| Message ID | 1449448108-9867-1-git-send-email-alistair@popple.id.au (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
Subject should start "powerpc/", so probably "powerpc/opal-irqchip:", I can fix it up here. On Mon, 2015-07-12 at 00:28:28 UTC, Alistair Popple wrote: > The OPAL event calls return a mask of events that are active in big > endian format. This is checked when unmasking the events in the > irqchip by comparison with a cached value. The cached value was stored > in big endian format but should've been converted to CPU endian > first. This bug leads to OPAL event delivery being delayed or dropped > on some systems. > > The bug is fixed by calling opal_handle_events(...) instead of > duplicating code in opal_event_unmask(...). > > Signed-off-by: Alistair Popple <alistair@popple.id.au> > Reported-by: Douglas L Lehr <dllehr@us.ibm.com> > Cc: stable@vger.kernel.org Which commit does it fix? Looks like: Fixes: 9f0fd0499d30 ("powerpc/powernv: Add a virtual irqchip for opal events") > diff --git a/arch/powerpc/platforms/powernv/opal-irqchip.c b/arch/powerpc/platforms/powernv/opal-irqchip.c > index 2c91ee7..a387d18 100644 > --- a/arch/powerpc/platforms/powernv/opal-irqchip.c > +++ b/arch/powerpc/platforms/powernv/opal-irqchip.c > @@ -43,11 +43,34 @@ static unsigned int opal_irq_count; > static unsigned int *opal_irqs; > > static void opal_handle_irq_work(struct irq_work *work); > -static __be64 last_outstanding_events; > +static u64 last_outstanding_events; > static struct irq_work opal_event_irq_work = { > .func = opal_handle_irq_work, > }; > > +void opal_handle_events(uint64_t events) > +{ > + int virq, hwirq = 0; > + u64 mask = opal_event_irqchip.mask; > + > + if (!in_irq() && (events & mask)) { > + last_outstanding_events = events; > + irq_work_queue(&opal_event_irq_work); > + return; > + } > + > + while (events & mask) { > + hwirq = fls64(events) - 1; > + if (BIT_ULL(hwirq) & mask) { > + virq = irq_find_mapping(opal_event_irqchip.domain, > + hwirq); > + if (virq) > + generic_handle_irq(virq); > + } > + events &= ~BIT_ULL(hwirq); > + } > +} So this is 100% just a move of the code from down below right? > @@ -55,12 +78,12 @@ static void opal_event_mask(struct irq_data *d) > > static void opal_event_unmask(struct irq_data *d) > { > + __be64 events; > + > set_bit(d->hwirq, &opal_event_irqchip.mask); > > - opal_poll_events(&last_outstanding_events); > - if (last_outstanding_events & opal_event_irqchip.mask) > - /* Need to retrigger the interrupt */ > - irq_work_queue(&opal_event_irq_work); > + opal_poll_events(&events); > + opal_handle_events(be64_to_cpu(events)); > } > > static int opal_event_set_type(struct irq_data *d, unsigned int flow_type) > @@ -96,29 +119,6 @@ static int opal_event_map(struct irq_domain *d, unsigned int irq, > return 0; > } > > -void opal_handle_events(uint64_t events) > -{ > - int virq, hwirq = 0; > - u64 mask = opal_event_irqchip.mask; > - > - if (!in_irq() && (events & mask)) { > - last_outstanding_events = events; > - irq_work_queue(&opal_event_irq_work); > - return; > - } > - > - while (events & mask) { > - hwirq = fls64(events) - 1; > - if (BIT_ULL(hwirq) & mask) { > - virq = irq_find_mapping(opal_event_irqchip.domain, > - hwirq); > - if (virq) > - generic_handle_irq(virq); > - } > - events &= ~BIT_ULL(hwirq); > - } > -} > - > static irqreturn_t opal_interrupt(int irq, void *data) > { > __be64 events; > @@ -131,7 +131,7 @@ static irqreturn_t opal_interrupt(int irq, void *data) > > static void opal_handle_irq_work(struct irq_work *work) > { > - opal_handle_events(be64_to_cpu(last_outstanding_events)); > + opal_handle_events(last_outstanding_events); > } > > static int opal_event_match(struct irq_domain *h, struct device_node *node, cheers
