* [PATCH iwl-next v7 11/14] ixd: add basic driver framework for Intel(R) Control Plane Function
From: Larysa Zaremba @ 2026-04-28 8:26 UTC (permalink / raw)
To: intel-wired-lan, Tony Nguyen
Cc: aleksander.lobakin, sridhar.samudrala, Michal Swiatkowski,
Larysa Zaremba, Fijalkowski, Maciej, Emil Tantilov, Madhu Chittim,
Josh Hay, Keller, Jacob E, jayaprakash.shanmugam, Jiri Pirko,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
Simon Horman, Jonathan Corbet, Richard Cochran, Przemek Kitszel,
Andrew Lunn, netdev, linux-doc, linux-kernel, Bharath R
In-Reply-To: <20260428082654.44364-1-larysa.zaremba@intel.com>
Add module register and probe functionality. Add the required support to
register IXD PCI driver, as well as probe and remove call backs. Enable the
PCI device and request the kernel to reserve the memory resources that will
be used by the driver. Finally map the BAR0 address space.
For now, use devm_alloc() to allocate adapter, as it requires the least
amount of code. In a later commit, it will be replaced with a devlink
alternative.
Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
.../device_drivers/ethernet/index.rst | 1 +
.../device_drivers/ethernet/intel/ixd.rst | 39 ++++++
drivers/net/ethernet/intel/Kconfig | 2 +
drivers/net/ethernet/intel/Makefile | 1 +
drivers/net/ethernet/intel/ixd/Kconfig | 13 ++
drivers/net/ethernet/intel/ixd/Makefile | 8 ++
drivers/net/ethernet/intel/ixd/ixd.h | 28 +++++
drivers/net/ethernet/intel/ixd/ixd_lan_regs.h | 28 +++++
drivers/net/ethernet/intel/ixd/ixd_main.c | 112 ++++++++++++++++++
9 files changed, 232 insertions(+)
create mode 100644 Documentation/networking/device_drivers/ethernet/intel/ixd.rst
create mode 100644 drivers/net/ethernet/intel/ixd/Kconfig
create mode 100644 drivers/net/ethernet/intel/ixd/Makefile
create mode 100644 drivers/net/ethernet/intel/ixd/ixd.h
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_main.c
diff --git a/Documentation/networking/device_drivers/ethernet/index.rst b/Documentation/networking/device_drivers/ethernet/index.rst
index 5f3f06111911..5b6846f387f8 100644
--- a/Documentation/networking/device_drivers/ethernet/index.rst
+++ b/Documentation/networking/device_drivers/ethernet/index.rst
@@ -38,6 +38,7 @@ Contents:
intel/igbvf
intel/ixgbe
intel/ixgbevf
+ intel/ixd
intel/i40e
intel/iavf
intel/ice
diff --git a/Documentation/networking/device_drivers/ethernet/intel/ixd.rst b/Documentation/networking/device_drivers/ethernet/intel/ixd.rst
new file mode 100644
index 000000000000..1387626e5d20
--- /dev/null
+++ b/Documentation/networking/device_drivers/ethernet/intel/ixd.rst
@@ -0,0 +1,39 @@
+.. SPDX-License-Identifier: GPL-2.0+
+
+==========================================================================
+iXD Linux* Base Driver for the Intel(R) Control Plane Function
+==========================================================================
+
+Intel iXD Linux driver.
+Copyright(C) 2025 Intel Corporation.
+
+.. contents::
+
+For questions related to hardware requirements, refer to the documentation
+supplied with your Intel adapter. All hardware requirements listed apply to use
+with Linux.
+
+
+Identifying Your Adapter
+========================
+For information on how to identify your adapter, and for the latest Intel
+network drivers, refer to the Intel Support website:
+http://www.intel.com/support
+
+
+Support
+=======
+For general information, go to the Intel support website at:
+http://www.intel.com/support/
+
+If an issue is identified with the released source code on a supported kernel
+with a supported adapter, email the specific information related to the issue
+to intel-wired-lan@lists.osuosl.org.
+
+
+Trademarks
+==========
+Intel is a trademark or registered trademark of Intel Corporation or its
+subsidiaries in the United States and/or other countries.
+
+* Other names and brands may be claimed as the property of others.
diff --git a/drivers/net/ethernet/intel/Kconfig b/drivers/net/ethernet/intel/Kconfig
index 288fa8ce53af..780f113986ea 100644
--- a/drivers/net/ethernet/intel/Kconfig
+++ b/drivers/net/ethernet/intel/Kconfig
@@ -398,4 +398,6 @@ config IGC_LEDS
source "drivers/net/ethernet/intel/idpf/Kconfig"
+source "drivers/net/ethernet/intel/ixd/Kconfig"
+
endif # NET_VENDOR_INTEL
diff --git a/drivers/net/ethernet/intel/Makefile b/drivers/net/ethernet/intel/Makefile
index 9a37dc76aef0..08b29f3b6801 100644
--- a/drivers/net/ethernet/intel/Makefile
+++ b/drivers/net/ethernet/intel/Makefile
@@ -19,3 +19,4 @@ obj-$(CONFIG_IAVF) += iavf/
obj-$(CONFIG_FM10K) += fm10k/
obj-$(CONFIG_ICE) += ice/
obj-$(CONFIG_IDPF) += idpf/
+obj-$(CONFIG_IXD) += ixd/
diff --git a/drivers/net/ethernet/intel/ixd/Kconfig b/drivers/net/ethernet/intel/ixd/Kconfig
new file mode 100644
index 000000000000..f5594efe292c
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/Kconfig
@@ -0,0 +1,13 @@
+# SPDX-License-Identifier: GPL-2.0-only
+# Copyright (C) 2025 Intel Corporation
+
+config IXD
+ tristate "Intel(R) Control Plane Function Support"
+ depends on PCI_MSI
+ select LIBETH
+ select LIBIE_PCI
+ help
+ This driver supports Intel(R) Control Plane PCI Function
+ of Intel E2100 and later IPUs and FNICs.
+ It facilitates a centralized control over multiple IDPF PFs/VFs/SFs
+ exposed by the same card.
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
new file mode 100644
index 000000000000..3849bc240600
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -0,0 +1,8 @@
+# SPDX-License-Identifier: GPL-2.0-only
+# Copyright (C) 2025 Intel Corporation
+
+# Intel(R) Control Plane Function Linux Driver
+
+obj-$(CONFIG_IXD) += ixd.o
+
+ixd-y := ixd_main.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd.h b/drivers/net/ethernet/intel/ixd/ixd.h
new file mode 100644
index 000000000000..d813c27941a5
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd.h
@@ -0,0 +1,28 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_H_
+#define _IXD_H_
+
+#include <linux/intel/libie/pci.h>
+
+/**
+ * struct ixd_adapter - Data structure representing a CPF
+ * @hw: Device access data
+ */
+struct ixd_adapter {
+ struct libie_mmio_info hw;
+};
+
+/**
+ * ixd_to_dev - Get the corresponding device struct from an adapter
+ * @adapter: PCI device driver-specific private data
+ *
+ * Return: struct device corresponding to the given adapter
+ */
+static inline struct device *ixd_to_dev(struct ixd_adapter *adapter)
+{
+ return &adapter->hw.pdev->dev;
+}
+
+#endif /* _IXD_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
new file mode 100644
index 000000000000..fbb88929d0de
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
@@ -0,0 +1,28 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_LAN_REGS_H_
+#define _IXD_LAN_REGS_H_
+
+/* Control Plane Function PCI ID */
+#define IXD_DEV_ID_CPF 0x1efe
+
+/* Control Queue (Mailbox) */
+#define PF_FW_MBX_REG_LEN 4096
+#define PF_FW_MBX 0x08400000
+
+/* Reset registers */
+#define PFGEN_RTRIG_REG_LEN 2048
+#define PFGEN_RTRIG 0x08407000 /* Device resets */
+
+/**
+ * struct ixd_bar_region - BAR region description
+ * @offset: BAR region offset
+ * @size: BAR region size
+ */
+struct ixd_bar_region {
+ resource_size_t offset;
+ resource_size_t size;
+};
+
+#endif /* _IXD_LAN_REGS_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
new file mode 100644
index 000000000000..75ee53152e61
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -0,0 +1,112 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_lan_regs.h"
+
+MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
+MODULE_IMPORT_NS("LIBIE_PCI");
+MODULE_LICENSE("GPL");
+
+/**
+ * ixd_remove - remove a CPF PCI device
+ * @pdev: PCI device being removed
+ */
+static void ixd_remove(struct pci_dev *pdev)
+{
+ struct ixd_adapter *adapter = pci_get_drvdata(pdev);
+
+ libie_pci_unmap_all_mmio_regions(&adapter->hw);
+}
+
+/**
+ * ixd_shutdown - shut down a CPF PCI device
+ * @pdev: PCI device being shut down
+ */
+static void ixd_shutdown(struct pci_dev *pdev)
+{
+ ixd_remove(pdev);
+
+ if (system_state == SYSTEM_POWER_OFF)
+ pci_set_power_state(pdev, PCI_D3hot);
+}
+
+/**
+ * ixd_iomap_regions - iomap PCI BARs
+ * @adapter: adapter to map memory regions for
+ *
+ * Returns: %0 on success, negative on failure
+ */
+static int ixd_iomap_regions(struct ixd_adapter *adapter)
+{
+ const struct ixd_bar_region regions[] = {
+ {
+ .offset = PFGEN_RTRIG,
+ .size = PFGEN_RTRIG_REG_LEN,
+ },
+ {
+ .offset = PF_FW_MBX,
+ .size = PF_FW_MBX_REG_LEN,
+ },
+ };
+
+ for (int i = 0; i < ARRAY_SIZE(regions); i++) {
+ struct libie_mmio_info *mmio_info = &adapter->hw;
+ bool map_ok;
+
+ map_ok = libie_pci_map_mmio_region(mmio_info,
+ regions[i].offset,
+ regions[i].size);
+ if (!map_ok) {
+ dev_err(ixd_to_dev(adapter),
+ "Failed to map PCI device MMIO region\n");
+
+ libie_pci_unmap_all_mmio_regions(mmio_info);
+ return -EIO;
+ }
+ }
+
+ return 0;
+}
+
+/**
+ * ixd_probe - probe a CPF PCI device
+ * @pdev: corresponding PCI device
+ * @ent: entry in ixd_pci_tbl
+ *
+ * Returns: %0 on success, negative errno code on failure
+ */
+static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+ struct ixd_adapter *adapter;
+ int err;
+
+ adapter = devm_kzalloc(&pdev->dev, sizeof(*adapter), GFP_KERNEL);
+ if (!adapter)
+ return -ENOMEM;
+ adapter->hw.pdev = pdev;
+ INIT_LIST_HEAD(&adapter->hw.mmio_list);
+
+ err = libie_pci_init_dev(pdev);
+ if (err)
+ return err;
+
+ pci_set_drvdata(pdev, adapter);
+
+ return ixd_iomap_regions(adapter);
+}
+
+static const struct pci_device_id ixd_pci_tbl[] = {
+ { PCI_VDEVICE(INTEL, IXD_DEV_ID_CPF) },
+ { }
+};
+MODULE_DEVICE_TABLE(pci, ixd_pci_tbl);
+
+static struct pci_driver ixd_driver = {
+ .name = KBUILD_MODNAME,
+ .id_table = ixd_pci_tbl,
+ .probe = ixd_probe,
+ .remove = ixd_remove,
+ .shutdown = ixd_shutdown,
+};
+module_pci_driver(ixd_driver);
--
2.47.0
^ permalink raw reply related
* [PATCH iwl-next v7 12/14] ixd: add reset checks and initialize the mailbox
From: Larysa Zaremba @ 2026-04-28 8:26 UTC (permalink / raw)
To: intel-wired-lan, Tony Nguyen
Cc: aleksander.lobakin, sridhar.samudrala, Michal Swiatkowski,
Larysa Zaremba, Fijalkowski, Maciej, Emil Tantilov, Madhu Chittim,
Josh Hay, Keller, Jacob E, jayaprakash.shanmugam, Jiri Pirko,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
Simon Horman, Jonathan Corbet, Richard Cochran, Przemek Kitszel,
Andrew Lunn, netdev, linux-doc, linux-kernel, Aleksandr Loktionov,
Bharath R
In-Reply-To: <20260428082654.44364-1-larysa.zaremba@intel.com>
At the end of the probe, trigger hard reset, initialize and schedule the
after-reset task. If the reset is complete in a pre-determined time,
initialize the default mailbox, through which other resources will be
negotiated.
Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/ixd/Kconfig | 1 +
drivers/net/ethernet/intel/ixd/Makefile | 2 +
drivers/net/ethernet/intel/ixd/ixd.h | 28 +++-
drivers/net/ethernet/intel/ixd/ixd_dev.c | 89 +++++++++++
drivers/net/ethernet/intel/ixd/ixd_lan_regs.h | 40 +++++
drivers/net/ethernet/intel/ixd/ixd_lib.c | 143 ++++++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_main.c | 32 +++-
7 files changed, 326 insertions(+), 9 deletions(-)
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_dev.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_lib.c
diff --git a/drivers/net/ethernet/intel/ixd/Kconfig b/drivers/net/ethernet/intel/ixd/Kconfig
index f5594efe292c..24510c50070e 100644
--- a/drivers/net/ethernet/intel/ixd/Kconfig
+++ b/drivers/net/ethernet/intel/ixd/Kconfig
@@ -5,6 +5,7 @@ config IXD
tristate "Intel(R) Control Plane Function Support"
depends on PCI_MSI
select LIBETH
+ select LIBIE_CP
select LIBIE_PCI
help
This driver supports Intel(R) Control Plane PCI Function
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
index 3849bc240600..164b2c86952f 100644
--- a/drivers/net/ethernet/intel/ixd/Makefile
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -6,3 +6,5 @@
obj-$(CONFIG_IXD) += ixd.o
ixd-y := ixd_main.o
+ixd-y += ixd_dev.o
+ixd-y += ixd_lib.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd.h b/drivers/net/ethernet/intel/ixd/ixd.h
index d813c27941a5..99c44f2aa659 100644
--- a/drivers/net/ethernet/intel/ixd/ixd.h
+++ b/drivers/net/ethernet/intel/ixd/ixd.h
@@ -4,14 +4,25 @@
#ifndef _IXD_H_
#define _IXD_H_
-#include <linux/intel/libie/pci.h>
+#include <linux/intel/libie/controlq.h>
/**
* struct ixd_adapter - Data structure representing a CPF
- * @hw: Device access data
+ * @cp_ctx: Control plane communication context
+ * @init_task: Delayed initialization after reset
+ * @xnm: virtchnl transaction manager
+ * @asq: Send control queue info
+ * @arq: Receive control queue info
*/
struct ixd_adapter {
- struct libie_mmio_info hw;
+ struct libie_ctlq_ctx cp_ctx;
+ struct {
+ struct delayed_work init_work;
+ u8 reset_retries;
+ } init_task;
+ struct libie_ctlq_xn_manager *xnm;
+ struct libie_ctlq_info *asq;
+ struct libie_ctlq_info *arq;
};
/**
@@ -22,7 +33,16 @@ struct ixd_adapter {
*/
static inline struct device *ixd_to_dev(struct ixd_adapter *adapter)
{
- return &adapter->hw.pdev->dev;
+ return &adapter->cp_ctx.mmio_info.pdev->dev;
}
+void ixd_ctlq_reg_init(struct ixd_adapter *adapter,
+ struct libie_ctlq_reg *ctlq_reg_tx,
+ struct libie_ctlq_reg *ctlq_reg_rx);
+void ixd_trigger_reset(struct ixd_adapter *adapter);
+bool ixd_check_reset_complete(struct ixd_adapter *adapter);
+void ixd_init_task(struct work_struct *work);
+int ixd_init_dflt_mbx(struct ixd_adapter *adapter);
+void ixd_deinit_dflt_mbx(struct ixd_adapter *adapter);
+
#endif /* _IXD_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_dev.c b/drivers/net/ethernet/intel/ixd/ixd_dev.c
new file mode 100644
index 000000000000..cdd5477cc1f4
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_dev.c
@@ -0,0 +1,89 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_lan_regs.h"
+
+/**
+ * ixd_ctlq_reg_init - Initialize default mailbox registers
+ * @adapter: PCI device driver-specific private data
+ * @ctlq_reg_tx: Transmit queue registers info to be filled
+ * @ctlq_reg_rx: Receive queue registers info to be filled
+ */
+void ixd_ctlq_reg_init(struct ixd_adapter *adapter,
+ struct libie_ctlq_reg *ctlq_reg_tx,
+ struct libie_ctlq_reg *ctlq_reg_rx)
+{
+ struct libie_mmio_info *mmio_info = &adapter->cp_ctx.mmio_info;
+ *ctlq_reg_tx = (struct libie_ctlq_reg) {
+ .head = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQH),
+ .tail = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQT),
+ .len = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQLEN),
+ .addr_high = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQBAH),
+ .addr_low = libie_pci_get_mmio_addr(mmio_info, PF_FW_ATQBAL),
+ .len_mask = PF_FW_ATQLEN_ATQLEN_M,
+ .len_ena_mask = PF_FW_ATQLEN_ATQENABLE_M,
+ .head_mask = PF_FW_ATQH_ATQH_M,
+ };
+
+ *ctlq_reg_rx = (struct libie_ctlq_reg) {
+ .head = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQH),
+ .tail = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQT),
+ .len = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQLEN),
+ .addr_high = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQBAH),
+ .addr_low = libie_pci_get_mmio_addr(mmio_info, PF_FW_ARQBAL),
+ .len_mask = PF_FW_ARQLEN_ARQLEN_M,
+ .len_ena_mask = PF_FW_ARQLEN_ARQENABLE_M,
+ .head_mask = PF_FW_ARQH_ARQH_M,
+ };
+}
+
+static const struct ixd_reset_reg ixd_reset_reg = {
+ .rstat = PFGEN_RSTAT,
+ .rstat_m = PFGEN_RSTAT_PFR_STATE_M,
+ .rstat_ok_v = 0b01,
+ .rtrigger = PFGEN_CTRL,
+ .rtrigger_m = PFGEN_CTRL_PFSWR,
+};
+
+/**
+ * ixd_trigger_reset - Trigger PFR reset
+ * @adapter: the device with mapped reset register
+ */
+void ixd_trigger_reset(struct ixd_adapter *adapter)
+{
+ void __iomem *addr;
+ u32 reg_val;
+
+ addr = libie_pci_get_mmio_addr(&adapter->cp_ctx.mmio_info,
+ ixd_reset_reg.rtrigger);
+ reg_val = readl(addr);
+ writel(reg_val | ixd_reset_reg.rtrigger_m, addr);
+}
+
+/**
+ * ixd_check_reset_complete - Check if the PFR reset is completed
+ * @adapter: CPF being reset
+ *
+ * Return: %true if the register read indicates reset has been finished,
+ * %false otherwise
+ */
+bool ixd_check_reset_complete(struct ixd_adapter *adapter)
+{
+ u32 reg_val, reset_status;
+ void __iomem *addr;
+
+ addr = libie_pci_get_mmio_addr(&adapter->cp_ctx.mmio_info,
+ ixd_reset_reg.rstat);
+ reg_val = readl(addr);
+ reset_status = reg_val & ixd_reset_reg.rstat_m;
+
+ /* 0xFFFFFFFF might be read if the other side hasn't cleared
+ * the register for us yet.
+ */
+ if (reg_val != GENMASK(31, 0) &&
+ reset_status == ixd_reset_reg.rstat_ok_v)
+ return true;
+
+ return false;
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
index fbb88929d0de..58e58c75981b 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
+++ b/drivers/net/ethernet/intel/ixd/ixd_lan_regs.h
@@ -11,9 +11,33 @@
#define PF_FW_MBX_REG_LEN 4096
#define PF_FW_MBX 0x08400000
+#define PF_FW_ARQBAL (PF_FW_MBX)
+#define PF_FW_ARQBAH (PF_FW_MBX + 0x4)
+#define PF_FW_ARQLEN (PF_FW_MBX + 0x8)
+#define PF_FW_ARQLEN_ARQLEN_M GENMASK(12, 0)
+#define PF_FW_ARQLEN_ARQENABLE_S 31
+#define PF_FW_ARQLEN_ARQENABLE_M BIT(PF_FW_ARQLEN_ARQENABLE_S)
+#define PF_FW_ARQH_ARQH_M GENMASK(12, 0)
+#define PF_FW_ARQH (PF_FW_MBX + 0xC)
+#define PF_FW_ARQT (PF_FW_MBX + 0x10)
+
+#define PF_FW_ATQBAL (PF_FW_MBX + 0x14)
+#define PF_FW_ATQBAH (PF_FW_MBX + 0x18)
+#define PF_FW_ATQLEN (PF_FW_MBX + 0x1C)
+#define PF_FW_ATQLEN_ATQLEN_M GENMASK(9, 0)
+#define PF_FW_ATQLEN_ATQENABLE_S 31
+#define PF_FW_ATQLEN_ATQENABLE_M BIT(PF_FW_ATQLEN_ATQENABLE_S)
+#define PF_FW_ATQH_ATQH_M GENMASK(9, 0)
+#define PF_FW_ATQH (PF_FW_MBX + 0x20)
+#define PF_FW_ATQT (PF_FW_MBX + 0x24)
+
/* Reset registers */
#define PFGEN_RTRIG_REG_LEN 2048
#define PFGEN_RTRIG 0x08407000 /* Device resets */
+#define PFGEN_RSTAT 0x08407008 /* PFR status */
+#define PFGEN_RSTAT_PFR_STATE_M GENMASK(1, 0)
+#define PFGEN_CTRL 0x0840700C /* PFR trigger */
+#define PFGEN_CTRL_PFSWR BIT(0)
/**
* struct ixd_bar_region - BAR region description
@@ -25,4 +49,20 @@ struct ixd_bar_region {
resource_size_t size;
};
+/**
+ * struct ixd_reset_reg - structure for reset registers
+ * @rstat: offset of status in register
+ * @rstat_m: status mask
+ * @rstat_ok_v: value that indicates PFR completed status
+ * @rtrigger: offset of reset trigger in register
+ * @rtrigger_m: reset trigger mask
+ */
+struct ixd_reset_reg {
+ u32 rstat;
+ u32 rstat_m;
+ u32 rstat_ok_v;
+ u32 rtrigger;
+ u32 rtrigger_m;
+};
+
#endif /* _IXD_LAN_REGS_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lib.c b/drivers/net/ethernet/intel/ixd/ixd_lib.c
new file mode 100644
index 000000000000..afc413d3650f
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_lib.c
@@ -0,0 +1,143 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+
+#define IXD_DFLT_MBX_Q_LEN 64
+
+/**
+ * ixd_init_ctlq_create_info - Initialize control queue info for creation
+ * @info: destination
+ * @type: type of the queue to create
+ * @ctlq_reg: register assigned to the control queue
+ */
+static void ixd_init_ctlq_create_info(struct libie_ctlq_create_info *info,
+ enum virtchnl2_queue_type type,
+ const struct libie_ctlq_reg *ctlq_reg)
+{
+ *info = (struct libie_ctlq_create_info) {
+ .type = type,
+ .id = -1,
+ .reg = *ctlq_reg,
+ .len = IXD_DFLT_MBX_Q_LEN,
+ };
+}
+
+/**
+ * ixd_init_libie_xn_params - Initialize xn transaction manager creation info
+ * @params: destination
+ * @adapter: adapter info struct
+ * @ctlqs: list of the managed queues to create
+ * @num_queues: length of the queue list
+ */
+static void ixd_init_libie_xn_params(struct libie_ctlq_xn_init_params *params,
+ struct ixd_adapter *adapter,
+ struct libie_ctlq_create_info *ctlqs,
+ uint num_queues)
+{
+ *params = (struct libie_ctlq_xn_init_params){
+ .cctlq_info = ctlqs,
+ .ctx = &adapter->cp_ctx,
+ .num_qs = num_queues,
+ };
+}
+
+/**
+ * ixd_adapter_fill_dflt_ctlqs - Find default control queues and store them
+ * @adapter: adapter info struct
+ */
+static void ixd_adapter_fill_dflt_ctlqs(struct ixd_adapter *adapter)
+{
+ guard(spinlock)(&adapter->cp_ctx.ctlqs_lock);
+ struct libie_ctlq_info *cq;
+
+ list_for_each_entry(cq, &adapter->cp_ctx.ctlqs, list) {
+ if (cq->qid != -1)
+ continue;
+ if (cq->type == LIBIE_CTLQ_TYPE_RX)
+ adapter->arq = cq;
