* [PATCH net-next,rdma-next 0/5][pull request] Prepare for Intel IPU E2000 (GEN3)
@ 2025-05-05 21:20 Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 1/5] iidc/ice/irdma: Rename IDC header file Tony Nguyen
` (4 more replies)
0 siblings, 5 replies; 8+ messages in thread
From: Tony Nguyen @ 2025-05-05 21:20 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, jgg, leon,
linux-rdma, netdev
Cc: Tony Nguyen, tatyana.e.nikolova, david.m.ertman,
przemyslaw.kitszel
This is the first part in introducing RDMA support for idpf.
This shared pull request targets both net-next and rdma-next branches
and is based on tag v6.15-rc1.
IWL reviews:
[v5] https://lore.kernel.org/all/20250416021549.606-1-tatyana.e.nikolova@intel.com/
[v4] https://lore.kernel.org/all/20250225050428.2166-1-tatyana.e.nikolova@intel.com/
[v3] https://lore.kernel.org/all/20250207194931.1569-1-tatyana.e.nikolova@intel.com/
[v2] https://lore.kernel.org/all/20240824031924.421-1-tatyana.e.nikolova@intel.com/
[v1] https://lore.kernel.org/all/20240724233917.704-1-tatyana.e.nikolova@intel.com/
There are a few minor conflicts that exist with this series and the
respective trees:
net-next:
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@@ -9341,9 -9310,8 +9341,10 @@@ ice_setup_tc(struct net_device *netdev
void *type_data)
{
struct ice_netdev_priv *np = netdev_priv(netdev);
+ enum flow_block_binder_type binder_type;
+ struct iidc_rdma_core_dev_info *cdev;
struct ice_pf *pf = np->vsi->back;
+ flow_setup_cb_t *flower_handler;
bool locked = false;
int err;
RDMA for-rc series [1] (showing direct conflicts):
--- a/drivers/infiniband/hw/irdma/main.c
+++ b/drivers/infiniband/hw/irdma/main.c
@@@ -221,8 -221,8 +221,8 @@@ static int irdma_init_interrupts(struc
break;
if (i < IRDMA_MIN_MSIX) {
- for (; i > 0; i--)
+ while (--i >= 0)
- ice_free_rdma_qvector(pf, &rf->msix_entries[i]);
+ ice_free_rdma_qvector(cdev, &rf->msix_entries[i]);
kfree(rf->msix_entries);
return -ENOMEM;
@@@ -245,40 -245,35 +245,42 @@@ static void irdma_deinit_interrupts(str
static void irdma_remove(struct auxiliary_device *aux_dev)
{
...
+ kfree(iwdev->rf);
+
- pr_debug("INIT: Gen2 PF[%d] device remove success\n", PCI_FUNC(pf->pdev->devfn));
+ pr_debug("INIT: Gen2 PF[%d] device remove success\n", PCI_FUNC(cdev_info->pdev->devfn));
}
[1] https://lore.kernel.org/linux-rdma/174514658010.719262.13870226048967432907.b4-ty@kernel.org/
The following changes since commit 0af2f6be1b4281385b618cb86ad946eded089ac8:
Linux 6.15-rc1 (2025-04-06 13:11:33 -0700)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/linux.git for-next
for you to fetch changes up to 8d7be3651c9a6c2b5bc952356536cbd9f4a722b2:
iidc/ice/irdma: Update IDC to support multiple consumers (2025-04-30 13:22:27 -0700)
----------------------------------------------------------------
Tatyana Nikolova says:
To align with review comments, the patch series introducing RDMA
RoCEv2 support for the Intel Infrastructure Processing Unit (IPU)
E2000 line of products is going to be submitted in three parts:
1. Modify ice to use specific and common IIDC definitions and
pass a core device info to irdma.
2. Add RDMA support to idpf and modify idpf to use specific and
common IIDC definitions and pass a core device info to irdma.
3. Add RDMA RoCEv2 support for the E2000 products, referred to as
GEN3 to irdma.
This first part is a 5 patch series based on the original
"iidc/ice/irdma: Update IDC to support multiple consumers" patch
to allow for multiple CORE PCI drivers, using the auxbus.
Patches:
1) Move header file to new name for clarity and replace ice
specific DSCP define with a kernel equivalent one in irdma
2) Unify naming convention
3) Separate header file into common and driver specific info
4) Replace ice specific DSCP define with a kernel equivalent
one in ice
5) Implement core device info struct and update drivers to use it
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
----------------------------------------------------------------
Dave Ertman (4):
iidc/ice/irdma: Rename IDC header file
iidc/ice/irdma: Rename to iidc_* convention
iidc/ice/irdma: Break iidc.h into two headers
iidc/ice/irdma: Update IDC to support multiple consumers
Tatyana Nikolova (1):
ice: Replace ice specific DSCP mapping num with a kernel define
MAINTAINERS | 2 +-
drivers/infiniband/hw/irdma/main.c | 125 +++++++-------
drivers/infiniband/hw/irdma/main.h | 3 +-
drivers/infiniband/hw/irdma/osdep.h | 2 +-
drivers/infiniband/hw/irdma/type.h | 4 +-
drivers/net/ethernet/intel/ice/devlink/devlink.c | 45 +++--
drivers/net/ethernet/intel/ice/ice.h | 6 +-
drivers/net/ethernet/intel/ice/ice_dcb.c | 2 +-
drivers/net/ethernet/intel/ice/ice_dcb_lib.c | 47 +++++-
drivers/net/ethernet/intel/ice/ice_dcb_lib.h | 9 +
drivers/net/ethernet/intel/ice/ice_dcb_nl.c | 4 +-
drivers/net/ethernet/intel/ice/ice_ethtool.c | 8 +-
drivers/net/ethernet/intel/ice/ice_idc.c | 204 ++++++++++++++---------
drivers/net/ethernet/intel/ice/ice_idc_int.h | 5 +-
drivers/net/ethernet/intel/ice/ice_main.c | 18 +-
drivers/net/ethernet/intel/ice/ice_type.h | 6 +-
include/linux/net/intel/iidc.h | 109 ------------
include/linux/net/intel/iidc_rdma.h | 68 ++++++++
include/linux/net/intel/iidc_rdma_ice.h | 70 ++++++++
19 files changed, 449 insertions(+), 288 deletions(-)
delete mode 100644 include/linux/net/intel/iidc.h
create mode 100644 include/linux/net/intel/iidc_rdma.h
create mode 100644 include/linux/net/intel/iidc_rdma_ice.h
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH net-next,rdma-next 1/5] iidc/ice/irdma: Rename IDC header file
2025-05-05 21:20 [PATCH net-next,rdma-next 0/5][pull request] Prepare for Intel IPU E2000 (GEN3) Tony Nguyen
@ 2025-05-05 21:20 ` Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 2/5] iidc/ice/irdma: Rename to iidc_* convention Tony Nguyen
` (3 subsequent siblings)
4 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2025-05-05 21:20 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, jgg, leon,
linux-rdma, netdev
Cc: Dave Ertman, anthony.l.nguyen, tatyana.e.nikolova,
przemyslaw.kitszel
From: Dave Ertman <david.m.ertman@intel.com>
To prepare for the IDC upgrade to support different CORE
PCI drivers, rename header file from iidc.h to iidc_rdma.h
since this files functionality is specifically for RDMA support.
Use net/dscp.h include in irdma osdep.h and DSCP_MAX type.h,
instead of iidc header and define.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Dave Ertman <david.m.ertman@intel.com>
Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
MAINTAINERS | 2 +-
drivers/infiniband/hw/irdma/main.h | 2 +-
drivers/infiniband/hw/irdma/osdep.h | 2 +-
drivers/infiniband/hw/irdma/type.h | 4 ++--
drivers/net/ethernet/intel/ice/ice_idc_int.h | 2 +-
include/linux/net/intel/{iidc.h => iidc_rdma.h} | 6 +++---
6 files changed, 9 insertions(+), 9 deletions(-)
rename include/linux/net/intel/{iidc.h => iidc_rdma.h} (97%)
diff --git a/MAINTAINERS b/MAINTAINERS
index 96b827049501..136309587791 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -11890,7 +11890,7 @@ F: Documentation/networking/device_drivers/ethernet/intel/
F: drivers/net/ethernet/intel/
F: drivers/net/ethernet/intel/*/
F: include/linux/avf/virtchnl.h
-F: include/linux/net/intel/iidc.h
+F: include/linux/net/intel/*/
INTEL ETHERNET PROTOCOL DRIVER FOR RDMA
M: Mustafa Ismail <mustafa.ismail@intel.com>
diff --git a/drivers/infiniband/hw/irdma/main.h b/drivers/infiniband/hw/irdma/main.h
index bb0b6494ccb2..e8083b0c8cb2 100644
--- a/drivers/infiniband/hw/irdma/main.h
+++ b/drivers/infiniband/hw/irdma/main.h
@@ -29,7 +29,7 @@
#include <linux/io-64-nonatomic-lo-hi.h>
#endif
#include <linux/auxiliary_bus.h>
-#include <linux/net/intel/iidc.h>
+#include <linux/net/intel/iidc_rdma.h>
#include <rdma/ib_smi.h>
#include <rdma/ib_verbs.h>
#include <rdma/ib_pack.h>
diff --git a/drivers/infiniband/hw/irdma/osdep.h b/drivers/infiniband/hw/irdma/osdep.h
index 4b4f78288d12..3f73ceacccb6 100644
--- a/drivers/infiniband/hw/irdma/osdep.h
+++ b/drivers/infiniband/hw/irdma/osdep.h
@@ -5,8 +5,8 @@
#include <linux/pci.h>
#include <linux/bitfield.h>
-#include <linux/net/intel/iidc.h>
#include <rdma/ib_verbs.h>
+#include <net/dscp.h>
#define STATS_TIMER_DELAY 60000
diff --git a/drivers/infiniband/hw/irdma/type.h b/drivers/infiniband/hw/irdma/type.h
index 59b34afa867b..527c6da2c1ac 100644
--- a/drivers/infiniband/hw/irdma/type.h
+++ b/drivers/infiniband/hw/irdma/type.h
@@ -567,7 +567,7 @@ struct irdma_sc_vsi {
u8 qos_rel_bw;
u8 qos_prio_type;
u8 stats_idx;
- u8 dscp_map[IIDC_MAX_DSCP_MAPPING];
+ u8 dscp_map[DSCP_MAX];
struct irdma_qos qos[IRDMA_MAX_USER_PRIORITY];
u64 hw_stats_regs[IRDMA_HW_STAT_INDEX_MAX_GEN_1];
bool dscp_mode:1;
