[PATCH v3 0/4] scsi/block: NUMA-local scan allocations, shared-tag path cleanup, and SCSI I/O counters

[PATCH v3 0/4] scsi/block: NUMA-local scan allocations, shared-tag path cleanup, and SCSI I/O counters

[Sumit Saxena]

This series contains three performance improvements targeting the SCSI
and block layers on multi-socket NUMA and heavily loaded SMP systems.

On multi-socket NUMA systems we observed extreme I/O throughput variance
of 50-60% between runs.  This series identifies and fixes two root causes:
cross-node memory accesses due to NUMA-unaware allocations in the scan
path, and false sharing between hot atomic counters in struct request_queue
and struct scsi_device.

Performance notes:

Tested on a dual-socket NUMA system (2x 32-core, 256 GB/socket) with
an mpi3mr HBA, running fio (random read, 4K, QD 64, 16 jobs, 60 s,
direct I/O).  IOPS figures are in KIOPS (thousands of IOPS):

  Configuration                    Avg KIOPS   Range (KIOPS)   Spread
  Baseline                         6,255       4,200 - 6,700   ~37%
  Baseline + all patches           7,350       7,000 - 7,700    ~10%

Key findings:

These patches combinedly reduces the observed 50-60% run-to-run variance
to under 10%, significantly improving workload predictability and
improves IOPs by 16-18%.

No functional regressions observed.

Changes in v3
-------------
-Handled feedback from Bart Van Assche and John Garry.
-Added a patch for shost local NUMA allocation.
-Converted ioerr_cnt and iotmo_cnt atomic counters into per-cpu counters. 

Changes in v2
--------------

  Patch 1 — Same functional goal as v1 patch 1: NUMA-local scsi_device /
  scsi_target allocations in the scan path so steady-state I/O does not
  habitually touch remote memory when the host has a fixed DMA/NUMA
  affinity.

  Patch 2 — Replaces v1’s ____cacheline_aligned_in_smp on
  nr_active_requests_shared_tags with removal of the shared-tag fairness
  throttling machinery (including hctx_may_queue(), blk_mq_hw_ctx.nr_active,
  and request_queue.nr_active_requests_shared_tags and their updates).
  This follows the earlier standalone proposal by Bart Van Assche [1],
  rebased for the current tree; it removes the high-frequency atomic
  accounting that motivated the v1 false-sharing workaround and, in our
  testing, improves IOPS on the order of roughly 16–18% for the shared-tag
  workload exercised.

  Patch 3 — Replaces v1’s cache-line padding of iodone_cnt with
  percpu_counter for both iorequest_cnt and iodone_cnt, so submission and
  completion paths mostly update CPU-local state instead of bouncing a
  single cache line, without inflating struct scsi_device for SMP
  alignment.

Merge / review hints
--------------------

Patch 3 touches the block layer and should have block maintainer review;
rest of patches are SCSI-oriented.  Please route or Ack as your subsystem
workflow requires.

Bart Van Assche (1):
  block: drop shared-tag fairness throttling

James Rizzo (1):
  scsi: scan: allocate sdev and starget on the NUMA node of the host
    adapter

Sumit Saxena (2):
  scsi: host: allocate struct Scsi_Host on the NUMA node of the host
    adapter
  scsi: use percpu counters for iostat counters in struct scsi_device

 block/blk-core.c                          |   2 -
 block/blk-mq-debugfs.c                    |  22 ++++-
 block/blk-mq-tag.c                        |   4 -
 block/blk-mq.c                            |  17 +---
 block/blk-mq.h                            | 100 ----------------------
 drivers/scsi/3w-9xxx.c                    |   2 +-
 drivers/scsi/3w-sas.c                     |   2 +-
 drivers/scsi/3w-xxxx.c                    |   2 +-
 drivers/scsi/53c700.c                     |   2 +-
 drivers/scsi/BusLogic.c                   |   2 +-
 drivers/scsi/a100u2w.c                    |   2 +-
 drivers/scsi/a2091.c                      |   2 +-
 drivers/scsi/a3000.c                      |   2 +-
 drivers/scsi/aacraid/linit.c              |   2 +-
 drivers/scsi/advansys.c                   |   6 +-
 drivers/scsi/aha152x.c                    |   2 +-
 drivers/scsi/aha1542.c                    |   2 +-
 drivers/scsi/aha1740.c                    |   2 +-
 drivers/scsi/aic7xxx/aic79xx_osm.c        |   2 +-
 drivers/scsi/aic7xxx/aic7xxx_osm.c        |   2 +-
 drivers/scsi/aic94xx/aic94xx_init.c       |   2 +-
 drivers/scsi/am53c974.c                   |   2 +-
 drivers/scsi/arcmsr/arcmsr_hba.c          |   3 +-
 drivers/scsi/arm/acornscsi.c              |   2 +-
 drivers/scsi/arm/arxescsi.c               |   2 +-
 drivers/scsi/arm/cumana_1.c               |   2 +-
 drivers/scsi/arm/cumana_2.c               |   2 +-
 drivers/scsi/arm/eesox.c                  |   2 +-
 drivers/scsi/arm/oak.c                    |   2 +-
 drivers/scsi/arm/powertec.c               |   2 +-
 drivers/scsi/atari_scsi.c                 |   2 +-
 drivers/scsi/atp870u.c                    |   2 +-
 drivers/scsi/bfa/bfad_im.c                |   2 +-
 drivers/scsi/csiostor/csio_init.c         |   4 +-
 drivers/scsi/dc395x.c                     |   2 +-
 drivers/scsi/dmx3191d.c                   |   2 +-
 drivers/scsi/elx/efct/efct_xport.c        |   4 +-
 drivers/scsi/esas2r/esas2r_main.c         |   2 +-
 drivers/scsi/fdomain.c                    |   2 +-
 drivers/scsi/fnic/fnic_main.c             |   2 +-
 drivers/scsi/g_NCR5380.c                  |   2 +-
 drivers/scsi/gvp11.c                      |   2 +-
 drivers/scsi/hisi_sas/hisi_sas_main.c     |   2 +-
 drivers/scsi/hisi_sas/hisi_sas_v3_hw.c    |   2 +-
 drivers/scsi/hosts.c                      |   6 +-
 drivers/scsi/hpsa.c                       |   2 +-
 drivers/scsi/hptiop.c                     |   2 +-
 drivers/scsi/ibmvscsi/ibmvfc.c            |   2 +-
 drivers/scsi/ibmvscsi/ibmvscsi.c          |   2 +-
 drivers/scsi/imm.c                        |   2 +-
 drivers/scsi/initio.c                     |   2 +-
 drivers/scsi/ipr.c                        |   2 +-
 drivers/scsi/ips.c                        |   2 +-
 drivers/scsi/isci/init.c                  |   2 +-
 drivers/scsi/jazz_esp.c                   |   2 +-
 drivers/scsi/libiscsi.c                   |   2 +-
 drivers/scsi/lpfc/lpfc_init.c             |   2 +-
 drivers/scsi/mac53c94.c                   |   2 +-
 drivers/scsi/mac_esp.c                    |   2 +-
 drivers/scsi/mac_scsi.c                   |   2 +-
 drivers/scsi/megaraid.c                   |   2 +-
 drivers/scsi/megaraid/megaraid_mbox.c     |   2 +-
 drivers/scsi/megaraid/megaraid_sas_base.c |   2 +-
 drivers/scsi/mesh.c                       |   2 +-
 drivers/scsi/mpi3mr/mpi3mr_os.c           |   2 +-
 drivers/scsi/mpt3sas/mpt3sas_scsih.c      |   4 +-
 drivers/scsi/mvme147.c                    |   2 +-
 drivers/scsi/mvsas/mv_init.c              |   2 +-
 drivers/scsi/mvumi.c                      |   2 +-
 drivers/scsi/myrb.c                       |   2 +-
 drivers/scsi/myrs.c                       |   2 +-
 drivers/scsi/ncr53c8xx.c                  |   2 +-
 drivers/scsi/nsp32.c                      |   2 +-
 drivers/scsi/pcmcia/nsp_cs.c              |   2 +-
 drivers/scsi/pcmcia/qlogic_stub.c         |   2 +-
 drivers/scsi/pcmcia/sym53c500_cs.c        |   2 +-
 drivers/scsi/pm8001/pm8001_init.c         |   2 +-
 drivers/scsi/pmcraid.c                    |   2 +-
 drivers/scsi/ppa.c                        |   2 +-
 drivers/scsi/ps3rom.c                     |   2 +-
 drivers/scsi/qla1280.c                    |   2 +-
 drivers/scsi/qla2xxx/qla_mid.c            |   2 +-
 drivers/scsi/qla2xxx/qla_os.c             |   2 +-
 drivers/scsi/qlogicfas.c                  |   2 +-
 drivers/scsi/qlogicpti.c                  |   2 +-
 drivers/scsi/scsi_debug.c                 |   2 +-
 drivers/scsi/scsi_error.c                 |   4 +-
 drivers/scsi/scsi_lib.c                   |  10 +--
 drivers/scsi/scsi_scan.c                  |  15 +++-
 drivers/scsi/scsi_sysfs.c                 |  23 +++--
 drivers/scsi/sd.c                         |   2 +-
 drivers/scsi/sgiwd93.c                    |   2 +-
 drivers/scsi/smartpqi/smartpqi_init.c     |   2 +-
 drivers/scsi/snic/snic_main.c             |   2 +-
 drivers/scsi/stex.c                       |   2 +-
 drivers/scsi/storvsc_drv.c                |   2 +-
 drivers/scsi/sun3_scsi.c                  |   2 +-
 drivers/scsi/sun3x_esp.c                  |   2 +-
 drivers/scsi/sun_esp.c                    |   2 +-
 drivers/scsi/sym53c8xx_2/sym_glue.c       |   2 +-
 drivers/scsi/virtio_scsi.c                |   2 +-
 drivers/scsi/vmw_pvscsi.c                 |   2 +-
 drivers/scsi/wd719x.c                     |   2 +-
 drivers/scsi/xen-scsifront.c              |   2 +-
 drivers/scsi/zorro_esp.c                  |   2 +-
 include/linux/blk-mq.h                    |   6 --
 include/linux/blkdev.h                    |   2 -
 include/scsi/libfc.h                      |   2 +-
 include/scsi/scsi_device.h                |   9 +-
 include/scsi/scsi_host.h                  |   3 +-
 110 files changed, 168 insertions(+), 258 deletions(-)

-- 
2.43.7

[PATCH v3 1/4] scsi: scan: allocate sdev and starget on the NUMA node of the host adapter

[Sumit Saxena]

From: James Rizzo <james.rizzo@broadcom.com>

When a host adapter is attached to a specific NUMA node, allocating
scsi_device and scsi_target via kzalloc() may place them on a remote
node.  All hot-path I/O accesses to these structures then cross the NUMA
interconnect, adding latency and consuming inter-node bandwidth.

Use kzalloc_node() with dev_to_node(shost->dma_dev) so allocations land
on the same node as the HBA, reducing cross-node traffic and improving
I/O performance on NUMA systems.

Signed-off-by: James Rizzo <james.rizzo@broadcom.com>
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
---
 drivers/scsi/scsi_scan.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index e27da038603a..121a14d5fdb8 100644
--- a/drivers/scsi/scsi_scan.c
+++ b/drivers/scsi/scsi_scan.c
@@ -34,6 +34,7 @@
 #include <linux/kthread.h>
 #include <linux/spinlock.h>
 #include <linux/async.h>
+#include <linux/topology.h>
 #include <linux/slab.h>
 #include <linux/unaligned.h>
 
@@ -287,8 +288,8 @@ static struct scsi_device *scsi_alloc_sdev(struct scsi_target *starget,
 	struct Scsi_Host *shost = dev_to_shost(starget->dev.parent);
 	struct queue_limits lim;
 
-	sdev = kzalloc(sizeof(*sdev) + shost->transportt->device_size,
-		       GFP_KERNEL);
+	sdev = kzalloc_node(sizeof(*sdev) + shost->transportt->device_size,
+		       GFP_KERNEL, dev_to_node(shost->dma_dev));
 	if (!sdev)
 		goto out;
 
@@ -502,7 +503,7 @@ static struct scsi_target *scsi_alloc_target(struct device *parent,
 	struct scsi_target *found_target;
 	int error, ref_got;
 
-	starget = kzalloc(size, GFP_KERNEL);
+	starget = kzalloc_node(size, GFP_KERNEL, dev_to_node(shost->dma_dev));
 	if (!starget) {
 		printk(KERN_ERR "%s: allocation failure\n", __func__);
 		return NULL;
-- 
2.43.7

[PATCH v3 2/4] scsi: host: allocate struct Scsi_Host on the NUMA node of the host adapter

[Sumit Saxena]

scsi_host_alloc() used kzalloc(), which always picks an arbitrary node.
Extend the function to accept a 'struct device *dev' parameter and use
kzalloc_node() with dev_to_node(dev) so the Scsi_Host struct lands on
the same NUMA node as the HBA, mirroring the treatment already applied
to struct scsi_device, struct scsi_target, and shost_data.

When dev is NULL (legacy ISA/platform drivers without a dma_dev) the
allocation falls back to NUMA_NO_NODE, preserving existing behaviour.

Update all in-tree callers:
  - PCI-based HBA drivers pass &pdev->dev (or the equivalent struct
    member such as &phba->pcidev->dev, &h->pdev->dev, &ha->pdev->dev)
    so their host struct is placed on the adapter's node.
  - Non-PCI drivers (ISA, Amiga, ARM PCMCIA, virtio, Hyper-V, PS3, …)
    pass NULL.
  - libfc's libfc_host_alloc() inline helper passes NULL; FC drivers
    that want NUMA awareness can open-code the call with their pdev.

