[PATCH v7 08/16] perf: RISC-V: Introduce Andes PMU for perf event sampling
Yu Chien Peter Lin
peterlin at andestech.com
Tue Jan 9 23:39:09 PST 2024
The Andes PMU extension provides the same mechanism as Sscofpmf,
allowing us to reuse the SBI PMU driver to support event sampling
and mode filtering.
To make use of this custom PMU extension, "xandespmu" needs
to be appended to the riscv,isa-extensions for each cpu node
in device-tree, and enable CONFIG_ANDES_CUSTOM_PMU.
Signed-off-by: Yu Chien Peter Lin <peterlin at andestech.com>
Reviewed-by: Charles Ci-Jyun Wu <dminus at andestech.com>
Reviewed-by: Leo Yu-Chi Liang <ycliang at andestech.com>
Co-developed-by: Locus Wei-Han Chen <locus84 at andestech.com>
Signed-off-by: Locus Wei-Han Chen <locus84 at andestech.com>
Reviewed-by: Lad Prabhakar <prabhakar.mahadev-lad.rj at bp.renesas.com>
Tested-by: Lad Prabhakar <prabhakar.mahadev-lad.rj at bp.renesas.com>
---
Changes v1 -> v2:
- New patch
Changes v2 -> v3:
- Reordered list in riscv_isa_ext[]
- Removed mvendorid check in pmu_sbi_setup_irqs()
Changes v3 -> v4:
- No change
Changes v4 -> v5:
- Let ANDES_CUSTOM_PMU depend on ARCH_RENESAS
- Include Prabhakar's Reviewed/Tested-by
Changes v5 -> v6:
- No change
Changes v6 -> v7:
- No change
---
arch/riscv/include/asm/hwcap.h | 1 +
arch/riscv/kernel/cpufeature.c | 1 +
drivers/perf/Kconfig | 14 ++++++++++++++
drivers/perf/riscv_pmu_sbi.c | 30 +++++++++++++++++++++++++-----
4 files changed, 41 insertions(+), 5 deletions(-)
diff --git a/arch/riscv/include/asm/hwcap.h b/arch/riscv/include/asm/hwcap.h
index 480f9da7fba7..c77da1c9f32b 100644
--- a/arch/riscv/include/asm/hwcap.h
+++ b/arch/riscv/include/asm/hwcap.h
@@ -81,6 +81,7 @@
#define RISCV_ISA_EXT_ZTSO 72
#define RISCV_ISA_EXT_ZACAS 73
#define RISCV_ISA_EXT_XTHEADPMU 74
+#define RISCV_ISA_EXT_XANDESPMU 75
#define RISCV_ISA_EXT_MAX 128
#define RISCV_ISA_EXT_INVALID U32_MAX
diff --git a/arch/riscv/kernel/cpufeature.c b/arch/riscv/kernel/cpufeature.c
index 4aded5bf8fc3..bdf8d2be3ea6 100644
--- a/arch/riscv/kernel/cpufeature.c
+++ b/arch/riscv/kernel/cpufeature.c
@@ -303,6 +303,7 @@ const struct riscv_isa_ext_data riscv_isa_ext[] = {
__RISCV_ISA_EXT_DATA(svinval, RISCV_ISA_EXT_SVINVAL),
__RISCV_ISA_EXT_DATA(svnapot, RISCV_ISA_EXT_SVNAPOT),
__RISCV_ISA_EXT_DATA(svpbmt, RISCV_ISA_EXT_SVPBMT),
+ __RISCV_ISA_EXT_DATA(xandespmu, RISCV_ISA_EXT_XANDESPMU),
__RISCV_ISA_EXT_DATA(xtheadpmu, RISCV_ISA_EXT_XTHEADPMU),
};
diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig
index 6cef15ec7c25..18f4c62eafc2 100644
--- a/drivers/perf/Kconfig
+++ b/drivers/perf/Kconfig
@@ -86,6 +86,20 @@ config RISCV_PMU_SBI
full perf feature support i.e. counter overflow, privilege mode
filtering, counter configuration.
+config ANDES_CUSTOM_PMU
+ bool "Andes custom PMU support"
+ depends on ARCH_RENESAS && RISCV_ALTERNATIVE && RISCV_PMU_SBI
+ default y
+ help
+ The Andes cores implement a PMU overflow extension very
+ similar to the core SSCOFPMF extension.
+
+ This will patch the overflow/pending CSR and handle the
+ non-standard behaviour via the regular SBI PMU driver and
+ interface.
+
+ If you don't know what to do here, say "Y".