@@ -695,7 +695,7 @@ struct irdma_l2params {
u16 qs_handle_list[IRDMA_MAX_USER_PRIORITY];
u16 mtu;
u8 up2tc[IRDMA_MAX_USER_PRIORITY];
- u8 dscp_map[IIDC_MAX_DSCP_MAPPING];
+ u8 dscp_map[DSCP_MAX];
u8 num_tc;
u8 vsi_rel_bw;
u8 vsi_prio_type;
diff --git a/drivers/net/ethernet/intel/ice/ice_idc_int.h b/drivers/net/ethernet/intel/ice/ice_idc_int.h
index 4b0c86757df9..b0c504a6408e 100644
--- a/drivers/net/ethernet/intel/ice/ice_idc_int.h
+++ b/drivers/net/ethernet/intel/ice/ice_idc_int.h
@@ -4,7 +4,7 @@
#ifndef _ICE_IDC_INT_H_
#define _ICE_IDC_INT_H_
-#include <linux/net/intel/iidc.h>
+#include <linux/net/intel/iidc_rdma.h>
struct ice_pf;
diff --git a/include/linux/net/intel/iidc.h b/include/linux/net/intel/iidc_rdma.h
similarity index 97%
rename from include/linux/net/intel/iidc.h
rename to include/linux/net/intel/iidc_rdma.h
index 13274c3def66..0cd75404e459 100644
--- a/include/linux/net/intel/iidc.h
+++ b/include/linux/net/intel/iidc_rdma.h
@@ -1,8 +1,8 @@
/* SPDX-License-Identifier: GPL-2.0 */
/* Copyright (C) 2021, Intel Corporation. */
-#ifndef _IIDC_H_
-#define _IIDC_H_
+#ifndef _IIDC_RDMA_H_
+#define _IIDC_RDMA_H_
#include <linux/auxiliary_bus.h>
#include <linux/dcbnl.h>
@@ -106,4 +106,4 @@ struct iidc_auxiliary_drv {
void (*event_handler)(struct ice_pf *pf, struct iidc_event *event);
};
-#endif /* _IIDC_H_*/
+#endif /* _IIDC_RDMA_H_*/
--
2.47.1
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net-next,rdma-next 2/5] iidc/ice/irdma: Rename to iidc_* convention
2025-05-05 21:20 [PATCH net-next,rdma-next 0/5][pull request] Prepare for Intel IPU E2000 (GEN3) Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 1/5] iidc/ice/irdma: Rename IDC header file Tony Nguyen
@ 2025-05-05 21:20 ` Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 3/5] iidc/ice/irdma: Break iidc.h into two headers Tony Nguyen
` (2 subsequent siblings)
4 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2025-05-05 21:20 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, jgg, leon,
linux-rdma, netdev
Cc: Dave Ertman, anthony.l.nguyen, tatyana.e.nikolova,
przemyslaw.kitszel
From: Dave Ertman <david.m.ertman@intel.com>
In preparation of supporting more than a single core PCI driver
for RDMA, homogenize naming to iidc_rdma_* and IIDC_RDMA_*
form.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Dave Ertman <david.m.ertman@intel.com>
Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/infiniband/hw/irdma/main.c | 41 +++++++++++---------
drivers/net/ethernet/intel/ice/ice_dcb_lib.c | 8 ++--
drivers/net/ethernet/intel/ice/ice_idc.c | 23 ++++++-----
drivers/net/ethernet/intel/ice/ice_idc_int.h | 2 +-
drivers/net/ethernet/intel/ice/ice_main.c | 8 ++--
include/linux/net/intel/iidc_rdma.h | 38 +++++++++---------
6 files changed, 64 insertions(+), 56 deletions(-)
diff --git a/drivers/infiniband/hw/irdma/main.c b/drivers/infiniband/hw/irdma/main.c
index 1ee8969595d3..46489c0ab511 100644
--- a/drivers/infiniband/hw/irdma/main.c
+++ b/drivers/infiniband/hw/irdma/main.c
@@ -61,7 +61,7 @@ static void irdma_log_invalid_mtu(u16 mtu, struct irdma_sc_dev *dev)
}
static void irdma_fill_qos_info(struct irdma_l2params *l2params,
- struct iidc_qos_params *qos_info)
+ struct iidc_rdma_qos_params *qos_info)
{
int i;
@@ -85,12 +85,12 @@ static void irdma_fill_qos_info(struct irdma_l2params *l2params,
}
}
-static void irdma_iidc_event_handler(struct ice_pf *pf, struct iidc_event *event)
+static void irdma_iidc_event_handler(struct ice_pf *pf, struct iidc_rdma_event *event)
{
struct irdma_device *iwdev = dev_get_drvdata(&pf->adev->dev);
struct irdma_l2params l2params = {};
- if (*event->type & BIT(IIDC_EVENT_AFTER_MTU_CHANGE)) {
+ if (*event->type & BIT(IIDC_RDMA_EVENT_AFTER_MTU_CHANGE)) {
ibdev_dbg(&iwdev->ibdev, "CLNT: new MTU = %d\n", iwdev->netdev->mtu);
if (iwdev->vsi.mtu != iwdev->netdev->mtu) {
l2params.mtu = iwdev->netdev->mtu;
@@ -98,13 +98,13 @@ static void irdma_iidc_event_handler(struct ice_pf *pf, struct iidc_event *event
irdma_log_invalid_mtu(l2params.mtu, &iwdev->rf->sc_dev);
irdma_change_l2params(&iwdev->vsi, &l2params);
}
- } else if (*event->type & BIT(IIDC_EVENT_BEFORE_TC_CHANGE)) {
+ } else if (*event->type & BIT(IIDC_RDMA_EVENT_BEFORE_TC_CHANGE)) {
if (iwdev->vsi.tc_change_pending)
return;
irdma_prep_tc_change(iwdev);
- } else if (*event->type & BIT(IIDC_EVENT_AFTER_TC_CHANGE)) {
- struct iidc_qos_params qos_info = {};
+ } else if (*event->type & BIT(IIDC_RDMA_EVENT_AFTER_TC_CHANGE)) {
+ struct iidc_rdma_qos_params qos_info = {};
if (!iwdev->vsi.tc_change_pending)
return;
@@ -116,7 +116,7 @@ static void irdma_iidc_event_handler(struct ice_pf *pf, struct iidc_event *event
if (iwdev->rf->protocol_used != IRDMA_IWARP_PROTOCOL_ONLY)
iwdev->dcb_vlan_mode = qos_info.num_tc > 1 && !l2params.dscp_mode;
irdma_change_l2params(&iwdev->vsi, &l2params);
- } else if (*event->type & BIT(IIDC_EVENT_CRIT_ERR)) {
+ } else if (*event->type & BIT(IIDC_RDMA_EVENT_CRIT_ERR)) {
ibdev_warn(&iwdev->ibdev, "ICE OICR event notification: oicr = 0x%08x\n",
event->reg);
if (event->reg & IRDMAPFINT_OICR_PE_CRITERR_M) {
@@ -245,11 +245,12 @@ static void irdma_deinit_interrupts(struct irdma_pci_f *rf, struct ice_pf *pf)
static void irdma_remove(struct auxiliary_device *aux_dev)
{
- struct iidc_auxiliary_dev *iidc_adev = container_of(aux_dev,
- struct iidc_auxiliary_dev,
- adev);
- struct ice_pf *pf = iidc_adev->pf;
struct irdma_device *iwdev = auxiliary_get_drvdata(aux_dev);
+ struct iidc_rdma_core_auxiliary_dev *iidc_adev;
+ struct ice_pf *pf;
+
+ iidc_adev = container_of(aux_dev, struct iidc_rdma_core_auxiliary_dev, adev);
+ pf = iidc_adev->pf;
irdma_ib_unregister_device(iwdev);
ice_rdma_update_vsi_filter(pf, iwdev->vsi_num, false);
@@ -292,17 +293,19 @@ static void irdma_fill_device_info(struct irdma_device *iwdev, struct ice_pf *pf
static int irdma_probe(struct auxiliary_device *aux_dev, const struct auxiliary_device_id *id)
{
- struct iidc_auxiliary_dev *iidc_adev = container_of(aux_dev,
- struct iidc_auxiliary_dev,
- adev);
- struct ice_pf *pf = iidc_adev->pf;
- struct ice_vsi *vsi = ice_get_main_vsi(pf);
- struct iidc_qos_params qos_info = {};
+ struct iidc_rdma_core_auxiliary_dev *iidc_adev;
+ struct iidc_rdma_qos_params qos_info = {};
+ struct irdma_l2params l2params = {};
struct irdma_device *iwdev;
struct irdma_pci_f *rf;
- struct irdma_l2params l2params = {};
+ struct ice_vsi *vsi;
+ struct ice_pf *pf;
int err;
+ iidc_adev = container_of(aux_dev, struct iidc_rdma_core_auxiliary_dev, adev);
+ pf = iidc_adev->pf;
+ vsi = ice_get_main_vsi(pf);
+
if (!vsi)
return -EIO;
iwdev = ib_alloc_device(irdma_device, ibdev);
@@ -367,7 +370,7 @@ static const struct auxiliary_device_id irdma_auxiliary_id_table[] = {
MODULE_DEVICE_TABLE(auxiliary, irdma_auxiliary_id_table);
-static struct iidc_auxiliary_drv irdma_auxiliary_drv = {
+static struct iidc_rdma_core_auxiliary_drv irdma_auxiliary_drv = {
.adrv = {
.id_table = irdma_auxiliary_id_table,
.probe = irdma_probe,
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
index a7c510832824..fe16c59796db 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
@@ -352,8 +352,8 @@ int ice_pf_dcb_cfg(struct ice_pf *pf, struct ice_dcbx_cfg *new_cfg, bool locked)
struct ice_aqc_port_ets_elem buf = { 0 };
struct ice_dcbx_cfg *old_cfg, *curr_cfg;
struct device *dev = ice_pf_to_dev(pf);
+ struct iidc_rdma_event *event;
int ret = ICE_DCB_NO_HW_CHG;
- struct iidc_event *event;
struct ice_vsi *pf_vsi;
curr_cfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
@@ -405,7 +405,7 @@ int ice_pf_dcb_cfg(struct ice_pf *pf, struct ice_dcbx_cfg *new_cfg, bool locked)
goto free_cfg;
}
- set_bit(IIDC_EVENT_BEFORE_TC_CHANGE, event->type);
+ set_bit(IIDC_RDMA_EVENT_BEFORE_TC_CHANGE, event->type);
ice_send_event_to_aux(pf, event);
kfree(event);
@@ -740,7 +740,7 @@ static int ice_dcb_noncontig_cfg(struct ice_pf *pf)
void ice_pf_dcb_recfg(struct ice_pf *pf, bool locked)
{
struct ice_dcbx_cfg *dcbcfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
- struct iidc_event *event;
+ struct iidc_rdma_event *event;
u8 tc_map = 0;
int v, ret;
@@ -789,7 +789,7 @@ void ice_pf_dcb_recfg(struct ice_pf *pf, bool locked)
if (!event)
return;
- set_bit(IIDC_EVENT_AFTER_TC_CHANGE, event->type);
+ set_bit(IIDC_RDMA_EVENT_AFTER_TC_CHANGE, event->type);
ice_send_event_to_aux(pf, event);
kfree(event);
}
diff --git a/drivers/net/ethernet/intel/ice/ice_idc.c b/drivers/net/ethernet/intel/ice/ice_idc.c
index bab3e81cad5d..d55107077d8c 100644
--- a/drivers/net/ethernet/intel/ice/ice_idc.c
+++ b/drivers/net/ethernet/intel/ice/ice_idc.c
@@ -15,7 +15,8 @@ static DEFINE_XARRAY_ALLOC1(ice_aux_id);
* This function has to be called with a device_lock on the
* pf->adev.dev to avoid race conditions.