Suggested-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
---
 drivers/scsi/3w-9xxx.c                    | 2 +-
 drivers/scsi/3w-sas.c                     | 2 +-
 drivers/scsi/3w-xxxx.c                    | 2 +-
 drivers/scsi/53c700.c                     | 2 +-
 drivers/scsi/BusLogic.c                   | 2 +-
 drivers/scsi/a100u2w.c                    | 2 +-
 drivers/scsi/a2091.c                      | 2 +-
 drivers/scsi/a3000.c                      | 2 +-
 drivers/scsi/aacraid/linit.c              | 2 +-
 drivers/scsi/advansys.c                   | 6 +++---
 drivers/scsi/aha152x.c                    | 2 +-
 drivers/scsi/aha1542.c                    | 2 +-
 drivers/scsi/aha1740.c                    | 2 +-
 drivers/scsi/aic7xxx/aic79xx_osm.c        | 2 +-
 drivers/scsi/aic7xxx/aic7xxx_osm.c        | 2 +-
 drivers/scsi/aic94xx/aic94xx_init.c       | 2 +-
 drivers/scsi/am53c974.c                   | 2 +-
 drivers/scsi/arcmsr/arcmsr_hba.c          | 3 ++-
 drivers/scsi/arm/acornscsi.c              | 2 +-
 drivers/scsi/arm/arxescsi.c               | 2 +-
 drivers/scsi/arm/cumana_1.c               | 2 +-
 drivers/scsi/arm/cumana_2.c               | 2 +-
 drivers/scsi/arm/eesox.c                  | 2 +-
 drivers/scsi/arm/oak.c                    | 2 +-
 drivers/scsi/arm/powertec.c               | 2 +-
 drivers/scsi/atari_scsi.c                 | 2 +-
 drivers/scsi/atp870u.c                    | 2 +-
 drivers/scsi/bfa/bfad_im.c                | 2 +-
 drivers/scsi/csiostor/csio_init.c         | 4 ++--
 drivers/scsi/dc395x.c                     | 2 +-
 drivers/scsi/dmx3191d.c                   | 2 +-
 drivers/scsi/elx/efct/efct_xport.c        | 4 ++--
 drivers/scsi/esas2r/esas2r_main.c         | 2 +-
 drivers/scsi/fdomain.c                    | 2 +-
 drivers/scsi/fnic/fnic_main.c             | 2 +-
 drivers/scsi/g_NCR5380.c                  | 2 +-
 drivers/scsi/gvp11.c                      | 2 +-
 drivers/scsi/hisi_sas/hisi_sas_main.c     | 2 +-
 drivers/scsi/hisi_sas/hisi_sas_v3_hw.c    | 2 +-
 drivers/scsi/hosts.c                      | 6 ++++--
 drivers/scsi/hpsa.c                       | 2 +-
 drivers/scsi/hptiop.c                     | 2 +-
 drivers/scsi/ibmvscsi/ibmvfc.c            | 2 +-
 drivers/scsi/ibmvscsi/ibmvscsi.c          | 2 +-
 drivers/scsi/imm.c                        | 2 +-
 drivers/scsi/initio.c                     | 2 +-
 drivers/scsi/ipr.c                        | 2 +-
 drivers/scsi/ips.c                        | 2 +-
 drivers/scsi/isci/init.c                  | 2 +-
 drivers/scsi/jazz_esp.c                   | 2 +-
 drivers/scsi/libiscsi.c                   | 2 +-
 drivers/scsi/lpfc/lpfc_init.c             | 2 +-
 drivers/scsi/mac53c94.c                   | 2 +-
 drivers/scsi/mac_esp.c                    | 2 +-
 drivers/scsi/mac_scsi.c                   | 2 +-
 drivers/scsi/megaraid.c                   | 2 +-
 drivers/scsi/megaraid/megaraid_mbox.c     | 2 +-
 drivers/scsi/megaraid/megaraid_sas_base.c | 2 +-
 drivers/scsi/mesh.c                       | 2 +-
 drivers/scsi/mpi3mr/mpi3mr_os.c           | 2 +-
 drivers/scsi/mpt3sas/mpt3sas_scsih.c      | 4 ++--
 drivers/scsi/mvme147.c                    | 2 +-
 drivers/scsi/mvsas/mv_init.c              | 2 +-
 drivers/scsi/mvumi.c                      | 2 +-
 drivers/scsi/myrb.c                       | 2 +-
 drivers/scsi/myrs.c                       | 2 +-
 drivers/scsi/ncr53c8xx.c                  | 2 +-
 drivers/scsi/nsp32.c                      | 2 +-
 drivers/scsi/pcmcia/nsp_cs.c              | 2 +-
 drivers/scsi/pcmcia/qlogic_stub.c         | 2 +-
 drivers/scsi/pcmcia/sym53c500_cs.c        | 2 +-
 drivers/scsi/pm8001/pm8001_init.c         | 2 +-
 drivers/scsi/pmcraid.c                    | 2 +-
 drivers/scsi/ppa.c                        | 2 +-
 drivers/scsi/ps3rom.c                     | 2 +-
 drivers/scsi/qla1280.c                    | 2 +-
 drivers/scsi/qla2xxx/qla_mid.c            | 2 +-
 drivers/scsi/qla2xxx/qla_os.c             | 2 +-
 drivers/scsi/qlogicfas.c                  | 2 +-
 drivers/scsi/qlogicpti.c                  | 2 +-
 drivers/scsi/scsi_debug.c                 | 2 +-
 drivers/scsi/sgiwd93.c                    | 2 +-
 drivers/scsi/smartpqi/smartpqi_init.c     | 2 +-
 drivers/scsi/snic/snic_main.c             | 2 +-
 drivers/scsi/stex.c                       | 2 +-
 drivers/scsi/storvsc_drv.c                | 2 +-
 drivers/scsi/sun3_scsi.c                  | 2 +-
 drivers/scsi/sun3x_esp.c                  | 2 +-
 drivers/scsi/sun_esp.c                    | 2 +-
 drivers/scsi/sym53c8xx_2/sym_glue.c       | 2 +-
 drivers/scsi/virtio_scsi.c                | 2 +-
 drivers/scsi/vmw_pvscsi.c                 | 2 +-
 drivers/scsi/wd719x.c                     | 2 +-
 drivers/scsi/xen-scsifront.c              | 2 +-
 drivers/scsi/zorro_esp.c                  | 2 +-
 include/scsi/libfc.h                      | 2 +-
 include/scsi/scsi_host.h                  | 3 ++-
 97 files changed, 107 insertions(+), 103 deletions(-)

diff --git a/drivers/scsi/3w-9xxx.c b/drivers/scsi/3w-9xxx.c
index 9b93a2440af8..444578ee8070 100644
--- a/drivers/scsi/3w-9xxx.c
+++ b/drivers/scsi/3w-9xxx.c
@@ -2021,7 +2021,7 @@ static int twa_probe(struct pci_dev *pdev, const struct pci_device_id *dev_id)
 		goto out_disable_device;
 	}
 
-	host = scsi_host_alloc(&driver_template, sizeof(TW_Device_Extension));
+	host = scsi_host_alloc(&driver_template, sizeof(TW_Device_Extension), &pdev->dev);
 	if (!host) {
 		TW_PRINTK(host, TW_DRIVER, 0x24, "Failed to allocate memory for device extension");
 		retval = -ENOMEM;
diff --git a/drivers/scsi/3w-sas.c b/drivers/scsi/3w-sas.c
index 52dc1aa639f7..d063d39faf4f 100644
--- a/drivers/scsi/3w-sas.c
+++ b/drivers/scsi/3w-sas.c
@@ -1576,7 +1576,7 @@ static int twl_probe(struct pci_dev *pdev, const struct pci_device_id *dev_id)
 		goto out_disable_device;
 	}
 
-	host = scsi_host_alloc(&driver_template, sizeof(TW_Device_Extension));
+	host = scsi_host_alloc(&driver_template, sizeof(TW_Device_Extension), &pdev->dev);
 	if (!host) {
 		TW_PRINTK(host, TW_DRIVER, 0x19, "Failed to allocate memory for device extension");
 		retval = -ENOMEM;
diff --git a/drivers/scsi/3w-xxxx.c b/drivers/scsi/3w-xxxx.c
index c68678fa72c1..0ccb5f1f8805 100644
--- a/drivers/scsi/3w-xxxx.c
+++ b/drivers/scsi/3w-xxxx.c
@@ -2268,7 +2268,7 @@ static int tw_probe(struct pci_dev *pdev, const struct pci_device_id *dev_id)
 		goto out_disable_device;
 	}
 
-	host = scsi_host_alloc(&driver_template, sizeof(TW_Device_Extension));
+	host = scsi_host_alloc(&driver_template, sizeof(TW_Device_Extension), &pdev->dev);
 	if (!host) {
 		printk(KERN_WARNING "3w-xxxx: Failed to allocate memory for device extension.");
 		retval = -ENOMEM;
diff --git a/drivers/scsi/53c700.c b/drivers/scsi/53c700.c
index c78f74b8f45c..e30d55ab5dea 100644
--- a/drivers/scsi/53c700.c
+++ b/drivers/scsi/53c700.c
@@ -341,7 +341,7 @@ NCR_700_detect(struct scsi_host_template *tpnt,
 	if(tpnt->proc_name == NULL)
 		tpnt->proc_name = "53c700";
 
-	host = scsi_host_alloc(tpnt, 4);
+	host = scsi_host_alloc(tpnt, 4, NULL);
 	if (!host)
 		return NULL;
 	memset(hostdata->slots, 0, sizeof(struct NCR_700_command_slot)
diff --git a/drivers/scsi/BusLogic.c b/drivers/scsi/BusLogic.c
index 5304d2febd63..f865fdec4136 100644
--- a/drivers/scsi/BusLogic.c
+++ b/drivers/scsi/BusLogic.c
@@ -2302,7 +2302,7 @@ static int __init blogic_init(void)
 		 */
 
 		host = scsi_host_alloc(&blogic_template,
-				sizeof(struct blogic_adapter));
+				sizeof(struct blogic_adapter), NULL);
 		if (host == NULL) {
 			release_region(myadapter->io_addr,
 					myadapter->addr_count);
diff --git a/drivers/scsi/a100u2w.c b/drivers/scsi/a100u2w.c
index 4365b896f5c4..9124c6103902 100644
--- a/drivers/scsi/a100u2w.c
+++ b/drivers/scsi/a100u2w.c
@@ -1106,7 +1106,7 @@ static int inia100_probe_one(struct pci_dev *pdev,
 	bios = inw(port + 0x50);
 
 
-	shost = scsi_host_alloc(&inia100_template, sizeof(struct orc_host));
+	shost = scsi_host_alloc(&inia100_template, sizeof(struct orc_host), &pdev->dev);
 	if (!shost)
 		goto out_release_region;
 
diff --git a/drivers/scsi/a2091.c b/drivers/scsi/a2091.c
index 204448bfd04b..51effb2edefb 100644
--- a/drivers/scsi/a2091.c
+++ b/drivers/scsi/a2091.c
@@ -214,7 +214,7 @@ static int a2091_probe(struct zorro_dev *z, const struct zorro_device_id *ent)
 		return -EBUSY;
 
 	instance = scsi_host_alloc(&a2091_scsi_template,
-				   sizeof(struct a2091_hostdata));
+				   sizeof(struct a2091_hostdata), NULL);
 	if (!instance) {
 		error = -ENOMEM;
 		goto fail_alloc;
diff --git a/drivers/scsi/a3000.c b/drivers/scsi/a3000.c
index bf054dd7682b..5b3d25b8ad37 100644
--- a/drivers/scsi/a3000.c
+++ b/drivers/scsi/a3000.c
@@ -235,7 +235,7 @@ static int __init amiga_a3000_scsi_probe(struct platform_device *pdev)
 		return -EBUSY;
 
 	instance = scsi_host_alloc(&amiga_a3000_scsi_template,
-				   sizeof(struct a3000_hostdata));
+				   sizeof(struct a3000_hostdata), NULL);
 	if (!instance) {
 		error = -ENOMEM;
 		goto fail_alloc;
diff --git a/drivers/scsi/aacraid/linit.c b/drivers/scsi/aacraid/linit.c
index 2fa8f7ddb703..d003667007f7 100644
--- a/drivers/scsi/aacraid/linit.c
+++ b/drivers/scsi/aacraid/linit.c
@@ -1636,7 +1636,7 @@ static int aac_probe_one(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	pci_set_master(pdev);
 
-	shost = scsi_host_alloc(&aac_driver_template, sizeof(struct aac_dev));
+	shost = scsi_host_alloc(&aac_driver_template, sizeof(struct aac_dev), &pdev->dev);
 	if (!shost) {
 		error = -ENOMEM;
 		goto out_disable_pdev;
diff --git a/drivers/scsi/advansys.c b/drivers/scsi/advansys.c
index 5cdbf2bdb13d..e7ef433778a1 100644
--- a/drivers/scsi/advansys.c
+++ b/drivers/scsi/advansys.c
@@ -11237,7 +11237,7 @@ static int advansys_vlb_probe(struct device *dev, unsigned int id)
 		goto release_region;
 
 	err = -ENOMEM;
-	shost = scsi_host_alloc(&advansys_template, sizeof(*board));
+	shost = scsi_host_alloc(&advansys_template, sizeof(*board), NULL);
 	if (!shost)
 		goto release_region;
 
@@ -11345,7 +11345,7 @@ static int advansys_eisa_probe(struct device *dev)
 			irq = advansys_eisa_irq_no(edev);
 
 		err = -ENOMEM;
-		shost = scsi_host_alloc(&advansys_template, sizeof(*board));
+		shost = scsi_host_alloc(&advansys_template, sizeof(*board), NULL);
 		if (!shost)
 			goto release_region;
 
@@ -11462,7 +11462,7 @@ static int advansys_pci_probe(struct pci_dev *pdev,
 	ioport = pci_resource_start(pdev, 0);
 
 	err = -ENOMEM;
-	shost = scsi_host_alloc(&advansys_template, sizeof(*board));
+	shost = scsi_host_alloc(&advansys_template, sizeof(*board), &pdev->dev);
 	if (!shost)
 		goto release_region;
 
diff --git a/drivers/scsi/aha152x.c b/drivers/scsi/aha152x.c
index e3ccb6bb62c0..d82ce80de098 100644
--- a/drivers/scsi/aha152x.c
+++ b/drivers/scsi/aha152x.c
@@ -734,7 +734,7 @@ struct Scsi_Host *aha152x_probe_one(struct aha152x_setup *setup)
 {
 	struct Scsi_Host *shpnt;
 
-	shpnt = scsi_host_alloc(&aha152x_driver_template, sizeof(struct aha152x_hostdata));
+	shpnt = scsi_host_alloc(&aha152x_driver_template, sizeof(struct aha152x_hostdata), NULL);
 	if (!shpnt) {
 		printk(KERN_ERR "aha152x: scsi_host_alloc failed\n");
 		return NULL;
diff --git a/drivers/scsi/aha1542.c b/drivers/scsi/aha1542.c
index fd766282d4a4..1a109c850785 100644
--- a/drivers/scsi/aha1542.c
+++ b/drivers/scsi/aha1542.c
@@ -752,7 +752,7 @@ static struct Scsi_Host *aha1542_hw_init(const struct scsi_host_template *tpnt,
 	if (!request_region(base_io, AHA1542_REGION_SIZE, "aha1542"))
 		return NULL;
 
-	sh = scsi_host_alloc(tpnt, sizeof(struct aha1542_hostdata));
+	sh = scsi_host_alloc(tpnt, sizeof(struct aha1542_hostdata), NULL);
 	if (!sh)
 		goto release;
 	aha1542 = shost_priv(sh);
diff --git a/drivers/scsi/aha1740.c b/drivers/scsi/aha1740.c
index c435769359f2..31a52edf0748 100644
--- a/drivers/scsi/aha1740.c
+++ b/drivers/scsi/aha1740.c
@@ -583,7 +583,7 @@ static int aha1740_probe (struct device *dev)
 	printk(KERN_INFO "aha174x: Extended translation %sabled.\n",
 	       translation ? "en" : "dis");
 	shpnt = scsi_host_alloc(&aha1740_template,
-			      sizeof(struct aha1740_hostdata));
+			      sizeof(struct aha1740_hostdata), NULL);
 	if(shpnt == NULL)
 		goto err_release_region;
 
diff --git a/drivers/scsi/aic7xxx/aic79xx_osm.c b/drivers/scsi/aic7xxx/aic79xx_osm.c
index feb1707feb7e..76e30b0784b9 100644
--- a/drivers/scsi/aic7xxx/aic79xx_osm.c
+++ b/drivers/scsi/aic7xxx/aic79xx_osm.c
@@ -1214,7 +1214,7 @@ ahd_linux_register_host(struct ahd_softc *ahd, struct scsi_host_template *templa
 	int	retval;
 
 	template->name = ahd->description;
-	host = scsi_host_alloc(template, sizeof(struct ahd_softc *));
+	host = scsi_host_alloc(template, sizeof(struct ahd_softc *), NULL);
 	if (host == NULL)
 		return (ENOMEM);
 
diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm.c b/drivers/scsi/aic7xxx/aic7xxx_osm.c
index d93b522695eb..0169509abd76 100644
--- a/drivers/scsi/aic7xxx/aic7xxx_osm.c
+++ b/drivers/scsi/aic7xxx/aic7xxx_osm.c
@@ -1083,7 +1083,7 @@ ahc_linux_register_host(struct ahc_softc *ahc, struct scsi_host_template *templa
 	int	retval;
 
 	template->name = ahc->description;
-	host = scsi_host_alloc(template, sizeof(struct ahc_softc *));
+	host = scsi_host_alloc(template, sizeof(struct ahc_softc *), NULL);
 	if (host == NULL)
 		return -ENOMEM;
 
diff --git a/drivers/scsi/aic94xx/aic94xx_init.c b/drivers/scsi/aic94xx/aic94xx_init.c
index 4400a3661d90..1336e5e38f8d 100644
--- a/drivers/scsi/aic94xx/aic94xx_init.c
+++ b/drivers/scsi/aic94xx/aic94xx_init.c
@@ -704,7 +704,7 @@ static int asd_pci_probe(struct pci_dev *dev, const struct pci_device_id *id)
 
 	err = -ENOMEM;
 
-	shost = scsi_host_alloc(&aic94xx_sht, sizeof(void *));
+	shost = scsi_host_alloc(&aic94xx_sht, sizeof(void *), &dev->dev);
 	if (!shost)
 		goto Err;
 
diff --git a/drivers/scsi/am53c974.c b/drivers/scsi/am53c974.c
index f972a3c90a2f..4ca73e801232 100644
--- a/drivers/scsi/am53c974.c
+++ b/drivers/scsi/am53c974.c
@@ -388,7 +388,7 @@ static int pci_esp_probe_one(struct pci_dev *pdev,
 		goto fail_disable_device;
 	}
 
