Re: [PATCH v2 2/4] powerpc/interrupt: Refactor prep_irq_for_user_exit()
flat view
From: Nicholas Piggin <npiggin@gmail.com>
Date: 2021-06-11 02:31:51
Also in:
lkml
Excerpts from Christophe Leroy's message of June 5, 2021 12:56 am:
prep_irq_for_user_exit() is a superset of prep_irq_for_kernel_enabled_exit(). Refactor it.
I like the refactoring, but now prep_irq_for_user_exit() is calling prep_irq_for_kernel_enabled_exit(), which seems like the wrong naming. You could re-name prep_irq_for_kernel_enabled_exit() to prep_irq_for_enabled_exit() maybe? Or it could be __prep_irq_for_enabled_exit() then prep_irq_for_kernel_enabled_exit() and prep_irq_for_user_exit() would both call it. Otherwise Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
quoted hunk ↗ jump to hunk
Signed-off-by: Christophe Leroy <redacted> --- arch/powerpc/kernel/interrupt.c | 25 +++++-------------------- 1 file changed, 5 insertions(+), 20 deletions(-)diff --git a/arch/powerpc/kernel/interrupt.c b/arch/powerpc/kernel/interrupt.c index 539455c62c5b..b6aa80930733 100644 --- a/arch/powerpc/kernel/interrupt.c +++ b/arch/powerpc/kernel/interrupt.c@@ -78,29 +78,14 @@ static notrace __always_inline bool prep_irq_for_kernel_enabled_exit(bool restar */ static notrace __always_inline bool prep_irq_for_user_exit(void) { - user_enter_irqoff(); - /* This must be done with RI=1 because tracing may touch vmaps */ - trace_hardirqs_on(); - -#ifdef CONFIG_PPC32 - __hard_EE_RI_disable(); -#else - if (exit_must_hard_disable()) - __hard_EE_RI_disable(); + bool ret; - /* This pattern matches prep_irq_for_idle */ - if (unlikely(lazy_irq_pending_nocheck())) { - if (exit_must_hard_disable()) { - local_paca->irq_happened |= PACA_IRQ_HARD_DIS; - __hard_RI_enable(); - } - trace_hardirqs_off(); + user_enter_irqoff(); + ret = prep_irq_for_kernel_enabled_exit(true); + if (!ret) user_exit_irqoff(); - return false; - } -#endif - return true; + return ret; } /* Has to run notrace because it is entered not completely "reconciled" */-- 2.25.0