+ else if (cq->type == LIBIE_CTLQ_TYPE_TX)
+ adapter->asq = cq;
+ }
+}
+
+/**
+ * ixd_deinit_dflt_mbx - Deinitialize default mailbox
+ * @adapter: adapter info struct
+ */
+void ixd_deinit_dflt_mbx(struct ixd_adapter *adapter)
+{
+ if (adapter->xnm)
+ libie_ctlq_xn_deinit(adapter->xnm, &adapter->cp_ctx);
+
+ adapter->arq = NULL;
+ adapter->asq = NULL;
+ adapter->xnm = NULL;
+}
+
+/**
+ * ixd_init_dflt_mbx - Setup default mailbox parameters and make request
+ * @adapter: adapter info struct
+ *
+ * Return: %0 on success, negative errno code on failure
+ */
+int ixd_init_dflt_mbx(struct ixd_adapter *adapter)
+{
+ struct libie_ctlq_create_info ctlqs_info[2];
+ struct libie_ctlq_xn_init_params xn_params;
+ struct libie_ctlq_reg ctlq_reg_tx;
+ struct libie_ctlq_reg ctlq_reg_rx;
+ int err;
+
+ ixd_ctlq_reg_init(adapter, &ctlq_reg_tx, &ctlq_reg_rx);
+ ixd_init_ctlq_create_info(&ctlqs_info[0], LIBIE_CTLQ_TYPE_TX,
+ &ctlq_reg_tx);
+ ixd_init_ctlq_create_info(&ctlqs_info[1], LIBIE_CTLQ_TYPE_RX,
+ &ctlq_reg_rx);
+ ixd_init_libie_xn_params(&xn_params, adapter, ctlqs_info,
+ ARRAY_SIZE(ctlqs_info));
+ err = libie_ctlq_xn_init(&xn_params);
+ if (err)
+ return err;
+ adapter->xnm = xn_params.xnm;
+
+ ixd_adapter_fill_dflt_ctlqs(adapter);
+
+ if (!adapter->asq || !adapter->arq) {
+ ixd_deinit_dflt_mbx(adapter);
+ return -ENOENT;
+ }
+
+ return 0;
+}
+
+/**
+ * ixd_init_task - Initialize after reset
+ * @work: init work struct
+ */
+void ixd_init_task(struct work_struct *work)
+{
+ struct ixd_adapter *adapter;
+ int err;
+
+ adapter = container_of(work, struct ixd_adapter,
+ init_task.init_work.work);
+
+ if (!ixd_check_reset_complete(adapter)) {
+ if (++adapter->init_task.reset_retries < 10)
+ queue_delayed_work(system_unbound_wq,
+ &adapter->init_task.init_work,
+ msecs_to_jiffies(500));
+ else
+ dev_err(ixd_to_dev(adapter),
+ "Device reset failed. The driver was unable to contact the device's firmware. Check that the FW is running.\n");
+ return;
+ }
+
+ adapter->init_task.reset_retries = 0;
+ err = ixd_init_dflt_mbx(adapter);
+ if (err)
+ dev_err(ixd_to_dev(adapter),
+ "Failed to initialize the default mailbox: %pe\n",
+ ERR_PTR(err));
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
index 75ee53152e61..b4d4000b63ed 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_main.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -5,6 +5,7 @@
#include "ixd_lan_regs.h"
MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
+MODULE_IMPORT_NS("LIBIE_CP");
MODULE_IMPORT_NS("LIBIE_PCI");
MODULE_LICENSE("GPL");
@@ -16,7 +17,13 @@ static void ixd_remove(struct pci_dev *pdev)
{
struct ixd_adapter *adapter = pci_get_drvdata(pdev);
- libie_pci_unmap_all_mmio_regions(&adapter->hw);
+ /* Do not mix removal with (re)initialization */
+ cancel_delayed_work_sync(&adapter->init_task.init_work);
+ /* Leave the device clean on exit */
+ ixd_trigger_reset(adapter);
+ ixd_deinit_dflt_mbx(adapter);
+
+ libie_pci_unmap_all_mmio_regions(&adapter->cp_ctx.mmio_info);
}
/**
@@ -51,7 +58,7 @@ static int ixd_iomap_regions(struct ixd_adapter *adapter)
};
for (int i = 0; i < ARRAY_SIZE(regions); i++) {
- struct libie_mmio_info *mmio_info = &adapter->hw;
+ struct libie_mmio_info *mmio_info = &adapter->cp_ctx.mmio_info;
bool map_ok;
map_ok = libie_pci_map_mmio_region(mmio_info,
@@ -81,11 +88,15 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
struct ixd_adapter *adapter;
int err;
+ if (WARN_ON(ent->device != IXD_DEV_ID_CPF))
+ return -EINVAL;
+
adapter = devm_kzalloc(&pdev->dev, sizeof(*adapter), GFP_KERNEL);
if (!adapter)
return -ENOMEM;
- adapter->hw.pdev = pdev;
- INIT_LIST_HEAD(&adapter->hw.mmio_list);
+
+ adapter->cp_ctx.mmio_info.pdev = pdev;
+ INIT_LIST_HEAD(&adapter->cp_ctx.mmio_info.mmio_list);
err = libie_pci_init_dev(pdev);
if (err)
@@ -93,7 +104,18 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
pci_set_drvdata(pdev, adapter);
- return ixd_iomap_regions(adapter);
+ err = ixd_iomap_regions(adapter);
+ if (err)
+ return err;
+
+ INIT_DELAYED_WORK(&adapter->init_task.init_work,
+ ixd_init_task);
+
+ ixd_trigger_reset(adapter);
+ queue_delayed_work(system_unbound_wq, &adapter->init_task.init_work,
+ msecs_to_jiffies(500));
+
+ return 0;
}
static const struct pci_device_id ixd_pci_tbl[] = {
--
2.47.0
^ permalink raw reply related
* [PATCH iwl-next v7 13/14] ixd: add the core initialization
From: Larysa Zaremba @ 2026-04-28 8:26 UTC (permalink / raw)
To: intel-wired-lan, Tony Nguyen
Cc: aleksander.lobakin, sridhar.samudrala, Michal Swiatkowski,
Larysa Zaremba, Fijalkowski, Maciej, Emil Tantilov, Madhu Chittim,
Josh Hay, Keller, Jacob E, jayaprakash.shanmugam, Jiri Pirko,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
Simon Horman, Jonathan Corbet, Richard Cochran, Przemek Kitszel,
Andrew Lunn, netdev, linux-doc, linux-kernel, Bharath R
In-Reply-To: <20260428082654.44364-1-larysa.zaremba@intel.com>
As the mailbox is setup, initialize the core. This makes use of the send
and receive mailbox message framework for virtchnl communication between
the driver and device Control Plane (CP).
To start with, driver confirms the virtchnl version with the CP. Once that
is done, it requests and gets the required capabilities and resources
needed such as max vectors, queues, vports etc.
Use a unified way of handling the virtchnl messages, where a single
function handles all related memory management and the caller only provides
the callbacks to fill the send buffer and to handle the response.
Place generic control queue message handling separately to facilitate the
addition of protocols other than virtchannel in the future.
Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/ixd/Makefile | 2 +
drivers/net/ethernet/intel/ixd/ixd.h | 10 +
drivers/net/ethernet/intel/ixd/ixd_ctlq.c | 149 +++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_ctlq.h | 33 ++++
drivers/net/ethernet/intel/ixd/ixd_lib.c | 25 ++-
drivers/net/ethernet/intel/ixd/ixd_main.c | 3 +
drivers/net/ethernet/intel/ixd/ixd_virtchnl.c | 178 ++++++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_virtchnl.h | 12 ++
8 files changed, 411 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_ctlq.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_ctlq.h
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_virtchnl.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_virtchnl.h
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
index 164b2c86952f..90abf231fb16 100644
--- a/drivers/net/ethernet/intel/ixd/Makefile
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -6,5 +6,7 @@
obj-$(CONFIG_IXD) += ixd.o
ixd-y := ixd_main.o
+ixd-y += ixd_ctlq.o
ixd-y += ixd_dev.o
ixd-y += ixd_lib.o
+ixd-y += ixd_virtchnl.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd.h b/drivers/net/ethernet/intel/ixd/ixd.h
index 99c44f2aa659..98d1f22534b5 100644
--- a/drivers/net/ethernet/intel/ixd/ixd.h
+++ b/drivers/net/ethernet/intel/ixd/ixd.h
@@ -10,19 +10,29 @@
* struct ixd_adapter - Data structure representing a CPF
* @cp_ctx: Control plane communication context
* @init_task: Delayed initialization after reset
+ * @mbx_task: Control queue Rx handling
* @xnm: virtchnl transaction manager
* @asq: Send control queue info
* @arq: Receive control queue info
+ * @vc_ver: Negotiated virtchnl version
+ * @caps: Negotiated virtchnl capabilities
*/
struct ixd_adapter {
struct libie_ctlq_ctx cp_ctx;
struct {
struct delayed_work init_work;
u8 reset_retries;
+ u8 vc_retries;
} init_task;
+ struct delayed_work mbx_task;
struct libie_ctlq_xn_manager *xnm;
struct libie_ctlq_info *asq;
struct libie_ctlq_info *arq;
+ struct {
+ u32 major;
+ u32 minor;
+ } vc_ver;
+ struct virtchnl2_get_capabilities caps;
};
/**
diff --git a/drivers/net/ethernet/intel/ixd/ixd_ctlq.c b/drivers/net/ethernet/intel/ixd/ixd_ctlq.c
new file mode 100644
index 000000000000..216aa5c02122
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_ctlq.c
@@ -0,0 +1,149 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_ctlq.h"
+#include "ixd_virtchnl.h"
+
+/**
+ * ixd_ctlq_clean_sq - Clean the send control queue after sending the message
+ * @adapter: The adapter that sent the messages
+ * @num_sent: Number of sent messages to be released
+ *
+ * Free the libie send resources after sending the message and handling
+ * the response.
+ */
+static void ixd_ctlq_clean_sq(struct ixd_adapter *adapter, u16 num_sent)
+{
+ if (!num_sent)
+ return;
+
+ struct libie_ctlq_xn_clean_params params = {
+ .ctlq = adapter->asq,
+ .ctx = &adapter->cp_ctx,
+ .num_msgs = num_sent,
+ .rel_tx_buf = kfree,
+ };
+
+ libie_ctlq_xn_send_clean(¶ms);
+}
+
+/**
+ * ixd_ctlq_init_sparams - Initialize control queue send parameters
+ * @adapter: The adapter with initialized mailbox
+ * @sparams: Parameters to initialize
+ * @msg_buf: DMA-mappable pointer to the message being sent
+ * @msg_size: Message size
+ */
+static void ixd_ctlq_init_sparams(struct ixd_adapter *adapter,
+ struct libie_ctlq_xn_send_params *sparams,
+ void *msg_buf, size_t msg_size)
+{
+ *sparams = (struct libie_ctlq_xn_send_params) {
+ .rel_tx_buf = kfree,
+ .xnm = adapter->xnm,
+ .ctlq = adapter->asq,
+ .timeout_ms = IXD_CTLQ_TIMEOUT,
+ .send_buf = (struct kvec) {
+ .iov_base = msg_buf,
+ .iov_len = msg_size,
+ },
+ };
+}
+
+/**
+ * ixd_ctlq_do_req - Perform a standard virtchnl request
+ * @adapter: The adapter with initialized mailbox
+ * @req: virtchnl request description
+ *
+ * Return: %0 if a message was sent and received a response
+ * that was successfully handled by the custom callback,
+ * negative error otherwise.