*/
-static struct iidc_auxiliary_drv *ice_get_auxiliary_drv(struct ice_pf *pf)
+static
+struct iidc_rdma_core_auxiliary_drv *ice_get_auxiliary_drv(struct ice_pf *pf)
{
struct auxiliary_device *adev;
@@ -23,8 +24,8 @@ static struct iidc_auxiliary_drv *ice_get_auxiliary_drv(struct ice_pf *pf)
if (!adev || !adev->dev.driver)
return NULL;
- return container_of(adev->dev.driver, struct iidc_auxiliary_drv,
- adrv.driver);
+ return container_of(adev->dev.driver,
+ struct iidc_rdma_core_auxiliary_drv, adrv.driver);
}
/**
@@ -32,9 +33,9 @@ static struct iidc_auxiliary_drv *ice_get_auxiliary_drv(struct ice_pf *pf)
* @pf: pointer to PF struct
* @event: event struct
*/
-void ice_send_event_to_aux(struct ice_pf *pf, struct iidc_event *event)
+void ice_send_event_to_aux(struct ice_pf *pf, struct iidc_rdma_event *event)
{
- struct iidc_auxiliary_drv *iadrv;
+ struct iidc_rdma_core_auxiliary_drv *iadrv;
if (WARN_ON_ONCE(!in_task()))
return;
@@ -141,7 +142,8 @@ EXPORT_SYMBOL_GPL(ice_del_rdma_qset);
* @pf: struct for PF
* @reset_type: type of reset
*/
-int ice_rdma_request_reset(struct ice_pf *pf, enum iidc_reset_type reset_type)
+int ice_rdma_request_reset(struct ice_pf *pf,
+ enum iidc_rdma_reset_type reset_type)
{
enum ice_reset_req reset;
@@ -205,7 +207,7 @@ EXPORT_SYMBOL_GPL(ice_rdma_update_vsi_filter);
* @pf: pointer to PF struct
* @qos: set of QoS values
*/
-void ice_get_qos_params(struct ice_pf *pf, struct iidc_qos_params *qos)
+void ice_get_qos_params(struct ice_pf *pf, struct iidc_rdma_qos_params *qos)
{
struct ice_dcbx_cfg *dcbx_cfg;
unsigned int i;
@@ -263,9 +265,10 @@ EXPORT_SYMBOL_GPL(ice_free_rdma_qvector);
*/
static void ice_adev_release(struct device *dev)
{
- struct iidc_auxiliary_dev *iadev;
+ struct iidc_rdma_core_auxiliary_dev *iadev;
- iadev = container_of(dev, struct iidc_auxiliary_dev, adev.dev);
+ iadev = container_of(dev, struct iidc_rdma_core_auxiliary_dev,
+ adev.dev);
kfree(iadev);
}
@@ -275,7 +278,7 @@ static void ice_adev_release(struct device *dev)
*/
int ice_plug_aux_dev(struct ice_pf *pf)
{
- struct iidc_auxiliary_dev *iadev;
+ struct iidc_rdma_core_auxiliary_dev *iadev;
struct auxiliary_device *adev;
int ret;
diff --git a/drivers/net/ethernet/intel/ice/ice_idc_int.h b/drivers/net/ethernet/intel/ice/ice_idc_int.h
index b0c504a6408e..03cd7d8d1aaa 100644
--- a/drivers/net/ethernet/intel/ice/ice_idc_int.h
+++ b/drivers/net/ethernet/intel/ice/ice_idc_int.h
@@ -8,6 +8,6 @@
struct ice_pf;
-void ice_send_event_to_aux(struct ice_pf *pf, struct iidc_event *event);
+void ice_send_event_to_aux(struct ice_pf *pf, struct iidc_rdma_event *event);
#endif /* !_ICE_IDC_INT_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index d390157b59fe..6a20f9f2e608 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -2401,11 +2401,11 @@ static void ice_service_task(struct work_struct *work)
}
if (test_and_clear_bit(ICE_AUX_ERR_PENDING, pf->state)) {
- struct iidc_event *event;
+ struct iidc_rdma_event *event;
event = kzalloc(sizeof(*event), GFP_KERNEL);
if (event) {
- set_bit(IIDC_EVENT_CRIT_ERR, event->type);
+ set_bit(IIDC_RDMA_EVENT_CRIT_ERR, event->type);
/* report the entire OICR value to AUX driver */
swap(event->reg, pf->oicr_err_reg);
ice_send_event_to_aux(pf, event);
@@ -2424,11 +2424,11 @@ static void ice_service_task(struct work_struct *work)
ice_plug_aux_dev(pf);
if (test_and_clear_bit(ICE_FLAG_MTU_CHANGED, pf->flags)) {
- struct iidc_event *event;
+ struct iidc_rdma_event *event;
event = kzalloc(sizeof(*event), GFP_KERNEL);
if (event) {
- set_bit(IIDC_EVENT_AFTER_MTU_CHANGE, event->type);
+ set_bit(IIDC_RDMA_EVENT_AFTER_MTU_CHANGE, event->type);
ice_send_event_to_aux(pf, event);
kfree(event);
}
diff --git a/include/linux/net/intel/iidc_rdma.h b/include/linux/net/intel/iidc_rdma.h
index 0cd75404e459..2b24a9912fa0 100644
--- a/include/linux/net/intel/iidc_rdma.h
+++ b/include/linux/net/intel/iidc_rdma.h
@@ -11,16 +11,16 @@
#include <linux/kernel.h>
#include <linux/netdevice.h>
-enum iidc_event_type {
- IIDC_EVENT_BEFORE_MTU_CHANGE,
- IIDC_EVENT_AFTER_MTU_CHANGE,
- IIDC_EVENT_BEFORE_TC_CHANGE,
- IIDC_EVENT_AFTER_TC_CHANGE,
- IIDC_EVENT_CRIT_ERR,
- IIDC_EVENT_NBITS /* must be last */
+enum iidc_rdma_event_type {
+ IIDC_RDMA_EVENT_BEFORE_MTU_CHANGE,
+ IIDC_RDMA_EVENT_AFTER_MTU_CHANGE,
+ IIDC_RDMA_EVENT_BEFORE_TC_CHANGE,
+ IIDC_RDMA_EVENT_AFTER_TC_CHANGE,
+ IIDC_RDMA_EVENT_CRIT_ERR,
+ IIDC_RDMA_EVENT_NBITS /* must be last */
};
-enum iidc_reset_type {
+enum iidc_rdma_reset_type {
IIDC_PFR,
IIDC_CORER,
IIDC_GLOBR,
@@ -47,7 +47,7 @@ struct iidc_rdma_qset_params {
u8 tc; /* TC branch the Qset should belong to */
};
-struct iidc_qos_info {
+struct iidc_rdma_qos_info {
u64 tc_ctx;
u8 rel_bw;
u8 prio_type;
@@ -56,8 +56,8 @@ struct iidc_qos_info {
};
/* Struct to pass QoS info */
-struct iidc_qos_params {
- struct iidc_qos_info tc_info[IEEE_8021QAZ_MAX_TCS];
+struct iidc_rdma_qos_params {
+ struct iidc_rdma_qos_info tc_info[IEEE_8021QAZ_MAX_TCS];
u8 up2tc[IIDC_MAX_USER_PRIORITY];
u8 vport_relative_bw;
u8 vport_priority_type;
@@ -66,8 +66,8 @@ struct iidc_qos_params {
u8 dscp_map[IIDC_MAX_DSCP_MAPPING];
};
-struct iidc_event {
- DECLARE_BITMAP(type, IIDC_EVENT_NBITS);
+struct iidc_rdma_event {
+ DECLARE_BITMAP(type, IIDC_RDMA_EVENT_NBITS);
u32 reg;
};
@@ -75,9 +75,11 @@ struct ice_pf;
int ice_add_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
int ice_del_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
-int ice_rdma_request_reset(struct ice_pf *pf, enum iidc_reset_type reset_type);
+int ice_rdma_request_reset(struct ice_pf *pf,
+ enum iidc_rdma_reset_type reset_type);
int ice_rdma_update_vsi_filter(struct ice_pf *pf, u16 vsi_id, bool enable);
-void ice_get_qos_params(struct ice_pf *pf, struct iidc_qos_params *qos);
+void ice_get_qos_params(struct ice_pf *pf,
+ struct iidc_rdma_qos_params *qos);
int ice_alloc_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
void ice_free_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
@@ -86,7 +88,7 @@ void ice_free_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
* instance of this struct dedicated to it.
*/
-struct iidc_auxiliary_dev {
+struct iidc_rdma_core_auxiliary_dev {
struct auxiliary_device adev;
struct ice_pf *pf;
};
@@ -96,14 +98,14 @@ struct iidc_auxiliary_dev {
* driver will access these ops by performing a container_of on the
* auxiliary_device->dev.driver.
*/
-struct iidc_auxiliary_drv {
+struct iidc_rdma_core_auxiliary_drv {
struct auxiliary_driver adrv;
/* This event_handler is meant to be a blocking call. For instance,
* when a BEFORE_MTU_CHANGE event comes in, the event_handler will not
* return until the auxiliary driver is ready for the MTU change to
* happen.
*/
- void (*event_handler)(struct ice_pf *pf, struct iidc_event *event);
+ void (*event_handler)(struct ice_pf *pf, struct iidc_rdma_event *event);
};
#endif /* _IIDC_RDMA_H_*/
--
2.47.1
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net-next,rdma-next 3/5] iidc/ice/irdma: Break iidc.h into two headers
2025-05-05 21:20 [PATCH net-next,rdma-next 0/5][pull request] Prepare for Intel IPU E2000 (GEN3) Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 1/5] iidc/ice/irdma: Rename IDC header file Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 2/5] iidc/ice/irdma: Rename to iidc_* convention Tony Nguyen
@ 2025-05-05 21:20 ` Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 4/5] ice: Replace ice specific DSCP mapping num with a kernel define Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to support multiple consumers Tony Nguyen
4 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2025-05-05 21:20 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, jgg, leon,
linux-rdma, netdev
Cc: Dave Ertman, anthony.l.nguyen, tatyana.e.nikolova,
przemyslaw.kitszel
From: Dave Ertman <david.m.ertman@intel.com>
In preparation of supporting more than a single core PCI driver
for RDMA, break the iidc_rdma.h header file into two more focused
headers.
Only the elements universal to all Intel drivers will remain in
the generic iidc_rdma.h header. Move the ice specific information
to an ice specific header file named iidc_rdma_ice.h.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Dave Ertman <david.m.ertman@intel.com>
Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/ice/ice_idc_int.h | 1 +
include/linux/net/intel/iidc_rdma.h | 14 +-------------
include/linux/net/intel/iidc_rdma_ice.h | 19 +++++++++++++++++++
3 files changed, 21 insertions(+), 13 deletions(-)
create mode 100644 include/linux/net/intel/iidc_rdma_ice.h
diff --git a/drivers/net/ethernet/intel/ice/ice_idc_int.h b/drivers/net/ethernet/intel/ice/ice_idc_int.h
index 03cd7d8d1aaa..17dbfcfb6a2a 100644
--- a/drivers/net/ethernet/intel/ice/ice_idc_int.h
+++ b/drivers/net/ethernet/intel/ice/ice_idc_int.h
@@ -5,6 +5,7 @@
#define _ICE_IDC_INT_H_
#include <linux/net/intel/iidc_rdma.h>
+#include <linux/net/intel/iidc_rdma_ice.h>
struct ice_pf;
diff --git a/include/linux/net/intel/iidc_rdma.h b/include/linux/net/intel/iidc_rdma.h
index 2b24a9912fa0..1e8136395154 100644
--- a/include/linux/net/intel/iidc_rdma.h
+++ b/include/linux/net/intel/iidc_rdma.h
@@ -1,5 +1,5 @@
/* SPDX-License-Identifier: GPL-2.0 */
-/* Copyright (C) 2021, Intel Corporation. */
+/* Copyright (C) 2021-2025, Intel Corporation. */
#ifndef _IIDC_RDMA_H_
#define _IIDC_RDMA_H_
@@ -71,18 +71,6 @@ struct iidc_rdma_event {
u32 reg;
};
-struct ice_pf;
-
-int ice_add_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
-int ice_del_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
-int ice_rdma_request_reset(struct ice_pf *pf,
- enum iidc_rdma_reset_type reset_type);
-int ice_rdma_update_vsi_filter(struct ice_pf *pf, u16 vsi_id, bool enable);
-void ice_get_qos_params(struct ice_pf *pf,
- struct iidc_rdma_qos_params *qos);
-int ice_alloc_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
-void ice_free_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
-
/* Structure representing auxiliary driver tailored information about the core
* PCI dev, each auxiliary driver using the IIDC interface will have an
* instance of this struct dedicated to it.
diff --git a/include/linux/net/intel/iidc_rdma_ice.h b/include/linux/net/intel/iidc_rdma_ice.h
new file mode 100644
index 000000000000..78d10003d776
--- /dev/null
+++ b/include/linux/net/intel/iidc_rdma_ice.h
@@ -0,0 +1,19 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2021-2025, Intel Corporation. */
+
+#ifndef _IIDC_RDMA_ICE_H_
+#define _IIDC_RDMA_ICE_H_
+
+struct ice_pf;
+
+int ice_add_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
+int ice_del_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
+int ice_rdma_request_reset(struct ice_pf *pf,
+ enum iidc_rdma_reset_type reset_type);
+int ice_rdma_update_vsi_filter(struct ice_pf *pf, u16 vsi_id, bool enable);
+void ice_get_qos_params(struct ice_pf *pf,
+ struct iidc_rdma_qos_params *qos);
+int ice_alloc_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
+void ice_free_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
+
+#endif /* _IIDC_RDMA_ICE_H_*/
--
2.47.1
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net-next,rdma-next 4/5] ice: Replace ice specific DSCP mapping num with a kernel define
2025-05-05 21:20 [PATCH net-next,rdma-next 0/5][pull request] Prepare for Intel IPU E2000 (GEN3) Tony Nguyen
` (2 preceding siblings ...)