On Tue, 8 Dec 2015 09:10:19 Michael Ellerman wrote: > Subject should start "powerpc/", so probably "powerpc/opal-irqchip:", I can fix > it up here. Ok, thanks. > On Mon, 2015-07-12 at 00:28:28 UTC, Alistair Popple wrote: > > The OPAL event calls return a mask of events that are active in big > > endian format. This is checked when unmasking the events in the > > irqchip by comparison with a cached value. The cached value was stored > > in big endian format but should've been converted to CPU endian > > first. This bug leads to OPAL event delivery being delayed or dropped > > on some systems. > > > > The bug is fixed by calling opal_handle_events(...) instead of > > duplicating code in opal_event_unmask(...). > > > > Signed-off-by: Alistair Popple <alistair@popple.id.au> > > Reported-by: Douglas L Lehr <dllehr@us.ibm.com> > > Cc: stable@vger.kernel.org > > Which commit does it fix? Looks like: > > Fixes: 9f0fd0499d30 ("powerpc/powernv: Add a virtual irqchip for opal events") Yep. > > diff --git a/arch/powerpc/platforms/powernv/opal-irqchip.c b/arch/powerpc/platforms/powernv/opal-irqchip.c > > index 2c91ee7..a387d18 100644 > > --- a/arch/powerpc/platforms/powernv/opal-irqchip.c > > +++ b/arch/powerpc/platforms/powernv/opal-irqchip.c > > @@ -43,11 +43,34 @@ static unsigned int opal_irq_count; > > static unsigned int *opal_irqs; > > > > static void opal_handle_irq_work(struct irq_work *work); > > -static __be64 last_outstanding_events; > > +static u64 last_outstanding_events; > > static struct irq_work opal_event_irq_work = { > > .func = opal_handle_irq_work, > > }; > > > > +void opal_handle_events(uint64_t events) > > +{ > > + int virq, hwirq = 0; > > + u64 mask = opal_event_irqchip.mask; > > + > > + if (!in_irq() && (events & mask)) { > > + last_outstanding_events = events; > > + irq_work_queue(&opal_event_irq_work); > > + return; > > + } > > + > > + while (events & mask) { > > + hwirq = fls64(events) - 1; > > + if (BIT_ULL(hwirq) & mask) { > > + virq = irq_find_mapping(opal_event_irqchip.domain, > > + hwirq); > > + if (virq) > > + generic_handle_irq(virq); > > + } > > + events &= ~BIT_ULL(hwirq); > > + } > > +} > > So this is 100% just a move of the code from down below right? That is correct - it should be identical. Thanks. - Alistair > > @@ -55,12 +78,12 @@ static void opal_event_mask(struct irq_data *d) > > > > static void opal_event_unmask(struct irq_data *d) > > { > > + __be64 events; > > + > > set_bit(d->hwirq, &opal_event_irqchip.mask); > > > > - opal_poll_events(&last_outstanding_events); > > - if (last_outstanding_events & opal_event_irqchip.mask) > > - /* Need to retrigger the interrupt */ > > - irq_work_queue(&opal_event_irq_work); > > + opal_poll_events(&events); > > + opal_handle_events(be64_to_cpu(events)); > > } > > > > static int opal_event_set_type(struct irq_data *d, unsigned int flow_type) > > @@ -96,29 +119,6 @@ static int opal_event_map(struct irq_domain *d, unsigned int irq, > > return 0; > > } > > > > -void opal_handle_events(uint64_t events) > > -{ > > - int virq, hwirq = 0; > > - u64 mask = opal_event_irqchip.mask; > > - > > - if (!in_irq() && (events & mask)) { > > - last_outstanding_events = events; > > - irq_work_queue(&opal_event_irq_work); > > - return; > > - } > > - > > - while (events & mask) { > > - hwirq = fls64(events) - 1; > > - if (BIT_ULL(hwirq) & mask) { > > - virq = irq_find_mapping(opal_event_irqchip.domain, > > - hwirq); > > - if (virq) > > - generic_handle_irq(virq); > > - } > > - events &= ~BIT_ULL(hwirq); > > - } > > -} > > - > > static irqreturn_t opal_interrupt(int irq, void *data) > > { > > __be64 events; > > @@ -131,7 +131,7 @@ static irqreturn_t opal_interrupt(int irq, void *data) > > > > static void opal_handle_irq_work(struct irq_work *work) > > { > > - opal_handle_events(be64_to_cpu(last_outstanding_events)); > > + opal_handle_events(last_outstanding_events); > > } > > > > static int opal_event_match(struct irq_domain *h, struct device_node *node, > > > cheers