+ */
+int ixd_ctlq_do_req(struct ixd_adapter *adapter, const struct ixd_ctlq_req *req)
+{
+ struct libie_ctlq_xn_send_params send_params = {};
+ u8 onstack_send_buff[LIBIE_CP_TX_COPYBREAK] = {};
+ struct kvec *recv_mem;
+ void *send_buff;
+ int err;
+
+ send_buff = libie_cp_can_send_onstack(req->send_size) ?
+ &onstack_send_buff : kzalloc(req->send_size, GFP_KERNEL);
+ if (!send_buff)
+ return -ENOMEM;
+
+ ixd_ctlq_init_sparams(adapter, &send_params, send_buff,
+ req->send_size);
+
+ send_params.chnl_opcode = req->opcode;
+
+ if (req->send_buff_init)
+ req->send_buff_init(adapter, send_buff, req->ctx);
+
+ err = libie_ctlq_xn_send(&send_params);
+ if (err)
+ return err;
+
+ recv_mem = &send_params.recv_mem;
+ if (req->recv_process)
+ err = req->recv_process(adapter, recv_mem->iov_base,
+ recv_mem->iov_len, req->ctx);
+
+ ixd_ctlq_clean_sq(adapter, 1);
+ libie_ctlq_release_rx_buf(recv_mem);
+
+ return err;
+}
+
+/**
+ * ixd_ctlq_handle_msg - Default control queue message handler
+ * @ctx: Control plane communication context
+ * @msg: Message received
+ */
+static void ixd_ctlq_handle_msg(struct libie_ctlq_ctx *ctx,
+ struct libie_ctlq_msg *msg)
+{
+ struct ixd_adapter *adapter = pci_get_drvdata(ctx->mmio_info.pdev);
+
+ if (ixd_vc_can_handle_msg(msg))
+ ixd_vc_recv_event_msg(adapter, msg);
+ else
+ dev_dbg_ratelimited(ixd_to_dev(adapter),
+ "Received an unsupported opcode 0x%x from the CP\n",
+ msg->chnl_opcode);
+
+ libie_ctlq_release_rx_buf(&msg->recv_mem);
+}
+
+/**
+ * ixd_ctlq_recv_mb_msg - Receive a potential message over mailbox periodically
+ * @adapter: The adapter with initialized mailbox
+ */
+static void ixd_ctlq_recv_mb_msg(struct ixd_adapter *adapter)
+{
+ struct libie_ctlq_xn_recv_params xn_params = {
+ .xnm = adapter->xnm,
+ .ctlq = adapter->arq,
+ .ctlq_msg_handler = ixd_ctlq_handle_msg,
+ .budget = LIBIE_CTLQ_MAX_XN_ENTRIES,
+ };
+
+ libie_ctlq_xn_recv(&xn_params);
+}
+
+/**
+ * ixd_ctlq_rx_task - Periodically check for mailbox responses and events
+ * @work: work handle
+ */
+void ixd_ctlq_rx_task(struct work_struct *work)
+{
+ struct ixd_adapter *adapter;
+
+ adapter = container_of(work, struct ixd_adapter, mbx_task.work);
+
+ queue_delayed_work(system_unbound_wq, &adapter->mbx_task,
+ msecs_to_jiffies(300));
+
+ ixd_ctlq_recv_mb_msg(adapter);
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_ctlq.h b/drivers/net/ethernet/intel/ixd/ixd_ctlq.h
new file mode 100644
index 000000000000..f450a3a0828f
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_ctlq.h
@@ -0,0 +1,33 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_CTLQ_H_
+#define _IXD_CTLQ_H_
+
+#include "linux/intel/virtchnl2.h"
+
+#define IXD_CTLQ_TIMEOUT 2000
+
+/**
+ * struct ixd_ctlq_req - Standard virtchnl request description
+ * @opcode: protocol opcode, only virtchnl2 is needed for now
+ * @send_size: required length of the send buffer
+ * @send_buff_init: function to initialize the allocated send buffer
+ * @recv_process: function to handle the CP response
+ * @ctx: additional context for callbacks
+ */
+struct ixd_ctlq_req {
+ enum virtchnl2_op opcode;
+ size_t send_size;
+ void (*send_buff_init)(struct ixd_adapter *adapter, void *send_buff,
+ void *ctx);
+ int (*recv_process)(struct ixd_adapter *adapter, void *recv_buff,
+ size_t recv_size, void *ctx);
+ void *ctx;
+};
+
+int ixd_ctlq_do_req(struct ixd_adapter *adapter,
+ const struct ixd_ctlq_req *req);
+void ixd_ctlq_rx_task(struct work_struct *work);
+
+#endif /* _IXD_CTLQ_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_lib.c b/drivers/net/ethernet/intel/ixd/ixd_lib.c
index afc413d3650f..24080cb30c43 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_lib.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_lib.c
@@ -2,6 +2,7 @@
/* Copyright (C) 2025 Intel Corporation */
#include "ixd.h"
+#include "ixd_virtchnl.h"
#define IXD_DFLT_MBX_Q_LEN 64
@@ -67,6 +68,8 @@ static void ixd_adapter_fill_dflt_ctlqs(struct ixd_adapter *adapter)
*/
void ixd_deinit_dflt_mbx(struct ixd_adapter *adapter)
{
+ cancel_delayed_work_sync(&adapter->mbx_task);
+
if (adapter->xnm)
libie_ctlq_xn_deinit(adapter->xnm, &adapter->cp_ctx);
@@ -108,6 +111,8 @@ int ixd_init_dflt_mbx(struct ixd_adapter *adapter)
return -ENOENT;
}
+ queue_delayed_work(system_unbound_wq, &adapter->mbx_task, 0);
+
return 0;
}
@@ -136,8 +141,26 @@ void ixd_init_task(struct work_struct *work)
adapter->init_task.reset_retries = 0;
err = ixd_init_dflt_mbx(adapter);
- if (err)
+ if (err) {
dev_err(ixd_to_dev(adapter),
"Failed to initialize the default mailbox: %pe\n",
ERR_PTR(err));
+ return;
+ }
+
+ if (!ixd_vc_dev_init(adapter)) {
+ adapter->init_task.vc_retries = 0;
+ return;
+ }
+
+ ixd_deinit_dflt_mbx(adapter);
+ if (++adapter->init_task.vc_retries > 5) {
+ dev_err(ixd_to_dev(adapter),
+ "Failed to establish mailbox communications with the hardware\n");
+ return;
+ }
+
+ ixd_trigger_reset(adapter);
+ queue_delayed_work(system_unbound_wq, &adapter->init_task.init_work,
+ msecs_to_jiffies(500));
}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
index b4d4000b63ed..6d5e6aca77df 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_main.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -2,6 +2,7 @@
/* Copyright (C) 2025 Intel Corporation */
#include "ixd.h"
+#include "ixd_ctlq.h"
#include "ixd_lan_regs.h"
MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
@@ -19,6 +20,7 @@ static void ixd_remove(struct pci_dev *pdev)
/* Do not mix removal with (re)initialization */
cancel_delayed_work_sync(&adapter->init_task.init_work);
+
/* Leave the device clean on exit */
ixd_trigger_reset(adapter);
ixd_deinit_dflt_mbx(adapter);
@@ -110,6 +112,7 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
INIT_DELAYED_WORK(&adapter->init_task.init_work,
ixd_init_task);
+ INIT_DELAYED_WORK(&adapter->mbx_task, ixd_ctlq_rx_task);
ixd_trigger_reset(adapter);
queue_delayed_work(system_unbound_wq, &adapter->init_task.init_work,
diff --git a/drivers/net/ethernet/intel/ixd/ixd_virtchnl.c b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.c
new file mode 100644
index 000000000000..66049d1b1d15
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.c
@@ -0,0 +1,178 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2025 Intel Corporation */
+
+#include "ixd.h"
+#include "ixd_ctlq.h"
+#include "ixd_virtchnl.h"
+
+/**
+ * ixd_vc_recv_event_msg - Handle virtchnl event message
+ * @adapter: The adapter handling the message
+ * @ctlq_msg: Message received
+ */
+void ixd_vc_recv_event_msg(struct ixd_adapter *adapter,
+ struct libie_ctlq_msg *ctlq_msg)
+{
+ int payload_size = ctlq_msg->data_len;
+ struct virtchnl2_event *v2e;
+
+ if (payload_size < sizeof(*v2e)) {
+ dev_warn_ratelimited(ixd_to_dev(adapter),
+ "Failed to receive valid payload for event msg (op 0x%X len %u)\n",
+ ctlq_msg->chnl_opcode,
+ payload_size);
+ return;
+ }
+
+ v2e = (struct virtchnl2_event *)ctlq_msg->recv_mem.iov_base;
+
+ dev_dbg(ixd_to_dev(adapter), "Got event 0x%X from the CP\n",
+ le32_to_cpu(v2e->event));
+}
+
+/**
+ * ixd_vc_can_handle_msg - Decide if an event has to be handled by virtchnl code
+ * @ctlq_msg: Message received
+ *
+ * Return: %true if virtchnl code can handle the event, %false otherwise
+ */
+bool ixd_vc_can_handle_msg(struct libie_ctlq_msg *ctlq_msg)
+{
+ return ctlq_msg->chnl_opcode == VIRTCHNL2_OP_EVENT;
+}
+
+/**
+ * ixd_handle_caps - Handle VIRTCHNL2_OP_GET_CAPS response
+ * @adapter: The adapter for which the capabilities are being updated
+ * @recv_buff: Buffer containing the response
+ * @recv_size: Response buffer size
+ * @ctx: unused
+ *
+ * Return: %0 if the response format is correct and was handled as expected,
+ * negative error otherwise.
+ */
+static int ixd_handle_caps(struct ixd_adapter *adapter, void *recv_buff,
+ size_t recv_size, void *ctx)
+{
+ if (recv_size < sizeof(adapter->caps))
+ return -EBADMSG;
+
+ adapter->caps = *(typeof(adapter->caps) *)recv_buff;
+
+ return 0;
+}
+
+/**
+ * ixd_req_vc_caps - Request and save device capability
+ * @adapter: The adapter to get the capabilities for
+ *
+ * Return: success or error if sending the get capability message fails
+ */
+static int ixd_req_vc_caps(struct ixd_adapter *adapter)
+{
+ const struct ixd_ctlq_req req = {
+ .opcode = VIRTCHNL2_OP_GET_CAPS,
+ .send_size = sizeof(struct virtchnl2_get_capabilities),
+ .ctx = NULL,
+ .send_buff_init = NULL,
+ .recv_process = ixd_handle_caps,
+ };
+
+ return ixd_ctlq_do_req(adapter, &req);
+}
+
+/**
+ * ixd_get_vc_ver - Get version info from adapter
+ *
+ * Return: filled in virtchannel2 version info, ready for sending
+ */
+static struct virtchnl2_version_info ixd_get_vc_ver(void)
+{
+ return (struct virtchnl2_version_info) {
+ .major = cpu_to_le32(VIRTCHNL2_VERSION_MAJOR_2),
+ .minor = cpu_to_le32(VIRTCHNL2_VERSION_MINOR_0),
+ };
+}
+
+static void ixd_fill_vc_ver(struct ixd_adapter *adapter, void *send_buff,
+ void *ctx)
+{
+ *(struct virtchnl2_version_info *)send_buff = ixd_get_vc_ver();
+}
+
+/**
+ * ixd_handle_vc_ver - Handle VIRTCHNL2_OP_VERSION response
+ * @adapter: The adapter for which the version is being updated
+ * @recv_buff: Buffer containing the response
+ * @recv_size: Response buffer size
+ * @ctx: Unused
+ *
+ * Return: %0 if the response format is correct and was handled as expected,
+ * negative error otherwise.