2025-05-05 21:20 ` [PATCH net-next,rdma-next 3/5] iidc/ice/irdma: Break iidc.h into two headers Tony Nguyen
@ 2025-05-05 21:20 ` Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to support multiple consumers Tony Nguyen
4 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2025-05-05 21:20 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, jgg, leon,
linux-rdma, netdev
Cc: Tatyana Nikolova, anthony.l.nguyen, david.m.ertman,
przemyslaw.kitszel, Simon Horman
From: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Replace ice driver specific DSCP mapping number defines
ICE_DSCP_NUM_VAL and IIDC_MAX_DSCP_MAPPING with
an equivalent kernel define DSCP_MAX.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Dave Ertman <david.m.ertman@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/ice/ice_dcb.c | 2 +-
drivers/net/ethernet/intel/ice/ice_dcb_nl.c | 4 ++--
drivers/net/ethernet/intel/ice/ice_idc.c | 2 +-
drivers/net/ethernet/intel/ice/ice_type.h | 6 +++---
include/linux/net/intel/iidc_rdma.h | 4 ++--
5 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb.c b/drivers/net/ethernet/intel/ice/ice_dcb.c
index 74418c445cc4..64737fc62306 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb.c
@@ -1288,7 +1288,7 @@ ice_add_dscp_up_tlv(struct ice_lldp_org_tlv *tlv, struct ice_dcbx_cfg *dcbcfg)
tlv->ouisubtype = htonl(ouisubtype);
/* bytes 0 - 63 - IPv4 DSCP2UP LUT */
- for (i = 0; i < ICE_DSCP_NUM_VAL; i++) {
+ for (i = 0; i < DSCP_MAX; i++) {
/* IPv4 mapping */
buf[i] = dcbcfg->dscp_map[i];
/* IPv6 mapping */
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_nl.c b/drivers/net/ethernet/intel/ice/ice_dcb_nl.c
index 6d50b90a7359..a10c1c8d8697 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb_nl.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb_nl.c
@@ -754,7 +754,7 @@ static int ice_dcbnl_setapp(struct net_device *netdev, struct dcb_app *app)
if (!ice_is_feature_supported(pf, ICE_F_DSCP))
return -EOPNOTSUPP;
- if (app->protocol >= ICE_DSCP_NUM_VAL) {
+ if (app->protocol >= DSCP_MAX) {
netdev_err(netdev, "DSCP value 0x%04X out of range\n",
app->protocol);
return -EINVAL;
@@ -931,7 +931,7 @@ static int ice_dcbnl_delapp(struct net_device *netdev, struct dcb_app *app)
/* if the last DSCP mapping just got deleted, need to switch
* to L2 VLAN QoS mode
*/
- if (bitmap_empty(new_cfg->dscp_mapped, ICE_DSCP_NUM_VAL) &&
+ if (bitmap_empty(new_cfg->dscp_mapped, DSCP_MAX) &&
new_cfg->pfc_mode == ICE_QOS_MODE_DSCP) {
ret = ice_aq_set_pfc_mode(&pf->hw,
ICE_AQC_PFC_VLAN_BASED_PFC,
diff --git a/drivers/net/ethernet/intel/ice/ice_idc.c b/drivers/net/ethernet/intel/ice/ice_idc.c
index d55107077d8c..4421ba024396 100644
--- a/drivers/net/ethernet/intel/ice/ice_idc.c
+++ b/drivers/net/ethernet/intel/ice/ice_idc.c
@@ -225,7 +225,7 @@ void ice_get_qos_params(struct ice_pf *pf, struct iidc_rdma_qos_params *qos)
qos->pfc_mode = dcbx_cfg->pfc_mode;
if (qos->pfc_mode == IIDC_DSCP_PFC_MODE)
- for (i = 0; i < IIDC_MAX_DSCP_MAPPING; i++)
+ for (i = 0; i < DSCP_MAX; i++)
qos->dscp_map[i] = dcbx_cfg->dscp_map[i];
}
EXPORT_SYMBOL_GPL(ice_get_qos_params);
diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h
index 0aab21113cc4..529f978ea45a 100644
--- a/drivers/net/ethernet/intel/ice/ice_type.h
+++ b/drivers/net/ethernet/intel/ice/ice_type.h
@@ -19,6 +19,7 @@
#include "ice_vlan_mode.h"
#include "ice_fwlog.h"
#include <linux/wait.h>
+#include <net/dscp.h>
static inline bool ice_is_tc_ena(unsigned long bitmap, u8 tc)
{
@@ -695,7 +696,6 @@ struct ice_dcb_app_priority_table {
#define ICE_MAX_USER_PRIORITY 8
#define ICE_DCBX_MAX_APPS 64
-#define ICE_DSCP_NUM_VAL 64
#define ICE_LLDPDU_SIZE 1500
#define ICE_TLV_STATUS_OPER 0x1
#define ICE_TLV_STATUS_SYNC 0x2
@@ -718,9 +718,9 @@ struct ice_dcbx_cfg {
u8 pfc_mode;
struct ice_dcb_app_priority_table app[ICE_DCBX_MAX_APPS];
/* when DSCP mapping defined by user set its bit to 1 */
- DECLARE_BITMAP(dscp_mapped, ICE_DSCP_NUM_VAL);
+ DECLARE_BITMAP(dscp_mapped, DSCP_MAX);
/* array holding DSCP -> UP/TC values for DSCP L3 QoS mode */
- u8 dscp_map[ICE_DSCP_NUM_VAL];
+ u8 dscp_map[DSCP_MAX];
u8 dcbx_mode;
#define ICE_DCBX_MODE_CEE 0x1
#define ICE_DCBX_MODE_IEEE 0x2
diff --git a/include/linux/net/intel/iidc_rdma.h b/include/linux/net/intel/iidc_rdma.h
index 1e8136395154..7f1910289534 100644
--- a/include/linux/net/intel/iidc_rdma.h
+++ b/include/linux/net/intel/iidc_rdma.h
@@ -10,6 +10,7 @@
#include <linux/if_ether.h>
#include <linux/kernel.h>
#include <linux/netdevice.h>
+#include <net/dscp.h>
enum iidc_rdma_event_type {
IIDC_RDMA_EVENT_BEFORE_MTU_CHANGE,
@@ -32,7 +33,6 @@ enum iidc_rdma_protocol {
};
#define IIDC_MAX_USER_PRIORITY 8
-#define IIDC_MAX_DSCP_MAPPING 64
#define IIDC_DSCP_PFC_MODE 0x1
/* Struct to hold per RDMA Qset info */
@@ -63,7 +63,7 @@ struct iidc_rdma_qos_params {
u8 vport_priority_type;
u8 num_tc;
u8 pfc_mode;
- u8 dscp_map[IIDC_MAX_DSCP_MAPPING];
+ u8 dscp_map[DSCP_MAX];
};
struct iidc_rdma_event {
--
2.47.1
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to support multiple consumers
2025-05-05 21:20 [PATCH net-next,rdma-next 0/5][pull request] Prepare for Intel IPU E2000 (GEN3) Tony Nguyen
` (3 preceding siblings ...)
2025-05-05 21:20 ` [PATCH net-next,rdma-next 4/5] ice: Replace ice specific DSCP mapping num with a kernel define Tony Nguyen
@ 2025-05-05 21:20 ` Tony Nguyen
2025-05-08 2:43 ` Jakub Kicinski
4 siblings, 1 reply; 8+ messages in thread
From: Tony Nguyen @ 2025-05-05 21:20 UTC (permalink / raw)
To: davem, kuba, pabeni, edumazet, andrew+netdev, jgg, leon,
linux-rdma, netdev
Cc: Dave Ertman, anthony.l.nguyen, tatyana.e.nikolova,
przemyslaw.kitszel
From: Dave Ertman <david.m.ertman@intel.com>
In preparation of supporting more than a single core PCI driver
for RDMA, move ice specific structs like qset_params, qos_info
and qos_params from iidc_rdma.h to iidc_rdma_ice.h.
Previously, the ice driver was just exporting its entire PF struct
to the auxiliary driver, but since each core driver will have its own
different PF struct, implement a universal struct that all core drivers
can provide to the auxiliary driver through the probe call.
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Dave Ertman <david.m.ertman@intel.com>
Co-developed-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Co-developed-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Co-developed-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/infiniband/hw/irdma/main.c | 102 +++++-----
drivers/infiniband/hw/irdma/main.h | 1 +
.../net/ethernet/intel/ice/devlink/devlink.c | 45 ++++-
drivers/net/ethernet/intel/ice/ice.h | 6 +-
drivers/net/ethernet/intel/ice/ice_dcb_lib.c | 39 +++-
drivers/net/ethernet/intel/ice/ice_dcb_lib.h | 9 +
drivers/net/ethernet/intel/ice/ice_ethtool.c | 8 +-
drivers/net/ethernet/intel/ice/ice_idc.c | 189 +++++++++++-------
drivers/net/ethernet/intel/ice/ice_main.c | 10 +-
include/linux/net/intel/iidc_rdma.h | 67 ++-----
include/linux/net/intel/iidc_rdma_ice.h | 69 ++++++-
11 files changed, 345 insertions(+), 200 deletions(-)
diff --git a/drivers/infiniband/hw/irdma/main.c b/drivers/infiniband/hw/irdma/main.c
index 46489c0ab511..abb532bc8ce4 100644
--- a/drivers/infiniband/hw/irdma/main.c
+++ b/drivers/infiniband/hw/irdma/main.c
@@ -1,10 +1,8 @@
// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
/* Copyright (c) 2015 - 2021 Intel Corporation */
#include "main.h"
-#include "../../../net/ethernet/intel/ice/ice.h"
MODULE_ALIAS("i40iw");
-MODULE_AUTHOR("Intel Corporation, <e1000-rdma@lists.sourceforge.net>");
MODULE_DESCRIPTION("Intel(R) Ethernet Protocol Driver for RDMA");
MODULE_LICENSE("Dual BSD/GPL");
@@ -85,9 +83,10 @@ static void irdma_fill_qos_info(struct irdma_l2params *l2params,
}
}
-static void irdma_iidc_event_handler(struct ice_pf *pf, struct iidc_rdma_event *event)
+static void irdma_iidc_event_handler(struct iidc_rdma_core_dev_info *cdev_info,
+ struct iidc_rdma_event *event)
{
- struct irdma_device *iwdev = dev_get_drvdata(&pf->adev->dev);
+ struct irdma_device *iwdev = dev_get_drvdata(&cdev_info->adev->dev);
struct irdma_l2params l2params = {};
if (*event->type & BIT(IIDC_RDMA_EVENT_AFTER_MTU_CHANGE)) {
@@ -104,17 +103,18 @@ static void irdma_iidc_event_handler(struct ice_pf *pf, struct iidc_rdma_event *
irdma_prep_tc_change(iwdev);
} else if (*event->type & BIT(IIDC_RDMA_EVENT_AFTER_TC_CHANGE)) {
- struct iidc_rdma_qos_params qos_info = {};
+ struct iidc_rdma_priv_dev_info *iidc_priv = cdev_info->iidc_priv;
if (!iwdev->vsi.tc_change_pending)
return;
l2params.tc_changed = true;
ibdev_dbg(&iwdev->ibdev, "CLNT: TC Change\n");
- ice_get_qos_params(pf, &qos_info);
- irdma_fill_qos_info(&l2params, &qos_info);
+
+ irdma_fill_qos_info(&l2params, &iidc_priv->qos_info);
if (iwdev->rf->protocol_used != IRDMA_IWARP_PROTOCOL_ONLY)
- iwdev->dcb_vlan_mode = qos_info.num_tc > 1 && !l2params.dscp_mode;
+ iwdev->dcb_vlan_mode =
+ l2params.num_tc > 1 && !l2params.dscp_mode;
irdma_change_l2params(&iwdev->vsi, &l2params);
} else if (*event->type & BIT(IIDC_RDMA_EVENT_CRIT_ERR)) {
ibdev_warn(&iwdev->ibdev, "ICE OICR event notification: oicr = 0x%08x\n",
@@ -151,10 +151,8 @@ static void irdma_iidc_event_handler(struct ice_pf *pf, struct iidc_rdma_event *
*/
static void irdma_request_reset(struct irdma_pci_f *rf)
{
- struct ice_pf *pf = rf->cdev;
-
ibdev_warn(&rf->iwdev->ibdev, "Requesting a reset\n");
- ice_rdma_request_reset(pf, IIDC_PFR);
+ ice_rdma_request_reset(rf->cdev, IIDC_FUNC_RESET);
}
/**
@@ -166,14 +164,15 @@ static int irdma_lan_register_qset(struct irdma_sc_vsi *vsi,
struct irdma_ws_node *tc_node)
{
struct irdma_device *iwdev = vsi->back_vsi;
- struct ice_pf *pf = iwdev->rf->cdev;
+ struct iidc_rdma_core_dev_info *cdev_info;
struct iidc_rdma_qset_params qset = {};
int ret;
+ cdev_info = iwdev->rf->cdev;
qset.qs_handle = tc_node->qs_handle;
qset.tc = tc_node->traffic_class;
qset.vport_id = vsi->vsi_idx;
- ret = ice_add_rdma_qset(pf, &qset);
+ ret = ice_add_rdma_qset(cdev_info, &qset);
if (ret) {
ibdev_dbg(&iwdev->ibdev, "WS: LAN alloc_res for rdma qset failed.\n");
return ret;
@@ -194,19 +193,20 @@ static void irdma_lan_unregister_qset(struct irdma_sc_vsi *vsi,
struct irdma_ws_node *tc_node)
{
struct irdma_device *iwdev = vsi->back_vsi;
- struct ice_pf *pf = iwdev->rf->cdev;
+ struct iidc_rdma_core_dev_info *cdev_info;
struct iidc_rdma_qset_params qset = {};
+ cdev_info = iwdev->rf->cdev;
qset.qs_handle = tc_node->qs_handle;
qset.tc = tc_node->traffic_class;
qset.vport_id = vsi->vsi_idx;
qset.teid = tc_node->l2_sched_node_id;
- if (ice_del_rdma_qset(pf, &qset))
+ if (ice_del_rdma_qset(cdev_info, &qset))
ibdev_dbg(&iwdev->ibdev, "WS: LAN free_res for rdma qset failed.\n");
}
-static int irdma_init_interrupts(struct irdma_pci_f *rf, struct ice_pf *pf)
+static int irdma_init_interrupts(struct irdma_pci_f *rf, struct iidc_rdma_core_dev_info *cdev)
{
int i;
@@ -217,12 +217,12 @@ static int irdma_init_interrupts(struct irdma_pci_f *rf, struct ice_pf *pf)
return -ENOMEM;
for (i = 0; i < rf->msix_count; i++)
- if (ice_alloc_rdma_qvector(pf, &rf->msix_entries[i]))
+ if (ice_alloc_rdma_qvector(cdev, &rf->msix_entries[i]))
break;
if (i < IRDMA_MIN_MSIX) {
for (; i > 0; i--)
- ice_free_rdma_qvector(pf, &rf->msix_entries[i]);
+ ice_free_rdma_qvector(cdev, &rf->msix_entries[i]);
kfree(rf->msix_entries);
return -ENOMEM;
@@ -233,12 +233,12 @@ static int irdma_init_interrupts(struct irdma_pci_f *rf, struct ice_pf *pf)
return 0;
}
-static void irdma_deinit_interrupts(struct irdma_pci_f *rf, struct ice_pf *pf)
+static void irdma_deinit_interrupts(struct irdma_pci_f *rf, struct iidc_rdma_core_dev_info *cdev)
{
int i;
for (i = 0; i < rf->msix_count; i++)
- ice_free_rdma_qvector(pf, &rf->msix_entries[i]);
+ ice_free_rdma_qvector(cdev, &rf->msix_entries[i]);
kfree(rf->msix_entries);
}
@@ -247,41 +247,49 @@ static void irdma_remove(struct auxiliary_device *aux_dev)
{
struct irdma_device *iwdev = auxiliary_get_drvdata(aux_dev);
struct iidc_rdma_core_auxiliary_dev *iidc_adev;
- struct ice_pf *pf;
+ struct iidc_rdma_core_dev_info *cdev_info;
iidc_adev = container_of(aux_dev, struct iidc_rdma_core_auxiliary_dev, adev);
- pf = iidc_adev->pf;
+ cdev_info = iidc_adev->cdev_info;
+ ice_rdma_update_vsi_filter(cdev_info, iwdev->vsi_num, false);
irdma_ib_unregister_device(iwdev);
- ice_rdma_update_vsi_filter(pf, iwdev->vsi_num, false);
- irdma_deinit_interrupts(iwdev->rf, pf);
+ irdma_deinit_interrupts(iwdev->rf, cdev_info);
- pr_debug("INIT: Gen2 PF[%d] device remove success\n", PCI_FUNC(pf->pdev->devfn));
+ pr_debug("INIT: Gen2 PF[%d] device remove success\n", PCI_FUNC(cdev_info->pdev->devfn));
}
-static void irdma_fill_device_info(struct irdma_device *iwdev, struct ice_pf *pf,
- struct ice_vsi *vsi)
+static void irdma_fill_device_info(struct irdma_device *iwdev,
+ struct iidc_rdma_core_dev_info *cdev_info)
{
+ struct iidc_rdma_priv_dev_info *iidc_priv = cdev_info->iidc_priv;
struct irdma_pci_f *rf = iwdev->rf;
- rf->cdev = pf;
+ rf->sc_dev.hw = &rf->hw;
+ rf->iwdev = iwdev;
+ rf->cdev = cdev_info;
+ rf->hw.hw_addr = iidc_priv->hw_addr;
+ rf->pcidev = cdev_info->pdev;
+ rf->hw.device = &rf->pcidev->dev;
+ rf->pf_id = iidc_priv->pf_id;
rf->gen_ops.register_qset = irdma_lan_register_qset;
rf->gen_ops.unregister_qset = irdma_lan_unregister_qset;
- rf->hw.hw_addr = pf->hw.hw_addr;
- rf->pcidev = pf->pdev;
- rf->pf_id = pf->hw.pf_id;
- rf->default_vsi.vsi_idx = vsi->vsi_num;
- rf->protocol_used = pf->rdma_mode & IIDC_RDMA_PROTOCOL_ROCEV2 ?
