The PMU on T-Head C9xx cores is quite similar to the SSCOFPMF extension
but not completely identical, so this series
changes in v5:
- add received Reviews
- fix sbi caching wrt. negative values (Drew)
- add comment about specific c9xx arch- and imp-ids (Conor)
changes in v4:
- add new patch to cache sbi mvendor, march and mimp-ids (Atish)
- errata dependencies in one line (Conor)
- make driver detection conditional on CONFIG_ERRATA_THEAD_PMU too (Atish)
changes in v3:
- improve commit message (Atish, Conor)
- IS_ENABLED and BIT() in errata probe (Conor)
The change depends on my cpufeature/t-head errata probe cleanup series [1].
changes in v2:
- use alternatives for the CSR access
- make the irq num selection a bit nicer
There is of course a matching opensbi-part whose most recent implementation
can be found on [0].
[0] https://patchwork.ozlabs.org/project/opensbi/cover/[email protected]
[1] https://lore.kernel.org/all/[email protected]/
Heiko Stuebner (2):
RISC-V: Cache SBI vendor values
drivers/perf: riscv_pmu_sbi: add support for PMU variant on T-Head
C9xx cores
arch/riscv/Kconfig.erratas | 13 +++++++++++
arch/riscv/errata/thead/errata.c | 19 ++++++++++++++++
arch/riscv/include/asm/errata_list.h | 16 +++++++++++++-
arch/riscv/kernel/sbi.c | 30 ++++++++++++++++++++++---
drivers/perf/riscv_pmu_sbi.c | 33 +++++++++++++++++++---------
5 files changed, 97 insertions(+), 14 deletions(-)
--
2.35.1
sbi_get_mvendorid(), sbi_get_marchid() and sbi_get_mimpid() might get
called multiple times, though the values of these CSRs should not change
during the runtime of a specific machine.
So cache the values in the functions and prevent multiple ecalls
to read these values.
As Andrew Jones noted, at least marchid and mimpid may be negative
values when viewed as a long, so we use a separate static bool to
indiciate the cached status.
Suggested-by: Atish Patra <[email protected]>
Signed-off-by: Heiko Stuebner <[email protected]>
---
arch/riscv/kernel/sbi.c | 30 +++++++++++++++++++++++++++---
1 file changed, 27 insertions(+), 3 deletions(-)
diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c
index 775d3322b422..cc618aaa9d11 100644
--- a/arch/riscv/kernel/sbi.c
+++ b/arch/riscv/kernel/sbi.c
@@ -625,17 +625,41 @@ static inline long sbi_get_firmware_version(void)
long sbi_get_mvendorid(void)
{
- return __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
+ static long id;
+ static bool cached;
+
+ if (!cached) {
+ id = __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
+ cached = true;
+ }
+
+ return id;
}
long sbi_get_marchid(void)
{
- return __sbi_base_ecall(SBI_EXT_BASE_GET_MARCHID);
+ static long id;
+ static bool cached;
+
+ if (!cached) {
+ id = __sbi_base_ecall(SBI_EXT_BASE_GET_MARCHID);
+ cached = true;
+ }
+
+ return id;
}
long sbi_get_mimpid(void)
{
- return __sbi_base_ecall(SBI_EXT_BASE_GET_MIMPID);
+ static long id;
+ static bool cached;
+
+ if (!cached) {
+ id = __sbi_base_ecall(SBI_EXT_BASE_GET_MIMPID);
+ cached = true;
+ }
+
+ return id;
}
static void sbi_send_cpumask_ipi(const struct cpumask *target)
--
2.35.1
With the T-HEAD C9XX cores being designed before or during the ratification
to the SSCOFPMF extension, it implements functionality very similar but
not equal to it.
It implements overflow handling and also some privilege-mode filtering.
While SSCOFPMF supports this for all modes, the C9XX only implements the
filtering for M-mode and S-mode but not user-mode.
So add some adaptions to allow the C9XX to still handle
its PMU through the regular SBI PMU interface instead of defining new
interfaces or drivers.
To work properly, this requires a matching change in SBI, though the actual
interface between kernel and SBI does not change.
The main differences are a the overflow CSR and irq number.
As the reading of the overflow-csr is in the hot-path during irq handling,
use an errata and alternatives to not introduce new conditionals there.
