It turns out, it's just not worth to add bits to the core just for
hisi_sas_v2 driver. I dropped those changes.
BTW, this series is based on linux-block/for-next from 2024-11-11. Not
sure if this is the right base, if not please let me know which branch I
should use.
Original cover letter:
These patches were part of 'honor isolcpus configuration' [1] series. To
simplify the review process I decided to send this as separate series
because I think it's a nice cleanup independent of the isolcpus feature.
Signed-off-by: Daniel Wagner <wagi@kernel.org>
---
Changes in v5:
- dropped the irq_get_affinity callback from struct device_driver
again.
- s/blk_mq_hctx_map_queues/blk_mq_map_hw_queues/
- collected tags
- Link to v4: https://lore.kernel.org/r/20241113-refactor-blk-affinity-helpers-v4-0-dd3baa1e267f@kernel.org
Changes in v4:
- added irq_get_affinity callback to struct device_driver
- hisi_sas use dev pointer directly from hisi_hba.
- blk_mq_hctx_map_queues: ty irq_get_affinity callback
from device_driver first then from bus_type
- collected tags
- fixed typos
- Link to v3: https://lore.kernel.org/r/20241112-refactor-blk-affinity-helpers-v3-0-573bfca0cbd8@kernel.org
Changes in v3:
- dropped the additinal argument in blk_mq_hctx_map_queues.
leave open coded version in hisi_sas_v2.
- splitted "blk-mp: introduce blk_mq_hctx_map_queues" patch into
three patches.
- dropped local variable in pci_device_irq_get_affinity
- Link to v2: https://lore.kernel.org/r/20241111-refactor-blk-affinity-helpers-v2-0-f360ddad231a@kernel.org
Changes in v2:
- added new callback to struct bus_type and call directly the affinity
helpers from there.
- Link to v1: https://lore.kernel.org/r/20240913-refactor-blk-affinity-helpers-v1-0-8e058f77af12@suse.de
Changes in v1:
- renamed blk_mq_dev_map_queues to blk_mq_hctx_map_queues
- squased 'virito: add APIs for retrieving vq affinity' into
'blk-mq: introduce blk_mq_hctx_map_queues'
- moved hisi_sas changed into a new patch
- hisi_sas use define instead of hard coded value
- moved helpers into their matching subsystem, removed
blk-mq-pci and blk-mq-virtio files
- fix spelling/typos
- fixed long lines in docu (yep new lines in brief descriptions are
supported, tested ti)
- based on the first part of
[1] https://lore.kernel.org/all/20240806-isolcpus-io-queues-v3-0-da0eecfeaf8b@suse.de
---
Daniel Wagner (8):
driver core: bus: add irq_get_affinity callback to bus_type
PCI: hookup irq_get_affinity callback
virtio: hookup irq_get_affinity callback
blk-mq: introduce blk_mq_map_hw_queues
scsi: replace blk_mq_pci_map_queues with blk_mq_map_hw_queues
nvme: replace blk_mq_pci_map_queues with blk_mq_map_hw_queues
virtio: blk/scsi: replace blk_mq_virtio_map_queues with blk_mq_map_hw_queues
blk-mq: remove unused queue mapping helpers
block/Makefile | 2 --
block/blk-mq-cpumap.c | 37 +++++++++++++++++++++++++
block/blk-mq-pci.c | 46 -------------------------------
block/blk-mq-virtio.c | 46 -------------------------------
drivers/block/virtio_blk.c | 4 +--
drivers/nvme/host/fc.c | 1 -
drivers/nvme/host/pci.c | 3 +-
drivers/pci/pci-driver.c | 14 ++++++++++
drivers/scsi/fnic/fnic_main.c | 3 +-
drivers/scsi/hisi_sas/hisi_sas.h | 1 -
drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 4 +--
drivers/scsi/megaraid/megaraid_sas_base.c | 3 +-
drivers/scsi/mpi3mr/mpi3mr.h | 1 -
drivers/scsi/mpi3mr/mpi3mr_os.c | 2 +-
drivers/scsi/mpt3sas/mpt3sas_scsih.c | 3 +-
drivers/scsi/pm8001/pm8001_init.c | 2 +-
drivers/scsi/pm8001/pm8001_sas.h | 1 -
drivers/scsi/qla2xxx/qla_nvme.c | 3 +-
drivers/scsi/qla2xxx/qla_os.c | 4 +--
drivers/scsi/smartpqi/smartpqi_init.c | 7 ++---
drivers/scsi/virtio_scsi.c | 3 +-
drivers/virtio/virtio.c | 19 +++++++++++++
include/linux/blk-mq-pci.h | 11 --------
include/linux/blk-mq-virtio.h | 11 --------
include/linux/blk-mq.h | 2 ++
include/linux/device/bus.h | 3 ++
26 files changed, 92 insertions(+), 144 deletions(-)
---
base-commit: c9af98a7e8af266bae73e9d662b8341da1ec5824
change-id: 20240912-refactor-blk-affinity-helpers-7089b95b4b10
Best regards,
--
Daniel Wagner <wagi@kernel.org>