u16 devid = _GET_DEVID(ev);
u8 bank = _GET_BANK(ev);
@@ -253,30 +248,34 @@ static void perf_iommu_enable_event(struct perf_event *ev)
u64 reg = 0ULL;
reg = csource;
- amd_iommu_pc_set_reg(0, bank, cntr,
+ amd_iommu_pc_set_reg(hwc->idx, bank, cntr,
IOMMU_PC_COUNTER_SRC_REG, ®);
reg = devid | (_GET_DEVID_MASK(ev) << 32);
if (reg)
reg |= BIT(31);
- amd_iommu_pc_set_reg(0, bank, cntr, IOMMU_PC_DEVID_MATCH_REG, ®);
+ amd_iommu_pc_set_reg(hwc->idx, bank, cntr,
+ IOMMU_PC_DEVID_MATCH_REG, ®);
reg = _GET_PASID(ev) | (_GET_PASID_MASK(ev) << 32);
if (reg)
reg |= BIT(31);
- amd_iommu_pc_set_reg(0, bank, cntr, IOMMU_PC_PASID_MATCH_REG, ®);
+ amd_iommu_pc_set_reg(hwc->idx, bank, cntr,
+ IOMMU_PC_PASID_MATCH_REG, ®);
reg = _GET_DOMID(ev) | (_GET_DOMID_MASK(ev) << 32);
if (reg)
reg |= BIT(31);
- amd_iommu_pc_set_reg(0, bank, cntr, IOMMU_PC_DOMID_MATCH_REG, ®);
+ amd_iommu_pc_set_reg(hwc->idx, bank, cntr,
+ IOMMU_PC_DOMID_MATCH_REG, ®);
}
static void perf_iommu_disable_event(struct perf_event *event)
{
+ struct hw_perf_event *hwc = &event->hw;
u64 reg = 0ULL;
- amd_iommu_pc_set_reg(0, _GET_BANK(event), _GET_CNTR(event),
+ amd_iommu_pc_set_reg(hwc->idx, _GET_BANK(event), _GET_CNTR(event),
IOMMU_PC_COUNTER_SRC_REG, ®);
}
@@ -295,7 +294,7 @@ static void perf_iommu_start(struct perf_event *event, int flags)
return;
val = local64_read(&hwc->prev_count) & GENMASK_ULL(48, 0);
- if (amd_iommu_pc_set_reg(0, _GET_BANK(event), _GET_CNTR(event),
+ if (amd_iommu_pc_set_reg(hwc->idx, _GET_BANK(event), _GET_CNTR(event),
IOMMU_PC_COUNTER_REG, &val))
return;
@@ -309,7 +308,7 @@ static void perf_iommu_read(struct perf_event *event)
s64 delta;
struct hw_perf_event *hwc = &event->hw;
- if (amd_iommu_pc_get_reg(0, _GET_BANK(event), _GET_CNTR(event),
+ if (amd_iommu_pc_get_reg(hwc->idx, _GET_BANK(event), _GET_CNTR(event),
IOMMU_PC_COUNTER_REG, &count))
return;
@@ -407,6 +406,13 @@ static __init int _init_events_attrs(void)
static __init void amd_iommu_pc_exit(void)
{
+ struct perf_amd_iommu *pi, *next;
+
+ list_for_each_entry_safe(pi, next, &perf_amd_iommu_list, list) {
+ list_del(&pi->list);
+ kfree(pi);
+ }
+
if (amd_iommu_events_group.attrs) {
kfree(amd_iommu_events_group.attrs);
amd_iommu_events_group.attrs = NULL;
@@ -421,46 +427,46 @@ static __init void amd_iommu_pc_exit(void)
};
static __init int
-_init_perf_amd_iommu(struct perf_amd_iommu *perf_iommu, char *name)
+init_one_perf_amd_iommu(struct perf_amd_iommu *perf_iommu, unsigned int idx)
{
int ret;
raw_spin_lock_init(&perf_iommu->lock);
- /* Init cpumask attributes to only core 0 */
- cpumask_set_cpu(0, &iommu_cpumask);
-
- perf_iommu->max_banks = amd_iommu_pc_get_max_banks(0);
- perf_iommu->max_counters = amd_iommu_pc_get_max_counters(0);
+ perf_iommu->idx = idx;
+ perf_iommu->max_banks = amd_iommu_pc_get_max_banks(idx);
+ perf_iommu->max_counters = amd_iommu_pc_get_max_counters(idx);
if (!perf_iommu->max_banks || !perf_iommu->max_counters)
return -EINVAL;
+ snprintf(perf_iommu->name, PERF_AMD_IOMMU_NAME_SZ, "amd_iommu_%u", idx);
+
+ perf_iommu->pmu.event_init = perf_iommu_event_init,
+ perf_iommu->pmu.add = perf_iommu_add,
+ perf_iommu->pmu.del = perf_iommu_del,
+ perf_iommu->pmu.start = perf_iommu_start,
+ perf_iommu->pmu.stop = perf_iommu_stop,
+ perf_iommu->pmu.read = perf_iommu_read,
+ perf_iommu->pmu.task_ctx_nr = perf_invalid_context;
perf_iommu->pmu.attr_groups = amd_iommu_attr_groups;
- ret = perf_pmu_register(&perf_iommu->pmu, name, -1);
+
+ ret = perf_pmu_register(&perf_iommu->pmu, perf_iommu->name, -1);
if (ret)
pr_err("Error initializing AMD IOMMU perf counters.\n");
else