-	shost = scsi_host_alloc(hostt, sizeof(struct esp));
+	shost = scsi_host_alloc(hostt, sizeof(struct esp), &pdev->dev);
 	if (!shost) {
 		dev_printk(KERN_INFO, &pdev->dev,
 			   "failed to allocate scsi host\n");
diff --git a/drivers/scsi/arcmsr/arcmsr_hba.c b/drivers/scsi/arcmsr/arcmsr_hba.c
index 8aa948f06cac..f0cc59e756dc 100644
--- a/drivers/scsi/arcmsr/arcmsr_hba.c
+++ b/drivers/scsi/arcmsr/arcmsr_hba.c
@@ -1087,7 +1087,8 @@ static int arcmsr_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	if(error){
 		return -ENODEV;
 	}
-	host = scsi_host_alloc(&arcmsr_scsi_host_template, sizeof(struct AdapterControlBlock));
+	host = scsi_host_alloc(&arcmsr_scsi_host_template,
+			       sizeof(struct AdapterControlBlock), &pdev->dev);
 	if(!host){
     		goto pci_disable_dev;
 	}
diff --git a/drivers/scsi/arm/acornscsi.c b/drivers/scsi/arm/acornscsi.c
index 79d7d7336b6a..97e3db7e6a7c 100644
--- a/drivers/scsi/arm/acornscsi.c
+++ b/drivers/scsi/arm/acornscsi.c
@@ -2806,7 +2806,7 @@ static int acornscsi_probe(struct expansion_card *ec, const struct ecard_id *id)
 	if (ret)
 		goto out;
 
-	host = scsi_host_alloc(&acornscsi_template, sizeof(AS_Host));
+	host = scsi_host_alloc(&acornscsi_template, sizeof(AS_Host), NULL);
 	if (!host) {
 		ret = -ENOMEM;
 		goto out_release;
diff --git a/drivers/scsi/arm/arxescsi.c b/drivers/scsi/arm/arxescsi.c
index 925d0bd68aa5..32f0a3aefb44 100644
--- a/drivers/scsi/arm/arxescsi.c
+++ b/drivers/scsi/arm/arxescsi.c
@@ -272,7 +272,7 @@ static int arxescsi_probe(struct expansion_card *ec, const struct ecard_id *id)
 		goto out_region;
 	}
 
-	host = scsi_host_alloc(&arxescsi_template, sizeof(struct arxescsi_info));
+	host = scsi_host_alloc(&arxescsi_template, sizeof(struct arxescsi_info), NULL);
 	if (!host) {
 		ret = -ENOMEM;
 		goto out_region;
diff --git a/drivers/scsi/arm/cumana_1.c b/drivers/scsi/arm/cumana_1.c
index d1a2a22ffe8c..d47ff9353c1b 100644
--- a/drivers/scsi/arm/cumana_1.c
+++ b/drivers/scsi/arm/cumana_1.c
@@ -238,7 +238,7 @@ static int cumanascsi1_probe(struct expansion_card *ec,
 	if (ret)
 		goto out;
 
-	host = scsi_host_alloc(&cumanascsi_template, sizeof(struct NCR5380_hostdata));
+	host = scsi_host_alloc(&cumanascsi_template, sizeof(struct NCR5380_hostdata), NULL);
 	if (!host) {
 		ret = -ENOMEM;
 		goto out_release;
diff --git a/drivers/scsi/arm/cumana_2.c b/drivers/scsi/arm/cumana_2.c
index e460068f6834..e35afe3a1fe4 100644
--- a/drivers/scsi/arm/cumana_2.c
+++ b/drivers/scsi/arm/cumana_2.c
@@ -394,7 +394,7 @@ static int cumanascsi2_probe(struct expansion_card *ec,
 	}
 
 	host = scsi_host_alloc(&cumanascsi2_template,
-			       sizeof(struct cumanascsi2_info));
+			       sizeof(struct cumanascsi2_info), NULL);
 	if (!host) {
 		ret = -ENOMEM;
 		goto out_region;
diff --git a/drivers/scsi/arm/eesox.c b/drivers/scsi/arm/eesox.c
index 99be9da8757f..de4d457f8ce7 100644
--- a/drivers/scsi/arm/eesox.c
+++ b/drivers/scsi/arm/eesox.c
@@ -510,7 +510,7 @@ static int eesoxscsi_probe(struct expansion_card *ec, const struct ecard_id *id)
 	}
 
 	host = scsi_host_alloc(&eesox_template,
-			       sizeof(struct eesoxscsi_info));
+			       sizeof(struct eesoxscsi_info), NULL);
 	if (!host) {
 		ret = -ENOMEM;
 		goto out_region;
diff --git a/drivers/scsi/arm/oak.c b/drivers/scsi/arm/oak.c
index d69245007096..b2ff8616f963 100644
--- a/drivers/scsi/arm/oak.c
+++ b/drivers/scsi/arm/oak.c
@@ -126,7 +126,7 @@ static int oakscsi_probe(struct expansion_card *ec, const struct ecard_id *id)
 	if (ret)
 		goto out;
 
-	host = scsi_host_alloc(&oakscsi_template, sizeof(struct NCR5380_hostdata));
+	host = scsi_host_alloc(&oakscsi_template, sizeof(struct NCR5380_hostdata), NULL);
 	if (!host) {
 		ret = -ENOMEM;
 		goto release;
diff --git a/drivers/scsi/arm/powertec.c b/drivers/scsi/arm/powertec.c
index 823c65ff6c12..045f35e50eff 100644
--- a/drivers/scsi/arm/powertec.c
+++ b/drivers/scsi/arm/powertec.c
@@ -318,7 +318,7 @@ static int powertecscsi_probe(struct expansion_card *ec,
 	}
 
 	host = scsi_host_alloc(&powertecscsi_template,
-			       sizeof (struct powertec_info));
+			       sizeof(struct powertec_info), NULL);
 	if (!host) {
 		ret = -ENOMEM;
 		goto out_region;
diff --git a/drivers/scsi/atari_scsi.c b/drivers/scsi/atari_scsi.c
index 85055677666c..9a469cf3991f 100644
--- a/drivers/scsi/atari_scsi.c
+++ b/drivers/scsi/atari_scsi.c
@@ -785,7 +785,7 @@ static int __init atari_scsi_probe(struct platform_device *pdev)
 	}
 
 	instance = scsi_host_alloc(&atari_scsi_template,
-	                           sizeof(struct NCR5380_hostdata));
+				   sizeof(struct NCR5380_hostdata), NULL);
 	if (!instance) {
 		error = -ENOMEM;
 		goto fail_alloc;
diff --git a/drivers/scsi/atp870u.c b/drivers/scsi/atp870u.c
index 67459d81f479..57f0b4a11ba7 100644
--- a/drivers/scsi/atp870u.c
+++ b/drivers/scsi/atp870u.c
@@ -1579,7 +1579,7 @@ static int atp870u_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	pci_set_master(pdev);
 
 	err = -ENOMEM;
-	shpnt = scsi_host_alloc(&atp870u_template, sizeof(struct atp_unit));
+	shpnt = scsi_host_alloc(&atp870u_template, sizeof(struct atp_unit), &pdev->dev);
 	if (!shpnt)
 		goto release_region;
 
diff --git a/drivers/scsi/bfa/bfad_im.c b/drivers/scsi/bfa/bfad_im.c
index 97990b285e17..bd14aee64886 100644
--- a/drivers/scsi/bfa/bfad_im.c
+++ b/drivers/scsi/bfa/bfad_im.c
@@ -740,7 +740,7 @@ bfad_scsi_host_alloc(struct bfad_im_port_s *im_port, struct bfad_s *bfad)
 
 	sht->sg_tablesize = bfad->cfg_data.io_max_sge;
 
-	return scsi_host_alloc(sht, sizeof(struct bfad_im_port_pointer));
+	return scsi_host_alloc(sht, sizeof(struct bfad_im_port_pointer), NULL);
 }
 
 void
diff --git a/drivers/scsi/csiostor/csio_init.c b/drivers/scsi/csiostor/csio_init.c
index 238431524801..a4bf1ba03248 100644
--- a/drivers/scsi/csiostor/csio_init.c
+++ b/drivers/scsi/csiostor/csio_init.c
@@ -606,11 +606,11 @@ csio_shost_init(struct csio_hw *hw, struct device *dev,
 	if (dev == &hw->pdev->dev)
 		shost = scsi_host_alloc(
 				&csio_fcoe_shost_template,
-				sizeof(struct csio_lnode));
+				sizeof(struct csio_lnode), &hw->pdev->dev);
 	else
 		shost = scsi_host_alloc(
 				&csio_fcoe_shost_vport_template,
-				sizeof(struct csio_lnode));
+				sizeof(struct csio_lnode), &hw->pdev->dev);
 
 	if (!shost)
 		goto err;
diff --git a/drivers/scsi/dc395x.c b/drivers/scsi/dc395x.c
index 6183ce05d8cf..16adeac93aac 100644
--- a/drivers/scsi/dc395x.c
+++ b/drivers/scsi/dc395x.c
@@ -3984,7 +3984,7 @@ static int dc395x_init_one(struct pci_dev *dev, const struct pci_device_id *id)
 
 	/* allocate scsi host information (includes out adapter) */
 	scsi_host = scsi_host_alloc(&dc395x_driver_template,
-				    sizeof(struct AdapterCtlBlk));
+				    sizeof(struct AdapterCtlBlk), &dev->dev);
 	if (!scsi_host)
 		goto fail;
 
diff --git a/drivers/scsi/dmx3191d.c b/drivers/scsi/dmx3191d.c
index d6d091b2f3c7..8ba17e3eefe3 100644
--- a/drivers/scsi/dmx3191d.c
+++ b/drivers/scsi/dmx3191d.c
@@ -74,7 +74,7 @@ static int dmx3191d_probe_one(struct pci_dev *pdev,
 	}
 
 	shost = scsi_host_alloc(&dmx3191d_driver_template,
-			sizeof(struct NCR5380_hostdata));
+			sizeof(struct NCR5380_hostdata), &pdev->dev);
 	if (!shost)
 		goto out_release_region;       
 
diff --git a/drivers/scsi/elx/efct/efct_xport.c b/drivers/scsi/elx/efct/efct_xport.c
index 9dcaef6fc188..74ef76e00eb5 100644
--- a/drivers/scsi/elx/efct/efct_xport.c
+++ b/drivers/scsi/elx/efct/efct_xport.c
@@ -378,7 +378,7 @@ efct_scsi_new_device(struct efct *efct)
 	int error = 0;
 	struct efct_vport *vport = NULL;
 
-	shost = scsi_host_alloc(&efct_template, sizeof(*vport));
+	shost = scsi_host_alloc(&efct_template, sizeof(*vport), NULL);
 	if (!shost) {
 		efc_log_err(efct, "failed to allocate Scsi_Host struct\n");
 		return -ENOMEM;
@@ -902,7 +902,7 @@ efct_scsi_new_vport(struct efct *efct, struct device *dev)
 	int error = 0;
 	struct efct_vport *vport = NULL;
 
-	shost = scsi_host_alloc(&efct_template, sizeof(*vport));
+	shost = scsi_host_alloc(&efct_template, sizeof(*vport), NULL);
 	if (!shost) {
 		efc_log_err(efct, "failed to allocate Scsi_Host struct\n");
 		return NULL;
diff --git a/drivers/scsi/esas2r/esas2r_main.c b/drivers/scsi/esas2r/esas2r_main.c
index ada278c24c51..4aac1f6db5e9 100644
--- a/drivers/scsi/esas2r/esas2r_main.c
+++ b/drivers/scsi/esas2r/esas2r_main.c
@@ -382,7 +382,7 @@ static int esas2r_probe(struct pci_dev *pcid,
 		       "after pci_enable_device() enable_cnt: %d",
 		       pcid->enable_cnt.counter);
 
-	host = scsi_host_alloc(&driver_template, host_alloc_size);
+	host = scsi_host_alloc(&driver_template, host_alloc_size, &pcid->dev);
 	if (host == NULL) {
 		esas2r_log(ESAS2R_LOG_CRIT, "scsi_host_alloc() FAIL");
 		return -ENODEV;
diff --git a/drivers/scsi/fdomain.c b/drivers/scsi/fdomain.c
index 22fbb0222f07..66ba4551def8 100644
--- a/drivers/scsi/fdomain.c
+++ b/drivers/scsi/fdomain.c
@@ -537,7 +537,7 @@ struct Scsi_Host *fdomain_create(int base, int irq, int this_id,
 		return NULL;
 	}
 
-	sh = scsi_host_alloc(&fdomain_template, sizeof(struct fdomain));
+	sh = scsi_host_alloc(&fdomain_template, sizeof(struct fdomain), NULL);
 	if (!sh)
 		return NULL;
 
diff --git a/drivers/scsi/fnic/fnic_main.c b/drivers/scsi/fnic/fnic_main.c
index 24d62c0874ac..688d85bc3f01 100644
--- a/drivers/scsi/fnic/fnic_main.c
+++ b/drivers/scsi/fnic/fnic_main.c
@@ -847,7 +847,7 @@ static int fnic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		{
 			host =
 				scsi_host_alloc(&fnic_host_template,
-								sizeof(struct fnic *));
+								sizeof(struct fnic *), &pdev->dev);
 			if (!host) {
 				dev_err(&fnic->pdev->dev, "Unable to allocate scsi host\n");
 				err = -ENOMEM;
diff --git a/drivers/scsi/g_NCR5380.c b/drivers/scsi/g_NCR5380.c
index 270eae7ac427..8b9076d6a964 100644
--- a/drivers/scsi/g_NCR5380.c
+++ b/drivers/scsi/g_NCR5380.c
@@ -312,7 +312,7 @@ static int generic_NCR5380_init_one(const struct scsi_host_template *tpnt,
 		goto out_release;
 	}
 
-	instance = scsi_host_alloc(tpnt, sizeof(struct NCR5380_hostdata));
+	instance = scsi_host_alloc(tpnt, sizeof(struct NCR5380_hostdata), NULL);
 	if (instance == NULL) {
 		ret = -ENOMEM;
 		goto out_unmap;
diff --git a/drivers/scsi/gvp11.c b/drivers/scsi/gvp11.c
index 0420bfe9bd42..ad5052db5a2e 100644
--- a/drivers/scsi/gvp11.c
+++ b/drivers/scsi/gvp11.c
@@ -353,7 +353,7 @@ static int gvp11_probe(struct zorro_dev *z, const struct zorro_device_id *ent)
 		goto fail_check_or_alloc;
 
 	instance = scsi_host_alloc(&gvp11_scsi_template,
-				   sizeof(struct gvp11_hostdata));
+				   sizeof(struct gvp11_hostdata), NULL);
 	if (!instance) {
 		error = -ENOMEM;
 		goto fail_check_or_alloc;
diff --git a/drivers/scsi/hisi_sas/hisi_sas_main.c b/drivers/scsi/hisi_sas/hisi_sas_main.c
index 944ce19ae2fc..5696da8da6c7 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_main.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_main.c
@@ -2483,7 +2483,7 @@ static struct Scsi_Host *hisi_sas_shost_alloc(struct platform_device *pdev,
 	struct device *dev = &pdev->dev;
 	int error;
 
-	shost = scsi_host_alloc(hw->sht, sizeof(*hisi_hba));
+	shost = scsi_host_alloc(hw->sht, sizeof(*hisi_hba), NULL);
 	if (!shost) {
 		dev_err(dev, "scsi host alloc failed\n");
 		return NULL;
diff --git a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
index c7430f7c4048..44e584496ed5 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
@@ -3469,7 +3469,7 @@ hisi_sas_shost_alloc_pci(struct pci_dev *pdev)
 	struct hisi_hba *hisi_hba;
 	struct device *dev = &pdev->dev;
 
-	shost = scsi_host_alloc(&sht_v3_hw, sizeof(*hisi_hba));
+	shost = scsi_host_alloc(&sht_v3_hw, sizeof(*hisi_hba), &pdev->dev);
 	if (!shost) {
 		dev_err(dev, "shost alloc failed\n");
 		return NULL;
diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c
index e047747d4ecf..e1f42be79729 100644
--- a/drivers/scsi/hosts.c
+++ b/drivers/scsi/hosts.c
@@ -403,12 +403,14 @@ static const struct device_type scsi_host_type = {
  * Return value:
  * 	Pointer to a new Scsi_Host
  **/
-struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *sht, int privsize)
+struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *sht, int privsize,
+				  struct device *dev)
 {
 	struct Scsi_Host *shost;
 	int index;
 
-	shost = kzalloc(sizeof(struct Scsi_Host) + privsize, GFP_KERNEL);
+	shost = kzalloc_node(sizeof(struct Scsi_Host) + privsize, GFP_KERNEL,
+			     dev ? dev_to_node(dev) : NUMA_NO_NODE);
 	if (!shost)
 		return NULL;
 
diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c
index a1b116cd4723..b9f9f18bd985 100644
--- a/drivers/scsi/hpsa.c
+++ b/drivers/scsi/hpsa.c
@@ -5837,7 +5837,7 @@ static int hpsa_scsi_host_alloc(struct ctlr_info *h)
 {
 	struct Scsi_Host *sh;
 
