diff --git a/arch/arm/src/armv7-a/arm_sigdeliver.c b/arch/arm/src/armv7-a/arm_sigdeliver.c index 1dc74093cb..f6afdb4500 100644 --- a/arch/arm/src/armv7-a/arm_sigdeliver.c +++ b/arch/arm/src/armv7-a/arm_sigdeliver.c @@ -116,9 +116,9 @@ void up_sigdeliver(void) * run with interrupts enabled. * * REVISIT: 'irqcount' could still be greater than zero in the SMP case. - * This would be an issue if the signal handler were to suspend because - * the critical section would be re-established when the signal handler - * resumes. + * This is an issue because (1) global spinlocks are still held and (2) if + * the signal handler were to suspend the the critical section would be + * re-established when the signal handler resumes. */ up_irq_enable(); diff --git a/arch/arm/src/armv7-m/up_sigdeliver.c b/arch/arm/src/armv7-m/up_sigdeliver.c index eedb1f22f6..577dff8ab4 100644 --- a/arch/arm/src/armv7-m/up_sigdeliver.c +++ b/arch/arm/src/armv7-m/up_sigdeliver.c @@ -129,9 +129,9 @@ void up_sigdeliver(void) * run with interrupts enabled. * * REVISIT: 'irqcount' could still be greater than zero in the SMP case. - * This would be an issue if the signal handler were to suspend because - * the critical section would be re-established when the signal handler - * resumes. + * This is an issue because (1) global spinlocks are still held and (2) if + * the signal handler were to suspend the the critical section would be + * re-established when the signal handler resumes. */ up_irq_enable(); diff --git a/arch/xtensa/src/common/xtensa_sigdeliver.c b/arch/xtensa/src/common/xtensa_sigdeliver.c index 315903a11a..b298ac0ad9 100644 --- a/arch/xtensa/src/common/xtensa_sigdeliver.c +++ b/arch/xtensa/src/common/xtensa_sigdeliver.c @@ -115,9 +115,9 @@ void xtensa_sig_deliver(void) * run with interrupts enabled. * * REVISIT: 'irqcount' could still be greater than zero in the SMP case. - * This would be an issue if the signal handler were to suspend because - * the critical section would be re-established when the signal handler - * resumes. + * This is an issue because (1) global spinlocks are still held and (2) if + * the signal handler were to suspend the the critical section would be + * re-established when the signal handler resumes. */ up_irq_enable();