+
config THEAD_CUSTOM_PMU
bool "T-Head custom PMU support"
depends on ARCH_THEAD && RISCV_ALTERNATIVE && RISCV_PMU_SBI
diff --git a/drivers/perf/riscv_pmu_sbi.c b/drivers/perf/riscv_pmu_sbi.c
index 31ca79846399..4d5d478430c8 100644
--- a/drivers/perf/riscv_pmu_sbi.c
+++ b/drivers/perf/riscv_pmu_sbi.c
@@ -19,6 +19,7 @@
#include <linux/of.h>
#include <linux/cpu_pm.h>
#include <linux/sched/clock.h>
+#include <linux/soc/andes/irq.h>
#include <asm/sbi.h>
#include <asm/cpufeature.h>
@@ -27,14 +28,26 @@
#define THEAD_C9XX_CSR_SCOUNTEROF 0x5c5
#define ALT_SBI_PMU_OVERFLOW(__ovl) \
-asm volatile(ALTERNATIVE( \
+asm volatile(ALTERNATIVE_2( \
"csrr %0, " __stringify(CSR_SSCOUNTOVF), \
"csrr %0, " __stringify(THEAD_C9XX_CSR_SCOUNTEROF), \
0, RISCV_ISA_EXT_XTHEADPMU, \
- CONFIG_THEAD_CUSTOM_PMU) \
+ CONFIG_THEAD_CUSTOM_PMU, \
+ "csrr %0, " __stringify(ANDES_CSR_SCOUNTEROF), \
+ 0, RISCV_ISA_EXT_XANDESPMU, \
+ CONFIG_ANDES_CUSTOM_PMU) \
: "=r" (__ovl) : \
: "memory")
+#define ALT_SBI_PMU_OVF_CLEAR_PENDING(__irq_mask) \
+asm volatile(ALTERNATIVE( \
+ "csrc " __stringify(CSR_IP) ", %0\n\t", \
+ "csrc " __stringify(ANDES_CSR_SLIP) ", %0\n\t", \
+ 0, RISCV_ISA_EXT_XANDESPMU, \
+ CONFIG_ANDES_CUSTOM_PMU) \
+ : : "r"(__irq_mask) \
+ : "memory")
+
#define SYSCTL_NO_USER_ACCESS 0
#define SYSCTL_USER_ACCESS 1
#define SYSCTL_LEGACY 2
@@ -72,6 +85,7 @@ static int sysctl_perf_user_access __read_mostly = SYSCTL_USER_ACCESS;
static union sbi_pmu_ctr_info *pmu_ctr_list;
static bool riscv_pmu_use_irq;
static unsigned int riscv_pmu_irq_num;
+static unsigned int riscv_pmu_irq_mask;
static unsigned int riscv_pmu_irq;
/* Cache the available counters in a bitmask */
@@ -705,7 +719,7 @@ static irqreturn_t pmu_sbi_ovf_handler(int irq, void *dev)
event = cpu_hw_evt->events[fidx];
if (!event) {
- csr_clear(CSR_SIP, BIT(riscv_pmu_irq_num));
+ ALT_SBI_PMU_OVF_CLEAR_PENDING(riscv_pmu_irq_mask);
return IRQ_NONE;
}
@@ -719,7 +733,7 @@ static irqreturn_t pmu_sbi_ovf_handler(int irq, void *dev)
* Overflow interrupt pending bit should only be cleared after stopping
* all the counters to avoid any race condition.
*/
- csr_clear(CSR_SIP, BIT(riscv_pmu_irq_num));
+ ALT_SBI_PMU_OVF_CLEAR_PENDING(riscv_pmu_irq_mask);
/* No overflow bit is set */
if (!overflow)
@@ -791,7 +805,7 @@ static int pmu_sbi_starting_cpu(unsigned int cpu, struct hlist_node *node)
if (riscv_pmu_use_irq) {
cpu_hw_evt->irq = riscv_pmu_irq;
- csr_clear(CSR_IP, BIT(riscv_pmu_irq_num));
+ ALT_SBI_PMU_OVF_CLEAR_PENDING(riscv_pmu_irq_mask);
enable_percpu_irq(riscv_pmu_irq, IRQ_TYPE_NONE);
}
@@ -823,8 +837,14 @@ static int pmu_sbi_setup_irqs(struct riscv_pmu *pmu, struct platform_device *pde
IS_ENABLED(CONFIG_THEAD_CUSTOM_PMU)) {
riscv_pmu_irq_num = THEAD_C9XX_RV_IRQ_PMU;
riscv_pmu_use_irq = true;
+ } else if (riscv_isa_extension_available(NULL, XANDESPMU) &&
+ IS_ENABLED(CONFIG_ANDES_CUSTOM_PMU)) {
+ riscv_pmu_irq_num = ANDES_SLI_CAUSE_BASE + ANDES_RV_IRQ_PMOVI;
+ riscv_pmu_use_irq = true;
}
+ riscv_pmu_irq_mask = BIT(riscv_pmu_irq_num % BITS_PER_LONG);
+
if (!riscv_pmu_use_irq)
return -EOPNOTSUPP;
--
2.34.1
More information about the linux-riscv
mailing list