-	sh = scsi_host_alloc(&hpsa_driver_template, sizeof(struct ctlr_info *));
+	sh = scsi_host_alloc(&hpsa_driver_template, sizeof(struct ctlr_info *), &h->pdev->dev);
 	if (sh == NULL) {
 		dev_err(&h->pdev->dev, "scsi_host_alloc failed\n");
 		return -ENOMEM;
diff --git a/drivers/scsi/hptiop.c b/drivers/scsi/hptiop.c
index 7083c14c5302..7d79357be265 100644
--- a/drivers/scsi/hptiop.c
+++ b/drivers/scsi/hptiop.c
@@ -1311,7 +1311,7 @@ static int hptiop_probe(struct pci_dev *pcidev, const struct pci_device_id *id)
 		goto disable_pci_device;
 	}
 
-	host = scsi_host_alloc(&driver_template, sizeof(struct hptiop_hba));
+	host = scsi_host_alloc(&driver_template, sizeof(struct hptiop_hba), &pcidev->dev);
 	if (!host) {
 		printk(KERN_ERR "hptiop: fail to alloc scsi host\n");
 		goto free_pci_regions;
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index 3dd2adda195e..b11d564a21d9 100644
--- a/drivers/scsi/ibmvscsi/ibmvfc.c
+++ b/drivers/scsi/ibmvscsi/ibmvfc.c
@@ -6325,7 +6325,7 @@ static int ibmvfc_probe(struct vio_dev *vdev, const struct vio_device_id *id)
 	unsigned int max_scsi_queues = min((unsigned int)IBMVFC_MAX_SCSI_QUEUES, online_cpus);
 
 	ENTER;
-	shost = scsi_host_alloc(&driver_template, sizeof(*vhost));
+	shost = scsi_host_alloc(&driver_template, sizeof(*vhost), NULL);
 	if (!shost) {
 		dev_err(dev, "Couldn't allocate host data\n");
 		goto out;
diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index 609bda730b3a..e8342e581246 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c
@@ -2235,7 +2235,7 @@ static int ibmvscsi_probe(struct vio_dev *vdev, const struct vio_device_id *id)
 
 	dev_set_drvdata(&vdev->dev, NULL);
 
-	host = scsi_host_alloc(&driver_template, sizeof(*hostdata));
+	host = scsi_host_alloc(&driver_template, sizeof(*hostdata), NULL);
 	if (!host) {
 		dev_err(&vdev->dev, "couldn't allocate host data\n");
 		goto scsi_host_alloc_failed;
diff --git a/drivers/scsi/imm.c b/drivers/scsi/imm.c
index 0535252e77e3..a6131f87fcaf 100644
--- a/drivers/scsi/imm.c
+++ b/drivers/scsi/imm.c
@@ -1221,7 +1221,7 @@ static int __imm_attach(struct parport *pb)
 	INIT_DELAYED_WORK(&dev->imm_tq, imm_interrupt);
 
 	err = -ENOMEM;
-	host = scsi_host_alloc(&imm_template, sizeof(imm_struct *));
+	host = scsi_host_alloc(&imm_template, sizeof(imm_struct *), NULL);
 	if (!host)
 		goto out1;
 	host->io_port = pb->base;
diff --git a/drivers/scsi/initio.c b/drivers/scsi/initio.c
index 06fbe85dccfa..294f7f8d5dbb 100644
--- a/drivers/scsi/initio.c
+++ b/drivers/scsi/initio.c
@@ -2824,7 +2824,7 @@ static int initio_probe_one(struct pci_dev *pdev,
 		error = -ENODEV;
 		goto out_disable_device;
 	}
-	shost = scsi_host_alloc(&initio_template, sizeof(struct initio_host));
+	shost = scsi_host_alloc(&initio_template, sizeof(struct initio_host), &pdev->dev);
 	if (!shost) {
 		printk(KERN_WARNING "initio: Could not allocate host structure.\n");
 		error = -ENOMEM;
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index d207e5e81afe..85608804ff39 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -9379,7 +9379,7 @@ static int ipr_probe_ioa(struct pci_dev *pdev,
 	ENTER;
 
 	dev_info(&pdev->dev, "Found IOA with IRQ: %d\n", pdev->irq);
-	host = scsi_host_alloc(&driver_template, sizeof(*ioa_cfg));
+	host = scsi_host_alloc(&driver_template, sizeof(*ioa_cfg), &pdev->dev);
 
 	if (!host) {
 		dev_err(&pdev->dev, "call to scsi_host_alloc failed!\n");
diff --git a/drivers/scsi/ips.c b/drivers/scsi/ips.c
index 41ed73966a48..709a2a799f3e 100644
--- a/drivers/scsi/ips.c
+++ b/drivers/scsi/ips.c
@@ -6638,7 +6638,7 @@ ips_register_scsi(int index)
 {
 	struct Scsi_Host *sh;
 	ips_ha_t *ha, *oldha = ips_ha[index];
-	sh = scsi_host_alloc(&ips_driver_template, sizeof (ips_ha_t));
+	sh = scsi_host_alloc(&ips_driver_template, sizeof(ips_ha_t), &oldha->pcidev->dev);
 	if (!sh) {
 		IPS_PRINTK(KERN_WARNING, oldha->pcidev,
 			   "Unable to register controller with SCSI subsystem\n");
diff --git a/drivers/scsi/isci/init.c b/drivers/scsi/isci/init.c
index acf0c2038d20..7da06ace20ad 100644
--- a/drivers/scsi/isci/init.c
+++ b/drivers/scsi/isci/init.c
@@ -538,7 +538,7 @@ static struct isci_host *isci_host_alloc(struct pci_dev *pdev, int id)
 		INIT_LIST_HEAD(&idev->node);
 	}
 
-	shost = scsi_host_alloc(&isci_sht, sizeof(void *));
+	shost = scsi_host_alloc(&isci_sht, sizeof(void *), &pdev->dev);
 	if (!shost)
 		return NULL;
 
diff --git a/drivers/scsi/jazz_esp.c b/drivers/scsi/jazz_esp.c
index 35137f5cfb3a..1817246e4cc6 100644
--- a/drivers/scsi/jazz_esp.c
+++ b/drivers/scsi/jazz_esp.c
@@ -110,7 +110,7 @@ static int esp_jazz_probe(struct platform_device *dev)
 	struct resource *res;
 	int err;
 
-	host = scsi_host_alloc(tpnt, sizeof(struct esp));
+	host = scsi_host_alloc(tpnt, sizeof(struct esp), NULL);
 
 	err = -ENOMEM;
 	if (!host)
diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
index 160f02f2f51d..458955dfc0aa 100644
--- a/drivers/scsi/libiscsi.c
+++ b/drivers/scsi/libiscsi.c
@@ -2903,7 +2903,7 @@ struct Scsi_Host *iscsi_host_alloc(const struct scsi_host_template *sht,
 	struct Scsi_Host *shost;
 	struct iscsi_host *ihost;
 
-	shost = scsi_host_alloc(sht, sizeof(struct iscsi_host) + dd_data_size);
+	shost = scsi_host_alloc(sht, sizeof(struct iscsi_host) + dd_data_size, NULL);
 	if (!shost)
 		return NULL;
 	ihost = shost_priv(shost);
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 82af59c913e9..25264866075f 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -4745,7 +4745,7 @@ lpfc_create_port(struct lpfc_hba *phba, int instance, struct device *dev)
 		template->sg_tablesize = lpfc_get_sg_tablesize(phba);
 	}
 
-	shost = scsi_host_alloc(template, sizeof(struct lpfc_vport));
+	shost = scsi_host_alloc(template, sizeof(struct lpfc_vport), &phba->pcidev->dev);
 	if (!shost)
 		goto out;
 
diff --git a/drivers/scsi/mac53c94.c b/drivers/scsi/mac53c94.c
index de2bd860b9d7..737e5f2fef6f 100644
--- a/drivers/scsi/mac53c94.c
+++ b/drivers/scsi/mac53c94.c
@@ -426,7 +426,7 @@ static int mac53c94_probe(struct macio_dev *mdev, const struct of_device_id *mat
 		return -EBUSY;
 	}
 
-       	host = scsi_host_alloc(&mac53c94_template, sizeof(struct fsc_state));
+	host = scsi_host_alloc(&mac53c94_template, sizeof(struct fsc_state), NULL);
 	if (host == NULL) {
 		printk(KERN_ERR "mac53c94: couldn't register host");
 		rc = -ENOMEM;
diff --git a/drivers/scsi/mac_esp.c b/drivers/scsi/mac_esp.c
index a0ceaa2428c2..c8652bfdb3b8 100644
--- a/drivers/scsi/mac_esp.c
+++ b/drivers/scsi/mac_esp.c
@@ -301,7 +301,7 @@ static int esp_mac_probe(struct platform_device *dev)
 	if (dev->id > 1)
 		return -ENODEV;
 
-	host = scsi_host_alloc(tpnt, sizeof(struct esp));
+	host = scsi_host_alloc(tpnt, sizeof(struct esp), NULL);
 
 	err = -ENOMEM;
 	if (!host)
diff --git a/drivers/scsi/mac_scsi.c b/drivers/scsi/mac_scsi.c
index a86bd839d08e..eeb00ee30aaa 100644
--- a/drivers/scsi/mac_scsi.c
+++ b/drivers/scsi/mac_scsi.c
@@ -474,7 +474,7 @@ static int __init mac_scsi_probe(struct platform_device *pdev)
 		mac_scsi_template.sg_tablesize = 1;
 
 	instance = scsi_host_alloc(&mac_scsi_template,
-	                           sizeof(struct NCR5380_hostdata));
+				   sizeof(struct NCR5380_hostdata), NULL);
 	if (!instance)
 		return -ENOMEM;
 
diff --git a/drivers/scsi/megaraid.c b/drivers/scsi/megaraid.c
index 9476a0d2c72d..701e54843193 100644
--- a/drivers/scsi/megaraid.c
+++ b/drivers/scsi/megaraid.c
@@ -4203,7 +4203,7 @@ megaraid_probe_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	}
 
 	/* Initialize SCSI Host structure */
-	host = scsi_host_alloc(&megaraid_template, sizeof(adapter_t));
+	host = scsi_host_alloc(&megaraid_template, sizeof(adapter_t), &pdev->dev);
 	if (!host)
 		goto out_iounmap;
 
diff --git a/drivers/scsi/megaraid/megaraid_mbox.c b/drivers/scsi/megaraid/megaraid_mbox.c
index ce89032a5a74..17b015b3d35f 100644
--- a/drivers/scsi/megaraid/megaraid_mbox.c
+++ b/drivers/scsi/megaraid/megaraid_mbox.c
@@ -620,7 +620,7 @@ megaraid_io_attach(adapter_t *adapter)
 	struct Scsi_Host	*host;
 
 	// Initialize SCSI Host structure
-	host = scsi_host_alloc(&megaraid_template_g, 8);
+	host = scsi_host_alloc(&megaraid_template_g, 8, &pdev->dev);
 	if (!host) {
 		con_log(CL_ANN, (KERN_WARNING
 			"megaraid mbox: scsi_host_alloc failed\n"));
diff --git a/drivers/scsi/megaraid/megaraid_sas_base.c b/drivers/scsi/megaraid/megaraid_sas_base.c
index ecd365d78ae3..bae1070371d5 100644
--- a/drivers/scsi/megaraid/megaraid_sas_base.c
+++ b/drivers/scsi/megaraid/megaraid_sas_base.c
@@ -7512,7 +7512,7 @@ static int megasas_probe_one(struct pci_dev *pdev,
 	pci_set_master(pdev);
 
 	host = scsi_host_alloc(&megasas_template,
-			       sizeof(struct megasas_instance));
+			       sizeof(struct megasas_instance), &pdev->dev);
 
 	if (!host) {
 		dev_printk(KERN_DEBUG, &pdev->dev, "scsi_host_alloc failed\n");
diff --git a/drivers/scsi/mesh.c b/drivers/scsi/mesh.c
index dc1402b321da..a4ba6bc49d23 100644
--- a/drivers/scsi/mesh.c
+++ b/drivers/scsi/mesh.c
@@ -1877,7 +1877,7 @@ static int mesh_probe(struct macio_dev *mdev, const struct of_device_id *match)
        		printk(KERN_ERR "mesh: unable to request memory resources");
 		return -EBUSY;
 	}
-       	mesh_host = scsi_host_alloc(&mesh_template, sizeof(struct mesh_state));
+	mesh_host = scsi_host_alloc(&mesh_template, sizeof(struct mesh_state), NULL);
 	if (mesh_host == NULL) {
 		printk(KERN_ERR "mesh: couldn't register host");
 		goto out_release;
diff --git a/drivers/scsi/mpi3mr/mpi3mr_os.c b/drivers/scsi/mpi3mr/mpi3mr_os.c
index 402d1f35d214..c74e2addc77d 100644
--- a/drivers/scsi/mpi3mr/mpi3mr_os.c
+++ b/drivers/scsi/mpi3mr/mpi3mr_os.c
@@ -5468,7 +5468,7 @@ mpi3mr_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	}
 
 	shost = scsi_host_alloc(&mpi3mr_driver_template,
-	    sizeof(struct mpi3mr_ioc));
+	    sizeof(struct mpi3mr_ioc), &pdev->dev);
 	if (!shost) {
 		retval = -ENODEV;
 		goto shost_failed;
diff --git a/drivers/scsi/mpt3sas/mpt3sas_scsih.c b/drivers/scsi/mpt3sas/mpt3sas_scsih.c
index 6ff788557294..06c8df6261d4 100644
--- a/drivers/scsi/mpt3sas/mpt3sas_scsih.c
+++ b/drivers/scsi/mpt3sas/mpt3sas_scsih.c
@@ -13367,7 +13367,7 @@ _scsih_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 			PCIE_LINK_STATE_L1 | PCIE_LINK_STATE_CLKPM);
 		/* Use mpt2sas driver host template for SAS 2.0 HBA's */
 		shost = scsi_host_alloc(&mpt2sas_driver_template,
-		  sizeof(struct MPT3SAS_ADAPTER));
+		  sizeof(struct MPT3SAS_ADAPTER), &pdev->dev);
 		if (!shost)
 			return -ENODEV;
 		ioc = shost_priv(shost);
@@ -13399,7 +13399,7 @@ _scsih_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	case MPI26_VERSION:
 		/* Use mpt3sas driver host template for SAS 3.0 HBA's */
 		shost = scsi_host_alloc(&mpt3sas_driver_template,
-		  sizeof(struct MPT3SAS_ADAPTER));
+		  sizeof(struct MPT3SAS_ADAPTER), &pdev->dev);
 		if (!shost)
 			return -ENODEV;
 		ioc = shost_priv(shost);
diff --git a/drivers/scsi/mvme147.c b/drivers/scsi/mvme147.c
index 98b99c0f5bc7..4d61e25de2cf 100644
--- a/drivers/scsi/mvme147.c
+++ b/drivers/scsi/mvme147.c
@@ -97,7 +97,7 @@ static int __init mvme147_init(void)
 		return 0;
 
 	mvme147_shost = scsi_host_alloc(&mvme147_host_template,
-			sizeof(struct WD33C93_hostdata));
+			sizeof(struct WD33C93_hostdata), NULL);
 	if (!mvme147_shost)
 		goto err_out;
 	mvme147_shost->base = 0xfffe4000;
diff --git a/drivers/scsi/mvsas/mv_init.c b/drivers/scsi/mvsas/mv_init.c
index 5abc17a2e261..fd90b5eec0b4 100644
--- a/drivers/scsi/mvsas/mv_init.c
+++ b/drivers/scsi/mvsas/mv_init.c
@@ -494,7 +494,7 @@ static int mvs_pci_init(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		goto err_out_regions;
 
-	shost = scsi_host_alloc(&mvs_sht, sizeof(void *));
+	shost = scsi_host_alloc(&mvs_sht, sizeof(void *), &pdev->dev);
 	if (!shost) {
 		rc = -ENOMEM;
 		goto err_out_regions;
diff --git a/drivers/scsi/mvumi.c b/drivers/scsi/mvumi.c
index e70d336b4ab3..d12b33a32a09 100644
--- a/drivers/scsi/mvumi.c
+++ b/drivers/scsi/mvumi.c
@@ -2468,7 +2468,7 @@ static int mvumi_probe_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	if (ret)
 		goto fail_set_dma_mask;
 
