A static key is used to select between SBI and Sstc timer usage in
riscv_clock_next_event(), but currently the direction is resolved
after cpuhp_setup_state() is called (which sets the next event). The
first event will therefore fall through the sbi_set_timer() path; this
breaks Sstc-only systems. So, apply the jump patching before first
use.
Fixes: 9f7a8ff6391f ("RISC-V: Prefer sstc extension if available")
Signed-off-by: Matt Evans <[email protected]>
---
V1 -> V2: Commit msg tweak.
drivers/clocksource/timer-riscv.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/clocksource/timer-riscv.c b/drivers/clocksource/timer-riscv.c
index 969a552da8d2..a36d173fd6cd 100644
--- a/drivers/clocksource/timer-riscv.c
+++ b/drivers/clocksource/timer-riscv.c
@@ -177,6 +177,11 @@ static int __init riscv_timer_init_dt(struct device_node *n)
return error;
}
+ if (riscv_isa_extension_available(NULL, SSTC)) {
+ pr_info("Timer interrupt in S-mode is available via sstc extension\n");
+ static_branch_enable(&riscv_sstc_available);
+ }
+
error = cpuhp_setup_state(CPUHP_AP_RISCV_TIMER_STARTING,
"clockevents/riscv/timer:starting",
riscv_timer_starting_cpu, riscv_timer_dying_cpu);
@@ -184,11 +189,6 @@ static int __init riscv_timer_init_dt(struct device_node *n)
pr_err("cpu hp setup state failed for RISCV timer [%d]\n",
error);
- if (riscv_isa_extension_available(NULL, SSTC)) {
- pr_info("Timer interrupt in S-mode is available via sstc extension\n");
- static_branch_enable(&riscv_sstc_available);
- }
-
return error;
}
--
2.30.2
On Thu, Feb 2, 2023 at 1:19 AM Matt Evans <[email protected]> wrote:
>
> A static key is used to select between SBI and Sstc timer usage in
> riscv_clock_next_event(), but currently the direction is resolved
> after cpuhp_setup_state() is called (which sets the next event). The
> first event will therefore fall through the sbi_set_timer() path; this
> breaks Sstc-only systems. So, apply the jump patching before first
> use.
>
> Fixes: 9f7a8ff6391f ("RISC-V: Prefer sstc extension if available")
> Signed-off-by: Matt Evans <[email protected]>
Looks good to me.
Reviewed-by: Anup Patel <[email protected]>
Regards,
Anup
> ---
>
> V1 -> V2: Commit msg tweak.
>
> drivers/clocksource/timer-riscv.c | 10 +++++-----
> 1 file changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/clocksource/timer-riscv.c b/drivers/clocksource/timer-riscv.c
> index 969a552da8d2..a36d173fd6cd 100644
> --- a/drivers/clocksource/timer-riscv.c
> +++ b/drivers/clocksource/timer-riscv.c
> @@ -177,6 +177,11 @@ static int __init riscv_timer_init_dt(struct device_node *n)
> return error;
> }
>
> + if (riscv_isa_extension_available(NULL, SSTC)) {
> + pr_info("Timer interrupt in S-mode is available via sstc extension\n");
> + static_branch_enable(&riscv_sstc_available);
> + }
> +
> error = cpuhp_setup_state(CPUHP_AP_RISCV_TIMER_STARTING,
> "clockevents/riscv/timer:starting",
> riscv_timer_starting_cpu, riscv_timer_dying_cpu);
> @@ -184,11 +189,6 @@ static int __init riscv_timer_init_dt(struct device_node *n)
> pr_err("cpu hp setup state failed for RISCV timer [%d]\n",
> error);
>
> - if (riscv_isa_extension_available(NULL, SSTC)) {
> - pr_info("Timer interrupt in S-mode is available via sstc extension\n");
> - static_branch_enable(&riscv_sstc_available);
> - }
> -
> return error;
> }
>
> --
> 2.30.2
>
>
On Wed, 01 Feb 2023 11:49:42 PST (-0800), Matt Evans wrote:
> A static key is used to select between SBI and Sstc timer usage in
> riscv_clock_next_event(), but currently the direction is resolved
> after cpuhp_setup_state() is called (which sets the next event). The
> first event will therefore fall through the sbi_set_timer() path; this
> breaks Sstc-only systems. So, apply the jump patching before first
> use.
>
> Fixes: 9f7a8ff6391f ("RISC-V: Prefer sstc extension if available")
> Signed-off-by: Matt Evans <[email protected]>
> ---
>
> V1 -> V2: Commit msg tweak.