Reviewed-by: Andrew Jones <[email protected]>
Reviewed-by: Conor Dooley <[email protected]>
Signed-off-by: Heiko Stuebner <[email protected]>
---
arch/riscv/Kconfig.erratas | 13 +++++++++++
arch/riscv/errata/thead/errata.c | 19 ++++++++++++++++
arch/riscv/include/asm/errata_list.h | 16 +++++++++++++-
drivers/perf/riscv_pmu_sbi.c | 33 +++++++++++++++++++---------
4 files changed, 70 insertions(+), 11 deletions(-)
diff --git a/arch/riscv/Kconfig.erratas b/arch/riscv/Kconfig.erratas
index f3623df23b5f..69621ae6d647 100644
--- a/arch/riscv/Kconfig.erratas
+++ b/arch/riscv/Kconfig.erratas
@@ -66,4 +66,17 @@ config ERRATA_THEAD_CMO
If you don't know what to do here, say "Y".
+config ERRATA_THEAD_PMU
+ bool "Apply T-Head PMU errata"
+ depends on ERRATA_THEAD && RISCV_PMU_SBI
+ default y
+ help
+ The T-Head C9xx cores implement a PMU overflow extension very
+ similar to the core SSCOFPMF extension.
+
+ This will apply the overflow errata to handle the non-standard
+ behaviour via the regular SBI PMU driver and interface.
+
+ If you don't know what to do here, say "Y".
+
endmenu # "CPU errata selection"
diff --git a/arch/riscv/errata/thead/errata.c b/arch/riscv/errata/thead/errata.c
index 21546937db39..fac5742d1c1e 100644
--- a/arch/riscv/errata/thead/errata.c
+++ b/arch/riscv/errata/thead/errata.c
@@ -47,6 +47,22 @@ static bool errata_probe_cmo(unsigned int stage,
return true;
}
+static bool errata_probe_pmu(unsigned int stage,
+ unsigned long arch_id, unsigned long impid)
+{
+ if (!IS_ENABLED(CONFIG_ERRATA_THEAD_PMU))
+ return false;
+
+ /* target-c9xx cores report arch_id and impid as 0 */
+ if (arch_id != 0 || impid != 0)
+ return false;
+
+ if (stage == RISCV_ALTERNATIVES_EARLY_BOOT)
+ return false;
+
+ return true;
+}
+
static u32 thead_errata_probe(unsigned int stage,
unsigned long archid, unsigned long impid)
{
@@ -58,6 +74,9 @@ static u32 thead_errata_probe(unsigned int stage,
if (errata_probe_cmo(stage, archid, impid))
cpu_req_errata |= BIT(ERRATA_THEAD_CMO);
+ if (errata_probe_pmu(stage, archid, impid))
+ cpu_req_errata |= BIT(ERRATA_THEAD_PMU);
+
return cpu_req_errata;
}
diff --git a/arch/riscv/include/asm/errata_list.h b/arch/riscv/include/asm/errata_list.h
index 19a771085781..4180312d2a70 100644
--- a/arch/riscv/include/asm/errata_list.h
+++ b/arch/riscv/include/asm/errata_list.h
@@ -6,6 +6,7 @@
#define ASM_ERRATA_LIST_H
#include <asm/alternative.h>
+#include <asm/csr.h>
#include <asm/vendorid_list.h>
#ifdef CONFIG_ERRATA_SIFIVE
@@ -17,7 +18,8 @@
#ifdef CONFIG_ERRATA_THEAD
#define ERRATA_THEAD_PBMT 0
#define ERRATA_THEAD_CMO 1
-#define ERRATA_THEAD_NUMBER 2
+#define ERRATA_THEAD_PMU 2
+#define ERRATA_THEAD_NUMBER 3
#endif
#define CPUFEATURE_SVPBMT 0
@@ -142,6 +144,18 @@ asm volatile(ALTERNATIVE_2( \
"r"((unsigned long)(_start) + (_size)) \
: "a0")
+#define THEAD_C9XX_RV_IRQ_PMU 17
+#define THEAD_C9XX_CSR_SCOUNTEROF 0x5c5
+
+#define ALT_SBI_PMU_OVERFLOW(__ovl) \
+asm volatile(ALTERNATIVE( \
+ "csrr %0, " __stringify(CSR_SSCOUNTOVF), \
+ "csrr %0, " __stringify(THEAD_C9XX_CSR_SCOUNTEROF), \