-	host = scsi_host_alloc(&mvumi_template, sizeof(*mhba));
+	host = scsi_host_alloc(&mvumi_template, sizeof(*mhba), &pdev->dev);
 	if (!host) {
 		dev_err(&pdev->dev, "scsi_host_alloc failed\n");
 		ret = -ENOMEM;
diff --git a/drivers/scsi/myrb.c b/drivers/scsi/myrb.c
index 3678b66310ed..f28c29b41cf6 100644
--- a/drivers/scsi/myrb.c
+++ b/drivers/scsi/myrb.c
@@ -3401,7 +3401,7 @@ static struct myrb_hba *myrb_detect(struct pci_dev *pdev,
 	struct Scsi_Host *shost;
 	struct myrb_hba *cb = NULL;
 
-	shost = scsi_host_alloc(&myrb_template, sizeof(struct myrb_hba));
+	shost = scsi_host_alloc(&myrb_template, sizeof(struct myrb_hba), &pdev->dev);
 	if (!shost) {
 		dev_err(&pdev->dev, "Unable to allocate Controller\n");
 		return NULL;
diff --git a/drivers/scsi/myrs.c b/drivers/scsi/myrs.c
index afd68225221a..a8ce488e6520 100644
--- a/drivers/scsi/myrs.c
+++ b/drivers/scsi/myrs.c
@@ -1937,7 +1937,7 @@ static struct myrs_hba *myrs_alloc_host(struct pci_dev *pdev,
 	struct Scsi_Host *shost;
 	struct myrs_hba *cs;
 
-	shost = scsi_host_alloc(&myrs_template, sizeof(struct myrs_hba));
+	shost = scsi_host_alloc(&myrs_template, sizeof(struct myrs_hba), &pdev->dev);
 	if (!shost)
 		return NULL;
 
diff --git a/drivers/scsi/ncr53c8xx.c b/drivers/scsi/ncr53c8xx.c
index 5369ca3fe4fd..009d4c55054e 100644
--- a/drivers/scsi/ncr53c8xx.c
+++ b/drivers/scsi/ncr53c8xx.c
@@ -8108,7 +8108,7 @@ struct Scsi_Host * __init ncr_attach(struct scsi_host_template *tpnt,
 	printk(KERN_INFO "ncr53c720-%d: rev 0x%x irq %d\n",
 		unit, device->chip.revision_id, device->slot.irq);
 
-	instance = scsi_host_alloc(tpnt, sizeof(*host_data));
+	instance = scsi_host_alloc(tpnt, sizeof(*host_data), NULL);
 	if (!instance)
 	        goto attach_error;
 	host_data = (struct host_data *) instance->hostdata;
diff --git a/drivers/scsi/nsp32.c b/drivers/scsi/nsp32.c
index e893d5677241..681e1d554657 100644
--- a/drivers/scsi/nsp32.c
+++ b/drivers/scsi/nsp32.c
@@ -2556,7 +2556,7 @@ static int nsp32_detect(struct pci_dev *pdev)
 	/*
 	 * register this HBA as SCSI device
 	 */
-	host = scsi_host_alloc(&nsp32_template, sizeof(nsp32_hw_data));
+	host = scsi_host_alloc(&nsp32_template, sizeof(nsp32_hw_data), &pdev->dev);
 	if (host == NULL) {
 		nsp32_msg (KERN_ERR, "failed to scsi register");
 		goto err;
diff --git a/drivers/scsi/pcmcia/nsp_cs.c b/drivers/scsi/pcmcia/nsp_cs.c
index ae70fda96ae9..32ca7872b7f8 100644
--- a/drivers/scsi/pcmcia/nsp_cs.c
+++ b/drivers/scsi/pcmcia/nsp_cs.c
@@ -1326,7 +1326,7 @@ static struct Scsi_Host *nsp_detect(struct scsi_host_template *sht)
 	nsp_hw_data *data_b = &nsp_data_base, *data;
 
 	nsp_dbg(NSP_DEBUG_INIT, "this_id=%d", sht->this_id);
-	host = scsi_host_alloc(&nsp_driver_template, sizeof(nsp_hw_data));
+	host = scsi_host_alloc(&nsp_driver_template, sizeof(nsp_hw_data), NULL);
 	if (host == NULL) {
 		nsp_dbg(NSP_DEBUG_INIT, "host failed");
 		return NULL;
diff --git a/drivers/scsi/pcmcia/qlogic_stub.c b/drivers/scsi/pcmcia/qlogic_stub.c
index 5d8a434d3f66..b417b39ab723 100644
--- a/drivers/scsi/pcmcia/qlogic_stub.c
+++ b/drivers/scsi/pcmcia/qlogic_stub.c
@@ -106,7 +106,7 @@ static struct Scsi_Host *qlogic_detect(struct scsi_host_template *host,
 	qlogicfas408_setup(qbase, qinitid, INT_TYPE);
 
 	host->name = qlogic_name;
-	shost = scsi_host_alloc(host, sizeof(struct qlogicfas408_priv));
+	shost = scsi_host_alloc(host, sizeof(struct qlogicfas408_priv), NULL);
 	if (!shost)
 		goto err;
 	shost->io_port = qbase;
diff --git a/drivers/scsi/pcmcia/sym53c500_cs.c b/drivers/scsi/pcmcia/sym53c500_cs.c
index 1530c1ad5d36..83aab6c69a62 100644
--- a/drivers/scsi/pcmcia/sym53c500_cs.c
+++ b/drivers/scsi/pcmcia/sym53c500_cs.c
@@ -752,7 +752,7 @@ SYM53C500_config(struct pcmcia_device *link)
 
 	chip_init(port_base);
 
-	host = scsi_host_alloc(tpnt, sizeof(struct sym53c500_data));
+	host = scsi_host_alloc(tpnt, sizeof(struct sym53c500_data), NULL);
 	if (!host) {
 		printk("SYM53C500: Unable to register host, giving up.\n");
 		goto err_release;
diff --git a/drivers/scsi/pm8001/pm8001_init.c b/drivers/scsi/pm8001/pm8001_init.c
index e93ea76b565e..873810c6853c 100644
--- a/drivers/scsi/pm8001/pm8001_init.c
+++ b/drivers/scsi/pm8001/pm8001_init.c
@@ -1142,7 +1142,7 @@ static int pm8001_pci_probe(struct pci_dev *pdev,
 	if (rc)
 		goto err_out_regions;
 
-	shost = scsi_host_alloc(&pm8001_sht, sizeof(void *));
+	shost = scsi_host_alloc(&pm8001_sht, sizeof(void *), &pdev->dev);
 	if (!shost) {
 		rc = -ENOMEM;
 		goto err_out_regions;
diff --git a/drivers/scsi/pmcraid.c b/drivers/scsi/pmcraid.c
index 942a99393204..a26c747806ef 100644
--- a/drivers/scsi/pmcraid.c
+++ b/drivers/scsi/pmcraid.c
@@ -5236,7 +5236,7 @@ static int pmcraid_probe(struct pci_dev *pdev,
 	}
 
 	host = scsi_host_alloc(&pmcraid_host_template,
-				sizeof(struct pmcraid_instance));
+				sizeof(struct pmcraid_instance), &pdev->dev);
 
 	if (!host) {
 		dev_err(&pdev->dev, "scsi_host_alloc failed!\n");
diff --git a/drivers/scsi/ppa.c b/drivers/scsi/ppa.c
index 8a4e910d5758..40fe9c6acc3b 100644
--- a/drivers/scsi/ppa.c
+++ b/drivers/scsi/ppa.c
@@ -1101,7 +1101,7 @@ static int __ppa_attach(struct parport *pb)
 	INIT_DELAYED_WORK(&dev->ppa_tq, ppa_interrupt);
 
 	err = -ENOMEM;
-	host = scsi_host_alloc(&ppa_template, sizeof(ppa_struct *));
+	host = scsi_host_alloc(&ppa_template, sizeof(ppa_struct *), NULL);
 	if (!host)
 		goto out1;
 	host->io_port = pb->base;
diff --git a/drivers/scsi/ps3rom.c b/drivers/scsi/ps3rom.c
index a9c727d22931..3542a35b137e 100644
--- a/drivers/scsi/ps3rom.c
+++ b/drivers/scsi/ps3rom.c
@@ -361,7 +361,7 @@ static int ps3rom_probe(struct ps3_system_bus_device *_dev)
 		goto fail_free_bounce;
 
 	host = scsi_host_alloc(&ps3rom_host_template,
-			       sizeof(struct ps3rom_private));
+			       sizeof(struct ps3rom_private), NULL);
 	if (!host) {
 		dev_err(&dev->sbd.core, "%s:%u: scsi_host_alloc failed\n",
 			__func__, __LINE__);
diff --git a/drivers/scsi/qla1280.c b/drivers/scsi/qla1280.c
index cdd6fe002c32..f88f2e659baa 100644
--- a/drivers/scsi/qla1280.c
+++ b/drivers/scsi/qla1280.c
@@ -4142,7 +4142,7 @@ qla1280_probe_one(struct pci_dev *pdev, const struct pci_device_id *id)
 	pci_set_master(pdev);
 
 	error = -ENOMEM;
-	host = scsi_host_alloc(&qla1280_driver_template, sizeof(*ha));
+	host = scsi_host_alloc(&qla1280_driver_template, sizeof(*ha), &pdev->dev);
 	if (!host) {
 		printk(KERN_WARNING
 		       "qla1280: Failed to register host, aborting.\n");
diff --git a/drivers/scsi/qla2xxx/qla_mid.c b/drivers/scsi/qla2xxx/qla_mid.c
index c563133f751e..4bafc367e21d 100644
--- a/drivers/scsi/qla2xxx/qla_mid.c
+++ b/drivers/scsi/qla2xxx/qla_mid.c
@@ -502,7 +502,7 @@ qla24xx_create_vhost(struct fc_vport *fc_vport)
 	vha = qla2x00_create_host(sht, ha);
 	if (!vha) {
 		ql_log(ql_log_warn, vha, 0xa005,
-		    "scsi_host_alloc() failed for vport.\n");
+		    "scsi_host_alloc() failed for vport.\n", NULL);
 		return(NULL);
 	}
 
diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
index 72b1c28e4dae..ce0d097f3317 100644
--- a/drivers/scsi/qla2xxx/qla_os.c
+++ b/drivers/scsi/qla2xxx/qla_os.c
@@ -5046,7 +5046,7 @@ struct scsi_qla_host *qla2x00_create_host(const struct scsi_host_template *sht,
 	struct Scsi_Host *host;
 	struct scsi_qla_host *vha = NULL;
 
-	host = scsi_host_alloc(sht, sizeof(scsi_qla_host_t));
+	host = scsi_host_alloc(sht, sizeof(scsi_qla_host_t), &ha->pdev->dev);
 	if (!host) {
 		ql_log_pci(ql_log_fatal, ha->pdev, 0x0107,
 		    "Failed to allocate host from the scsi layer, aborting.\n");
diff --git a/drivers/scsi/qlogicfas.c b/drivers/scsi/qlogicfas.c
index 8f05e3707d69..b9ead7dc371c 100644
--- a/drivers/scsi/qlogicfas.c
+++ b/drivers/scsi/qlogicfas.c
@@ -95,7 +95,7 @@ static struct Scsi_Host *__qlogicfas_detect(struct scsi_host_template *host,
 
 	qlogicfas408_setup(qbase, qinitid, INT_TYPE);
 
-	hreg = scsi_host_alloc(host, sizeof(struct qlogicfas408_priv));
+	hreg = scsi_host_alloc(host, sizeof(struct qlogicfas408_priv), NULL);
 	if (!hreg)
 		goto err_release_mem;
 	priv = get_priv_by_host(hreg);
diff --git a/drivers/scsi/qlogicpti.c b/drivers/scsi/qlogicpti.c
index ea0a2b5a0a42..f67a9b400100 100644
--- a/drivers/scsi/qlogicpti.c
+++ b/drivers/scsi/qlogicpti.c
@@ -1316,7 +1316,7 @@ static int qpti_sbus_probe(struct platform_device *op)
 	if (op->archdata.irqs[0] == 0)
 		return -ENODEV;
 
-	host = scsi_host_alloc(&qpti_template, sizeof(struct qlogicpti));
+	host = scsi_host_alloc(&qpti_template, sizeof(struct qlogicpti), NULL);
 	if (!host)
 		return -ENOMEM;
 
diff --git a/drivers/scsi/scsi_debug.c b/drivers/scsi/scsi_debug.c
index bb6b0e7fb910..59488bf74ce0 100644
--- a/drivers/scsi/scsi_debug.c
+++ b/drivers/scsi/scsi_debug.c
@@ -9548,7 +9548,7 @@ static int sdebug_driver_probe(struct device *dev)
 
 	sdbg_host = dev_to_sdebug_host(dev);
 
-	hpnt = scsi_host_alloc(&sdebug_driver_template, 0);
+	hpnt = scsi_host_alloc(&sdebug_driver_template, 0, NULL);
 	if (NULL == hpnt) {
 		pr_err("scsi_host_alloc failed\n");
 		error = -ENODEV;
diff --git a/drivers/scsi/sgiwd93.c b/drivers/scsi/sgiwd93.c
index 6594661db5f4..07fbe6fda7c2 100644
--- a/drivers/scsi/sgiwd93.c
+++ b/drivers/scsi/sgiwd93.c
@@ -231,7 +231,7 @@ static int sgiwd93_probe(struct platform_device *pdev)
 	unsigned int irq = pd->irq;
 	int err;
 
-	host = scsi_host_alloc(&sgiwd93_template, sizeof(struct ip22_hostdata));
+	host = scsi_host_alloc(&sgiwd93_template, sizeof(struct ip22_hostdata), NULL);
 	if (!host) {
 		err = -ENOMEM;
 		goto out;
diff --git a/drivers/scsi/smartpqi/smartpqi_init.c b/drivers/scsi/smartpqi/smartpqi_init.c
index 65ff50982978..a3163c06b3f8 100644
--- a/drivers/scsi/smartpqi/smartpqi_init.c
+++ b/drivers/scsi/smartpqi/smartpqi_init.c
@@ -7619,7 +7619,7 @@ static int pqi_register_scsi(struct pqi_ctrl_info *ctrl_info)
 	int rc;
 	struct Scsi_Host *shost;
 
-	shost = scsi_host_alloc(&pqi_driver_template, sizeof(ctrl_info));
+	shost = scsi_host_alloc(&pqi_driver_template, sizeof(ctrl_info), &ctrl_info->pci_dev->dev);
 	if (!shost) {
 		dev_err(&ctrl_info->pci_dev->dev, "scsi_host_alloc failed\n");
 		return -ENOMEM;
diff --git a/drivers/scsi/snic/snic_main.c b/drivers/scsi/snic/snic_main.c
index 82953e6a0915..9edf6661e6f1 100644
--- a/drivers/scsi/snic/snic_main.c
+++ b/drivers/scsi/snic/snic_main.c
@@ -363,7 +363,7 @@ snic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	/*
 	 * Allocate SCSI Host and setup association between host, and snic
 	 */
-	shost = scsi_host_alloc(&snic_host_template, sizeof(struct snic));
+	shost = scsi_host_alloc(&snic_host_template, sizeof(struct snic), &pdev->dev);
 	if (!shost) {
 		SNIC_ERR("Unable to alloc scsi_host\n");
 		ret = -ENOMEM;
diff --git a/drivers/scsi/stex.c b/drivers/scsi/stex.c
index 6aeeb338633d..7d6b851fef24 100644
--- a/drivers/scsi/stex.c
+++ b/drivers/scsi/stex.c
@@ -1667,7 +1667,7 @@ static int stex_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	S6flag = 0;
 	register_reboot_notifier(&stex_notifier);
 
-	host = scsi_host_alloc(&driver_template, sizeof(struct st_hba));
+	host = scsi_host_alloc(&driver_template, sizeof(struct st_hba), &pdev->dev);
 
 	if (!host) {
 		printk(KERN_ERR DRV_NAME "(%s): scsi_host_alloc failed\n",
diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c
index 571ea549152b..fc4c05127dc4 100644
--- a/drivers/scsi/storvsc_drv.c
+++ b/drivers/scsi/storvsc_drv.c
@@ -1969,7 +1969,7 @@ static int storvsc_probe(struct hv_device *device,
 				(100 - ring_avail_percent_lowater) / 100;
 
 	host = scsi_host_alloc(&scsi_driver,
-			       sizeof(struct hv_host_device));
+			       sizeof(struct hv_host_device), NULL);
 	if (!host)
 		return -ENOMEM;
 
diff --git a/drivers/scsi/sun3_scsi.c b/drivers/scsi/sun3_scsi.c
index ca9cd691cc32..ed41b605328e 100644
--- a/drivers/scsi/sun3_scsi.c
+++ b/drivers/scsi/sun3_scsi.c
@@ -578,7 +578,7 @@ static int __init sun3_scsi_probe(struct platform_device *pdev)
 #endif
 
 	instance = scsi_host_alloc(&sun3_scsi_template,
-	                           sizeof(struct NCR5380_hostdata));
+				   sizeof(struct NCR5380_hostdata), NULL);
 	if (!instance) {
 		error = -ENOMEM;
 		goto fail_alloc;
diff --git a/drivers/scsi/sun3x_esp.c b/drivers/scsi/sun3x_esp.c
index 365406885b8e..f7e48f4c5444 100644
--- a/drivers/scsi/sun3x_esp.c
+++ b/drivers/scsi/sun3x_esp.c
@@ -175,7 +175,7 @@ static int esp_sun3x_probe(struct platform_device *dev)
 	struct resource *res;
 	int err = -ENOMEM;
 