>
> drivers/clocksource/timer-riscv.c | 10 +++++-----
> 1 file changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/clocksource/timer-riscv.c b/drivers/clocksource/timer-riscv.c
> index 969a552da8d2..a36d173fd6cd 100644
> --- a/drivers/clocksource/timer-riscv.c
> +++ b/drivers/clocksource/timer-riscv.c
> @@ -177,6 +177,11 @@ static int __init riscv_timer_init_dt(struct device_node *n)
> return error;
> }
>
> + if (riscv_isa_extension_available(NULL, SSTC)) {
> + pr_info("Timer interrupt in S-mode is available via sstc extension\n");
> + static_branch_enable(&riscv_sstc_available);
> + }
> +
> error = cpuhp_setup_state(CPUHP_AP_RISCV_TIMER_STARTING,
> "clockevents/riscv/timer:starting",
> riscv_timer_starting_cpu, riscv_timer_dying_cpu);
> @@ -184,11 +189,6 @@ static int __init riscv_timer_init_dt(struct device_node *n)
> pr_err("cpu hp setup state failed for RISCV timer [%d]\n",
> error);
>
> - if (riscv_isa_extension_available(NULL, SSTC)) {
> - pr_info("Timer interrupt in S-mode is available via sstc extension\n");
> - static_branch_enable(&riscv_sstc_available);
> - }
> -
> return error;
> }
>
> --
> 2.30.2
Reviewed-by: Palmer Dabbelt <[email protected]>
Looks like I've taken stuff here before, but I'm also happy to have the
clocksource folks pick this up so
Acked-by: Palmer Dabbelt <[email protected]>
I don't think this is a particularly critical fix, as IIUC there's still
no Sstc-only hardware in the real world. Given how late we are I'd err
towards putting this on for-next anyway, so I'll give the clocksource
folks a bit more time to chime in. Otherwise I'll put it on for-next
before the merge window (assuming we're got an rc8 coming).
Thanks!
The following commit has been merged into the timers/core branch of tip:
Commit-ID: 225b9596cb0227c1c1b1e4a836dad43595c3e61a
Gitweb: https://git.kernel.org/tip/225b9596cb0227c1c1b1e4a836dad43595c3e61a
Author: Matt Evans <[email protected]>
AuthorDate: Wed, 01 Feb 2023 19:49:42
Committer: Daniel Lezcano <[email protected]>
CommitterDate: Mon, 13 Feb 2023 13:10:17 +01:00
clocksource/drivers/riscv: Patch riscv_clock_next_event() jump before first use
A static key is used to select between SBI and Sstc timer usage in
riscv_clock_next_event(), but currently the direction is resolved
after cpuhp_setup_state() is called (which sets the next event). The
first event will therefore fall through the sbi_set_timer() path; this
breaks Sstc-only systems. So, apply the jump patching before first
use.
Fixes: 9f7a8ff6391f ("RISC-V: Prefer sstc extension if available")
Signed-off-by: Matt Evans <[email protected]>
Reviewed-by: Palmer Dabbelt <[email protected]>
Acked-by: Palmer Dabbelt <[email protected]>
Reviewed-by: Anup Patel <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Daniel Lezcano <[email protected]>
---
drivers/clocksource/timer-riscv.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/clocksource/timer-riscv.c b/drivers/clocksource/timer-riscv.c
index d8cb629..5f0f10c 100644
--- a/drivers/clocksource/timer-riscv.c
+++ b/drivers/clocksource/timer-riscv.c
@@ -192,6 +192,11 @@ static int __init riscv_timer_init_dt(struct device_node *n)
return error;
}
+ if (riscv_isa_extension_available(NULL, SSTC)) {
+ pr_info("Timer interrupt in S-mode is available via sstc extension\n");
+ static_branch_enable(&riscv_sstc_available);
+ }
+
error = cpuhp_setup_state(CPUHP_AP_RISCV_TIMER_STARTING,
"clockevents/riscv/timer:starting",
riscv_timer_starting_cpu, riscv_timer_dying_cpu);
@@ -199,11 +204,6 @@ static int __init riscv_timer_init_dt(struct device_node *n)
pr_err("cpu hp setup state failed for RISCV timer [%d]\n",
error);
- if (riscv_isa_extension_available(NULL, SSTC)) {
- pr_info("Timer interrupt in S-mode is available via sstc extension\n");
- static_branch_enable(&riscv_sstc_available);
- }
-
return error;
}