+ THEAD_VENDOR_ID, ERRATA_THEAD_PMU, \
+ CONFIG_ERRATA_THEAD_PMU) \
+ : "=r" (__ovl) : \
+ : "memory")
+
#endif /* __ASSEMBLY__ */
#endif
diff --git a/drivers/perf/riscv_pmu_sbi.c b/drivers/perf/riscv_pmu_sbi.c
index 8de4ca2fef21..ec0972c7c562 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 <asm/errata_list.h>
#include <asm/sbi.h>
#include <asm/hwcap.h>
@@ -46,6 +47,8 @@ static const struct attribute_group *riscv_pmu_attr_groups[] = {
* per_cpu in case of harts with different pmu counters
*/
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;
struct sbi_pmu_event_data {
@@ -575,7 +578,7 @@ static irqreturn_t pmu_sbi_ovf_handler(int irq, void *dev)
fidx = find_first_bit(cpu_hw_evt->used_hw_ctrs, RISCV_MAX_COUNTERS);
event = cpu_hw_evt->events[fidx];
if (!event) {
- csr_clear(CSR_SIP, SIP_LCOFIP);
+ csr_clear(CSR_SIP, BIT(riscv_pmu_irq_num));
return IRQ_NONE;
}
@@ -583,13 +586,13 @@ static irqreturn_t pmu_sbi_ovf_handler(int irq, void *dev)
pmu_sbi_stop_hw_ctrs(pmu);
/* Overflow status register should only be read after counter are stopped */
- overflow = csr_read(CSR_SSCOUNTOVF);
+ ALT_SBI_PMU_OVERFLOW(overflow);
/*
* Overflow interrupt pending bit should only be cleared after stopping
* all the counters to avoid any race condition.
*/
- csr_clear(CSR_SIP, SIP_LCOFIP);
+ csr_clear(CSR_SIP, BIT(riscv_pmu_irq_num));
/* No overflow bit is set */
if (!overflow)
@@ -651,10 +654,10 @@ static int pmu_sbi_starting_cpu(unsigned int cpu, struct hlist_node *node)
/* Stop all the counters so that they can be enabled from perf */
pmu_sbi_stop_all(pmu);
- if (riscv_isa_extension_available(NULL, SSCOFPMF)) {
+ if (riscv_pmu_use_irq) {
cpu_hw_evt->irq = riscv_pmu_irq;
- csr_clear(CSR_IP, BIT(RV_IRQ_PMU));
- csr_set(CSR_IE, BIT(RV_IRQ_PMU));
+ csr_clear(CSR_IP, BIT(riscv_pmu_irq_num));
+ csr_set(CSR_IE, BIT(riscv_pmu_irq_num));
enable_percpu_irq(riscv_pmu_irq, IRQ_TYPE_NONE);
}
@@ -663,9 +666,9 @@ static int pmu_sbi_starting_cpu(unsigned int cpu, struct hlist_node *node)
static int pmu_sbi_dying_cpu(unsigned int cpu, struct hlist_node *node)
{
- if (riscv_isa_extension_available(NULL, SSCOFPMF)) {
+ if (riscv_pmu_use_irq) {
disable_percpu_irq(riscv_pmu_irq);
- csr_clear(CSR_IE, BIT(RV_IRQ_PMU));
+ csr_clear(CSR_IE, BIT(riscv_pmu_irq_num));
}
/* Disable all counters access for user mode now */
@@ -681,7 +684,17 @@ static int pmu_sbi_setup_irqs(struct riscv_pmu *pmu, struct platform_device *pde
struct device_node *cpu, *child;
struct irq_domain *domain = NULL;
- if (!riscv_isa_extension_available(NULL, SSCOFPMF))
+ if (riscv_isa_extension_available(NULL, SSCOFPMF)) {
+ riscv_pmu_irq_num = RV_IRQ_PMU;
+ riscv_pmu_use_irq = true;
+ } else if (IS_ENABLED(CONFIG_ERRATA_THEAD_PMU) &&
+ sbi_get_mvendorid() == THEAD_VENDOR_ID &&
+ sbi_get_marchid() == 0 && sbi_get_mimpid() == 0) {
+ riscv_pmu_irq_num = THEAD_C9XX_RV_IRQ_PMU;
+ riscv_pmu_use_irq = true;
+ }
+
+ if (!riscv_pmu_use_irq)
return -EOPNOTSUPP;
for_each_of_cpu_node(cpu) {
@@ -703,7 +716,7 @@ static int pmu_sbi_setup_irqs(struct riscv_pmu *pmu, struct platform_device *pde