-	host = scsi_host_alloc(tpnt, sizeof(struct esp));
+	host = scsi_host_alloc(tpnt, sizeof(struct esp), NULL);
 	if (!host)
 		goto fail;
 
diff --git a/drivers/scsi/sun_esp.c b/drivers/scsi/sun_esp.c
index aa430501f0c7..bc4e4030acb6 100644
--- a/drivers/scsi/sun_esp.c
+++ b/drivers/scsi/sun_esp.c
@@ -457,7 +457,7 @@ static int esp_sbus_probe_one(struct platform_device *op,
 	struct esp *esp;
 	int err;
 
-	host = scsi_host_alloc(tpnt, sizeof(struct esp));
+	host = scsi_host_alloc(tpnt, sizeof(struct esp), NULL);
 
 	err = -ENOMEM;
 	if (!host)
diff --git a/drivers/scsi/sym53c8xx_2/sym_glue.c b/drivers/scsi/sym53c8xx_2/sym_glue.c
index 27e22acaf1a7..16e821c3b59e 100644
--- a/drivers/scsi/sym53c8xx_2/sym_glue.c
+++ b/drivers/scsi/sym53c8xx_2/sym_glue.c
@@ -1300,7 +1300,7 @@ static struct Scsi_Host *sym_attach(const struct scsi_host_template *tpnt, int u
 	if (!fw)
 		goto attach_failed;
 
-	shost = scsi_host_alloc(tpnt, sizeof(*sym_data));
+	shost = scsi_host_alloc(tpnt, sizeof(*sym_data), &pdev->dev);
 	if (!shost)
 		goto attach_failed;
 	sym_data = shost_priv(shost);
diff --git a/drivers/scsi/virtio_scsi.c b/drivers/scsi/virtio_scsi.c
index 5fdaa71f0652..88375574cb18 100644
--- a/drivers/scsi/virtio_scsi.c
+++ b/drivers/scsi/virtio_scsi.c
@@ -929,7 +929,7 @@ static int virtscsi_probe(struct virtio_device *vdev)
 	num_targets = virtscsi_config_get(vdev, max_target) + 1;
 
 	shost = scsi_host_alloc(&virtscsi_host_template,
-				struct_size(vscsi, req_vqs, num_queues));
+				struct_size(vscsi, req_vqs, num_queues), NULL);
 	if (!shost)
 		return -ENOMEM;
 
diff --git a/drivers/scsi/vmw_pvscsi.c b/drivers/scsi/vmw_pvscsi.c
index 151cac9f9c2a..32c39c66c49b 100644
--- a/drivers/scsi/vmw_pvscsi.c
+++ b/drivers/scsi/vmw_pvscsi.c
@@ -1435,7 +1435,7 @@ static int pvscsi_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		PVSCSI_MAX_NUM_REQ_ENTRIES_PER_PAGE;
 	pvscsi_template.cmd_per_lun =
 		min(pvscsi_template.can_queue, pvscsi_cmd_per_lun);
-	host = scsi_host_alloc(&pvscsi_template, sizeof(struct pvscsi_adapter));
+	host = scsi_host_alloc(&pvscsi_template, sizeof(struct pvscsi_adapter), &pdev->dev);
 	if (!host) {
 		printk(KERN_ERR "vmw_pvscsi: failed to allocate host\n");
 		goto out_release_resources_and_disable;
diff --git a/drivers/scsi/wd719x.c b/drivers/scsi/wd719x.c
index 830d40f57f6a..0aa6bb093431 100644
--- a/drivers/scsi/wd719x.c
+++ b/drivers/scsi/wd719x.c
@@ -921,7 +921,7 @@ static int wd719x_pci_probe(struct pci_dev *pdev, const struct pci_device_id *d)
 		goto release_region;
 
 	err = -ENOMEM;
-	sh = scsi_host_alloc(&wd719x_template, sizeof(struct wd719x));
+	sh = scsi_host_alloc(&wd719x_template, sizeof(struct wd719x), &pdev->dev);
 	if (!sh)
 		goto release_region;
 
diff --git a/drivers/scsi/xen-scsifront.c b/drivers/scsi/xen-scsifront.c
index 989bcaee42ca..d4d57f33cc15 100644
--- a/drivers/scsi/xen-scsifront.c
+++ b/drivers/scsi/xen-scsifront.c
@@ -899,7 +899,7 @@ static int scsifront_probe(struct xenbus_device *dev,
 	int err = -ENOMEM;
 	char name[TASK_COMM_LEN];
 
-	host = scsi_host_alloc(&scsifront_sht, sizeof(*info));
+	host = scsi_host_alloc(&scsifront_sht, sizeof(*info), NULL);
 	if (!host) {
 		xenbus_dev_fatal(dev, err, "fail to allocate scsi host");
 		return err;
diff --git a/drivers/scsi/zorro_esp.c b/drivers/scsi/zorro_esp.c
index 1622285c9aec..5983015877a7 100644
--- a/drivers/scsi/zorro_esp.c
+++ b/drivers/scsi/zorro_esp.c
@@ -774,7 +774,7 @@ static int zorro_esp_probe(struct zorro_dev *z,
 		goto fail_free_zep;
 	}
 
-	host = scsi_host_alloc(tpnt, sizeof(struct esp));
+	host = scsi_host_alloc(tpnt, sizeof(struct esp), NULL);
 
 	if (!host) {
 		pr_err("No host detected; board configuration problem?\n");
diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h
index be0ffe1e3395..17e545fa5c7e 100644
--- a/include/scsi/libfc.h
+++ b/include/scsi/libfc.h
@@ -883,7 +883,7 @@ libfc_host_alloc(const struct scsi_host_template *sht, int priv_size)
 	struct fc_lport *lport;
 	struct Scsi_Host *shost;
 
-	shost = scsi_host_alloc(sht, sizeof(*lport) + priv_size);
+	shost = scsi_host_alloc(sht, sizeof(*lport) + priv_size, NULL);
 	if (!shost)
 		return NULL;
 	lport = shost_priv(shost);
diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h
index 7e2011830ba4..09c82a41b7a1 100644
--- a/include/scsi/scsi_host.h
+++ b/include/scsi/scsi_host.h
@@ -796,7 +796,8 @@ static inline int scsi_host_in_recovery(struct Scsi_Host *shost)
 extern int scsi_queue_work(struct Scsi_Host *, struct work_struct *);
 extern void scsi_flush_work(struct Scsi_Host *);
 
-extern struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *, int);
+extern struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *sht,
+					 int privsize, struct device *dev);
 extern int __must_check scsi_add_host_with_dma(struct Scsi_Host *,
 					       struct device *,
 					       struct device *);
-- 
2.43.7

[PATCH v3 3/4] block: drop shared-tag fairness throttling

[Sumit Saxena]

From: Bart Van Assche <bvanassche@acm.org>

Original patch [1] by Bart Van Assche; this version is rebased onto the
current tree.  In testing it improves IOPS by roughly 16-18% by removing
the fair-sharing throttle on shared tag queues.

This patch removes the following code and structure members:
- The function hctx_may_queue().
- blk_mq_hw_ctx.nr_active and request_queue.nr_active_requests_shared_tags
  and also all the code that modifies these two member variables.

[1]: https://lore.kernel.org/linux-block/20240529213921.3166462-1-bvanassche@acm.org/

Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
---
 block/blk-core.c       |   2 -
 block/blk-mq-debugfs.c |  22 ++++++++-
 block/blk-mq-tag.c     |   4 --
 block/blk-mq.c         |  17 +------
 block/blk-mq.h         | 100 -----------------------------------------
 include/linux/blk-mq.h |   6 ---
 include/linux/blkdev.h |   2 -
 7 files changed, 22 insertions(+), 131 deletions(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index 17450058ea6d..129acc1b27e5 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -421,8 +421,6 @@ struct request_queue *blk_alloc_queue(struct queue_limits *lim, int node_id)
 
 	q->node = node_id;
 
-	atomic_set(&q->nr_active_requests_shared_tags, 0);
-
 	timer_setup(&q->timeout, blk_rq_timed_out_timer, 0);
 	INIT_WORK(&q->timeout_work, blk_timeout_work);
 	INIT_LIST_HEAD(&q->icq_list);
diff --git a/block/blk-mq-debugfs.c b/block/blk-mq-debugfs.c
index 047ec887456b..8b85a7f8e987 100644
--- a/block/blk-mq-debugfs.c
+++ b/block/blk-mq-debugfs.c
@@ -468,11 +468,31 @@ static int hctx_sched_tags_bitmap_show(void *data, struct seq_file *m)
 	return 0;
 }
 
+struct count_active_params {
+	struct blk_mq_hw_ctx	*hctx;
+	int			*active;
+};
+
+static bool hctx_count_active(struct request *rq, void *data)
+{
+	const struct count_active_params *params = data;
+
+	if (rq->mq_hctx == params->hctx)
+		(*params->active)++;
+
+	return true;
+}
+
 static int hctx_active_show(void *data, struct seq_file *m)
 {
 	struct blk_mq_hw_ctx *hctx = data;
+	int active = 0;
+	struct count_active_params params = { .hctx = hctx, .active = &active };
+
+	blk_mq_all_tag_iter(hctx->sched_tags ?: hctx->tags, hctx_count_active,
+			    &params);
 
-	seq_printf(m, "%d\n", __blk_mq_active_requests(hctx));
+	seq_printf(m, "%d\n", active);
 	return 0;
 }
 
diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c
index 33946cdb5716..bfd27cc6249b 100644
--- a/block/blk-mq-tag.c
+++ b/block/blk-mq-tag.c
@@ -109,10 +109,6 @@ void __blk_mq_tag_idle(struct blk_mq_hw_ctx *hctx)
 static int __blk_mq_get_tag(struct blk_mq_alloc_data *data,
 			    struct sbitmap_queue *bt)
 {
-	if (!data->q->elevator && !(data->flags & BLK_MQ_REQ_RESERVED) &&
-			!hctx_may_queue(data->hctx, bt))
-		return BLK_MQ_NO_TAG;
-
 	if (data->shallow_depth)
 		return sbitmap_queue_get_shallow(bt, data->shallow_depth);
 	else
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 4c5c16cce4f8..bbac59a06044 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -489,8 +489,6 @@ __blk_mq_alloc_requests_batch(struct blk_mq_alloc_data *data)
 		}
 	} while (data->nr_tags > nr);
 
-	if (!(data->rq_flags & RQF_SCHED_TAGS))
-		blk_mq_add_active_requests(data->hctx, nr);
 	/* caller already holds a reference, add for remainder */
 	percpu_ref_get_many(&data->q->q_usage_counter, nr - 1);
 	data->nr_tags -= nr;
@@ -587,8 +585,6 @@ static struct request *__blk_mq_alloc_requests(struct blk_mq_alloc_data *data)
 		goto retry;
 	}
 
-	if (!(data->rq_flags & RQF_SCHED_TAGS))
-		blk_mq_inc_active_requests(data->hctx);
 	rq = blk_mq_rq_ctx_init(data, blk_mq_tags_from_data(data), tag);
 	blk_mq_rq_time_init(rq, alloc_time_ns);
 	return rq;
@@ -763,8 +759,6 @@ struct request *blk_mq_alloc_request_hctx(struct request_queue *q,
 	tag = blk_mq_get_tag(&data);
 	if (tag == BLK_MQ_NO_TAG)
 		goto out_queue_exit;
-	if (!(data.rq_flags & RQF_SCHED_TAGS))
-		blk_mq_inc_active_requests(data.hctx);
 	rq = blk_mq_rq_ctx_init(&data, blk_mq_tags_from_data(&data), tag);
 	blk_mq_rq_time_init(rq, alloc_time_ns);
 	rq->__data_len = 0;
@@ -807,10 +801,8 @@ static void __blk_mq_free_request(struct request *rq)
 	blk_pm_mark_last_busy(rq);
 	rq->mq_hctx = NULL;
 
-	if (rq->tag != BLK_MQ_NO_TAG) {
-		blk_mq_dec_active_requests(hctx);
+	if (rq->tag != BLK_MQ_NO_TAG)
 		blk_mq_put_tag(hctx->tags, ctx, rq->tag);
-	}
 	if (sched_tag != BLK_MQ_NO_TAG)
 		blk_mq_put_tag(hctx->sched_tags, ctx, sched_tag);
 	blk_mq_sched_restart(hctx);
@@ -1188,8 +1180,6 @@ static inline void blk_mq_flush_tag_batch(struct blk_mq_hw_ctx *hctx,
 {
 	struct request_queue *q = hctx->queue;
 
-	blk_mq_sub_active_requests(hctx, nr_tags);
-
 	blk_mq_put_tags(hctx->tags, tag_array, nr_tags);
 	percpu_ref_put_many(&q->q_usage_counter, nr_tags);
 }
@@ -1875,9 +1865,6 @@ bool __blk_mq_alloc_driver_tag(struct request *rq)
 	if (blk_mq_tag_is_reserved(rq->mq_hctx->sched_tags, rq->internal_tag)) {
 		bt = &rq->mq_hctx->tags->breserved_tags;
 		tag_offset = 0;
-	} else {
-		if (!hctx_may_queue(rq->mq_hctx, bt))
-			return false;
 	}
 
 	tag = __sbitmap_queue_get(bt);
@@ -1885,7 +1872,6 @@ bool __blk_mq_alloc_driver_tag(struct request *rq)
 		return false;
 
 	rq->tag = tag + tag_offset;
-	blk_mq_inc_active_requests(rq->mq_hctx);
 	return true;
 }
 
@@ -4058,7 +4044,6 @@ blk_mq_alloc_hctx(struct request_queue *q, struct blk_mq_tag_set *set,
 	if (!zalloc_cpumask_var_node(&hctx->cpumask, gfp, node))
 		goto free_hctx;
 
-	atomic_set(&hctx->nr_active, 0);
 	if (node == NUMA_NO_NODE)
 		node = set->numa_node;
 	hctx->numa_node = node;
diff --git a/block/blk-mq.h b/block/blk-mq.h
index aa15d31aaae9..8dfb67c55f5d 100644
--- a/block/blk-mq.h
+++ b/block/blk-mq.h
@@ -291,70 +291,9 @@ static inline int blk_mq_get_rq_budget_token(struct request *rq)
 	return -1;
 }
 