- pr_info("perf: amd_iommu: Detected. (%d banks, %d counters/bank)\n",
- amd_iommu_pc_get_max_banks(0),
- amd_iommu_pc_get_max_counters(0));
+ pr_info("Detected AMD IOMMU #%d (%d banks, %d counters/bank)\n",
+ idx, amd_iommu_pc_get_max_banks(idx),
+ amd_iommu_pc_get_max_counters(idx));
return ret;
}
-static struct perf_amd_iommu __perf_iommu = {
- .pmu = {
- .task_ctx_nr = perf_invalid_context,
- .event_init = perf_iommu_event_init,
- .add = perf_iommu_add,
- .del = perf_iommu_del,
- .start = perf_iommu_start,
- .stop = perf_iommu_stop,
- .read = perf_iommu_read,
- },
-};
-
static __init int amd_iommu_pc_init(void)
{
int ret;
+ unsigned int i;
+
+ /* Init cpumask attributes to only core 0 */
+ cpumask_set_cpu(0, &iommu_cpumask);
/* Make sure the IOMMU PC resource is available */
if (!amd_iommu_pc_supported())
@@ -470,7 +476,21 @@ static __init int amd_iommu_pc_init(void)
if (ret)
goto err_out;
- ret = _init_perf_amd_iommu(&__perf_iommu, "amd_iommu");
+ for (i = 0 ; i < amd_iommu_get_num_iommus(); i++) {
+ struct perf_amd_iommu *pi;
+
+ pi = kzalloc(sizeof(struct perf_amd_iommu), GFP_KERNEL);
+ if (!pi) {
+ ret = -ENOMEM;
+ break;
+ }
+
+ list_add_tail(&pi->list, &perf_amd_iommu_list);
+ ret = init_one_perf_amd_iommu(pi, i);
+ if (ret)
+ break;
+ }
+
if (ret)
goto err_out;
--
1.8.3.1
next prev parent reply other threads:[~2017-01-16 7:25 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-16 7:23 [PATCH v8 0/9] perf/amd/iommu: Enable multi-IOMMU support Suravee Suthikulpanit
2017-01-16 7:23 ` [PATCH v8 1/9] perf/amd/iommu: Declare pr_fmt and remove unnecessary pr_debug Suravee Suthikulpanit
2017-01-16 7:23 ` [PATCH v8 2/9] perf/amd/iommu: Clean up perf_iommu_enable_event Suravee Suthikulpanit
2017-01-18 18:20 ` Borislav Petkov
2017-01-16 7:23 ` [PATCH v8 3/9] perf/amd/iommu: Misc fix up perf_iommu_read Suravee Suthikulpanit
2017-01-19 10:01 ` Borislav Petkov
2017-01-23 12:33 ` Peter Zijlstra
2017-02-07 4:50 ` Suravee Suthikulpanit
2017-01-16 7:23 ` [PATCH v8 4/9] iommu/amd: Introduce amd_iommu_get_num_iommus() Suravee Suthikulpanit
2017-01-19 18:41 ` Borislav Petkov
2017-01-16 7:23 ` [PATCH v8 5/9] perf/amd/iommu: Modify functions to query max banks and counters Suravee Suthikulpanit
2017-01-22 19:53 ` Borislav Petkov
2017-01-16 7:23 ` [PATCH v8 6/9] perf/amd/iommu: Modify amd_iommu_pc_get_set_reg_val() API to allow specifying IOMMU index Suravee Suthikulpanit
2017-01-22 19:53 ` Borislav Petkov
2017-01-16 7:23 ` [PATCH v8 7/9] perf/amd/iommu: Check return value when set and get counter value Suravee Suthikulpanit
2017-01-22 19:53 ` Borislav Petkov
2017-01-23 12:31 ` Peter Zijlstra
2017-01-16 7:23 ` [PATCH v8 8/9] perf/amd/iommu: Fix sysfs perf attribute groups Suravee Suthikulpanit
2017-01-22 19:54 ` Borislav Petkov
2017-01-16 7:23 ` Suravee Suthikulpanit [this message]
2017-01-22 19:55 ` [PATCH v8 9/9] perf/amd/iommu: Enable support for multiple IOMMUs Borislav Petkov
2017-02-07 1:42 ` Suravee Suthikulpanit
2017-01-25 9:46 ` Peter Zijlstra
2017-01-25 9:55 ` Borislav Petkov
2017-02-07 1:58 ` Suravee Suthikulpanit
2017-02-07 1:57 ` Suravee Suthikulpanit
2017-02-14 12:31 ` Peter Zijlstra
2017-02-23 17:43 ` Suravee Suthikulpanit
2017-02-23 18:11 ` Peter Zijlstra
2017-02-23 18:20 ` Suravee Suthikulpanit
2017-01-17 15:36 ` [PATCH v8 0/9] perf/amd/iommu: Enable multi-IOMMU support Joerg Roedel
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1484551416-5440-10-git-send-email-Suravee.Suthikulpanit@amd.com \
--to=suravee.suthikulpanit@amd.com \
--cc=bp@alien8.de \
--cc=iommu@lists.linux-foundation.org \
--cc=joro@8bytes.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).