+ */
+static int ixd_handle_vc_ver(struct ixd_adapter *adapter, void *recv_buff,
+ size_t recv_size, void *ctx)
+{
+ struct virtchnl2_version_info need_ver = ixd_get_vc_ver();
+ struct virtchnl2_version_info *recv_ver;
+
+ if (recv_size < sizeof(need_ver))
+ return -EBADMSG;
+
+ recv_ver = recv_buff;
+ if (le32_to_cpu(need_ver.major) > le32_to_cpu(recv_ver->major))
+ return -EOPNOTSUPP;
+
+ adapter->vc_ver.major = le32_to_cpu(recv_ver->major);
+ adapter->vc_ver.minor = le32_to_cpu(recv_ver->minor);
+
+ return 0;
+}
+
+/**
+ * ixd_req_vc_version - Request and save Virtchannel2 version
+ * @adapter: The adapter to get the version for
+ *
+ * Return: success or error if sending fails or the response was not as expected
+ */
+static int ixd_req_vc_version(struct ixd_adapter *adapter)
+{
+ const struct ixd_ctlq_req req = {
+ .opcode = VIRTCHNL2_OP_VERSION,
+ .send_size = sizeof(struct virtchnl2_version_info),
+ .ctx = NULL,
+ .send_buff_init = ixd_fill_vc_ver,
+ .recv_process = ixd_handle_vc_ver,
+ };
+
+ return ixd_ctlq_do_req(adapter, &req);
+}
+
+/**
+ * ixd_vc_dev_init - virtchnl device core initialization
+ * @adapter: device information
+ *
+ * Return: %0 on success or error if any step of the initialization fails
+ */
+int ixd_vc_dev_init(struct ixd_adapter *adapter)
+{
+ int err;
+
+ err = ixd_req_vc_version(adapter);
+ if (err) {
+ dev_warn(ixd_to_dev(adapter),
+ "Getting virtchnl version failed, error=%pe\n",
+ ERR_PTR(err));
+ return err;
+ }
+
+ err = ixd_req_vc_caps(adapter);
+ if (err) {
+ dev_warn(ixd_to_dev(adapter),
+ "Getting virtchnl capabilities failed, error=%pe\n",
+ ERR_PTR(err));
+ return err;
+ }
+
+ return err;
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_virtchnl.h b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.h
new file mode 100644
index 000000000000..1a53da8b545c
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_virtchnl.h
@@ -0,0 +1,12 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2025 Intel Corporation */
+
+#ifndef _IXD_VIRTCHNL_H_
+#define _IXD_VIRTCHNL_H_
+
+int ixd_vc_dev_init(struct ixd_adapter *adapter);
+bool ixd_vc_can_handle_msg(struct libie_ctlq_msg *ctlq_msg);
+void ixd_vc_recv_event_msg(struct ixd_adapter *adapter,
+ struct libie_ctlq_msg *ctlq_msg);
+
+#endif /* _IXD_VIRTCHNL_H_ */
--
2.47.0
^ permalink raw reply related
* [PATCH iwl-next v7 08/14] idpf: refactor idpf to use libie control queues
From: Larysa Zaremba @ 2026-04-28 8:26 UTC (permalink / raw)
To: intel-wired-lan, Tony Nguyen
Cc: aleksander.lobakin, sridhar.samudrala, Michal Swiatkowski,
Larysa Zaremba, Fijalkowski, Maciej, Emil Tantilov, Madhu Chittim,
Josh Hay, Keller, Jacob E, jayaprakash.shanmugam, Jiri Pirko,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
Simon Horman, Jonathan Corbet, Richard Cochran, Przemek Kitszel,
Andrew Lunn, netdev, linux-doc, linux-kernel, Aleksandr Loktionov,
Samuel Salin
In-Reply-To: <20260428082654.44364-1-larysa.zaremba@intel.com>
From: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Support to initialize and configure controlqs, and manage their
transactions was introduced in libie. As part of it, most of the existing
controlq structures are renamed and modified. Use those APIs in idpf and
make all the necessary changes.
Previously for the send and receive virtchnl messages, there used to be a
memcpy involved in controlq code to copy the buffer info passed by the send
function into the controlq specific buffers. There was no restriction to
use automatic memory in that case. The new implementation in libie removed
copying of the send buffer info and introduced DMA mapping of the send
buffer itself. To accommodate it, use dynamic memory for the larger send
buffers. For smaller ones (<= 128 bytes) libie still can copy them into the
pre-allocated message memory.
In case of receive, idpf receives a page pool buffer allocated by the libie
and care should be taken to release it after use in the idpf.
The changes are fairly trivial and localized, with a notable exception
being the consolidation of idpf_vc_xn_shutdown and idpf_deinit_dflt_mbx
under the latter name. This has some additional consequences that are
addressed in the following patches.
This refactoring introduces roughly additional 40KB of module storage used
for systems that only run idpf, so idpf + libie_cp + libie_pci takes about
7% more storage than just idpf before refactoring.
We now pre-allocate small TX buffers, so that does increase the memory
usage, but reduces the need to allocate. This results in additional 256 *
128B of memory permanently used, increasing the worst-case memory usage by
32KB but our ctlq RX buffers need to be of size 4096B anyway (not changed
by the patchset), so this is hardly noticeable.
As for the timings, the fact that we are mostly limited by the HW response
time which is far from instant, is not changed by this refactor.
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
Co-developed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Tested-by: Samuel Salin <Samuel.salin@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/idpf/Makefile | 2 -
drivers/net/ethernet/intel/idpf/idpf.h | 28 +-
.../net/ethernet/intel/idpf/idpf_controlq.c | 631 -------
.../net/ethernet/intel/idpf/idpf_controlq.h | 142 --
.../ethernet/intel/idpf/idpf_controlq_api.h | 177 --
.../ethernet/intel/idpf/idpf_controlq_setup.c | 169 --
drivers/net/ethernet/intel/idpf/idpf_dev.c | 60 +-
.../net/ethernet/intel/idpf/idpf_ethtool.c | 28 +-
drivers/net/ethernet/intel/idpf/idpf_lib.c | 51 +-
drivers/net/ethernet/intel/idpf/idpf_main.c | 5 -
drivers/net/ethernet/intel/idpf/idpf_mem.h | 20 -
drivers/net/ethernet/intel/idpf/idpf_txrx.h | 2 +-
drivers/net/ethernet/intel/idpf/idpf_vf_dev.c | 64 +-
.../net/ethernet/intel/idpf/idpf_virtchnl.c | 1632 +++++++----------
.../net/ethernet/intel/idpf/idpf_virtchnl.h | 94 +-
.../ethernet/intel/idpf/idpf_virtchnl_ptp.c | 248 ++-
16 files changed, 832 insertions(+), 2521 deletions(-)
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq.c
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq.h
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq_api.h
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c
delete mode 100644 drivers/net/ethernet/intel/idpf/idpf_mem.h
diff --git a/drivers/net/ethernet/intel/idpf/Makefile b/drivers/net/ethernet/intel/idpf/Makefile
index 651ddee942bd..4aaafa175ec3 100644
--- a/drivers/net/ethernet/intel/idpf/Makefile
+++ b/drivers/net/ethernet/intel/idpf/Makefile
@@ -6,8 +6,6 @@
obj-$(CONFIG_IDPF) += idpf.o
idpf-y := \
- idpf_controlq.o \
- idpf_controlq_setup.o \
idpf_dev.o \
idpf_ethtool.o \
idpf_idc.o \
diff --git a/drivers/net/ethernet/intel/idpf/idpf.h b/drivers/net/ethernet/intel/idpf/idpf.h
index efdb58990a8b..679539a1b947 100644
--- a/drivers/net/ethernet/intel/idpf/idpf.h
+++ b/drivers/net/ethernet/intel/idpf/idpf.h
@@ -27,7 +27,6 @@ struct idpf_rss_data;
#include <linux/intel/virtchnl2.h>
#include "idpf_txrx.h"
-#include "idpf_controlq.h"
#define GETMAXVAL(num_bits) GENMASK((num_bits) - 1, 0)
@@ -37,11 +36,10 @@ struct idpf_rss_data;
#define IDPF_NUM_FILTERS_PER_MSG 20
#define IDPF_NUM_DFLT_MBX_Q 2 /* includes both TX and RX */
#define IDPF_DFLT_MBX_Q_LEN 64
-#define IDPF_DFLT_MBX_ID -1
/* maximum number of times to try before resetting mailbox */
#define IDPF_MB_MAX_ERR 20
#define IDPF_NUM_CHUNKS_PER_MSG(struct_sz, chunk_sz) \
- ((IDPF_CTLQ_MAX_BUF_LEN - (struct_sz)) / (chunk_sz))
+ ((LIBIE_CTLQ_MAX_BUF_LEN - (struct_sz)) / (chunk_sz))
#define IDPF_WAIT_FOR_MARKER_TIMEO 500
#define IDPF_MAX_WAIT 500
@@ -202,8 +200,8 @@ struct idpf_vport_max_q {
* @ptp_reg_init: PTP register initialization
*/
struct idpf_reg_ops {
- void (*ctlq_reg_init)(struct idpf_adapter *adapter,
- struct idpf_ctlq_create_info *cq);
+ void (*ctlq_reg_init)(struct libie_mmio_info *mmio,
+ struct libie_ctlq_create_info *cctlq_info);
int (*intr_reg_init)(struct idpf_vport *vport,
struct idpf_q_vec_rsrc *rsrc);
void (*mb_intr_reg_init)(struct idpf_adapter *adapter);
@@ -606,8 +604,6 @@ struct idpf_vport_config {
DECLARE_BITMAP(flags, IDPF_VPORT_CONFIG_FLAGS_NBITS);
};
-struct idpf_vc_xn_manager;
-
#define idpf_for_each_vport(adapter, iter) \
for (struct idpf_vport **__##iter = &(adapter)->vports[0], \
*iter = (adapter)->max_vports ? *__##iter : NULL; \
@@ -625,8 +621,10 @@ struct idpf_vc_xn_manager;
* @state: Init state machine
* @flags: See enum idpf_flags
* @reset_reg: See struct idpf_reset_reg
- * @hw: Device access data
* @ctlq_ctx: controlq context
+ * @asq: Send control queue info
+ * @arq: Receive control queue info
+ * @xnm: Xn transaction manager
* @num_avail_msix: Available number of MSIX vectors
* @num_msix_entries: Number of entries in MSIX table
* @msix_entries: MSIX table
@@ -659,7 +657,6 @@ struct idpf_vc_xn_manager;
* @stats_task: Periodic statistics retrieval task
* @stats_wq: Workqueue for statistics task
* @caps: Negotiated capabilities with device
- * @vcxn_mngr: Virtchnl transaction manager
* @dev_ops: See idpf_dev_ops
* @cdev_info: IDC core device info pointer
* @num_vfs: Number of allocated VFs through sysfs. PF does not directly talk
@@ -683,8 +680,10 @@ struct idpf_adapter {
enum idpf_state state;
DECLARE_BITMAP(flags, IDPF_FLAGS_NBITS);
struct idpf_reset_reg reset_reg;
- struct idpf_hw hw;
struct libie_ctlq_ctx ctlq_ctx;
+ struct libie_ctlq_info *asq;
+ struct libie_ctlq_info *arq;
+ struct libie_ctlq_xn_manager *xnm;
u16 num_avail_msix;
u16 num_msix_entries;
struct msix_entry *msix_entries;
@@ -721,7 +720,6 @@ struct idpf_adapter {
struct delayed_work stats_task;
struct workqueue_struct *stats_wq;
struct virtchnl2_get_capabilities caps;
- struct idpf_vc_xn_manager *vcxn_mngr;
struct idpf_dev_ops dev_ops;
struct iidc_rdma_core_dev_info *cdev_info;
@@ -881,12 +879,12 @@ static inline u8 idpf_get_min_tx_pkt_len(struct idpf_adapter *adapter)
*/
static inline bool idpf_is_reset_detected(struct idpf_adapter *adapter)
{
- if (!adapter->hw.arq)
+ struct libie_ctlq_info *arq = adapter->arq;
+
+ if (!arq)
return true;
- return !(readl(libie_pci_get_mmio_addr(&adapter->ctlq_ctx.mmio_info,
- adapter->hw.arq->reg.len)) &
- adapter->hw.arq->reg.len_mask);
+ return !(readl(arq->reg.len) & arq->reg.len_mask);
}
/**
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq.c b/drivers/net/ethernet/intel/idpf/idpf_controlq.c
deleted file mode 100644
index 020b08367e18..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq.c
+++ /dev/null
@@ -1,631 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-only
-/* Copyright (C) 2023 Intel Corporation */
-
-#include "idpf.h"
-
-/**
- * idpf_ctlq_setup_regs - initialize control queue registers
- * @cq: pointer to the specific control queue
- * @q_create_info: structs containing info for each queue to be initialized
- */
-static void idpf_ctlq_setup_regs(struct idpf_ctlq_info *cq,
- struct idpf_ctlq_create_info *q_create_info)
-{
- /* set control queue registers in our local struct */
- cq->reg.head = q_create_info->reg.head;
- cq->reg.tail = q_create_info->reg.tail;
- cq->reg.len = q_create_info->reg.len;
- cq->reg.bah = q_create_info->reg.bah;
- cq->reg.bal = q_create_info->reg.bal;
- cq->reg.len_mask = q_create_info->reg.len_mask;
- cq->reg.len_ena_mask = q_create_info->reg.len_ena_mask;
- cq->reg.head_mask = q_create_info->reg.head_mask;
-}
-
-/**
- * idpf_ctlq_init_regs - Initialize control queue registers
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- * @is_rxq: true if receive control queue, false otherwise
- *
- * Initialize registers. The caller is expected to have already initialized the
- * descriptor ring memory and buffer memory
- */
-static void idpf_ctlq_init_regs(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
- bool is_rxq)
-{
- struct libie_mmio_info *mmio = &hw->back->ctlq_ctx.mmio_info;
-
- /* Update tail to post pre-allocated buffers for rx queues */
- if (is_rxq)
- writel((u32)(cq->ring_size - 1),
- libie_pci_get_mmio_addr(mmio, cq->reg.tail));
-
- /* For non-Mailbox control queues only TAIL need to be set */
- if (cq->q_id != -1)
- return;
-
- /* Clear Head for both send or receive */
- writel(0, libie_pci_get_mmio_addr(mmio, cq->reg.head));
-
- /* set starting point */
- writel(lower_32_bits(cq->desc_ring.pa),
- libie_pci_get_mmio_addr(mmio, cq->reg.bal));
- writel(upper_32_bits(cq->desc_ring.pa),
- libie_pci_get_mmio_addr(mmio, cq->reg.bah));
- writel((cq->ring_size | cq->reg.len_ena_mask),
- libie_pci_get_mmio_addr(mmio, cq->reg.len));
-}
-
-/**
- * idpf_ctlq_init_rxq_bufs - populate receive queue descriptors with buf
- * @cq: pointer to the specific Control queue
- *
- * Record the address of the receive queue DMA buffers in the descriptors.
