The ACPI specification says OS shouldn't attempt to use GICC configuration
parameters if the flag ACPI_MADT_ENABLED is cleared. The ARM64-SMP code
skips the disabled GICC entries but not causing any issue. However the
current GICv3 driver probe bails out causing kernel panic() instead of
skipping the disabled GICC interfaces. This issues on happens on systems
where redistributor regions are not in the always-on power domain and
one of GICC interface marked with ACPI_MADT_ENABLED=0.
This patch does the two things to fix the panic.
- Don't return an error in gic_acpi_match_gicc() for disabled GICC entry.
- No need to keep GICR region information for disabled GICC entry.
Observed kernel crash on QDF2400 platform GICC entry is disabled.
Kernel crash traces:
Kernel panic - not syncing: No interrupt controller found.
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.13.5 #26
[<ffff000008087770>] dump_backtrace+0x0/0x218
[<ffff0000080879dc>] show_stack+0x14/0x20
[<ffff00000883b078>] dump_stack+0x98/0xb8
[<ffff0000080c5c14>] panic+0x118/0x26c
[<ffff000008b62348>] init_IRQ+0x24/0x2c
[<ffff000008b609fc>] start_kernel+0x230/0x394
[<ffff000008b601e4>] __primary_switched+0x64/0x6c
---[ end Kernel panic - not syncing: No interrupt controller found.
Signed-off-by: Shanker Donthineni <[email protected]>
---
changes since v1:
Added a new "if condition" to skip disabled GICC entry and edited commit.
drivers/irqchip/irq-gic-v3.c | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
index b56c3e2..606d4d2 100644
--- a/drivers/irqchip/irq-gic-v3.c
+++ b/drivers/irqchip/irq-gic-v3.c
@@ -1331,6 +1331,10 @@ static int __init gic_of_init(struct device_node *node, struct device_node *pare
u32 size = reg == GIC_PIDR2_ARCH_GICv4 ? SZ_64K * 4 : SZ_64K * 2;
void __iomem *redist_base;
+ /* GICC entry which has !ACPI_MADT_ENABLED is not unusable so skip */
+ if (!(gicc->flags & ACPI_MADT_ENABLED))
+ return 0;
+
redist_base = ioremap(gicc->gicr_base_address, size);
if (!redist_base)
return -ENOMEM;
@@ -1380,6 +1384,13 @@ static int __init gic_acpi_match_gicc(struct acpi_subtable_header *header,
if ((gicc->flags & ACPI_MADT_ENABLED) && gicc->gicr_base_address)
return 0;
+ /*
+ * It's perfectly valid firmware can pass disabled GICC entry, driver
+ * should treat an errors, skip the entry.
+ */
+ if (!(gicc->flags & ACPI_MADT_ENABLED))
+ return 0;
+
return -ENODEV;
}
--
Qualcomm Datacenter Technologies, Inc. on behalf of the Qualcomm Technologies, Inc.
Qualcomm Technologies, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project.
Hi Marc,
I messed-up patch contents please disregard this one and review v4 patch.
[v4] irqchip/gic-v3: Fix the driver probe() fail due to disabled GICC entry
https://patchwork.kernel.org/patch/10093653/
--
Shanker Donthineni
Qualcomm Datacenter Technologies, Inc. as an affiliate of Qualcomm Technologies, Inc.
Qualcomm Technologies, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project.