[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <9aa8c627-ad69-43e9-9331-40a5df53de7f@ghiti.fr>
Date: Tue, 29 Oct 2024 08:52:57 +0100
From: Alexandre Ghiti <alex@...ti.fr>
To: Sunil V L <sunilvl@...tanamicro.com>, linux-riscv@...ts.infradead.org,
linux-kernel@...r.kernel.org
Cc: Albert Ou <aou@...s.berkeley.edu>, Anup Patel <anup@...infault.org>,
"Rafael J . Wysocki" <rafael.j.wysocki@...el.com>,
Björn Töpel <bjorn@...nel.org>,
Palmer Dabbelt <palmer@...belt.com>, Paul Walmsley
<paul.walmsley@...ive.com>, Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [PATCH] irqchip/riscv-intc: Fix no-SMP boot with ACPI
On 29/10/2024 08:46, Alexandre Ghiti wrote:
> Hi Thomas,
>
> On 14/10/2024 08:57, Sunil V L wrote:
>> When CONFIG_SMP is disabled, the static array rintc_acpi_data with size
>> NR_CPUS will not be sufficient to hold all RINTC structures passed from
>> the firmware. All RINTC structures are required to configure
>> IMSIC/APLIC/PLIC properly irrespective of SMP in the OS. So, allocate
>> dynamic memory based on the number of RINTC structures in MADT to fix
>> this issue.
>>
>> Fixes: f8619b66bdb1 ("irqchip/riscv-intc: Add ACPI support for AIA")
>> Reported-by: Björn Töpel <bjorn@...nel.org>
>> Closes:
>> https://github.com/linux-riscv/linux-riscv/actions/runs/11280997511/job/31375229012
>> Signed-off-by: Sunil V L <sunilvl@...tanamicro.com>
>> ---
>> drivers/irqchip/irq-riscv-intc.c | 19 ++++++++++++++++++-
>> 1 file changed, 18 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/irqchip/irq-riscv-intc.c
>> b/drivers/irqchip/irq-riscv-intc.c
>> index 8c5411386220..f653c13de62b 100644
>> --- a/drivers/irqchip/irq-riscv-intc.c
>> +++ b/drivers/irqchip/irq-riscv-intc.c
>> @@ -265,7 +265,7 @@ struct rintc_data {
>> };
>> static u32 nr_rintc;
>> -static struct rintc_data *rintc_acpi_data[NR_CPUS];
>> +static struct rintc_data **rintc_acpi_data;
>> #define for_each_matching_plic(_plic_id) \
>> unsigned int _plic; \
>> @@ -329,13 +329,30 @@ int acpi_rintc_get_imsic_mmio_info(u32 index,
>> struct resource *res)
>> return 0;
>> }
>> +static int __init riscv_intc_acpi_match(union
>> acpi_subtable_headers *header,
>> + const unsigned long end)
>> +{
>> + return 0;
>> +}
>> +
>> static int __init riscv_intc_acpi_init(union acpi_subtable_headers
>> *header,
>> const unsigned long end)
>> {
>> struct acpi_madt_rintc *rintc;
>> struct fwnode_handle *fn;
>> + int count;
>> int rc;
>> + if (!rintc_acpi_data) {
>> + count = acpi_table_parse_madt(ACPI_MADT_TYPE_RINTC,
>> riscv_intc_acpi_match, 0);
>> + if (count <= 0)
>> + return -EINVAL;
>> +
>> + rintc_acpi_data = kcalloc(count, sizeof(*rintc_acpi_data),
>> GFP_KERNEL);
>> + if (!rintc_acpi_data)
>> + return -ENOMEM;
>> + }
>> +
>> rintc = (struct acpi_madt_rintc *)header;
>> rintc_acpi_data[nr_rintc] =
>> kzalloc(sizeof(*rintc_acpi_data[0]), GFP_KERNEL);
>> if (!rintc_acpi_data[nr_rintc])
>
>
> Do you think you can merge this patch for rc6? We still need it to fix
> our nosmp builds.
Please ignore this message, I found it with a different title.
Thanks,
Alex
>
> Thanks,
>
> Alex
>
>
> _______________________________________________
> linux-riscv mailing list
> linux-riscv@...ts.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-riscv
Powered by blists - more mailing lists