- IRDMA_ROCE_PROTOCOL_ONLY : IRDMA_IWARP_PROTOCOL_ONLY;
+
+ rf->default_vsi.vsi_idx = iidc_priv->vport_id;
+ rf->protocol_used =
+ cdev_info->rdma_protocol == IIDC_RDMA_PROTOCOL_ROCEV2 ?
+ IRDMA_ROCE_PROTOCOL_ONLY : IRDMA_IWARP_PROTOCOL_ONLY;
rf->rdma_ver = IRDMA_GEN_2;
rf->rsrc_profile = IRDMA_HMC_PROFILE_DEFAULT;
rf->rst_to = IRDMA_RST_TIMEOUT_HZ;
rf->gen_ops.request_reset = irdma_request_reset;
rf->limits_sel = 7;
rf->iwdev = iwdev;
+
mutex_init(&iwdev->ah_tbl_lock);
- iwdev->netdev = vsi->netdev;
- iwdev->vsi_num = vsi->vsi_num;
+
+ iwdev->netdev = iidc_priv->netdev;
+ iwdev->vsi_num = iidc_priv->vport_id;
iwdev->init_state = INITIAL_STATE;
iwdev->roce_cwnd = IRDMA_ROCE_CWND_DEFAULT;
iwdev->roce_ackcreds = IRDMA_ROCE_ACKCREDS_DEFAULT;
@@ -294,20 +302,17 @@ static void irdma_fill_device_info(struct irdma_device *iwdev, struct ice_pf *pf
static int irdma_probe(struct auxiliary_device *aux_dev, const struct auxiliary_device_id *id)
{
struct iidc_rdma_core_auxiliary_dev *iidc_adev;
- struct iidc_rdma_qos_params qos_info = {};
+ struct iidc_rdma_core_dev_info *cdev_info;
+ struct iidc_rdma_priv_dev_info *iidc_priv;
struct irdma_l2params l2params = {};
struct irdma_device *iwdev;
struct irdma_pci_f *rf;
- struct ice_vsi *vsi;
- struct ice_pf *pf;
int err;
iidc_adev = container_of(aux_dev, struct iidc_rdma_core_auxiliary_dev, adev);
- pf = iidc_adev->pf;
- vsi = ice_get_main_vsi(pf);
+ cdev_info = iidc_adev->cdev_info;
+ iidc_priv = cdev_info->iidc_priv;
- if (!vsi)
- return -EIO;
iwdev = ib_alloc_device(irdma_device, ibdev);
if (!iwdev)
return -ENOMEM;
@@ -317,10 +322,10 @@ static int irdma_probe(struct auxiliary_device *aux_dev, const struct auxiliary_
return -ENOMEM;
}
- irdma_fill_device_info(iwdev, pf, vsi);
+ irdma_fill_device_info(iwdev, cdev_info);
rf = iwdev->rf;
- err = irdma_init_interrupts(rf, pf);
+ err = irdma_init_interrupts(rf, cdev_info);
if (err)
goto err_init_interrupts;
@@ -329,8 +334,7 @@ static int irdma_probe(struct auxiliary_device *aux_dev, const struct auxiliary_
goto err_ctrl_init;
l2params.mtu = iwdev->netdev->mtu;
- ice_get_qos_params(pf, &qos_info);
- irdma_fill_qos_info(&l2params, &qos_info);
+ irdma_fill_qos_info(&l2params, &iidc_priv->qos_info);
if (iwdev->rf->protocol_used != IRDMA_IWARP_PROTOCOL_ONLY)
iwdev->dcb_vlan_mode = l2params.num_tc > 1 && !l2params.dscp_mode;
@@ -342,7 +346,7 @@ static int irdma_probe(struct auxiliary_device *aux_dev, const struct auxiliary_
if (err)
goto err_ibreg;
- ice_rdma_update_vsi_filter(pf, iwdev->vsi_num, true);
+ ice_rdma_update_vsi_filter(cdev_info, iwdev->vsi_num, true);
ibdev_dbg(&iwdev->ibdev, "INIT: Gen2 PF[%d] device probe success\n", PCI_FUNC(rf->pcidev->devfn));
auxiliary_set_drvdata(aux_dev, iwdev);
@@ -354,7 +358,7 @@ static int irdma_probe(struct auxiliary_device *aux_dev, const struct auxiliary_
err_rt_init:
irdma_ctrl_deinit_hw(rf);
err_ctrl_init:
- irdma_deinit_interrupts(rf, pf);
+ irdma_deinit_interrupts(rf, cdev_info);
err_init_interrupts:
kfree(iwdev->rf);
ib_dealloc_device(&iwdev->ibdev);
diff --git a/drivers/infiniband/hw/irdma/main.h b/drivers/infiniband/hw/irdma/main.h
index e8083b0c8cb2..674acc952168 100644
--- a/drivers/infiniband/hw/irdma/main.h
+++ b/drivers/infiniband/hw/irdma/main.h
@@ -30,6 +30,7 @@
#endif
#include <linux/auxiliary_bus.h>
#include <linux/net/intel/iidc_rdma.h>
+#include <linux/net/intel/iidc_rdma_ice.h>
#include <rdma/ib_smi.h>
#include <rdma/ib_verbs.h>
#include <rdma/ib_pack.h>
diff --git a/drivers/net/ethernet/intel/ice/devlink/devlink.c b/drivers/net/ethernet/intel/ice/devlink/devlink.c
index fcb199efbea5..4af60e2f37df 100644
--- a/drivers/net/ethernet/intel/ice/devlink/devlink.c
+++ b/drivers/net/ethernet/intel/ice/devlink/devlink.c
@@ -1339,8 +1339,13 @@ ice_devlink_enable_roce_get(struct devlink *devlink, u32 id,
struct devlink_param_gset_ctx *ctx)
{
struct ice_pf *pf = devlink_priv(devlink);
+ struct iidc_rdma_core_dev_info *cdev;
- ctx->val.vbool = pf->rdma_mode & IIDC_RDMA_PROTOCOL_ROCEV2 ? true : false;
+ cdev = pf->cdev_info;
+ if (!cdev)
+ return -ENODEV;
+
+ ctx->val.vbool = !!(cdev->rdma_protocol & IIDC_RDMA_PROTOCOL_ROCEV2);
return 0;
}
@@ -1350,19 +1355,24 @@ static int ice_devlink_enable_roce_set(struct devlink *devlink, u32 id,
struct netlink_ext_ack *extack)
{
struct ice_pf *pf = devlink_priv(devlink);
+ struct iidc_rdma_core_dev_info *cdev;
bool roce_ena = ctx->val.vbool;
int ret;
+ cdev = pf->cdev_info;
+ if (!cdev)
+ return -ENODEV;
+
if (!roce_ena) {
ice_unplug_aux_dev(pf);
- pf->rdma_mode &= ~IIDC_RDMA_PROTOCOL_ROCEV2;
+ cdev->rdma_protocol &= ~IIDC_RDMA_PROTOCOL_ROCEV2;
return 0;
}
- pf->rdma_mode |= IIDC_RDMA_PROTOCOL_ROCEV2;
+ cdev->rdma_protocol |= IIDC_RDMA_PROTOCOL_ROCEV2;
ret = ice_plug_aux_dev(pf);
if (ret)
- pf->rdma_mode &= ~IIDC_RDMA_PROTOCOL_ROCEV2;
+ cdev->rdma_protocol &= ~IIDC_RDMA_PROTOCOL_ROCEV2;
return ret;
}
@@ -1373,11 +1383,16 @@ ice_devlink_enable_roce_validate(struct devlink *devlink, u32 id,
struct netlink_ext_ack *extack)
{
struct ice_pf *pf = devlink_priv(devlink);
+ struct iidc_rdma_core_dev_info *cdev;
+
+ cdev = pf->cdev_info;
+ if (!cdev)
+ return -ENODEV;
if (!test_bit(ICE_FLAG_RDMA_ENA, pf->flags))
return -EOPNOTSUPP;
- if (pf->rdma_mode & IIDC_RDMA_PROTOCOL_IWARP) {
+ if (cdev->rdma_protocol & IIDC_RDMA_PROTOCOL_IWARP) {
NL_SET_ERR_MSG_MOD(extack, "iWARP is currently enabled. This device cannot enable iWARP and RoCEv2 simultaneously");
return -EOPNOTSUPP;
}
@@ -1390,8 +1405,13 @@ ice_devlink_enable_iw_get(struct devlink *devlink, u32 id,
struct devlink_param_gset_ctx *ctx)
{
struct ice_pf *pf = devlink_priv(devlink);
+ struct iidc_rdma_core_dev_info *cdev;
- ctx->val.vbool = pf->rdma_mode & IIDC_RDMA_PROTOCOL_IWARP;
+ cdev = pf->cdev_info;
+ if (!cdev)
+ return -ENODEV;
+
+ ctx->val.vbool = !!(cdev->rdma_protocol & IIDC_RDMA_PROTOCOL_IWARP);
return 0;
}
@@ -1401,19 +1421,24 @@ static int ice_devlink_enable_iw_set(struct devlink *devlink, u32 id,
struct netlink_ext_ack *extack)
{
struct ice_pf *pf = devlink_priv(devlink);
+ struct iidc_rdma_core_dev_info *cdev;
bool iw_ena = ctx->val.vbool;
int ret;
+ cdev = pf->cdev_info;
+ if (!cdev)
+ return -ENODEV;
+
if (!iw_ena) {
ice_unplug_aux_dev(pf);
- pf->rdma_mode &= ~IIDC_RDMA_PROTOCOL_IWARP;
+ cdev->rdma_protocol &= ~IIDC_RDMA_PROTOCOL_IWARP;
return 0;
}
- pf->rdma_mode |= IIDC_RDMA_PROTOCOL_IWARP;
+ cdev->rdma_protocol |= IIDC_RDMA_PROTOCOL_IWARP;
ret = ice_plug_aux_dev(pf);
if (ret)
- pf->rdma_mode &= ~IIDC_RDMA_PROTOCOL_IWARP;
+ cdev->rdma_protocol &= ~IIDC_RDMA_PROTOCOL_IWARP;
return ret;
}