return -ENODEV;
}
- riscv_pmu_irq = irq_create_mapping(domain, RV_IRQ_PMU);
+ riscv_pmu_irq = irq_create_mapping(domain, riscv_pmu_irq_num);
if (!riscv_pmu_irq) {
pr_err("Failed to map PMU interrupt for node\n");
return -ENODEV;
--
2.35.1
Am Montag, 10. Oktober 2022, 14:45:45 CEST schrieb Anup Patel:
> On Mon, Oct 10, 2022 at 5:57 PM Heiko Stuebner <[email protected]> wrote:
> >
> > sbi_get_mvendorid(), sbi_get_marchid() and sbi_get_mimpid() might get
> > called multiple times, though the values of these CSRs should not change
> > during the runtime of a specific machine.
> >
> > So cache the values in the functions and prevent multiple ecalls
> > to read these values.
> >
> > As Andrew Jones noted, at least marchid and mimpid may be negative
> > values when viewed as a long, so we use a separate static bool to
> > indiciate the cached status.
> >
> > Suggested-by: Atish Patra <[email protected]>
> > Signed-off-by: Heiko Stuebner <[email protected]>
> > ---
> > arch/riscv/kernel/sbi.c | 30 +++++++++++++++++++++++++++---
> > 1 file changed, 27 insertions(+), 3 deletions(-)
> >
> > diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c
> > index 775d3322b422..cc618aaa9d11 100644
> > --- a/arch/riscv/kernel/sbi.c
> > +++ b/arch/riscv/kernel/sbi.c
> > @@ -625,17 +625,41 @@ static inline long sbi_get_firmware_version(void)
> >
> > long sbi_get_mvendorid(void)
> > {
> > - return __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> > + static long id;
> > + static bool cached;
> > +
> > + if (!cached) {
> > + id = __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> > + cached = true;
> > + }
> > +
> > + return id;
> > }
> >
> > long sbi_get_marchid(void)
> > {
> > - return __sbi_base_ecall(SBI_EXT_BASE_GET_MARCHID);
> > + static long id;
> > + static bool cached;
>
> This breaks for heterogeneous SMP systems (similar to big.LITTLE)
> where HARTs will have different marchid even though they belong to
> same CPU Vendor.
>
> Due to the above rationale, the patch adding marchid, mvendorid, and
> mimpid in /proc/cpuinfo caches these values on a per-CPU basis.
For people reading along, I think you mean
https://lore.kernel.org/r/[email protected]
For my understanding, was there a reason in the past for doing the caching
only for cpuinfo and not for every invocation of the ecalls?
Thanks
Heiko
On Mon, Oct 10, 2022 at 6:25 PM Heiko Stuebner <[email protected]> wrote:
>
> Am Montag, 10. Oktober 2022, 14:45:45 CEST schrieb Anup Patel:
> > On Mon, Oct 10, 2022 at 5:57 PM Heiko Stuebner <[email protected]> wrote:
> > >
> > > sbi_get_mvendorid(), sbi_get_marchid() and sbi_get_mimpid() might get
> > > called multiple times, though the values of these CSRs should not change
> > > during the runtime of a specific machine.
> > >
> > > So cache the values in the functions and prevent multiple ecalls
> > > to read these values.
> > >
> > > As Andrew Jones noted, at least marchid and mimpid may be negative
> > > values when viewed as a long, so we use a separate static bool to
> > > indiciate the cached status.