- * The buffers must have been previously allocated.
- */
-static void idpf_ctlq_init_rxq_bufs(struct idpf_ctlq_info *cq)
-{
- int i;
-
- for (i = 0; i < cq->ring_size; i++) {
- struct idpf_ctlq_desc *desc = IDPF_CTLQ_DESC(cq, i);
- struct idpf_dma_mem *bi = cq->bi.rx_buff[i];
-
- /* No buffer to post to descriptor, continue */
- if (!bi)
- continue;
-
- desc->flags =
- cpu_to_le16(IDPF_CTLQ_FLAG_BUF | IDPF_CTLQ_FLAG_RD);
- desc->opcode = 0;
- desc->datalen = cpu_to_le16(bi->size);
- desc->ret_val = 0;
- desc->v_opcode_dtype = 0;
- desc->v_retval = 0;
- desc->params.indirect.addr_high =
- cpu_to_le32(upper_32_bits(bi->pa));
- desc->params.indirect.addr_low =
- cpu_to_le32(lower_32_bits(bi->pa));
- desc->params.indirect.param0 = 0;
- desc->params.indirect.sw_cookie = 0;
- desc->params.indirect.v_flags = 0;
- }
-}
-
-/**
- * idpf_ctlq_shutdown - shutdown the CQ
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * The main shutdown routine for any controq queue
- */
-static void idpf_ctlq_shutdown(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- spin_lock(&cq->cq_lock);
-
- /* free ring buffers and the ring itself */
- idpf_ctlq_dealloc_ring_res(hw, cq);
-
- /* Set ring_size to 0 to indicate uninitialized queue */
- cq->ring_size = 0;
-
- spin_unlock(&cq->cq_lock);
-}
-
-/**
- * idpf_ctlq_add - add one control queue
- * @hw: pointer to hardware struct
- * @qinfo: info for queue to be created
- * @cq_out: (output) double pointer to control queue to be created
- *
- * Allocate and initialize a control queue and add it to the control queue list.
- * The cq parameter will be allocated/initialized and passed back to the caller
- * if no errors occur.
- *
- * Note: idpf_ctlq_init must be called prior to any calls to idpf_ctlq_add
- */
-int idpf_ctlq_add(struct idpf_hw *hw,
- struct idpf_ctlq_create_info *qinfo,
- struct idpf_ctlq_info **cq_out)
-{
- struct idpf_ctlq_info *cq;
- bool is_rxq = false;
- int err;
-
- cq = kzalloc_obj(*cq);
- if (!cq)
- return -ENOMEM;
-
- cq->cq_type = qinfo->type;
- cq->q_id = qinfo->id;
- cq->buf_size = qinfo->buf_size;
- cq->ring_size = qinfo->len;
-
- cq->next_to_use = 0;
- cq->next_to_clean = 0;
- cq->next_to_post = cq->ring_size - 1;
-
- switch (qinfo->type) {
- case IDPF_CTLQ_TYPE_MAILBOX_RX:
- is_rxq = true;
- fallthrough;
- case IDPF_CTLQ_TYPE_MAILBOX_TX:
- err = idpf_ctlq_alloc_ring_res(hw, cq);
- break;
- default:
- err = -EBADR;
- break;
- }
-
- if (err)
- goto init_free_q;
-
- if (is_rxq) {
- idpf_ctlq_init_rxq_bufs(cq);
- } else {
- /* Allocate the array of msg pointers for TX queues */
- cq->bi.tx_msg = kzalloc_objs(struct idpf_ctlq_msg *, qinfo->len);
- if (!cq->bi.tx_msg) {
- err = -ENOMEM;
- goto init_dealloc_q_mem;
- }
- }
-
- idpf_ctlq_setup_regs(cq, qinfo);
-
- idpf_ctlq_init_regs(hw, cq, is_rxq);
-
- spin_lock_init(&cq->cq_lock);
-
- list_add(&cq->cq_list, &hw->cq_list_head);
-
- *cq_out = cq;
-
- return 0;
-
-init_dealloc_q_mem:
- /* free ring buffers and the ring itself */
- idpf_ctlq_dealloc_ring_res(hw, cq);
-init_free_q:
- kfree(cq);
-
- return err;
-}
-
-/**
- * idpf_ctlq_remove - deallocate and remove specified control queue
- * @hw: pointer to hardware struct
- * @cq: pointer to control queue to be removed
- */
-void idpf_ctlq_remove(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- list_del(&cq->cq_list);
- idpf_ctlq_shutdown(hw, cq);
- kfree(cq);
-}
-
-/**
- * idpf_ctlq_init - main initialization routine for all control queues
- * @hw: pointer to hardware struct
- * @num_q: number of queues to initialize
- * @q_info: array of structs containing info for each queue to be initialized
- *
- * This initializes any number and any type of control queues. This is an all
- * or nothing routine; if one fails, all previously allocated queues will be
- * destroyed. This must be called prior to using the individual add/remove
- * APIs.
- */
-int idpf_ctlq_init(struct idpf_hw *hw, u8 num_q,
- struct idpf_ctlq_create_info *q_info)
-{
- struct idpf_ctlq_info *cq, *tmp;
- int err;
- int i;
-
- INIT_LIST_HEAD(&hw->cq_list_head);
-
- for (i = 0; i < num_q; i++) {
- struct idpf_ctlq_create_info *qinfo = q_info + i;
-
- err = idpf_ctlq_add(hw, qinfo, &cq);
- if (err)
- goto init_destroy_qs;
- }
-
- return 0;
-
-init_destroy_qs:
- list_for_each_entry_safe(cq, tmp, &hw->cq_list_head, cq_list)
- idpf_ctlq_remove(hw, cq);
-
- return err;
-}
-
-/**
- * idpf_ctlq_deinit - destroy all control queues
- * @hw: pointer to hw struct
- */
-void idpf_ctlq_deinit(struct idpf_hw *hw)
-{
- struct idpf_ctlq_info *cq, *tmp;
-
- list_for_each_entry_safe(cq, tmp, &hw->cq_list_head, cq_list)
- idpf_ctlq_remove(hw, cq);
-}
-
-/**
- * idpf_ctlq_send - send command to Control Queue (CTQ)
- * @hw: pointer to hw struct
- * @cq: handle to control queue struct to send on
- * @num_q_msg: number of messages to send on control queue
- * @q_msg: pointer to array of queue messages to be sent
- *
- * The caller is expected to allocate DMAable buffers and pass them to the
- * send routine via the q_msg struct / control queue specific data struct.
- * The control queue will hold a reference to each send message until
- * the completion for that message has been cleaned.
- */
-int idpf_ctlq_send(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
- u16 num_q_msg, struct idpf_ctlq_msg q_msg[])
-{
- struct idpf_ctlq_desc *desc;
- int num_desc_avail;
- int err = 0;
- int i;
-
- spin_lock(&cq->cq_lock);
-
- /* Ensure there are enough descriptors to send all messages */
- num_desc_avail = IDPF_CTLQ_DESC_UNUSED(cq);
- if (num_desc_avail == 0 || num_desc_avail < num_q_msg) {
- err = -ENOSPC;
- goto err_unlock;
- }
-
- for (i = 0; i < num_q_msg; i++) {
- struct idpf_ctlq_msg *msg = &q_msg[i];
-
- desc = IDPF_CTLQ_DESC(cq, cq->next_to_use);
-
- desc->opcode = cpu_to_le16(msg->opcode);
- desc->pfid_vfid = cpu_to_le16(msg->func_id);
-
- desc->v_opcode_dtype = cpu_to_le32(msg->cookie.mbx.chnl_opcode);
- desc->v_retval = cpu_to_le32(msg->cookie.mbx.chnl_retval);
-
- desc->flags = cpu_to_le16((msg->host_id & IDPF_HOST_ID_MASK) <<
- IDPF_CTLQ_FLAG_HOST_ID_S);
- if (msg->data_len) {
- struct idpf_dma_mem *buff = msg->ctx.indirect.payload;
-
- desc->datalen |= cpu_to_le16(msg->data_len);
- desc->flags |= cpu_to_le16(IDPF_CTLQ_FLAG_BUF);
- desc->flags |= cpu_to_le16(IDPF_CTLQ_FLAG_RD);
-
- /* Update the address values in the desc with the pa
- * value for respective buffer
- */
- desc->params.indirect.addr_high =
- cpu_to_le32(upper_32_bits(buff->pa));
- desc->params.indirect.addr_low =
- cpu_to_le32(lower_32_bits(buff->pa));
-
- memcpy(&desc->params, msg->ctx.indirect.context,
- IDPF_INDIRECT_CTX_SIZE);
- } else {
- memcpy(&desc->params, msg->ctx.direct,
- IDPF_DIRECT_CTX_SIZE);
- }
-
- /* Store buffer info */
- cq->bi.tx_msg[cq->next_to_use] = msg;
-
- (cq->next_to_use)++;
- if (cq->next_to_use == cq->ring_size)
- cq->next_to_use = 0;
- }
-
- /* Force memory write to complete before letting hardware
- * know that there are new descriptors to fetch.