-static inline void __blk_mq_add_active_requests(struct blk_mq_hw_ctx *hctx,
-						int val)
-{
-	if (blk_mq_is_shared_tags(hctx->flags))
-		atomic_add(val, &hctx->queue->nr_active_requests_shared_tags);
-	else
-		atomic_add(val, &hctx->nr_active);
-}
-
-static inline void __blk_mq_inc_active_requests(struct blk_mq_hw_ctx *hctx)
-{
-	__blk_mq_add_active_requests(hctx, 1);
-}
-
-static inline void __blk_mq_sub_active_requests(struct blk_mq_hw_ctx *hctx,
-		int val)
-{
-	if (blk_mq_is_shared_tags(hctx->flags))
-		atomic_sub(val, &hctx->queue->nr_active_requests_shared_tags);
-	else
-		atomic_sub(val, &hctx->nr_active);
-}
-
-static inline void __blk_mq_dec_active_requests(struct blk_mq_hw_ctx *hctx)
-{
-	__blk_mq_sub_active_requests(hctx, 1);
-}
-
-static inline void blk_mq_add_active_requests(struct blk_mq_hw_ctx *hctx,
-					      int val)
-{
-	if (hctx->flags & BLK_MQ_F_TAG_QUEUE_SHARED)
-		__blk_mq_add_active_requests(hctx, val);
-}
-
-static inline void blk_mq_inc_active_requests(struct blk_mq_hw_ctx *hctx)
-{
-	if (hctx->flags & BLK_MQ_F_TAG_QUEUE_SHARED)
-		__blk_mq_inc_active_requests(hctx);
-}
-
-static inline void blk_mq_sub_active_requests(struct blk_mq_hw_ctx *hctx,
-					      int val)
-{
-	if (hctx->flags & BLK_MQ_F_TAG_QUEUE_SHARED)
-		__blk_mq_sub_active_requests(hctx, val);
-}
-
-static inline void blk_mq_dec_active_requests(struct blk_mq_hw_ctx *hctx)
-{
-	if (hctx->flags & BLK_MQ_F_TAG_QUEUE_SHARED)
-		__blk_mq_dec_active_requests(hctx);
-}
-
-static inline int __blk_mq_active_requests(struct blk_mq_hw_ctx *hctx)
-{
-	if (blk_mq_is_shared_tags(hctx->flags))
-		return atomic_read(&hctx->queue->nr_active_requests_shared_tags);
-	return atomic_read(&hctx->nr_active);
-}
 static inline void __blk_mq_put_driver_tag(struct blk_mq_hw_ctx *hctx,
 					   struct request *rq)
 {
-	blk_mq_dec_active_requests(hctx);
 	blk_mq_put_tag(hctx->tags, rq->mq_ctx, rq->tag);
 	rq->tag = BLK_MQ_NO_TAG;
 }
@@ -396,45 +335,6 @@ static inline void blk_mq_free_requests(struct list_head *list)
 	}
 }
 
-/*
- * For shared tag users, we track the number of currently active users
- * and attempt to provide a fair share of the tag depth for each of them.
- */
-static inline bool hctx_may_queue(struct blk_mq_hw_ctx *hctx,
-				  struct sbitmap_queue *bt)
-{
-	unsigned int depth, users;
-
-	if (!hctx || !(hctx->flags & BLK_MQ_F_TAG_QUEUE_SHARED))
-		return true;
-
-	/*
-	 * Don't try dividing an ant
-	 */
-	if (bt->sb.depth == 1)
-		return true;
-
-	if (blk_mq_is_shared_tags(hctx->flags)) {
-		struct request_queue *q = hctx->queue;
-
-		if (!test_bit(QUEUE_FLAG_HCTX_ACTIVE, &q->queue_flags))
-			return true;
-	} else {
-		if (!test_bit(BLK_MQ_S_TAG_ACTIVE, &hctx->state))
-			return true;
-	}
-
-	users = READ_ONCE(hctx->tags->active_queues);
-	if (!users)
-		return true;
-
-	/*
-	 * Allow at least some tags
-	 */
-	depth = max((bt->sb.depth + users - 1) / users, 4U);
-	return __blk_mq_active_requests(hctx) < depth;
-}
-
 /* run the code block in @dispatch_ops with rcu/srcu read lock held */
 #define __blk_mq_run_dispatch_ops(q, check_sleep, dispatch_ops)	\
 do {								\
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index 18a2388ba581..ccbb07559402 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -432,12 +432,6 @@ struct blk_mq_hw_ctx {
 	/** @queue_num: Index of this hardware queue. */
 	unsigned int		queue_num;
 
-	/**
-	 * @nr_active: Number of active requests. Only used when a tag set is
-	 * shared across request queues.
-	 */
-	atomic_t		nr_active;
-
 	/** @cpuhp_online: List to store request if CPU is going to die */
 	struct hlist_node	cpuhp_online;
 	/** @cpuhp_dead: List to store request if some CPU die. */
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 890128cdea1c..95525b1d7b74 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -567,8 +567,6 @@ struct request_queue {
 	struct timer_list	timeout;
 	struct work_struct	timeout_work;
 
-	atomic_t		nr_active_requests_shared_tags;
-
 	struct blk_mq_tags	*sched_shared_tags;
 
 	struct list_head	icq_list;
-- 
2.43.7

[PATCH v3 4/4] scsi: use percpu counters for iostat counters in struct scsi_device

[Sumit Saxena]

iorequest_cnt and iodone_cnt are updated on every command dispatch and
completion, often from different CPUs on high queue depth workloads.
Using adjacent atomic_t fields causes cache line contention between the
submission and completion paths.

Extend the same treatment to ioerr_cnt and iotmo_cnt so all four iostat
counters in struct scsi_device use struct percpu_counter.

Suggested-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
---
 drivers/scsi/scsi_error.c  |  4 ++--
 drivers/scsi/scsi_lib.c    | 10 +++++-----
 drivers/scsi/scsi_scan.c   |  8 ++++++++
 drivers/scsi/scsi_sysfs.c  | 23 ++++++++++++++---------
 drivers/scsi/sd.c          |  2 +-
 include/scsi/scsi_device.h |  9 +++++----
 6 files changed, 35 insertions(+), 21 deletions(-)

diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c
index 147127fb4db9..b1aa7da2ba7c 100644
--- a/drivers/scsi/scsi_error.c
+++ b/drivers/scsi/scsi_error.c
@@ -349,7 +349,7 @@ enum blk_eh_timer_return scsi_timeout(struct request *req)
 	trace_scsi_dispatch_cmd_timeout(scmd);
 	scsi_log_completion(scmd, TIMEOUT_ERROR);
 
-	atomic_inc(&scmd->device->iotmo_cnt);
+	percpu_counter_inc(&scmd->device->iotmo_cnt);
 	if (host->eh_deadline != -1 && !host->last_reset)
 		host->last_reset = jiffies;
 
@@ -370,7 +370,7 @@ enum blk_eh_timer_return scsi_timeout(struct request *req)
 	 */
 	if (test_and_set_bit(SCMD_STATE_COMPLETE, &scmd->state))
 		return BLK_EH_DONE;
-	atomic_inc(&scmd->device->iodone_cnt);
+	percpu_counter_inc(&scmd->device->iodone_cnt);
 	if (scsi_abort_command(scmd) != SUCCESS) {
 		set_host_byte(scmd, DID_TIME_OUT);
 		scsi_eh_scmd_add(scmd);
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 6e8c7a42603e..979fdace33ac 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -1554,9 +1554,9 @@ static void scsi_complete(struct request *rq)
 
 	INIT_LIST_HEAD(&cmd->eh_entry);
 
-	atomic_inc(&cmd->device->iodone_cnt);
+	percpu_counter_inc(&cmd->device->iodone_cnt);
 	if (cmd->result)
-		atomic_inc(&cmd->device->ioerr_cnt);
+		percpu_counter_inc(&cmd->device->ioerr_cnt);
 
 	disposition = scsi_decide_disposition(cmd);
 	if (disposition != SUCCESS && scsi_cmd_runtime_exceeced(cmd))
@@ -1592,7 +1592,7 @@ static enum scsi_qc_status scsi_dispatch_cmd(struct scsi_cmnd *cmd)
 	struct Scsi_Host *host = cmd->device->host;
 	int rtn = 0;
 
-	atomic_inc(&cmd->device->iorequest_cnt);
+	percpu_counter_inc(&cmd->device->iorequest_cnt);
 
 	/* check if the device is still usable */
 	if (unlikely(cmd->device->sdev_state == SDEV_DEL)) {
@@ -1614,7 +1614,7 @@ static enum scsi_qc_status scsi_dispatch_cmd(struct scsi_cmnd *cmd)
 		 */
 		SCSI_LOG_MLQUEUE(3, scmd_printk(KERN_INFO, cmd,
 			"queuecommand : device blocked\n"));
-		atomic_dec(&cmd->device->iorequest_cnt);
+		percpu_counter_dec(&cmd->device->iorequest_cnt);
 		return SCSI_MLQUEUE_DEVICE_BUSY;
 	}
 
@@ -1647,7 +1647,7 @@ static enum scsi_qc_status scsi_dispatch_cmd(struct scsi_cmnd *cmd)
 	trace_scsi_dispatch_cmd_start(cmd);
 	rtn = host->hostt->queuecommand(host, cmd);
 	if (rtn) {
-		atomic_dec(&cmd->device->iorequest_cnt);
+		percpu_counter_dec(&cmd->device->iorequest_cnt);
 		trace_scsi_dispatch_cmd_error(cmd, rtn);
 		if (rtn != SCSI_MLQUEUE_DEVICE_BUSY &&
 		    rtn != SCSI_MLQUEUE_TARGET_BUSY)
diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index 121a14d5fdb8..bc885c72f01e 100644
--- a/drivers/scsi/scsi_scan.c
+++ b/drivers/scsi/scsi_scan.c
@@ -350,6 +350,14 @@ static struct scsi_device *scsi_alloc_sdev(struct scsi_target *starget,
 
 	scsi_sysfs_device_initialize(sdev);
 
+	if (percpu_counter_init(&sdev->iorequest_cnt, 0, GFP_KERNEL) ||
+	    percpu_counter_init(&sdev->iodone_cnt, 0, GFP_KERNEL) ||
+	    percpu_counter_init(&sdev->ioerr_cnt, 0, GFP_KERNEL) ||
+	    percpu_counter_init(&sdev->iotmo_cnt, 0, GFP_KERNEL)) {
+		ret = -ENOMEM;
+		goto out_device_destroy;
+	}
+
 	if (scsi_device_is_pseudo_dev(sdev))
 		return sdev;
 
diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c
index dfc3559e7e04..f652edd16497 100644
--- a/drivers/scsi/scsi_sysfs.c
+++ b/drivers/scsi/scsi_sysfs.c
@@ -516,6 +516,10 @@ static void scsi_device_dev_release(struct device *dev)
 	if (vpd_pgb7)
 		kfree_rcu(vpd_pgb7, rcu);
 	kfree(sdev->inquiry);
+	percpu_counter_destroy(&sdev->iotmo_cnt);
+	percpu_counter_destroy(&sdev->ioerr_cnt);
+	percpu_counter_destroy(&sdev->iodone_cnt);
+	percpu_counter_destroy(&sdev->iorequest_cnt);
 	kfree(sdev);
 
 	if (parent)
@@ -936,26 +940,27 @@ static ssize_t
 show_iostat_counterbits(struct device *dev, struct device_attribute *attr,
 			char *buf)
 {
-	return snprintf(buf, 20, "%d\n", (int)sizeof(atomic_t) * 8);
+	/* iostat counters are per-CPU sums (s64).  Report width for tools. */
+	return sysfs_emit(buf, "%zu\n", sizeof(s64) * 8);
 }
 
 static DEVICE_ATTR(iocounterbits, S_IRUGO, show_iostat_counterbits, NULL);
 
-#define show_sdev_iostat(field)						\
+#define show_sdev_iostat_percpu(field)					\
 static ssize_t								\
 show_iostat_##field(struct device *dev, struct device_attribute *attr,	\
 		    char *buf)						\
 {									\
 	struct scsi_device *sdev = to_scsi_device(dev);			\
-	unsigned long long count = atomic_read(&sdev->field);		\
-	return snprintf(buf, 20, "0x%llx\n", count);			\
+	unsigned long long count = percpu_counter_sum(&sdev->field);	\
+	return sysfs_emit(buf, "0x%llx\n", count);			\
 }									\
-static DEVICE_ATTR(field, S_IRUGO, show_iostat_##field, NULL)
+static DEVICE_ATTR(field, 0444, show_iostat_##field, NULL)
 
-show_sdev_iostat(iorequest_cnt);
-show_sdev_iostat(iodone_cnt);
-show_sdev_iostat(ioerr_cnt);
-show_sdev_iostat(iotmo_cnt);
+show_sdev_iostat_percpu(iorequest_cnt);
+show_sdev_iostat_percpu(iodone_cnt);
+show_sdev_iostat_percpu(ioerr_cnt);
+show_sdev_iostat_percpu(iotmo_cnt);
 
 static ssize_t
 sdev_show_modalias(struct device *dev, struct device_attribute *attr, char *buf)
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index adc3fa55ca2c..b7ce01de17b3 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -4043,7 +4043,7 @@ static int sd_probe(struct scsi_device *sdp)
 	sdkp->index = index;
 	sdkp->max_retries = SD_MAX_RETRIES;
 	atomic_set(&sdkp->openers, 0);
-	atomic_set(&sdkp->device->ioerr_cnt, 0);
+	percpu_counter_set(&sdkp->device->ioerr_cnt, 0);
 
 	if (!sdp->request_queue->rq_timeout) {
 		if (sdp->type != TYPE_MOD)
diff --git a/include/scsi/scsi_device.h b/include/scsi/scsi_device.h
index 029f5115b2ea..4be36bf2a475 100644
--- a/include/scsi/scsi_device.h
+++ b/include/scsi/scsi_device.h
@@ -9,6 +9,7 @@
 #include <scsi/scsi.h>
 #include <scsi/scsi_common.h>
 #include <linux/atomic.h>
+#include <linux/percpu_counter.h>
 #include <linux/sbitmap.h>
 
 struct bsg_device;
@@ -272,10 +273,10 @@ struct scsi_device {
 	unsigned int max_device_blocked; /* what device_blocked counts down from  */
 #define SCSI_DEFAULT_DEVICE_BLOCKED	3
 
-	atomic_t iorequest_cnt;
-	atomic_t iodone_cnt;
-	atomic_t ioerr_cnt;
-	atomic_t iotmo_cnt;
+	struct percpu_counter iorequest_cnt;
+	struct percpu_counter iodone_cnt;
+	struct percpu_counter ioerr_cnt;
+	struct percpu_counter iotmo_cnt;
 
 	struct device		sdev_gendev,
 				sdev_dev;
-- 
2.43.7

Re: [PATCH v3 2/4] scsi: host: allocate struct Scsi_Host on the NUMA node of the host adapter

[John Garry]

On 09/06/2026 13:18, Sumit Saxena wrote:
> scsi_host_alloc() used kzalloc(), which always picks an arbitrary node.
> Extend the function to accept a 'struct device *dev' parameter and use
> kzalloc_node() with dev_to_node(dev) so the Scsi_Host struct lands on
> the same NUMA node as the HBA, mirroring the treatment already applied
> to struct scsi_device, struct scsi_target, and shost_data.
> 
> When dev is NULL (legacy ISA/platform drivers without a dma_dev) the
> allocation falls back to NUMA_NO_NODE, preserving existing behaviour.
> 
> Update all in-tree callers:
>    - PCI-based HBA drivers pass &pdev->dev (or the equivalent struct
>      member such as &phba->pcidev->dev, &h->pdev->dev, &ha->pdev->dev)
>      so their host struct is placed on the adapter's node.
>    - Non-PCI drivers (ISA, Amiga, ARM PCMCIA, virtio, Hyper-V, PS3, …)
>      pass NULL.
>    - libfc's libfc_host_alloc() inline helper passes NULL; FC drivers
>      that want NUMA awareness can open-code the call with their pdev.
> 
> Suggested-by: John Garry <john.g.garry@oracle.com>
> Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>

Wow ... I was not expecting such a large change, but admittedly I did 
not consider the implementation.

I did mention that pci-based adapters should already be effectively 
doing kzalloc_node() since the adapter driver is probed on the local 
NUMA node (and kmalloc first tries local NUMA allocations).

> ---

> diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c
> index e047747d4ecf..e1f42be79729 100644
> --- a/drivers/scsi/hosts.c
> +++ b/drivers/scsi/hosts.c
> @@ -403,12 +403,14 @@ static const struct device_type scsi_host_type = {
>    * Return value:
>    * 	Pointer to a new Scsi_Host
>    **/
> -struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *sht, int privsize)
> +struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *sht, int privsize,
> +				  struct device *dev)
>   {
>   	struct Scsi_Host *shost;
>   	int index;
>   
> -	shost = kzalloc(sizeof(struct Scsi_Host) + privsize, GFP_KERNEL);
> +	shost = kzalloc_node(sizeof(struct Scsi_Host) + privsize, GFP_KERNEL,
> +			     dev ? dev_to_node(dev) : NUMA_NO_NODE);
>   	if (!shost)
>   		return NULL;
>   

> -extern struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *, int);
> +extern struct Scsi_Host *scsi_host_alloc(const struct scsi_host_template *sht,
> +					 int privsize, struct device *dev);
>   extern int __must_check scsi_add_host_with_dma(struct Scsi_Host *,
>   					       struct device *,
>   					       struct device *);


scsi_add_host_with_dma() and scsi_add_host() do assignment of 
shost->dma_dev, so I think that could be moved to scsi_host_alloc().

I can imagine that we always know dev and dma_dev at Scsi_Host alloc 
time (and not just scsi_add_host()) time. However those would be very 
intrusive changes.

Let me consider this more. Maybe we can have a platform device version 
of shost alloc, as I can't imagine that we care about much more. Thanks!