> > >
> > > Suggested-by: Atish Patra <[email protected]>
> > > Signed-off-by: Heiko Stuebner <[email protected]>
> > > ---
> > > arch/riscv/kernel/sbi.c | 30 +++++++++++++++++++++++++++---
> > > 1 file changed, 27 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c
> > > index 775d3322b422..cc618aaa9d11 100644
> > > --- a/arch/riscv/kernel/sbi.c
> > > +++ b/arch/riscv/kernel/sbi.c
> > > @@ -625,17 +625,41 @@ static inline long sbi_get_firmware_version(void)
> > >
> > > long sbi_get_mvendorid(void)
> > > {
> > > - return __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> > > + static long id;
> > > + static bool cached;
> > > +
> > > + if (!cached) {
> > > + id = __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> > > + cached = true;
> > > + }
> > > +
> > > + return id;
> > > }
> > >
> > > long sbi_get_marchid(void)
> > > {
> > > - return __sbi_base_ecall(SBI_EXT_BASE_GET_MARCHID);
> > > + static long id;
> > > + static bool cached;
> >
> > This breaks for heterogeneous SMP systems (similar to big.LITTLE)
> > where HARTs will have different marchid even though they belong to
> > same CPU Vendor.
> >
> > Due to the above rationale, the patch adding marchid, mvendorid, and
> > mimpid in /proc/cpuinfo caches these values on a per-CPU basis.
>
> For people reading along, I think you mean
> https://lore.kernel.org/r/[email protected]
Yes, this is the patch I am referring.
>
> For my understanding, was there a reason in the past for doing the caching
> only for cpuinfo and not for every invocation of the ecalls?
The caching was done only for /proc/cpuinfo because at that point
time only "cat /proc/cpuinfo" would need these value at runtime.
Now that we have more uses of marchid, mvendorid, and mimpid,
at runtime, we should definitely have the cached values available
to other parts of kernel.
My suggestion is to keep sbi_get_xyz() functions unmodified and
instead add new functions in arch/riscv/kernel/cpu.c (on-top-of
/proc/cpuinfo patch) which allows users to read cached values of
any CPU.
For example, we could export following functions from
arch/riscv/kernel/cpu.c:
unsigned long riscv_cached_mvendorid(unsigned int cpu);
unsigned long riscv_cached_marchid(unsigned int cpu);
unsigned long riscv_cached_mimpid(unsigned int cpu);
Regards,
Anup
On Mon, Oct 10, 2022 at 5:57 PM Heiko Stuebner <[email protected]> wrote:
>
> sbi_get_mvendorid(), sbi_get_marchid() and sbi_get_mimpid() might get
> called multiple times, though the values of these CSRs should not change
> during the runtime of a specific machine.
>
> So cache the values in the functions and prevent multiple ecalls
> to read these values.
>
> As Andrew Jones noted, at least marchid and mimpid may be negative
> values when viewed as a long, so we use a separate static bool to
> indiciate the cached status.
>
> Suggested-by: Atish Patra <[email protected]>
> Signed-off-by: Heiko Stuebner <[email protected]>
> ---
> arch/riscv/kernel/sbi.c | 30 +++++++++++++++++++++++++++---
> 1 file changed, 27 insertions(+), 3 deletions(-)
>
> diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c
> index 775d3322b422..cc618aaa9d11 100644
> --- a/arch/riscv/kernel/sbi.c
> +++ b/arch/riscv/kernel/sbi.c
> @@ -625,17 +625,41 @@ static inline long sbi_get_firmware_version(void)
>
> long sbi_get_mvendorid(void)
> {
> - return __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> + static long id;
> + static bool cached;
> +
> + if (!cached) {
> + id = __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> + cached = true;
> + }
> +
> + return id;
> }
>
> long sbi_get_marchid(void)
> {
> - return __sbi_base_ecall(SBI_EXT_BASE_GET_MARCHID);
> + static long id;
> + static bool cached;
This breaks for heterogeneous SMP systems (similar to big.LITTLE)
where HARTs will have different marchid even though they belong to
same CPU Vendor.
Due to the above rationale, the patch adding marchid, mvendorid, and
mimpid in /proc/cpuinfo caches these values on a per-CPU basis.
