| Message ID | 20240912033844.3014051-1-ruanjinjie@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | soc: ti: knav_qmss_queue: Use IRQF_NO_AUTOEN flag in request_irq() | expand |
On 11:38-20240912, Jinjie Ruan wrote: > disable_irq() after request_irq() still has a time gap in which > interrupts can come. request_irq() with IRQF_NO_AUTOEN flag will > disable IRQ auto-enable when request IRQ. does it explicitly disable the irq? > > Fixes: 41f93af900a2 ("soc: ti: add Keystone Navigator QMSS driver") Not a fixes. at least not a fix when the code was introduced in 2014 and the flag was introduced in 2015. > Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com> > --- > drivers/soc/ti/knav_qmss_queue.c | 5 ++--- > 1 file changed, 2 insertions(+), 3 deletions(-) > > diff --git a/drivers/soc/ti/knav_qmss_queue.c b/drivers/soc/ti/knav_qmss_queue.c > index f2055a76f84c..9c8d2f13b947 100644 > --- a/drivers/soc/ti/knav_qmss_queue.c > +++ b/drivers/soc/ti/knav_qmss_queue.c > @@ -119,11 +119,10 @@ static int knav_queue_setup_irq(struct knav_range_info *range, > > if (range->flags & RANGE_HAS_IRQ) { > irq = range->irqs[queue].irq; > - ret = request_irq(irq, knav_queue_int_handler, 0, > - inst->irq_name, inst); > + ret = request_irq(irq, knav_queue_int_handler, IRQF_NO_AUTOEN, > + inst->irq_name, inst); > if (ret) > return ret; > - disable_irq(irq); > if (range->irqs[queue].cpu_mask) { > ret = irq_set_affinity_hint(irq, range->irqs[queue].cpu_mask); > if (ret) { > -- > 2.34.1 >
On 2024/9/14 2:14, Nishanth Menon wrote: > On 11:38-20240912, Jinjie Ruan wrote: >> disable_irq() after request_irq() still has a time gap in which >> interrupts can come. request_irq() with IRQF_NO_AUTOEN flag will >> disable IRQ auto-enable when request IRQ. > > does it explicitly disable the irq? Not really, it just not enable it at startup. 1506 static int 1507 __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new) [...] 1782 >------->-------if (!(new->flags & IRQF_NO_AUTOEN) && 1783 >------->------- irq_settings_can_autoenable(desc)) { 1784 >------->------->-------irq_startup(desc, IRQ_RESEND, IRQ_START_COND); 1785 >------->-------} else { 1786 >------->------->-------/* 1787 >------->------->------- * Shared interrupts do not go well with disabling 1788 >------->------->------- * auto enable. The sharing interrupt might request 1789 >------->------->------- * it while it's still disabled and then wait for 1790 >------->------->------- * interrupts forever. 1791 >------->------->------- */ 1792 >------->------->-------WARN_ON_ONCE(new->flags & IRQF_SHARED); 1793 >------->------->-------/* Undo nested disables: */ 1794 >------->------->-------desc->depth = 1; 1795 >------->-------} >> >> Fixes: 41f93af900a2 ("soc: ti: add Keystone Navigator QMSS driver") > > Not a fixes. at least not a fix when the code was introduced in 2014 and > the flag was introduced in 2015. OK, let me remove the fix tag and update the commit message and resend. > >> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com> >> --- >> drivers/soc/ti/knav_qmss_queue.c | 5 ++--- >> 1 file changed, 2 insertions(+), 3 deletions(-) >> >> diff --git a/drivers/soc/ti/knav_qmss_queue.c b/drivers/soc/ti/knav_qmss_queue.c >> index f2055a76f84c..9c8d2f13b947 100644 >> --- a/drivers/soc/ti/knav_qmss_queue.c >> +++ b/drivers/soc/ti/knav_qmss_queue.c >> @@ -119,11 +119,10 @@ static int knav_queue_setup_irq(struct knav_range_info *range, >> >> if (range->flags & RANGE_HAS_IRQ) { >> irq = range->irqs[queue].irq; >> - ret = request_irq(irq, knav_queue_int_handler, 0, >> - inst->irq_name, inst); >> + ret = request_irq(irq, knav_queue_int_handler, IRQF_NO_AUTOEN, >> + inst->irq_name, inst); >> if (ret) >> return ret; >> - disable_irq(irq); >> if (range->irqs[queue].cpu_mask) { >> ret = irq_set_affinity_hint(irq, range->irqs[queue].cpu_mask); >> if (ret) { >> -- >> 2.34.1 >> >
diff --git a/drivers/soc/ti/knav_qmss_queue.c b/drivers/soc/ti/knav_qmss_queue.c index f2055a76f84c..9c8d2f13b947 100644 --- a/drivers/soc/ti/knav_qmss_queue.c +++ b/drivers/soc/ti/knav_qmss_queue.c @@ -119,11 +119,10 @@ static int knav_queue_setup_irq(struct knav_range_info *range, if (range->flags & RANGE_HAS_IRQ) { irq = range->irqs[queue].irq; - ret = request_irq(irq, knav_queue_int_handler, 0, - inst->irq_name, inst); + ret = request_irq(irq, knav_queue_int_handler, IRQF_NO_AUTOEN, + inst->irq_name, inst); if (ret) return ret; - disable_irq(irq); if (range->irqs[queue].cpu_mask) { ret = irq_set_affinity_hint(irq, range->irqs[queue].cpu_mask); if (ret) {
disable_irq() after request_irq() still has a time gap in which interrupts can come. request_irq() with IRQF_NO_AUTOEN flag will disable IRQ auto-enable when request IRQ. Fixes: 41f93af900a2 ("soc: ti: add Keystone Navigator QMSS driver") Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com> --- drivers/soc/ti/knav_qmss_queue.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)