On Mon, 2015-07-12 at 00:28:28 UTC, Alistair Popple wrote: > The OPAL event calls return a mask of events that are active in big > endian format. This is checked when unmasking the events in the > irqchip by comparison with a cached value. The cached value was stored > in big endian format but should've been converted to CPU endian > first. This bug leads to OPAL event delivery being delayed or dropped > on some systems. > > The bug is fixed by calling opal_handle_events(...) instead of > duplicating code in opal_event_unmask(...). > > Signed-off-by: Alistair Popple <alistair@popple.id.au> > Reported-by: Douglas L Lehr <dllehr@us.ibm.com> > Cc: stable@vger.kernel.org Applied to powerpc fixes, thanks. https://git.kernel.org/powerpc/c/25642e1459ace29f6ce5a171 cheers
diff --git a/arch/powerpc/platforms/powernv/opal-irqchip.c b/arch/powerpc/platforms/powernv/opal-irqchip.c index 2c91ee7..a387d18 100644 --- a/arch/powerpc/platforms/powernv/opal-irqchip.c +++ b/arch/powerpc/platforms/powernv/opal-irqchip.c @@ -43,11 +43,34 @@ static unsigned int opal_irq_count; static unsigned int *opal_irqs; static void opal_handle_irq_work(struct irq_work *work); -static __be64 last_outstanding_events; +static u64 last_outstanding_events; static struct irq_work opal_event_irq_work = { .func = opal_handle_irq_work, }; +void opal_handle_events(uint64_t events) +{ + int virq, hwirq = 0; + u64 mask = opal_event_irqchip.mask; + + if (!in_irq() && (events & mask)) { + last_outstanding_events = events; + irq_work_queue(&opal_event_irq_work); + return; + } + + while (events & mask) { + hwirq = fls64(events) - 1; + if (BIT_ULL(hwirq) & mask) { + virq = irq_find_mapping(opal_event_irqchip.domain, + hwirq); + if (virq) + generic_handle_irq(virq); + } + events &= ~BIT_ULL(hwirq); + } +} + static void opal_event_mask(struct irq_data *d) { clear_bit(d->hwirq, &opal_event_irqchip.mask); @@ -55,12 +78,12 @@ static void opal_event_mask(struct irq_data *d) static void opal_event_unmask(struct irq_data *d) { + __be64 events; + set_bit(d->hwirq, &opal_event_irqchip.mask); - opal_poll_events(&last_outstanding_events); - if (last_outstanding_events & opal_event_irqchip.mask) - /* Need to retrigger the interrupt */ - irq_work_queue(&opal_event_irq_work); + opal_poll_events(&events); + opal_handle_events(be64_to_cpu(events)); } static int opal_event_set_type(struct irq_data *d, unsigned int flow_type) @@ -96,29 +119,6 @@ static int opal_event_map(struct irq_domain *d, unsigned int irq, return 0; } -void opal_handle_events(uint64_t events) -{ - int virq, hwirq = 0; - u64 mask = opal_event_irqchip.mask; - - if (!in_irq() && (events & mask)) { - last_outstanding_events = events; - irq_work_queue(&opal_event_irq_work); - return; - } - - while (events & mask) { - hwirq = fls64(events) - 1; - if (BIT_ULL(hwirq) & mask) { - virq = irq_find_mapping(opal_event_irqchip.domain, - hwirq); - if (virq) - generic_handle_irq(virq); - } - events &= ~BIT_ULL(hwirq); - } -} - static irqreturn_t opal_interrupt(int irq, void *data) { __be64 events; @@ -131,7 +131,7 @@ static irqreturn_t opal_interrupt(int irq, void *data) static void opal_handle_irq_work(struct irq_work *work) { - opal_handle_events(be64_to_cpu(last_outstanding_events)); + opal_handle_events(last_outstanding_events); } static int opal_event_match(struct irq_domain *h, struct device_node *node,
The OPAL event calls return a mask of events that are active in big endian format. This is checked when unmasking the events in the irqchip by comparison with a cached value. The cached value was stored in big endian format but should've been converted to CPU endian first. This bug leads to OPAL event delivery being delayed or dropped on some systems. The bug is fixed by calling opal_handle_events(...) instead of duplicating code in opal_event_unmask(...). Signed-off-by: Alistair Popple <alistair@popple.id.au> Reported-by: Douglas L Lehr <dllehr@us.ibm.com> Cc: stable@vger.kernel.org --- arch/powerpc/platforms/powernv/opal-irqchip.c | 58 +++++++++++++-------------- 1 file changed, 29 insertions(+), 29 deletions(-)