Regards,
Anup
> +
> + if (!cached) {
> + id = __sbi_base_ecall(SBI_EXT_BASE_GET_MARCHID);
> + cached = true;
> + }
> +
> + return id;
> }
>
> long sbi_get_mimpid(void)
> {
> - return __sbi_base_ecall(SBI_EXT_BASE_GET_MIMPID);
> + static long id;
> + static bool cached;
> +
> + if (!cached) {
> + id = __sbi_base_ecall(SBI_EXT_BASE_GET_MIMPID);
> + cached = true;
> + }
> +
> + return id;
> }
>
> static void sbi_send_cpumask_ipi(const struct cpumask *target)
> --
> 2.35.1
>
Hi Anup,
Am Montag, 10. Oktober 2022, 15:14:21 CEST schrieb Anup Patel:
> On Mon, Oct 10, 2022 at 6:25 PM Heiko Stuebner <[email protected]> wrote:
> >
> > Am Montag, 10. Oktober 2022, 14:45:45 CEST schrieb Anup Patel:
> > > On Mon, Oct 10, 2022 at 5:57 PM Heiko Stuebner <[email protected]> wrote:
> > > >
> > > > sbi_get_mvendorid(), sbi_get_marchid() and sbi_get_mimpid() might get
> > > > called multiple times, though the values of these CSRs should not change
> > > > during the runtime of a specific machine.
> > > >
> > > > So cache the values in the functions and prevent multiple ecalls
> > > > to read these values.
> > > >
> > > > As Andrew Jones noted, at least marchid and mimpid may be negative
> > > > values when viewed as a long, so we use a separate static bool to
> > > > indiciate the cached status.
> > > >
> > > > Suggested-by: Atish Patra <[email protected]>
> > > > Signed-off-by: Heiko Stuebner <[email protected]>
> > > > ---
> > > > arch/riscv/kernel/sbi.c | 30 +++++++++++++++++++++++++++---
> > > > 1 file changed, 27 insertions(+), 3 deletions(-)
> > > >
> > > > diff --git a/arch/riscv/kernel/sbi.c b/arch/riscv/kernel/sbi.c
> > > > index 775d3322b422..cc618aaa9d11 100644
> > > > --- a/arch/riscv/kernel/sbi.c
> > > > +++ b/arch/riscv/kernel/sbi.c
> > > > @@ -625,17 +625,41 @@ static inline long sbi_get_firmware_version(void)
> > > >
> > > > long sbi_get_mvendorid(void)
> > > > {
> > > > - return __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> > > > + static long id;
> > > > + static bool cached;
> > > > +
> > > > + if (!cached) {
> > > > + id = __sbi_base_ecall(SBI_EXT_BASE_GET_MVENDORID);
> > > > + cached = true;
> > > > + }
> > > > +
> > > > + return id;
> > > > }
> > > >
> > > > long sbi_get_marchid(void)
> > > > {
> > > > - return __sbi_base_ecall(SBI_EXT_BASE_GET_MARCHID);
> > > > + static long id;
> > > > + static bool cached;
> > >
> > > This breaks for heterogeneous SMP systems (similar to big.LITTLE)
> > > where HARTs will have different marchid even though they belong to
> > > same CPU Vendor.
> > >
> > > Due to the above rationale, the patch adding marchid, mvendorid, and
> > > mimpid in /proc/cpuinfo caches these values on a per-CPU basis.
> >
> > For people reading along, I think you mean
> > https://lore.kernel.org/r/[email protected]
>
> Yes, this is the patch I am referring.
>
> >
> > For my understanding, was there a reason in the past for doing the caching
> > only for cpuinfo and not for every invocation of the ecalls?
>
> The caching was done only for /proc/cpuinfo because at that point
> time only "cat /proc/cpuinfo" would need these value at runtime.
>
> Now that we have more uses of marchid, mvendorid, and mimpid,
> at runtime, we should definitely have the cached values available
> to other parts of kernel.
>
> My suggestion is to keep sbi_get_xyz() functions unmodified and
> instead add new functions in arch/riscv/kernel/cpu.c (on-top-of
> /proc/cpuinfo patch) which allows users to read cached values of
> any CPU.
>
> For example, we could export following functions from
> arch/riscv/kernel/cpu.c:
> unsigned long riscv_cached_mvendorid(unsigned int cpu);
> unsigned long riscv_cached_marchid(unsigned int cpu);
> unsigned long riscv_cached_mimpid(unsigned int cpu);
sounds sensible, so I'll go that way you suggested :-)
Thanks
Heiko