- */
- dma_wmb();
-
- writel(cq->next_to_use,
- libie_pci_get_mmio_addr(&hw->back->ctlq_ctx.mmio_info,
- cq->reg.tail));
-
-err_unlock:
- spin_unlock(&cq->cq_lock);
-
- return err;
-}
-
-/**
- * idpf_ctlq_clean_sq - reclaim send descriptors on HW write back for the
- * requested queue
- * @cq: pointer to the specific Control queue
- * @clean_count: (input|output) number of descriptors to clean as input, and
- * number of descriptors actually cleaned as output
- * @msg_status: (output) pointer to msg pointer array to be populated; needs
- * to be allocated by caller
- *
- * Returns an array of message pointers associated with the cleaned
- * descriptors. The pointers are to the original ctlq_msgs sent on the cleaned
- * descriptors. The status will be returned for each; any messages that failed
- * to send will have a non-zero status. The caller is expected to free original
- * ctlq_msgs and free or reuse the DMA buffers.
- */
-int idpf_ctlq_clean_sq(struct idpf_ctlq_info *cq, u16 *clean_count,
- struct idpf_ctlq_msg *msg_status[])
-{
- struct idpf_ctlq_desc *desc;
- u16 i, num_to_clean;
- u16 ntc, desc_err;
-
- if (*clean_count == 0)
- return 0;
- if (*clean_count > cq->ring_size)
- return -EBADR;
-
- spin_lock(&cq->cq_lock);
-
- ntc = cq->next_to_clean;
-
- num_to_clean = *clean_count;
-
- for (i = 0; i < num_to_clean; i++) {
- /* Fetch next descriptor and check if marked as done */
- desc = IDPF_CTLQ_DESC(cq, ntc);
- if (!(le16_to_cpu(desc->flags) & IDPF_CTLQ_FLAG_DD))
- break;
-
- /* Ensure no other fields are read until DD flag is checked */
- dma_rmb();
-
- /* strip off FW internal code */
- desc_err = le16_to_cpu(desc->ret_val) & 0xff;
-
- msg_status[i] = cq->bi.tx_msg[ntc];
- msg_status[i]->status = desc_err;
-
- cq->bi.tx_msg[ntc] = NULL;
-
- /* Zero out any stale data */
- memset(desc, 0, sizeof(*desc));
-
- ntc++;
- if (ntc == cq->ring_size)
- ntc = 0;
- }
-
- cq->next_to_clean = ntc;
-
- spin_unlock(&cq->cq_lock);
-
- /* Return number of descriptors actually cleaned */
- *clean_count = i;
-
- return 0;
-}
-
-/**
- * idpf_ctlq_post_rx_buffs - post buffers to descriptor ring
- * @hw: pointer to hw struct
- * @cq: pointer to control queue handle
- * @buff_count: (input|output) input is number of buffers caller is trying to
- * return; output is number of buffers that were not posted
- * @buffs: array of pointers to dma mem structs to be given to hardware
- *
- * Caller uses this function to return DMA buffers to the descriptor ring after
- * consuming them; buff_count will be the number of buffers.
- *
- * Note: this function needs to be called after a receive call even
- * if there are no DMA buffers to be returned, i.e. buff_count = 0,
- * buffs = NULL to support direct commands
- */
-int idpf_ctlq_post_rx_buffs(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
- u16 *buff_count, struct idpf_dma_mem **buffs)
-{
- struct idpf_ctlq_desc *desc;
- u16 ntp = cq->next_to_post;
- bool buffs_avail = false;
- u16 tbp = ntp + 1;
- int i = 0;
-
- if (*buff_count > cq->ring_size)
- return -EBADR;
-
- if (*buff_count > 0)
- buffs_avail = true;
-
- spin_lock(&cq->cq_lock);
-
- if (tbp >= cq->ring_size)
- tbp = 0;
-
- if (tbp == cq->next_to_clean)
- /* Nothing to do */
- goto post_buffs_out;
-
- /* Post buffers for as many as provided or up until the last one used */
- while (ntp != cq->next_to_clean) {
- desc = IDPF_CTLQ_DESC(cq, ntp);
-
- if (cq->bi.rx_buff[ntp])
- goto fill_desc;
- if (!buffs_avail) {
- /* If the caller hasn't given us any buffers or
- * there are none left, search the ring itself
- * for an available buffer to move to this
- * entry starting at the next entry in the ring
- */
- tbp = ntp + 1;
-
- /* Wrap ring if necessary */
- if (tbp >= cq->ring_size)
- tbp = 0;
-
- while (tbp != cq->next_to_clean) {
- if (cq->bi.rx_buff[tbp]) {
- cq->bi.rx_buff[ntp] =
- cq->bi.rx_buff[tbp];
- cq->bi.rx_buff[tbp] = NULL;
-
- /* Found a buffer, no need to
- * search anymore
- */
- break;
- }
-
- /* Wrap ring if necessary */
- tbp++;
- if (tbp >= cq->ring_size)
- tbp = 0;
- }
-
- if (tbp == cq->next_to_clean)
- goto post_buffs_out;
- } else {
- /* Give back pointer to DMA buffer */
- cq->bi.rx_buff[ntp] = buffs[i];
- i++;
-
- if (i >= *buff_count)
- buffs_avail = false;
- }
-
-fill_desc:
- desc->flags =
- cpu_to_le16(IDPF_CTLQ_FLAG_BUF | IDPF_CTLQ_FLAG_RD);
-
- /* Post buffers to descriptor */
- desc->datalen = cpu_to_le16(cq->bi.rx_buff[ntp]->size);
- desc->params.indirect.addr_high =
- cpu_to_le32(upper_32_bits(cq->bi.rx_buff[ntp]->pa));
- desc->params.indirect.addr_low =
- cpu_to_le32(lower_32_bits(cq->bi.rx_buff[ntp]->pa));
-
- ntp++;
- if (ntp == cq->ring_size)
- ntp = 0;
- }
-
-post_buffs_out:
- /* Only update tail if buffers were actually posted */
- if (cq->next_to_post != ntp) {
- if (ntp)
- /* Update next_to_post to ntp - 1 since current ntp
- * will not have a buffer
- */
- cq->next_to_post = ntp - 1;
- else
- /* Wrap to end of end ring since current ntp is 0 */
- cq->next_to_post = cq->ring_size - 1;
-
- dma_wmb();
-
- writel(cq->next_to_post,
- libie_pci_get_mmio_addr(&hw->back->ctlq_ctx.mmio_info,
- cq->reg.tail));
- }
-
- spin_unlock(&cq->cq_lock);
-
- /* return the number of buffers that were not posted */
- *buff_count = *buff_count - i;
-
- return 0;
-}
-
-/**
- * idpf_ctlq_recv - receive control queue message call back
- * @cq: pointer to control queue handle to receive on
- * @num_q_msg: (input|output) input number of messages that should be received;
- * output number of messages actually received
- * @q_msg: (output) array of received control queue messages on this q;
- * needs to be pre-allocated by caller for as many messages as requested
- *
- * Called by interrupt handler or polling mechanism. Caller is expected
- * to free buffers
- */
-int idpf_ctlq_recv(struct idpf_ctlq_info *cq, u16 *num_q_msg,
- struct idpf_ctlq_msg *q_msg)
-{
- u16 num_to_clean, ntc, flags;
- struct idpf_ctlq_desc *desc;
- int err = 0;
- u16 i;
-
- /* take the lock before we start messing with the ring */
- spin_lock(&cq->cq_lock);
-
- ntc = cq->next_to_clean;
-
- num_to_clean = *num_q_msg;
-
- for (i = 0; i < num_to_clean; i++) {
- /* Fetch next descriptor and check if marked as done */
- desc = IDPF_CTLQ_DESC(cq, ntc);
- flags = le16_to_cpu(desc->flags);
-
- if (!(flags & IDPF_CTLQ_FLAG_DD))
- break;
-
- /* Ensure no other fields are read until DD flag is checked */
- dma_rmb();
-
- q_msg[i].vmvf_type = (flags &
- (IDPF_CTLQ_FLAG_FTYPE_VM |
- IDPF_CTLQ_FLAG_FTYPE_PF)) >>
- IDPF_CTLQ_FLAG_FTYPE_S;
-
- if (flags & IDPF_CTLQ_FLAG_ERR)
- err = -EBADMSG;
-
- q_msg[i].cookie.mbx.chnl_opcode =
- le32_to_cpu(desc->v_opcode_dtype);
- q_msg[i].cookie.mbx.chnl_retval =
- le32_to_cpu(desc->v_retval);
-
- q_msg[i].opcode = le16_to_cpu(desc->opcode);
- q_msg[i].data_len = le16_to_cpu(desc->datalen);
- q_msg[i].status = le16_to_cpu(desc->ret_val);
-
- if (desc->datalen) {
- memcpy(q_msg[i].ctx.indirect.context,
- &desc->params.indirect, IDPF_INDIRECT_CTX_SIZE);
-
- /* Assign pointer to dma buffer to ctlq_msg array
- * to be given to upper layer
- */
- q_msg[i].ctx.indirect.payload = cq->bi.rx_buff[ntc];
-
- /* Zero out pointer to DMA buffer info;
- * will be repopulated by post buffers API
- */
- cq->bi.rx_buff[ntc] = NULL;
- } else {
- memcpy(q_msg[i].ctx.direct, desc->params.raw,
- IDPF_DIRECT_CTX_SIZE);
- }
-
- /* Zero out stale data in descriptor */
- memset(desc, 0, sizeof(struct idpf_ctlq_desc));
-
- ntc++;
- if (ntc == cq->ring_size)
- ntc = 0;
- }
-
- cq->next_to_clean = ntc;
-
- spin_unlock(&cq->cq_lock);
-
- *num_q_msg = i;
- if (*num_q_msg == 0)
- err = -ENOMSG;
-
- return err;
-}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq.h b/drivers/net/ethernet/intel/idpf/idpf_controlq.h
deleted file mode 100644
index acf595e9265f..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq.h
+++ /dev/null
@@ -1,142 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/* Copyright (C) 2023 Intel Corporation */
-
-#ifndef _IDPF_CONTROLQ_H_
-#define _IDPF_CONTROLQ_H_
-
-#include <linux/slab.h>
-
-#include "idpf_controlq_api.h"
-
-/* Maximum buffer length for all control queue types */
-#define IDPF_CTLQ_MAX_BUF_LEN 4096
-
-#define IDPF_CTLQ_DESC(R, i) \
- (&(((struct idpf_ctlq_desc *)((R)->desc_ring.va))[i]))
-
-#define IDPF_CTLQ_DESC_UNUSED(R) \
- ((u16)((((R)->next_to_clean > (R)->next_to_use) ? 0 : (R)->ring_size) + \
- (R)->next_to_clean - (R)->next_to_use - 1))
-
-/* Control Queue default settings */
-#define IDPF_CTRL_SQ_CMD_TIMEOUT 250 /* msecs */
-
-struct idpf_ctlq_desc {
- /* Control queue descriptor flags */
- __le16 flags;
- /* Control queue message opcode */
- __le16 opcode;
- __le16 datalen; /* 0 for direct commands */
- union {
- __le16 ret_val;
- __le16 pfid_vfid;
-#define IDPF_CTLQ_DESC_VF_ID_S 0
-#define IDPF_CTLQ_DESC_VF_ID_M (0x7FF << IDPF_CTLQ_DESC_VF_ID_S)
-#define IDPF_CTLQ_DESC_PF_ID_S 11
-#define IDPF_CTLQ_DESC_PF_ID_M (0x1F << IDPF_CTLQ_DESC_PF_ID_S)
- };
-
- /* Virtchnl message opcode and virtchnl descriptor type
- * v_opcode=[27:0], v_dtype=[31:28]
- */
- __le32 v_opcode_dtype;
- /* Virtchnl return value */
- __le32 v_retval;
- union {
- struct {
- __le32 param0;
- __le32 param1;
- __le32 param2;
- __le32 param3;
- } direct;
- struct {
- __le32 param0;
- __le16 sw_cookie;
- /* Virtchnl flags */
- __le16 v_flags;
- __le32 addr_high;
- __le32 addr_low;
- } indirect;
- u8 raw[16];
- } params;
-};
-
-/* Flags sub-structure
- * |0 |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |
- * |DD |CMP|ERR| * RSV * |FTYPE | *RSV* |RD |VFC|BUF| HOST_ID |
- */
-/* command flags and offsets */
-#define IDPF_CTLQ_FLAG_DD_S 0
-#define IDPF_CTLQ_FLAG_CMP_S 1
-#define IDPF_CTLQ_FLAG_ERR_S 2
-#define IDPF_CTLQ_FLAG_FTYPE_S 6
-#define IDPF_CTLQ_FLAG_RD_S 10
-#define IDPF_CTLQ_FLAG_VFC_S 11
-#define IDPF_CTLQ_FLAG_BUF_S 12
-#define IDPF_CTLQ_FLAG_HOST_ID_S 13
-
-#define IDPF_CTLQ_FLAG_DD BIT(IDPF_CTLQ_FLAG_DD_S) /* 0x1 */
-#define IDPF_CTLQ_FLAG_CMP BIT(IDPF_CTLQ_FLAG_CMP_S) /* 0x2 */
-#define IDPF_CTLQ_FLAG_ERR BIT(IDPF_CTLQ_FLAG_ERR_S) /* 0x4 */
-#define IDPF_CTLQ_FLAG_FTYPE_VM BIT(IDPF_CTLQ_FLAG_FTYPE_S) /* 0x40 */
-#define IDPF_CTLQ_FLAG_FTYPE_PF BIT(IDPF_CTLQ_FLAG_FTYPE_S + 1) /* 0x80 */
-#define IDPF_CTLQ_FLAG_RD BIT(IDPF_CTLQ_FLAG_RD_S) /* 0x400 */
-#define IDPF_CTLQ_FLAG_VFC BIT(IDPF_CTLQ_FLAG_VFC_S) /* 0x800 */
-#define IDPF_CTLQ_FLAG_BUF BIT(IDPF_CTLQ_FLAG_BUF_S) /* 0x1000 */
-
-/* Host ID is a special field that has 3b and not a 1b flag */
-#define IDPF_CTLQ_FLAG_HOST_ID_M MAKE_MASK(0x7000UL, IDPF_CTLQ_FLAG_HOST_ID_S)
-
-struct idpf_mbxq_desc {
- u8 pad[8]; /* CTLQ flags/opcode/len/retval fields */
- u32 chnl_opcode; /* avoid confusion with desc->opcode */
- u32 chnl_retval; /* ditto for desc->retval */
- u32 pf_vf_id; /* used by CP when sending to PF */
-};
-
-/* Max number of MMIO regions not including the mailbox and rstat regions in
- * the fallback case when the whole bar is mapped.
