[PATCH v11 01/14] RISC-V: Don't fail in riscv_of_parent_hartid() for disabled HARTs

Anup Patel posted 14 patches 2 years, 2 months ago
There is a newer version of this series
[PATCH v11 01/14] RISC-V: Don't fail in riscv_of_parent_hartid() for disabled HARTs
Posted by Anup Patel 2 years, 2 months ago
The riscv_of_processor_hartid() used by riscv_of_parent_hartid() fails
for HARTs disabled in the DT. This results in the following warning
thrown by the RISC-V INTC driver for the E-core on SiFive boards:

[    0.000000] riscv-intc: unable to find hart id for /cpus/cpu@0/interrupt-controller

The riscv_of_parent_hartid() is only expected to read the hartid from
the DT so we should directly call of_get_cpu_hwid() instead of calling
riscv_of_processor_hartid().

Fixes: ad635e723e17 ("riscv: cpu: Add 64bit hartid support on RV64")
Signed-off-by: Anup Patel <apatel@ventanamicro.com>
Reviewed-by: Atish Patra <atishp@rivosinc.com>
---
 arch/riscv/kernel/cpu.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/arch/riscv/kernel/cpu.c b/arch/riscv/kernel/cpu.c
index c17dacb1141c..157ace8b262c 100644
--- a/arch/riscv/kernel/cpu.c
+++ b/arch/riscv/kernel/cpu.c
@@ -125,13 +125,14 @@ int __init riscv_early_of_processor_hartid(struct device_node *node, unsigned lo
  */
 int riscv_of_parent_hartid(struct device_node *node, unsigned long *hartid)
 {
-	int rc;
-
 	for (; node; node = node->parent) {
 		if (of_device_is_compatible(node, "riscv")) {
-			rc = riscv_of_processor_hartid(node, hartid);
-			if (!rc)
-				return 0;
+			*hartid = (unsigned long)of_get_cpu_hwid(node, 0);
+			if (*hartid == ~0UL) {
+				pr_warn("Found CPU without hart ID\n");
+				return -ENODEV;
+			}
+			return 0;
 		}
 	}
 
-- 
2.34.1
Re: [PATCH v11 01/14] RISC-V: Don't fail in riscv_of_parent_hartid() for disabled HARTs
Posted by Björn Töpel 2 years, 2 months ago
Anup Patel <apatel@ventanamicro.com> writes:

> The riscv_of_processor_hartid() used by riscv_of_parent_hartid() fails
> for HARTs disabled in the DT. This results in the following warning
> thrown by the RISC-V INTC driver for the E-core on SiFive boards:
>
> [    0.000000] riscv-intc: unable to find hart id for /cpus/cpu@0/interrupt-controller
>
> The riscv_of_parent_hartid() is only expected to read the hartid from
> the DT so we should directly call of_get_cpu_hwid() instead of calling
> riscv_of_processor_hartid().
>
> Fixes: ad635e723e17 ("riscv: cpu: Add 64bit hartid support on RV64")

Patch 1 and 3: These fixes are stand alone, and doesn't have to be part
of the series.

Wouldn't it be better to pull these out of the long-going series, and
try to get in the fixes ASAP?


Björn
Re: [PATCH v11 01/14] RISC-V: Don't fail in riscv_of_parent_hartid() for disabled HARTs
Posted by Anup Patel 2 years, 2 months ago
On Tue, Oct 24, 2023 at 5:25 PM Björn Töpel <bjorn@kernel.org> wrote:
>
> Anup Patel <apatel@ventanamicro.com> writes:
>
> > The riscv_of_processor_hartid() used by riscv_of_parent_hartid() fails
> > for HARTs disabled in the DT. This results in the following warning
> > thrown by the RISC-V INTC driver for the E-core on SiFive boards:
> >
> > [    0.000000] riscv-intc: unable to find hart id for /cpus/cpu@0/interrupt-controller
> >
> > The riscv_of_parent_hartid() is only expected to read the hartid from
> > the DT so we should directly call of_get_cpu_hwid() instead of calling
> > riscv_of_processor_hartid().
> >
> > Fixes: ad635e723e17 ("riscv: cpu: Add 64bit hartid support on RV64")
>
> Patch 1 and 3: These fixes are stand alone, and doesn't have to be part
> of the series.
>
> Wouldn't it be better to pull these out of the long-going series, and
> try to get in the fixes ASAP?

Yes, that is correct. In fact, PATCH2 can also be taken for Linux-6.7.

I suggest PATCH1 to PATCH3 (3 patches) be taken for Linux-6.7
since the merge window is pretty close.

Regards,
Anup