[PATCH 1/1] arm: traps: remove fp underflow checking

[PATCH 1/1] arm: traps: remove fp underflow checking

[Maninder Singh]

with IRQSTACKS fp can point outside of thread's stack.
And when there is crash reported from IRQ side, it always report
"frame pointer underflow", which is not real underflow.

Unhandled fault: page domain fault (0x81b) at 0x00000000
..
PC is at handle_irq_desc+0x68/0xa4
LR is at generic_handle_domain_irq+0x18/0x1c
..

Call trace: frame pointer underflow
 handle_irq_desc from generic_handle_domain_irq+0x18/0x1c
 generic_handle_domain_irq from gic_handle_irq+0x78/0x8c
 gic_handle_irq from generic_handle_arch_irq+0x3c/0x4c
..

It is just warning print and there is no decision based on this
to stop backtrack or etc. So it can be removed.

Signed-off-by: Maninder Singh <maninder1.s@samsung.com>
---
 arch/arm/kernel/traps.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c
index afbd2ebe5c39..47f8c0738399 100644
--- a/arch/arm/kernel/traps.c
+++ b/arch/arm/kernel/traps.c
@@ -243,8 +243,8 @@ void dump_backtrace(struct pt_regs *regs, struct task_struct *tsk,
 	} else if (verify_stack(fp)) {
 		pr_cont("invalid frame pointer 0x%08x", fp);
 		ok = 0;
-	} else if (fp < (unsigned long)end_of_stack(tsk))
-		pr_cont("frame pointer underflow");
+	}
+
 	pr_cont("\n");
 
 	if (ok)
-- 
2.34.1

Re: [PATCH 1/1] arm: traps: remove fp underflow checking

[Russell King (Oracle)]

On Fri, Apr 24, 2026 at 10:00:29AM +0530, Maninder Singh wrote:
> with IRQSTACKS fp can point outside of thread's stack.
> And when there is crash reported from IRQ side, it always report
> "frame pointer underflow", which is not real underflow.
> 
> Unhandled fault: page domain fault (0x81b) at 0x00000000
> ..
> PC is at handle_irq_desc+0x68/0xa4
> LR is at generic_handle_domain_irq+0x18/0x1c
> ..
> 
> Call trace: frame pointer underflow
>  handle_irq_desc from generic_handle_domain_irq+0x18/0x1c
>  generic_handle_domain_irq from gic_handle_irq+0x78/0x8c
>  gic_handle_irq from generic_handle_arch_irq+0x3c/0x4c
> ..
> 
> It is just warning print and there is no decision based on this
> to stop backtrack or etc. So it can be removed.

Please find a way to check whether the frame pointer is within the
IRQ stack rather than removing this.

Thanks.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!