| Message ID | 20190213105041.13537-4-ming.lei@redhat.com (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
| Series | genirq/affinity: add .calc_sets for improving IRQ allocation & spread | expand |
On Wed, Feb 13, 2019 at 06:50:39PM +0800, Ming Lei wrote: > Currently pre-caculated set vectors are provided by driver for > allocating & spread vectors. This way only works when drivers passes > same 'max_vecs' and 'min_vecs' to pci_alloc_irq_vectors_affinity(), > also requires driver to retry the allocating & spread. s/pre-caculated/precalculated/ s/drivers/a driver/ s/also requires/which also requires the/ > As Bjorn and Keith mentioned, the current usage & interface for irq sets > is a bit awkward because the retrying should have been avoided by providing > one resonable 'min_vecs'. However, if 'min_vecs' isn't same with > 'max_vecs', number of the allocated vectors is unknown before calling > pci_alloc_irq_vectors_affinity(), then each set's vectors can't be > pre-caculated. s/resonable/reasonable/ s/isn't same with/isn't the same as/ s/number of/the number of/ s/pre-caculated/precalculated/ s/then each/and the/ s/irq/IRQ/ > Add a new callback of .calc_sets into 'struct irq_affinity' so that > driver can caculate set vectors after IRQ vector is allocated and > before spread IRQ vectors. Add 'priv' so that driver may retrieve > its private data via the 'struct irq_affinity'. s/caculate/calculate/ I *think* "set vectors" here has something to do with an affinity cpumask? "Set vectors" doesn't seem like quite the right terminology. > Suggested-by: Thomas Gleixner <tglx@linutronix.de> > Signed-off-by: Ming Lei <ming.lei@redhat.com> > --- > include/linux/interrupt.h | 4 ++++ > kernel/irq/affinity.c | 8 ++++++-- > 2 files changed, 10 insertions(+), 2 deletions(-) > > diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h > index a20150627a32..7a27f6ba1f2f 100644 > --- a/include/linux/interrupt.h > +++ b/include/linux/interrupt.h > @@ -269,12 +269,16 @@ struct irq_affinity_notify { > * the MSI(-X) vector space > * @nr_sets: Length of passed in *sets array > * @set_vectors: Number of affinitized sets > + * @calc_sets: Callback for caculating set vectors Possible s/set vectors/<something else>/ ? > + * @priv: Private data of @calc_sets > */ > struct irq_affinity { > int pre_vectors; > int post_vectors; > int nr_sets; > int set_vectors[IRQ_MAX_SETS]; > + void (*calc_sets)(struct irq_affinity *, int nvecs); > + void *priv; > }; > > /** > diff --git a/kernel/irq/affinity.c b/kernel/irq/affinity.c > index b868b9d3df7f..2341b1f005fa 100644 > --- a/kernel/irq/affinity.c > +++ b/kernel/irq/affinity.c > @@ -267,7 +267,9 @@ irq_create_affinity_masks(int nvecs, struct irq_affinity *affd) > * Spread on present CPUs starting from affd->pre_vectors. If we > * have multiple sets, build each sets affinity mask separately. > */ > - if (!affd->nr_sets) { > + if (affd->calc_sets) { > + affd->calc_sets(affd, nvecs); > + } else if (!affd->nr_sets) { > affd->nr_sets = 1; > affd->set_vectors[0] = affvecs; > } > @@ -316,7 +318,9 @@ int irq_calc_affinity_vectors(int minvec, int maxvec, const struct irq_affinity > if (resv > minvec) > return 0; > > - if (affd->nr_sets) { > + if (affd->calc_sets) { > + set_vecs = vecs; > + } else if (affd->nr_sets) { > int i; > > for (i = 0, set_vecs = 0; i < affd->nr_sets; i++) > -- > 2.9.5 >
On Wed, 13 Feb 2019, Bjorn Helgaas wrote: > > Add a new callback of .calc_sets into 'struct irq_affinity' so that > > driver can caculate set vectors after IRQ vector is allocated and > > before spread IRQ vectors. Add 'priv' so that driver may retrieve > > its private data via the 'struct irq_affinity'. > > s/caculate/calculate/ > > I *think* "set vectors" here has something to do with an affinity > cpumask? "Set vectors" doesn't seem like quite the right terminology. See previous mail. Let's stay with 'interrupt sets' and get rid of the vector naming completely. Thanks, tglx
diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index a20150627a32..7a27f6ba1f2f 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -269,12 +269,16 @@ struct irq_affinity_notify { * the MSI(-X) vector space * @nr_sets: Length of passed in *sets array * @set_vectors: Number of affinitized sets + * @calc_sets: Callback for caculating set vectors + * @priv: Private data of @calc_sets */ struct irq_affinity { int pre_vectors; int post_vectors; int nr_sets; int set_vectors[IRQ_MAX_SETS]; + void (*calc_sets)(struct irq_affinity *, int nvecs); + void *priv; }; /** diff --git a/kernel/irq/affinity.c b/kernel/irq/affinity.c index b868b9d3df7f..2341b1f005fa 100644 --- a/kernel/irq/affinity.c +++ b/kernel/irq/affinity.c @@ -267,7 +267,9 @@ irq_create_affinity_masks(int nvecs, struct irq_affinity *affd) * Spread on present CPUs starting from affd->pre_vectors. If we * have multiple sets, build each sets affinity mask separately. */ - if (!affd->nr_sets) { + if (affd->calc_sets) { + affd->calc_sets(affd, nvecs); + } else if (!affd->nr_sets) { affd->nr_sets = 1; affd->set_vectors[0] = affvecs; } @@ -316,7 +318,9 @@ int irq_calc_affinity_vectors(int minvec, int maxvec, const struct irq_affinity if (resv > minvec) return 0; - if (affd->nr_sets) { + if (affd->calc_sets) { + set_vecs = vecs; + } else if (affd->nr_sets) { int i; for (i = 0, set_vecs = 0; i < affd->nr_sets; i++)
Currently pre-caculated set vectors are provided by driver for allocating & spread vectors. This way only works when drivers passes same 'max_vecs' and 'min_vecs' to pci_alloc_irq_vectors_affinity(), also requires driver to retry the allocating & spread. As Bjorn and Keith mentioned, the current usage & interface for irq sets is a bit awkward because the retrying should have been avoided by providing one resonable 'min_vecs'. However, if 'min_vecs' isn't same with 'max_vecs', number of the allocated vectors is unknown before calling pci_alloc_irq_vectors_affinity(), then each set's vectors can't be pre-caculated. Add a new callback of .calc_sets into 'struct irq_affinity' so that driver can caculate set vectors after IRQ vector is allocated and before spread IRQ vectors. Add 'priv' so that driver may retrieve its private data via the 'struct irq_affinity'. Suggested-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ming Lei <ming.lei@redhat.com> --- include/linux/interrupt.h | 4 ++++ kernel/irq/affinity.c | 8 ++++++-- 2 files changed, 10 insertions(+), 2 deletions(-)