Re: [PATCH v3 2/4] scsi: host: allocate struct Scsi_Host on the NUMA node of the host adapter

[Hannes Reinecke]

On 6/9/26 14:18, Sumit Saxena wrote:
> scsi_host_alloc() used kzalloc(), which always picks an arbitrary node.
> Extend the function to accept a 'struct device *dev' parameter and use
> kzalloc_node() with dev_to_node(dev) so the Scsi_Host struct lands on
> the same NUMA node as the HBA, mirroring the treatment already applied
> to struct scsi_device, struct scsi_target, and shost_data.
> 
> When dev is NULL (legacy ISA/platform drivers without a dma_dev) the
> allocation falls back to NUMA_NO_NODE, preserving existing behaviour.
> 
> Update all in-tree callers:
>    - PCI-based HBA drivers pass &pdev->dev (or the equivalent struct
>      member such as &phba->pcidev->dev, &h->pdev->dev, &ha->pdev->dev)
>      so their host struct is placed on the adapter's node.
>    - Non-PCI drivers (ISA, Amiga, ARM PCMCIA, virtio, Hyper-V, PS3, …)
>      pass NULL.
>    - libfc's libfc_host_alloc() inline helper passes NULL; FC drivers
>      that want NUMA awareness can open-code the call with their pdev.
> 
> Suggested-by: John Garry <john.g.garry@oracle.com>
> Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
> ---
>   drivers/scsi/3w-9xxx.c                    | 2 +-
>   drivers/scsi/3w-sas.c                     | 2 +-
>   drivers/scsi/3w-xxxx.c                    | 2 +-
>   drivers/scsi/53c700.c                     | 2 +-
>   drivers/scsi/BusLogic.c                   | 2 +-
>   drivers/scsi/a100u2w.c                    | 2 +-
>   drivers/scsi/a2091.c                      | 2 +-
>   drivers/scsi/a3000.c                      | 2 +-
>   drivers/scsi/aacraid/linit.c              | 2 +-
>   drivers/scsi/advansys.c                   | 6 +++---
>   drivers/scsi/aha152x.c                    | 2 +-
>   drivers/scsi/aha1542.c                    | 2 +-
>   drivers/scsi/aha1740.c                    | 2 +-
>   drivers/scsi/aic7xxx/aic79xx_osm.c        | 2 +-
>   drivers/scsi/aic7xxx/aic7xxx_osm.c        | 2 +-
>   drivers/scsi/aic94xx/aic94xx_init.c       | 2 +-
>   drivers/scsi/am53c974.c                   | 2 +-
>   drivers/scsi/arcmsr/arcmsr_hba.c          | 3 ++-
>   drivers/scsi/arm/acornscsi.c              | 2 +-
>   drivers/scsi/arm/arxescsi.c               | 2 +-
>   drivers/scsi/arm/cumana_1.c               | 2 +-
>   drivers/scsi/arm/cumana_2.c               | 2 +-
>   drivers/scsi/arm/eesox.c                  | 2 +-
>   drivers/scsi/arm/oak.c                    | 2 +-
>   drivers/scsi/arm/powertec.c               | 2 +-
>   drivers/scsi/atari_scsi.c                 | 2 +-
>   drivers/scsi/atp870u.c                    | 2 +-
>   drivers/scsi/bfa/bfad_im.c                | 2 +-
>   drivers/scsi/csiostor/csio_init.c         | 4 ++--
>   drivers/scsi/dc395x.c                     | 2 +-
>   drivers/scsi/dmx3191d.c                   | 2 +-
>   drivers/scsi/elx/efct/efct_xport.c        | 4 ++--
>   drivers/scsi/esas2r/esas2r_main.c         | 2 +-
>   drivers/scsi/fdomain.c                    | 2 +-
>   drivers/scsi/fnic/fnic_main.c             | 2 +-
>   drivers/scsi/g_NCR5380.c                  | 2 +-
>   drivers/scsi/gvp11.c                      | 2 +-
>   drivers/scsi/hisi_sas/hisi_sas_main.c     | 2 +-
>   drivers/scsi/hisi_sas/hisi_sas_v3_hw.c    | 2 +-
>   drivers/scsi/hosts.c                      | 6 ++++--
>   drivers/scsi/hpsa.c                       | 2 +-
>   drivers/scsi/hptiop.c                     | 2 +-
>   drivers/scsi/ibmvscsi/ibmvfc.c            | 2 +-
>   drivers/scsi/ibmvscsi/ibmvscsi.c          | 2 +-
>   drivers/scsi/imm.c                        | 2 +-
>   drivers/scsi/initio.c                     | 2 +-
>   drivers/scsi/ipr.c                        | 2 +-
>   drivers/scsi/ips.c                        | 2 +-
>   drivers/scsi/isci/init.c                  | 2 +-
>   drivers/scsi/jazz_esp.c                   | 2 +-
>   drivers/scsi/libiscsi.c                   | 2 +-
>   drivers/scsi/lpfc/lpfc_init.c             | 2 +-
>   drivers/scsi/mac53c94.c                   | 2 +-
>   drivers/scsi/mac_esp.c                    | 2 +-
>   drivers/scsi/mac_scsi.c                   | 2 +-
>   drivers/scsi/megaraid.c                   | 2 +-
>   drivers/scsi/megaraid/megaraid_mbox.c     | 2 +-
>   drivers/scsi/megaraid/megaraid_sas_base.c | 2 +-
>   drivers/scsi/mesh.c                       | 2 +-
>   drivers/scsi/mpi3mr/mpi3mr_os.c           | 2 +-
>   drivers/scsi/mpt3sas/mpt3sas_scsih.c      | 4 ++--
>   drivers/scsi/mvme147.c                    | 2 +-
>   drivers/scsi/mvsas/mv_init.c              | 2 +-
>   drivers/scsi/mvumi.c                      | 2 +-
>   drivers/scsi/myrb.c                       | 2 +-
>   drivers/scsi/myrs.c                       | 2 +-
>   drivers/scsi/ncr53c8xx.c                  | 2 +-
>   drivers/scsi/nsp32.c                      | 2 +-
>   drivers/scsi/pcmcia/nsp_cs.c              | 2 +-
>   drivers/scsi/pcmcia/qlogic_stub.c         | 2 +-
>   drivers/scsi/pcmcia/sym53c500_cs.c        | 2 +-
>   drivers/scsi/pm8001/pm8001_init.c         | 2 +-
>   drivers/scsi/pmcraid.c                    | 2 +-
>   drivers/scsi/ppa.c                        | 2 +-
>   drivers/scsi/ps3rom.c                     | 2 +-
>   drivers/scsi/qla1280.c                    | 2 +-
>   drivers/scsi/qla2xxx/qla_mid.c            | 2 +-
>   drivers/scsi/qla2xxx/qla_os.c             | 2 +-
>   drivers/scsi/qlogicfas.c                  | 2 +-
>   drivers/scsi/qlogicpti.c                  | 2 +-
>   drivers/scsi/scsi_debug.c                 | 2 +-
>   drivers/scsi/sgiwd93.c                    | 2 +-
>   drivers/scsi/smartpqi/smartpqi_init.c     | 2 +-
>   drivers/scsi/snic/snic_main.c             | 2 +-
>   drivers/scsi/stex.c                       | 2 +-
>   drivers/scsi/storvsc_drv.c                | 2 +-
>   drivers/scsi/sun3_scsi.c                  | 2 +-
>   drivers/scsi/sun3x_esp.c                  | 2 +-
>   drivers/scsi/sun_esp.c                    | 2 +-
>   drivers/scsi/sym53c8xx_2/sym_glue.c       | 2 +-
>   drivers/scsi/virtio_scsi.c                | 2 +-
>   drivers/scsi/vmw_pvscsi.c                 | 2 +-
>   drivers/scsi/wd719x.c                     | 2 +-
>   drivers/scsi/xen-scsifront.c              | 2 +-
>   drivers/scsi/zorro_esp.c                  | 2 +-
>   include/scsi/libfc.h                      | 2 +-
>   include/scsi/scsi_host.h                  | 3 ++-
>   97 files changed, 107 insertions(+), 103 deletions(-)
> 
Quite a lot of churn for such a (relatively) simple change.

I think it might be better to introduce a new function
(scsi_host_alloc_node() ?) with the additional parameter,
and make scsi_host_alloc() a wrapper around that.

That will reduce the size of this patch immensely.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke                  Kernel Storage Architect
hare@suse.com                               +49 911 74053 688
SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich

Re: [PATCH v3 1/4] scsi: scan: allocate sdev and starget on the NUMA node of the host adapter

[Hannes Reinecke]

On 6/9/26 14:18, Sumit Saxena wrote:
> From: James Rizzo <james.rizzo@broadcom.com>
> 
> When a host adapter is attached to a specific NUMA node, allocating
> scsi_device and scsi_target via kzalloc() may place them on a remote
> node.  All hot-path I/O accesses to these structures then cross the NUMA
> interconnect, adding latency and consuming inter-node bandwidth.
> 
> Use kzalloc_node() with dev_to_node(shost->dma_dev) so allocations land
> on the same node as the HBA, reducing cross-node traffic and improving
> I/O performance on NUMA systems.
> 
> Signed-off-by: James Rizzo <james.rizzo@broadcom.com>
> Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
> ---
>   drivers/scsi/scsi_scan.c | 7 ++++---
>   1 file changed, 4 insertions(+), 3 deletions(-)
> 
Reviewed-by: Hannes Reinecke <hare@kernel.org>

Cheers,

Hannes
-- 
Dr. Hannes Reinecke                  Kernel Storage Architect
hare@suse.com                               +49 911 74053 688
SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich

Re: [PATCH v3 3/4] block: drop shared-tag fairness throttling

[Christoph Hellwig]

Just dropping the fairness was rejected before and there is no
explanation here on why any of that has changed.

On Tue, Jun 09, 2026 at 05:48:02PM +0530, Sumit Saxena wrote:
> From: Bart Van Assche <bvanassche@acm.org>
> 
> Original patch [1] by Bart Van Assche; this version is rebased onto the
> current tree.  In testing it improves IOPS by roughly 16-18% by removing
> the fair-sharing throttle on shared tag queues.
> 
> This patch removes the following code and structure members:
> - The function hctx_may_queue().
> - blk_mq_hw_ctx.nr_active and request_queue.nr_active_requests_shared_tags
>   and also all the code that modifies these two member variables.

.. and besides that, this commit message is still entirely useless
as it doesn't explain any of the thoughts of why this change is safe
and desirable.  While the mechanics above are totally obvious from
the code change itself.

Re: [PATCH v3 3/4] block: drop shared-tag fairness throttling

[Hannes Reinecke]

On 6/9/26 14:18, Sumit Saxena wrote:
> From: Bart Van Assche <bvanassche@acm.org>
> 
> Original patch [1] by Bart Van Assche; this version is rebased onto the
> current tree.  In testing it improves IOPS by roughly 16-18% by removing
> the fair-sharing throttle on shared tag queues.
> 
> This patch removes the following code and structure members:
> - The function hctx_may_queue().
> - blk_mq_hw_ctx.nr_active and request_queue.nr_active_requests_shared_tags
>    and also all the code that modifies these two member variables.
> 
> [1]: https://lore.kernel.org/linux-block/20240529213921.3166462-1-bvanassche@acm.org/
> 
> Signed-off-by: Bart Van Assche <bvanassche@acm.org>
> Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
> ---
>   block/blk-core.c       |   2 -
>   block/blk-mq-debugfs.c |  22 ++++++++-
>   block/blk-mq-tag.c     |   4 --
>   block/blk-mq.c         |  17 +------
>   block/blk-mq.h         | 100 -----------------------------------------
>   include/linux/blk-mq.h |   6 ---
>   include/linux/blkdev.h |   2 -
>   7 files changed, 22 insertions(+), 131 deletions(-)
> 
What tests did you perform?
I'm pretty sure you see an improvement when having just a few drives,
but what about having a lot of them (ie tens of drives)?
The whole point of this was to increase fairness between drives, so
of course removing it will make an individual drive going faster ...

Maybe it's an idea to move the fairness algorithm into an I/O scheduler;
that way we can keep the original behaviour yet get the performance
increase if people want it.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke                  Kernel Storage Architect
hare@suse.de                                +49 911 74053 688
SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich

Re: [PATCH v3 4/4] scsi: use percpu counters for iostat counters in struct scsi_device

[Hannes Reinecke]

On 6/9/26 14:18, Sumit Saxena wrote:
> iorequest_cnt and iodone_cnt are updated on every command dispatch and
> completion, often from different CPUs on high queue depth workloads.
> Using adjacent atomic_t fields causes cache line contention between the
> submission and completion paths.
> 
> Extend the same treatment to ioerr_cnt and iotmo_cnt so all four iostat
> counters in struct scsi_device use struct percpu_counter.
> 
> Suggested-by: John Garry <john.g.garry@oracle.com>
> Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
> ---
>   drivers/scsi/scsi_error.c  |  4 ++--
>   drivers/scsi/scsi_lib.c    | 10 +++++-----
>   drivers/scsi/scsi_scan.c   |  8 ++++++++
>   drivers/scsi/scsi_sysfs.c  | 23 ++++++++++++++---------
>   drivers/scsi/sd.c          |  2 +-
>   include/scsi/scsi_device.h |  9 +++++----
>   6 files changed, 35 insertions(+), 21 deletions(-)
> 
Good idea.

Reviewed-by: Hannes Reinecke <hare@kernel.org>

Cheers,

Hannes
-- 
Dr. Hannes Reinecke                  Kernel Storage Architect
hare@suse.de                                +49 911 74053 688
SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich

Re: [PATCH v3 3/4] block: drop shared-tag fairness throttling

[Sumit Saxena]

[-- Attachment #1.1: Type: text/plain, Size: 1889 bytes --]

On Wed, Jun 10, 2026 at 11:44 AM Christoph Hellwig <hch@infradead.org>
wrote:
>
> Just dropping the fairness was rejected before and there is no
> explanation here on why any of that has changed.

I missed the fact that this patch had been previously rejected.

The motivation for this change stems from performance issue we
encountered due to false sharing of the 'nr_active_requests_shared_tags'
counter
on certain CPU architectures. I initially submitted a patch to move that
counter to
its own cache line to avoid conflicts with 'nr_requests' and other hot
fields
(see:
https://patchwork.kernel.org/project/linux-scsi/patch/20260402074637.92417-3-sumit.saxena@broadcom.com/
).

During the review, Bart shared his work, which eliminates the
counter entirely by removing the fairness throttling. My testing confirmed
that
this approach resolved the performance issues and improved IOPS.
This patch is part of a larger set, and I have reported the cumulative
performance
improvements in the cover letter.

Thanks,
Sumit

>
> On Tue, Jun 09, 2026 at 05:48:02PM +0530, Sumit Saxena wrote:
> > From: Bart Van Assche <bvanassche@acm.org>
> >
> > Original patch [1] by Bart Van Assche; this version is rebased onto the
> > current tree.  In testing it improves IOPS by roughly 16-18% by removing
> > the fair-sharing throttle on shared tag queues.
> >
> > This patch removes the following code and structure members:
> > - The function hctx_may_queue().
> > - blk_mq_hw_ctx.nr_active and
request_queue.nr_active_requests_shared_tags
> >   and also all the code that modifies these two member variables.
>
> .. and besides that, this commit message is still entirely useless
> as it doesn't explain any of the thoughts of why this change is safe
> and desirable.  While the mechanics above are totally obvious from
> the code change itself.
>

[-- Attachment #1.2: Type: text/html, Size: 2485 bytes --]

[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 5469 bytes --]

Re: [PATCH v3 3/4] block: drop shared-tag fairness throttling

[Keith Busch]

On Wed, Jun 10, 2026 at 09:16:11PM +0530, Sumit Saxena wrote:
> The motivation for this change stems from performance issue we
> encountered due to false sharing of the 'nr_active_requests_shared_tags'
> counter
> on certain CPU architectures. I initially submitted a patch to move that
> counter to
> its own cache line to avoid conflicts with 'nr_requests' and other hot
> fields
> (see:
> https://patchwork.kernel.org/project/linux-scsi/patch/20260402074637.92417-3-sumit.saxena@broadcom.com/
> ).
> 
> During the review, Bart shared his work, which eliminates the
> counter entirely by removing the fairness throttling. My testing confirmed
> that
> this approach resolved the performance issues and improved IOPS.
> This patch is part of a larger set, and I have reported the cumulative
> performance
> improvements in the cover letter.

So the problem is just the atomic operation accounting overhead? I
previously thought the device just really needed to consume all the tags
to hit performance.