@@ -1428,7 +1453,7 @@ ice_devlink_enable_iw_validate(struct devlink *devlink, u32 id,
if (!test_bit(ICE_FLAG_RDMA_ENA, pf->flags))
return -EOPNOTSUPP;
- if (pf->rdma_mode & IIDC_RDMA_PROTOCOL_ROCEV2) {
+ if (pf->cdev_info->rdma_protocol & IIDC_RDMA_PROTOCOL_ROCEV2) {
NL_SET_ERR_MSG_MOD(extack, "RoCEv2 is currently enabled. This device cannot enable iWARP and RoCEv2 simultaneously");
return -EOPNOTSUPP;
}
diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h
index fd083647c14a..e27d9044bcb3 100644
--- a/drivers/net/ethernet/intel/ice/ice.h
+++ b/drivers/net/ethernet/intel/ice/ice.h
@@ -401,7 +401,6 @@ struct ice_vsi {
u16 req_rxq; /* User requested Rx queues */
u16 num_rx_desc;
u16 num_tx_desc;
- u16 qset_handle[ICE_MAX_TRAFFIC_CLASS];
struct ice_tc_cfg tc_cfg;
struct bpf_prog *xdp_prog;
struct ice_tx_ring **xdp_rings; /* XDP ring array */
@@ -557,7 +556,6 @@ struct ice_pf {
struct devlink_port devlink_port;
/* OS reserved IRQ details */
- struct msix_entry *msix_entries;
struct ice_irq_tracker irq_tracker;
struct ice_virt_irq_tracker virt_irq_tracker;
@@ -592,7 +590,6 @@ struct ice_pf {
struct gnss_serial *gnss_serial;
struct gnss_device *gnss_dev;
u16 num_rdma_msix; /* Total MSIX vectors for RDMA driver */
- u16 rdma_base_vector;
/* spinlock to protect the AdminQ wait list */
spinlock_t aq_wait_lock;
@@ -625,14 +622,12 @@ struct ice_pf {
struct ice_hw_port_stats stats_prev;
struct ice_hw hw;
u8 stat_prev_loaded:1; /* has previous stats been loaded */
- u8 rdma_mode;
u16 dcbx_cap;
u32 tx_timeout_count;
unsigned long tx_timeout_last_recovery;
u32 tx_timeout_recovery_level;
char int_name[ICE_INT_NAME_STR_LEN];
char int_name_ll_ts[ICE_INT_NAME_STR_LEN];
- struct auxiliary_device *adev;
int aux_idx;
u32 sw_int_count;
/* count of tc_flower filters specific to channel (aka where filter
@@ -664,6 +659,7 @@ struct ice_pf {
struct ice_dplls dplls;
struct device *hwmon_dev;
struct ice_health health_reporters;
+ struct iidc_rdma_core_dev_info *cdev_info;
u8 num_quanta_prof_used;
};
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
index fe16c59796db..c5ef33c100d6 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
@@ -740,6 +740,8 @@ static int ice_dcb_noncontig_cfg(struct ice_pf *pf)
void ice_pf_dcb_recfg(struct ice_pf *pf, bool locked)
{
struct ice_dcbx_cfg *dcbcfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
+ struct iidc_rdma_priv_dev_info *privd;
+ struct iidc_rdma_core_dev_info *cdev;
struct iidc_rdma_event *event;
u8 tc_map = 0;
int v, ret;
@@ -783,7 +785,11 @@ void ice_pf_dcb_recfg(struct ice_pf *pf, bool locked)
if (vsi->type == ICE_VSI_PF)
ice_dcbnl_set_all(vsi);
}
- if (!locked) {
+
+ cdev = pf->cdev_info;
+ if (cdev && !locked) {
+ privd = cdev->iidc_priv;
+ ice_setup_dcb_qos_info(pf, &privd->qos_info);
/* Notify the AUX drivers that TC change is finished */
event = kzalloc(sizeof(*event), GFP_KERNEL);
if (!event)
@@ -944,6 +950,37 @@ ice_tx_prepare_vlan_flags_dcb(struct ice_tx_ring *tx_ring,
}
}
+/**
+ * ice_setup_dcb_qos_info - Setup DCB QoS information
+ * @pf: ptr to ice_pf
+ * @qos_info: QoS param instance
+ */
+void ice_setup_dcb_qos_info(struct ice_pf *pf, struct iidc_rdma_qos_params *qos_info)
+{
+ struct ice_dcbx_cfg *dcbx_cfg;
+ unsigned int i;
+ u32 up2tc;
+
+ if (!pf || !qos_info)
+ return;
+
+ dcbx_cfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
+ up2tc = rd32(&pf->hw, PRTDCB_TUP2TC);
+
+ qos_info->num_tc = ice_dcb_get_num_tc(dcbx_cfg);
+
+ for (i = 0; i < IIDC_MAX_USER_PRIORITY; i++)
+ qos_info->up2tc[i] = (up2tc >> (i * 3)) & 0x7;
+
+ for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++)
+ qos_info->tc_info[i].rel_bw = dcbx_cfg->etscfg.tcbwtable[i];
+
+ qos_info->pfc_mode = dcbx_cfg->pfc_mode;
+ if (qos_info->pfc_mode == IIDC_DSCP_PFC_MODE)
+ for (i = 0; i < DSCP_MAX; i++)
+ qos_info->dscp_map[i] = dcbx_cfg->dscp_map[i];
+}
+
/**
* ice_dcb_is_mib_change_pending - Check if MIB change is pending
* @state: MIB change state
diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_lib.h b/drivers/net/ethernet/intel/ice/ice_dcb_lib.h
index 800879a88c5e..da9ba814b4e8 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb_lib.h
+++ b/drivers/net/ethernet/intel/ice/ice_dcb_lib.h
@@ -31,6 +31,9 @@ void
ice_tx_prepare_vlan_flags_dcb(struct ice_tx_ring *tx_ring,
struct ice_tx_buf *first);
void
+ice_setup_dcb_qos_info(struct ice_pf *pf,
+ struct iidc_rdma_qos_params *qos_info);
+void
ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf,
struct ice_rq_event_info *event);
/**
@@ -134,5 +137,11 @@ static inline void ice_update_dcb_stats(struct ice_pf *pf) { }
static inline void
ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf, struct ice_rq_event_info *event) { }
static inline void ice_set_cgd_num(struct ice_tlan_ctx *tlan_ctx, u8 dcb_tc) { }
+static inline void
+ice_setup_dcb_qos_info(struct ice_pf *pf, struct iidc_rdma_qos_params *qos_info)
+{
+ qos_info->num_tc = 1;
+ qos_info->tc_info[0].rel_bw = 100;
+}
#endif /* CONFIG_DCB */
#endif /* _ICE_DCB_LIB_H_ */
diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c
index 7c2dc347e4e5..46fbcd391a80 100644
--- a/drivers/net/ethernet/intel/ice/ice_ethtool.c
+++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c
@@ -3964,11 +3964,11 @@ static int ice_set_channels(struct net_device *dev, struct ethtool_channels *ch)
return -EINVAL;
}
- if (pf->adev) {
+ if (pf->cdev_info && pf->cdev_info->adev) {
mutex_lock(&pf->adev_mutex);
- device_lock(&pf->adev->dev);
+ device_lock(&pf->cdev_info->adev->dev);
locked = true;
- if (pf->adev->dev.driver) {
+ if (pf->cdev_info->adev->dev.driver) {
netdev_err(dev, "Cannot change channels when RDMA is active\n");
ret = -EBUSY;
goto adev_unlock;
@@ -3987,7 +3987,7 @@ static int ice_set_channels(struct net_device *dev, struct ethtool_channels *ch)
adev_unlock:
if (locked) {
- device_unlock(&pf->adev->dev);
+ device_unlock(&pf->cdev_info->adev->dev);
mutex_unlock(&pf->adev_mutex);
}
return ret;
diff --git a/drivers/net/ethernet/intel/ice/ice_idc.c b/drivers/net/ethernet/intel/ice/ice_idc.c
index 4421ba024396..121089e3b125 100644
--- a/drivers/net/ethernet/intel/ice/ice_idc.c
+++ b/drivers/net/ethernet/intel/ice/ice_idc.c
@@ -9,18 +9,20 @@
static DEFINE_XARRAY_ALLOC1(ice_aux_id);
/**
- * ice_get_auxiliary_drv - retrieve iidc_auxiliary_drv struct
- * @pf: pointer to PF struct
+ * ice_get_auxiliary_drv - retrieve iidc_rdma_core_auxiliary_drv struct
+ * @cdev: pointer to iidc_rdma_core_dev_info struct
*
* This function has to be called with a device_lock on the
- * pf->adev.dev to avoid race conditions.
+ * cdev->adev.dev to avoid race conditions.
+ *
+ * Return: pointer to the matched auxiliary driver struct
*/
-static
-struct iidc_rdma_core_auxiliary_drv *ice_get_auxiliary_drv(struct ice_pf *pf)
+static struct iidc_rdma_core_auxiliary_drv *
+ice_get_auxiliary_drv(struct iidc_rdma_core_dev_info *cdev)
{
struct auxiliary_device *adev;
- adev = pf->adev;
+ adev = cdev->adev;
if (!adev || !adev->dev.driver)
return NULL;
@@ -36,41 +38,51 @@ struct iidc_rdma_core_auxiliary_drv *ice_get_auxiliary_drv(struct ice_pf *pf)
void ice_send_event_to_aux(struct ice_pf *pf, struct iidc_rdma_event *event)
{
struct iidc_rdma_core_auxiliary_drv *iadrv;
+ struct iidc_rdma_core_dev_info *cdev;
if (WARN_ON_ONCE(!in_task()))
return;
+ cdev = pf->cdev_info;
+ if (!cdev)
+ return;
+
mutex_lock(&pf->adev_mutex);
- if (!pf->adev)
+ if (!cdev->adev)
goto finish;
- device_lock(&pf->adev->dev);
- iadrv = ice_get_auxiliary_drv(pf);
+ device_lock(&cdev->adev->dev);
+ iadrv = ice_get_auxiliary_drv(cdev);
if (iadrv && iadrv->event_handler)
- iadrv->event_handler(pf, event);
- device_unlock(&pf->adev->dev);
+ iadrv->event_handler(cdev, event);
+ device_unlock(&cdev->adev->dev);
finish:
mutex_unlock(&pf->adev_mutex);
}
/**
* ice_add_rdma_qset - Add Leaf Node for RDMA Qset
- * @pf: PF struct
+ * @cdev: pointer to iidc_rdma_core_dev_info struct
* @qset: Resource to be allocated
+ *
+ * Return: Zero on success or error code encountered
*/
-int ice_add_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset)
+int ice_add_rdma_qset(struct iidc_rdma_core_dev_info *cdev,
+ struct iidc_rdma_qset_params *qset)
{
u16 max_rdmaqs[ICE_MAX_TRAFFIC_CLASS];
struct ice_vsi *vsi;
struct device *dev;
+ struct ice_pf *pf;
u32 qset_teid;
u16 qs_handle;
int status;
int i;
- if (WARN_ON(!pf || !qset))
+ if (WARN_ON(!cdev || !qset))
return -EINVAL;
+ pf = pci_get_drvdata(cdev->pdev);
dev = ice_pf_to_dev(pf);
if (!ice_is_rdma_ena(pf))
@@ -101,7 +113,6 @@ int ice_add_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset)
dev_err(dev, "Failed VSI RDMA Qset enable\n");
return status;
}
- vsi->qset_handle[qset->tc] = qset->qs_handle;
qset->teid = qset_teid;
return 0;
@@ -110,18 +121,23 @@ EXPORT_SYMBOL_GPL(ice_add_rdma_qset);
/**
* ice_del_rdma_qset - Delete leaf node for RDMA Qset
- * @pf: PF struct
+ * @cdev: pointer to iidc_rdma_core_dev_info struct
* @qset: Resource to be freed
+ *
+ * Return: Zero on success, error code on failure
*/
-int ice_del_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset)
+int ice_del_rdma_qset(struct iidc_rdma_core_dev_info *cdev,
+ struct iidc_rdma_qset_params *qset)
{
struct ice_vsi *vsi;
+ struct ice_pf *pf;
u32 teid;
u16 q_id;
- if (WARN_ON(!pf || !qset))
+ if (WARN_ON(!cdev || !qset))
return -EINVAL;
+ pf = pci_get_drvdata(cdev->pdev);
vsi = ice_find_vsi(pf, qset->vport_id);
if (!vsi) {
dev_err(ice_pf_to_dev(pf), "RDMA Invalid VSI\n");
@@ -131,37 +147,36 @@ int ice_del_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset)
q_id = qset->qs_handle;
teid = qset->teid;
- vsi->qset_handle[qset->tc] = 0;
-
return ice_dis_vsi_rdma_qset(vsi->port_info, 1, &teid, &q_id);
}
EXPORT_SYMBOL_GPL(ice_del_rdma_qset);
/**
* ice_rdma_request_reset - accept request from RDMA to perform a reset
- * @pf: struct for PF
+ * @cdev: pointer to iidc_rdma_core_dev_info struct
* @reset_type: type of reset
+ *
+ * Return: Zero on success, error code on failure
*/
-int ice_rdma_request_reset(struct ice_pf *pf,
+int ice_rdma_request_reset(struct iidc_rdma_core_dev_info *cdev,
enum iidc_rdma_reset_type reset_type)
{
enum ice_reset_req reset;
+ struct ice_pf *pf;
- if (WARN_ON(!pf))
+ if (WARN_ON(!cdev))
return -EINVAL;
+ pf = pci_get_drvdata(cdev->pdev);
+
switch (reset_type) {
- case IIDC_PFR:
+ case IIDC_FUNC_RESET:
reset = ICE_RESET_PFR;
break;
- case IIDC_CORER:
+ case IIDC_DEV_RESET:
reset = ICE_RESET_CORER;
break;
- case IIDC_GLOBR:
- reset = ICE_RESET_GLOBR;
- break;
default:
- dev_err(ice_pf_to_dev(pf), "incorrect reset request\n");
return -EINVAL;
}
@@ -171,18 +186,23 @@ EXPORT_SYMBOL_GPL(ice_rdma_request_reset);
/**
* ice_rdma_update_vsi_filter - update main VSI filters for RDMA
- * @pf: pointer to struct for PF
+ * @cdev: pointer to iidc_rdma_core_dev_info struct
* @vsi_id: VSI HW idx to update filter on
* @enable: bool whether to enable or disable filters
+ *
+ * Return: Zero on success, error code on failure
*/
-int ice_rdma_update_vsi_filter(struct ice_pf *pf, u16 vsi_id, bool enable)
+int ice_rdma_update_vsi_filter(struct iidc_rdma_core_dev_info *cdev,
+ u16 vsi_id, bool enable)
{
struct ice_vsi *vsi;
+ struct ice_pf *pf;
int status;
- if (WARN_ON(!pf))
+ if (WARN_ON(!cdev))
return -EINVAL;
+ pf = pci_get_drvdata(cdev->pdev);
vsi = ice_find_vsi(pf, vsi_id);
if (!vsi)
return -EINVAL;
@@ -203,37 +223,23 @@ int ice_rdma_update_vsi_filter(struct ice_pf *pf, u16 vsi_id, bool enable)
EXPORT_SYMBOL_GPL(ice_rdma_update_vsi_filter);
/**
- * ice_get_qos_params - parse QoS params for RDMA consumption
- * @pf: pointer to PF struct
- * @qos: set of QoS values
+ * ice_alloc_rdma_qvector - alloc vector resources reserved for RDMA driver
+ * @cdev: pointer to iidc_rdma_core_dev_info struct
+ * @entry: MSI-X entry to be removed
+ *
+ * Return: Zero on success, error code on failure
*/
-void ice_get_qos_params(struct ice_pf *pf, struct iidc_rdma_qos_params *qos)
+int ice_alloc_rdma_qvector(struct iidc_rdma_core_dev_info *cdev,
+ struct msix_entry *entry)
{
- struct ice_dcbx_cfg *dcbx_cfg;
- unsigned int i;
- u32 up2tc;
-
- dcbx_cfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
- up2tc = rd32(&pf->hw, PRTDCB_TUP2TC);
-
- qos->num_tc = ice_dcb_get_num_tc(dcbx_cfg);
- for (i = 0; i < IIDC_MAX_USER_PRIORITY; i++)
- qos->up2tc[i] = (up2tc >> (i * 3)) & 0x7;
-
- for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++)
- qos->tc_info[i].rel_bw = dcbx_cfg->etscfg.tcbwtable[i];
-
- qos->pfc_mode = dcbx_cfg->pfc_mode;
- if (qos->pfc_mode == IIDC_DSCP_PFC_MODE)
- for (i = 0; i < DSCP_MAX; i++)
- qos->dscp_map[i] = dcbx_cfg->dscp_map[i];
-}
-EXPORT_SYMBOL_GPL(ice_get_qos_params);
+ struct msi_map map;
+ struct ice_pf *pf;
-int ice_alloc_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry)
-{
- struct msi_map map = ice_alloc_irq(pf, true);
+ if (WARN_ON(!cdev))
+ return -EINVAL;
+ pf = pci_get_drvdata(cdev->pdev);
+ map = ice_alloc_irq(pf, true);
if (map.index < 0)
return -ENOMEM;
@@ -246,12 +252,19 @@ EXPORT_SYMBOL_GPL(ice_alloc_rdma_qvector);
/**
* ice_free_rdma_qvector - free vector resources reserved for RDMA driver
- * @pf: board private structure to initialize
+ * @cdev: pointer to iidc_rdma_core_dev_info struct
* @entry: MSI-X entry to be removed
*/
-void ice_free_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry)
+void ice_free_rdma_qvector(struct iidc_rdma_core_dev_info *cdev,
+ struct msix_entry *entry)
{
struct msi_map map;
+ struct ice_pf *pf;
+
+ if (WARN_ON(!cdev || !entry))
+ return;
+
+ pf = pci_get_drvdata(cdev->pdev);
map.index = entry->entry;
map.virq = entry->vector;
@@ -275,10 +288,13 @@ static void ice_adev_release(struct device *dev)
/**
* ice_plug_aux_dev - allocate and register AUX device
* @pf: pointer to pf struct
+ *
+ * Return: Zero on success, error code on failure
*/
int ice_plug_aux_dev(struct ice_pf *pf)
{
struct iidc_rdma_core_auxiliary_dev *iadev;
+ struct iidc_rdma_core_dev_info *cdev;
struct auxiliary_device *adev;
int ret;
@@ -288,17 +304,22 @@ int ice_plug_aux_dev(struct ice_pf *pf)
if (!ice_is_rdma_ena(pf))
return 0;
+ cdev = pf->cdev_info;
+ if (!cdev)
+ return -ENODEV;
+
iadev = kzalloc(sizeof(*iadev), GFP_KERNEL);
if (!iadev)
return -ENOMEM;
adev = &iadev->adev;
- iadev->pf = pf;
+ iadev->cdev_info = cdev;
adev->id = pf->aux_idx;
adev->dev.release = ice_adev_release;
adev->dev.parent = &pf->pdev->dev;
- adev->name = pf->rdma_mode & IIDC_RDMA_PROTOCOL_ROCEV2 ? "roce" : "iwarp";
+ adev->name = cdev->rdma_protocol & IIDC_RDMA_PROTOCOL_ROCEV2 ?
+ "roce" : "iwarp";
ret = auxiliary_device_init(adev);
if (ret) {
@@ -313,7 +334,7 @@ int ice_plug_aux_dev(struct ice_pf *pf)
}
mutex_lock(&pf->adev_mutex);
- pf->adev = adev;
+ cdev->adev = adev;
mutex_unlock(&pf->adev_mutex);
return 0;
@@ -327,8 +348,8 @@ void ice_unplug_aux_dev(struct ice_pf *pf)
struct auxiliary_device *adev;
mutex_lock(&pf->adev_mutex);
- adev = pf->adev;
- pf->adev = NULL;
+ adev = pf->cdev_info->adev;
+ pf->cdev_info->adev = NULL;
mutex_unlock(&pf->adev_mutex);
if (adev) {
@@ -343,7 +364,9 @@ void ice_unplug_aux_dev(struct ice_pf *pf)
*/
int ice_init_rdma(struct ice_pf *pf)
{
+ struct iidc_rdma_priv_dev_info *privd;
struct device *dev = &pf->pdev->dev;
+ struct iidc_rdma_core_dev_info *cdev;
int ret;
if (!ice_is_rdma_ena(pf)) {
@@ -351,22 +374,50 @@ int ice_init_rdma(struct ice_pf *pf)
return 0;
}
+ cdev = kzalloc(sizeof(*cdev), GFP_KERNEL);
+ if (!cdev)
+ return -ENOMEM;
+
+ pf->cdev_info = cdev;
+
+ privd = kzalloc(sizeof(*privd), GFP_KERNEL);
+ if (!privd) {
+ ret = -ENOMEM;
+ goto err_privd_alloc;
+ }
+
+ privd->pf_id = pf->hw.pf_id;
ret = xa_alloc(&ice_aux_id, &pf->aux_idx, NULL, XA_LIMIT(1, INT_MAX),
GFP_KERNEL);
if (ret) {
dev_err(dev, "Failed to allocate device ID for AUX driver\n");
- return -ENOMEM;
+ ret = -ENOMEM;
+ goto err_alloc_xa;
}
- pf->rdma_mode |= IIDC_RDMA_PROTOCOL_ROCEV2;
+ cdev->iidc_priv = privd;
+ privd->netdev = pf->vsi[0]->netdev;
+
+ privd->hw_addr = (u8 __iomem *)pf->hw.hw_addr;
+ cdev->pdev = pf->pdev;
+ privd->vport_id = pf->vsi[0]->vsi_num;
+
+ pf->cdev_info->rdma_protocol |= IIDC_RDMA_PROTOCOL_ROCEV2;
+ ice_setup_dcb_qos_info(pf, &privd->qos_info);
ret = ice_plug_aux_dev(pf);
if (ret)
goto err_plug_aux_dev;
return 0;
err_plug_aux_dev:
- pf->adev = NULL;
+ pf->cdev_info->adev = NULL;
xa_erase(&ice_aux_id, pf->aux_idx);
+err_alloc_xa:
+ kfree(privd);
+err_privd_alloc:
+ kfree(cdev);
+ pf->cdev_info = NULL;
+
return ret;
}
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 6a20f9f2e608..eace0e3f15e8 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -9310,6 +9310,7 @@ ice_setup_tc(struct net_device *netdev, enum tc_setup_type type,
void *type_data)
{
struct ice_netdev_priv *np = netdev_priv(netdev);
+ struct iidc_rdma_core_dev_info *cdev;
struct ice_pf *pf = np->vsi->back;
bool locked = false;
int err;
@@ -9326,11 +9327,12 @@ ice_setup_tc(struct net_device *netdev, enum tc_setup_type type,
return -EOPNOTSUPP;
}
- if (pf->adev) {
+ cdev = pf->cdev_info;
+ if (cdev && cdev->adev) {
mutex_lock(&pf->adev_mutex);
- device_lock(&pf->adev->dev);
+ device_lock(&cdev->adev->dev);
locked = true;
- if (pf->adev->dev.driver) {
+ if (cdev->adev->dev.driver) {
netdev_err(netdev, "Cannot change qdisc when RDMA is active\n");
err = -EBUSY;
goto adev_unlock;
@@ -9344,7 +9346,7 @@ ice_setup_tc(struct net_device *netdev, enum tc_setup_type type,
adev_unlock:
if (locked) {
- device_unlock(&pf->adev->dev);
+ device_unlock(&cdev->adev->dev);
mutex_unlock(&pf->adev_mutex);
}
return err;
diff --git a/include/linux/net/intel/iidc_rdma.h b/include/linux/net/intel/iidc_rdma.h
index 7f1910289534..8baad1082042 100644
--- a/include/linux/net/intel/iidc_rdma.h
+++ b/include/linux/net/intel/iidc_rdma.h
@@ -5,7 +5,6 @@
#define _IIDC_RDMA_H_
#include <linux/auxiliary_bus.h>
-#include <linux/dcbnl.h>
#include <linux/device.h>
#include <linux/if_ether.h>
#include <linux/kernel.h>
@@ -17,14 +16,19 @@ enum iidc_rdma_event_type {
IIDC_RDMA_EVENT_AFTER_MTU_CHANGE,
IIDC_RDMA_EVENT_BEFORE_TC_CHANGE,
IIDC_RDMA_EVENT_AFTER_TC_CHANGE,
+ IIDC_RDMA_EVENT_WARN_RESET,
IIDC_RDMA_EVENT_CRIT_ERR,
IIDC_RDMA_EVENT_NBITS /* must be last */
};
+struct iidc_rdma_event {
+ DECLARE_BITMAP(type, IIDC_RDMA_EVENT_NBITS);
+ u32 reg;
+};
+
enum iidc_rdma_reset_type {
- IIDC_PFR,
- IIDC_CORER,
- IIDC_GLOBR,
+ IIDC_FUNC_RESET,
+ IIDC_DEV_RESET,
};
enum iidc_rdma_protocol {
@@ -32,53 +36,22 @@ enum iidc_rdma_protocol {
IIDC_RDMA_PROTOCOL_ROCEV2 = BIT(1),
};
-#define IIDC_MAX_USER_PRIORITY 8
-#define IIDC_DSCP_PFC_MODE 0x1
-
-/* Struct to hold per RDMA Qset info */
-struct iidc_rdma_qset_params {
- /* Qset TEID returned to the RDMA driver in
- * ice_add_rdma_qset and used by RDMA driver
- * for calls to ice_del_rdma_qset
- */
- u32 teid; /* Qset TEID */
- u16 qs_handle; /* RDMA driver provides this */
- u16 vport_id; /* VSI index */
- u8 tc; /* TC branch the Qset should belong to */
-};
-
-struct iidc_rdma_qos_info {
- u64 tc_ctx;
- u8 rel_bw;
- u8 prio_type;
- u8 egress_virt_up;
- u8 ingress_virt_up;
-};
-
-/* Struct to pass QoS info */
-struct iidc_rdma_qos_params {
- struct iidc_rdma_qos_info tc_info[IEEE_8021QAZ_MAX_TCS];
- u8 up2tc[IIDC_MAX_USER_PRIORITY];
- u8 vport_relative_bw;
- u8 vport_priority_type;
- u8 num_tc;
- u8 pfc_mode;
- u8 dscp_map[DSCP_MAX];
-};
-
-struct iidc_rdma_event {
- DECLARE_BITMAP(type, IIDC_RDMA_EVENT_NBITS);
- u32 reg;
+/* Structure to be populated by core LAN PCI driver */
+struct iidc_rdma_core_dev_info {
+ struct pci_dev *pdev; /* PCI device of corresponding to main function */
+ struct auxiliary_device *adev;
+ /* Current active RDMA protocol */
+ enum iidc_rdma_protocol rdma_protocol;
+ void *iidc_priv; /* elements unique to each driver */
};
/* Structure representing auxiliary driver tailored information about the core
* PCI dev, each auxiliary driver using the IIDC interface will have an
* instance of this struct dedicated to it.
*/
-
struct iidc_rdma_core_auxiliary_dev {
struct auxiliary_device adev;
- struct ice_pf *pf;
+ struct iidc_rdma_core_dev_info *cdev_info;
};
/* structure representing the auxiliary driver. This struct is to be
@@ -88,12 +61,8 @@ struct iidc_rdma_core_auxiliary_dev {
*/
struct iidc_rdma_core_auxiliary_drv {
struct auxiliary_driver adrv;
- /* This event_handler is meant to be a blocking call. For instance,
- * when a BEFORE_MTU_CHANGE event comes in, the event_handler will not
- * return until the auxiliary driver is ready for the MTU change to
- * happen.
- */
- void (*event_handler)(struct ice_pf *pf, struct iidc_rdma_event *event);
+ void (*event_handler)(struct iidc_rdma_core_dev_info *cdev,
+ struct iidc_rdma_event *event);
};
#endif /* _IIDC_RDMA_H_*/
diff --git a/include/linux/net/intel/iidc_rdma_ice.h b/include/linux/net/intel/iidc_rdma_ice.h
index 78d10003d776..b40eed0e13fe 100644
--- a/include/linux/net/intel/iidc_rdma_ice.h
+++ b/include/linux/net/intel/iidc_rdma_ice.h
@@ -4,16 +4,67 @@
#ifndef _IIDC_RDMA_ICE_H_
#define _IIDC_RDMA_ICE_H_
-struct ice_pf;
+#include <linux/dcbnl.h>
-int ice_add_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
-int ice_del_rdma_qset(struct ice_pf *pf, struct iidc_rdma_qset_params *qset);
-int ice_rdma_request_reset(struct ice_pf *pf,
+#define IIDC_MAX_USER_PRIORITY 8
+#define IIDC_DSCP_PFC_MODE 0x1
+
+/**
+ * struct iidc_rdma_qset_params - Struct to hold per RDMA Qset info
+ * @teid: TEID of the Qset node
+ * @qs_handle: SW index of the Qset, RDMA provides this
+ * @vport_id: VSI index
+ * @tc: Traffic Class branch the QSet should belong to
+ */
+struct iidc_rdma_qset_params {
+ /* Qset TEID returned to the RDMA driver in
+ * ice_add_rdma_qset and used by RDMA driver
+ * for calls to ice_del_rdma_qset
+ */
+ u32 teid;
+ u16 qs_handle;
+ u16 vport_id;
+ u8 tc;
+};
+
+struct iidc_rdma_qos_info {
+ u64 tc_ctx;
+ u8 rel_bw;
+ u8 prio_type;
+ u8 egress_virt_up;
+ u8 ingress_virt_up;
+};
+
+/* Struct to pass QoS info */
+struct iidc_rdma_qos_params {
+ struct iidc_rdma_qos_info tc_info[IEEE_8021QAZ_MAX_TCS];
+ u8 up2tc[IIDC_MAX_USER_PRIORITY];
+ u8 vport_relative_bw;
+ u8 vport_priority_type;
+ u8 num_tc;
+ u8 pfc_mode;
+ u8 dscp_map[DSCP_MAX];
+};
+
+struct iidc_rdma_priv_dev_info {
+ u8 pf_id;
+ u16 vport_id;
+ struct net_device *netdev;
+ struct iidc_rdma_qos_params qos_info;
+ u8 __iomem *hw_addr;
+};
+
+int ice_add_rdma_qset(struct iidc_rdma_core_dev_info *cdev,
+ struct iidc_rdma_qset_params *qset);
+int ice_del_rdma_qset(struct iidc_rdma_core_dev_info *cdev,
+ struct iidc_rdma_qset_params *qset);
+int ice_rdma_request_reset(struct iidc_rdma_core_dev_info *cdev,
enum iidc_rdma_reset_type reset_type);
-int ice_rdma_update_vsi_filter(struct ice_pf *pf, u16 vsi_id, bool enable);
-void ice_get_qos_params(struct ice_pf *pf,
- struct iidc_rdma_qos_params *qos);
-int ice_alloc_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
-void ice_free_rdma_qvector(struct ice_pf *pf, struct msix_entry *entry);
+int ice_rdma_update_vsi_filter(struct iidc_rdma_core_dev_info *cdev, u16 vsi_id,
+ bool enable);
+int ice_alloc_rdma_qvector(struct iidc_rdma_core_dev_info *cdev,
+ struct msix_entry *entry);
+void ice_free_rdma_qvector(struct iidc_rdma_core_dev_info *cdev,
+ struct msix_entry *entry);
#endif /* _IIDC_RDMA_ICE_H_*/
--
2.47.1
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to support multiple consumers
2025-05-05 21:20 ` [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to support multiple consumers Tony Nguyen
@ 2025-05-08 2:43 ` Jakub Kicinski
2025-05-08 23:20 ` Ertman, David M
0 siblings, 1 reply; 8+ messages in thread
From: Jakub Kicinski @ 2025-05-08 2:43 UTC (permalink / raw)
To: Tony Nguyen
Cc: davem, pabeni, edumazet, andrew+netdev, jgg, leon, linux-rdma,
netdev, Dave Ertman, tatyana.e.nikolova, przemyslaw.kitszel
On Mon, 5 May 2025 14:20:34 -0700 Tony Nguyen wrote:
> - pf->rdma_mode |= IIDC_RDMA_PROTOCOL_ROCEV2;
> + cdev->iidc_priv = privd;
> + privd->netdev = pf->vsi[0]->netdev;
> +
> + privd->hw_addr = (u8 __iomem *)pf->hw.hw_addr;
> + cdev->pdev = pf->pdev;
> + privd->vport_id = pf->vsi[0]->vsi_num;
> +
> + pf->cdev_info->rdma_protocol |= IIDC_RDMA_PROTOCOL_ROCEV2;
> + ice_setup_dcb_qos_info(pf, &privd->qos_info);
> ret = ice_plug_aux_dev(pf);
> if (ret)
> goto err_plug_aux_dev;
> return 0;
>
> err_plug_aux_dev:
> - pf->adev = NULL;
> + pf->cdev_info->adev = NULL;
> xa_erase(&ice_aux_id, pf->aux_idx);
> +err_alloc_xa:
> + kfree(privd);
> +err_privd_alloc:
> + kfree(cdev);
> + pf->cdev_info = NULL;
Where do privd and cdev get freed on normal device removal?
^ permalink raw reply [flat|nested] 8+ messages in thread
* RE: [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to support multiple consumers
2025-05-08 2:43 ` Jakub Kicinski
@ 2025-05-08 23:20 ` Ertman, David M
0 siblings, 0 replies; 8+ messages in thread
From: Ertman, David M @ 2025-05-08 23:20 UTC (permalink / raw)
To: Jakub Kicinski, Nguyen, Anthony L
Cc: davem@davemloft.net, pabeni@redhat.com, edumazet@google.com,
andrew+netdev@lunn.ch, jgg@ziepe.ca, leon@kernel.org,
linux-rdma@vger.kernel.org, netdev@vger.kernel.org,
Nikolova, Tatyana E, Kitszel, Przemyslaw
> -----Original Message-----
> From: Jakub Kicinski <kuba@kernel.org>
> Sent: Wednesday, May 7, 2025 7:43 PM
> To: Nguyen, Anthony L <anthony.l.nguyen@intel.com>
> Cc: davem@davemloft.net; pabeni@redhat.com; edumazet@google.com;
> andrew+netdev@lunn.ch; jgg@ziepe.ca; leon@kernel.org; linux-
> rdma@vger.kernel.org; netdev@vger.kernel.org; Ertman, David M
> <david.m.ertman@intel.com>; Nikolova, Tatyana E
> <tatyana.e.nikolova@intel.com>; Kitszel, Przemyslaw
> <przemyslaw.kitszel@intel.com>
> Subject: Re: [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to
> support multiple consumers
>
> On Mon, 5 May 2025 14:20:34 -0700 Tony Nguyen wrote:
> > - pf->rdma_mode |= IIDC_RDMA_PROTOCOL_ROCEV2;
> > + cdev->iidc_priv = privd;
> > + privd->netdev = pf->vsi[0]->netdev;
> > +
> > + privd->hw_addr = (u8 __iomem *)pf->hw.hw_addr;
> > + cdev->pdev = pf->pdev;
> > + privd->vport_id = pf->vsi[0]->vsi_num;
> > +
> > + pf->cdev_info->rdma_protocol |= IIDC_RDMA_PROTOCOL_ROCEV2;
> > + ice_setup_dcb_qos_info(pf, &privd->qos_info);
> > ret = ice_plug_aux_dev(pf);
> > if (ret)
> > goto err_plug_aux_dev;
> > return 0;
> >
> > err_plug_aux_dev:
> > - pf->adev = NULL;
> > + pf->cdev_info->adev = NULL;
> > xa_erase(&ice_aux_id, pf->aux_idx);
> > +err_alloc_xa:
> > + kfree(privd);
> > +err_privd_alloc:
> > + kfree(cdev);
> > + pf->cdev_info = NULL;
>
> Where do privd and cdev get freed on normal device removal?
I literally smacked myself in the head for missing this first time through.
Thank you very much for the catch. This should exist in ice_deinit_rdma
and a new version will be incoming.
DaveE
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2025-05-08 23:20 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-05-05 21:20 [PATCH net-next,rdma-next 0/5][pull request] Prepare for Intel IPU E2000 (GEN3) Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 1/5] iidc/ice/irdma: Rename IDC header file Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 2/5] iidc/ice/irdma: Rename to iidc_* convention Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 3/5] iidc/ice/irdma: Break iidc.h into two headers Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 4/5] ice: Replace ice specific DSCP mapping num with a kernel define Tony Nguyen
2025-05-05 21:20 ` [PATCH net-next,rdma-next 5/5] iidc/ice/irdma: Update IDC to support multiple consumers Tony Nguyen
2025-05-08 2:43 ` Jakub Kicinski
2025-05-08 23:20 ` Ertman, David M
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).