- */
-#define IDPF_MMIO_MAP_FALLBACK_MAX_REMAINING 3
-
-struct idpf_mmio_reg {
- void __iomem *vaddr;
- resource_size_t addr_start;
- resource_size_t addr_len;
-};
-
-/* Define the driver hardware struct to replace other control structs as needed
- * Align to ctlq_hw_info
- */
-struct idpf_hw {
- /* Array of remaining LAN BAR regions */
- int num_lan_regs;
- struct idpf_mmio_reg *lan_regs;
-
- struct idpf_adapter *back;
-
- /* control queue - send and receive */
- struct idpf_ctlq_info *asq;
- struct idpf_ctlq_info *arq;
-
- /* pci info */
- u16 device_id;
- u16 vendor_id;
- u16 subsystem_device_id;
- u16 subsystem_vendor_id;
- u8 revision_id;
- bool adapter_stopped;
-
- struct list_head cq_list_head;
-};
-
-int idpf_ctlq_alloc_ring_res(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq);
-
-void idpf_ctlq_dealloc_ring_res(struct idpf_hw *hw, struct idpf_ctlq_info *cq);
-
-/* prototype for functions used for dynamic memory allocation */
-void *idpf_alloc_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem,
- u64 size);
-void idpf_free_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem);
-#endif /* _IDPF_CONTROLQ_H_ */
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq_api.h b/drivers/net/ethernet/intel/idpf/idpf_controlq_api.h
deleted file mode 100644
index 3414c5f9a831..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq_api.h
+++ /dev/null
@@ -1,177 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/* Copyright (C) 2023 Intel Corporation */
-
-#ifndef _IDPF_CONTROLQ_API_H_
-#define _IDPF_CONTROLQ_API_H_
-
-#include "idpf_mem.h"
-
-struct idpf_hw;
-
-/* Used for queue init, response and events */
-enum idpf_ctlq_type {
- IDPF_CTLQ_TYPE_MAILBOX_TX = 0,
- IDPF_CTLQ_TYPE_MAILBOX_RX = 1,
- IDPF_CTLQ_TYPE_CONFIG_TX = 2,
- IDPF_CTLQ_TYPE_CONFIG_RX = 3,
- IDPF_CTLQ_TYPE_EVENT_RX = 4,
- IDPF_CTLQ_TYPE_RDMA_TX = 5,
- IDPF_CTLQ_TYPE_RDMA_RX = 6,
- IDPF_CTLQ_TYPE_RDMA_COMPL = 7
-};
-
-/* Generic Control Queue Structures */
-struct idpf_ctlq_reg {
- /* used for queue tracking */
- u32 head;
- u32 tail;
- /* Below applies only to default mb (if present) */
- u32 len;
- u32 bah;
- u32 bal;
- u32 len_mask;
- u32 len_ena_mask;
- u32 head_mask;
-};
-
-/* Generic queue msg structure */
-struct idpf_ctlq_msg {
- u8 vmvf_type; /* represents the source of the message on recv */
-#define IDPF_VMVF_TYPE_VF 0
-#define IDPF_VMVF_TYPE_VM 1
-#define IDPF_VMVF_TYPE_PF 2
- u8 host_id;
- /* 3b field used only when sending a message to CP - to be used in
- * combination with target func_id to route the message
- */
-#define IDPF_HOST_ID_MASK 0x7
-
- u16 opcode;
- u16 data_len; /* data_len = 0 when no payload is attached */
- union {
- u16 func_id; /* when sending a message */
- u16 status; /* when receiving a message */
- };
- union {
- struct {
- u32 chnl_opcode;
- u32 chnl_retval;
- } mbx;
- } cookie;
- union {
-#define IDPF_DIRECT_CTX_SIZE 16
-#define IDPF_INDIRECT_CTX_SIZE 8
- /* 16 bytes of context can be provided or 8 bytes of context
- * plus the address of a DMA buffer
- */
- u8 direct[IDPF_DIRECT_CTX_SIZE];
- struct {
- u8 context[IDPF_INDIRECT_CTX_SIZE];
- struct idpf_dma_mem *payload;
- } indirect;
- struct {
- u32 rsvd;
- u16 data;
- u16 flags;
- } sw_cookie;
- } ctx;
-};
-
-/* Generic queue info structures */
-/* MB, CONFIG and EVENT q do not have extended info */
-struct idpf_ctlq_create_info {
- enum idpf_ctlq_type type;
- int id; /* absolute queue offset passed as input
- * -1 for default mailbox if present
- */
- u16 len; /* Queue length passed as input */
- u16 buf_size; /* buffer size passed as input */
- u64 base_address; /* output, HPA of the Queue start */
- struct idpf_ctlq_reg reg; /* registers accessed by ctlqs */
-
- int ext_info_size;
- void *ext_info; /* Specific to q type */
-};
-
-/* Control Queue information */
-struct idpf_ctlq_info {
- struct list_head cq_list;
-
- enum idpf_ctlq_type cq_type;
- int q_id;
- spinlock_t cq_lock; /* control queue lock */
- /* used for interrupt processing */
- u16 next_to_use;
- u16 next_to_clean;
- u16 next_to_post; /* starting descriptor to post buffers
- * to after recev
- */
-
- struct idpf_dma_mem desc_ring; /* descriptor ring memory
- * idpf_dma_mem is defined in OSdep.h
- */
- union {
- struct idpf_dma_mem **rx_buff;
- struct idpf_ctlq_msg **tx_msg;
- } bi;
-
- u16 buf_size; /* queue buffer size */
- u16 ring_size; /* Number of descriptors */
- struct idpf_ctlq_reg reg; /* registers accessed by ctlqs */
-};
-
-/**
- * enum idpf_mbx_opc - PF/VF mailbox commands
- * @idpf_mbq_opc_send_msg_to_cp: used by PF or VF to send a message to its CP
- * @idpf_mbq_opc_send_msg_to_peer_drv: used by PF or VF to send a message to
- * any peer driver
- */
-enum idpf_mbx_opc {
- idpf_mbq_opc_send_msg_to_cp = 0x0801,
- idpf_mbq_opc_send_msg_to_peer_drv = 0x0804,
-};
-
-/* API supported for control queue management */
-/* Will init all required q including default mb. "q_info" is an array of
- * create_info structs equal to the number of control queues to be created.
- */
-int idpf_ctlq_init(struct idpf_hw *hw, u8 num_q,
- struct idpf_ctlq_create_info *q_info);
-
-/* Allocate and initialize a single control queue, which will be added to the
- * control queue list; returns a handle to the created control queue
- */
-int idpf_ctlq_add(struct idpf_hw *hw,
- struct idpf_ctlq_create_info *qinfo,
- struct idpf_ctlq_info **cq);
-
-/* Deinitialize and deallocate a single control queue */
-void idpf_ctlq_remove(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq);
-
-/* Sends messages to HW and will also free the buffer*/
-int idpf_ctlq_send(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq,
- u16 num_q_msg,
- struct idpf_ctlq_msg q_msg[]);
-
-/* Receives messages and called by interrupt handler/polling
- * initiated by app/process. Also caller is supposed to free the buffers
- */
-int idpf_ctlq_recv(struct idpf_ctlq_info *cq, u16 *num_q_msg,
- struct idpf_ctlq_msg *q_msg);
-
-/* Reclaims send descriptors on HW write back */
-int idpf_ctlq_clean_sq(struct idpf_ctlq_info *cq, u16 *clean_count,
- struct idpf_ctlq_msg *msg_status[]);
-
-/* Indicate RX buffers are done being processed */
-int idpf_ctlq_post_rx_buffs(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq,
- u16 *buff_count,
- struct idpf_dma_mem **buffs);
-
-/* Will destroy all q including the default mb */
-void idpf_ctlq_deinit(struct idpf_hw *hw);
-
-#endif /* _IDPF_CONTROLQ_API_H_ */
diff --git a/drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c b/drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c
deleted file mode 100644
index d4d488c7cfd6..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_controlq_setup.c
+++ /dev/null
@@ -1,169 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0-only
-/* Copyright (C) 2023 Intel Corporation */
-
-#include "idpf_controlq.h"
-
-/**
- * idpf_ctlq_alloc_desc_ring - Allocate Control Queue (CQ) rings
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- */
-static int idpf_ctlq_alloc_desc_ring(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- size_t size = cq->ring_size * sizeof(struct idpf_ctlq_desc);
-
- cq->desc_ring.va = idpf_alloc_dma_mem(hw, &cq->desc_ring, size);
- if (!cq->desc_ring.va)
- return -ENOMEM;
-
- return 0;
-}
-
-/**
- * idpf_ctlq_alloc_bufs - Allocate Control Queue (CQ) buffers
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * Allocate the buffer head for all control queues, and if it's a receive
- * queue, allocate DMA buffers
- */
-static int idpf_ctlq_alloc_bufs(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- int i;
-
- /* Do not allocate DMA buffers for transmit queues */
- if (cq->cq_type == IDPF_CTLQ_TYPE_MAILBOX_TX)
- return 0;
-
- /* We'll be allocating the buffer info memory first, then we can
- * allocate the mapped buffers for the event processing
- */
- cq->bi.rx_buff = kzalloc_objs(struct idpf_dma_mem *, cq->ring_size);
- if (!cq->bi.rx_buff)
- return -ENOMEM;
-
- /* allocate the mapped buffers (except for the last one) */
- for (i = 0; i < cq->ring_size - 1; i++) {
- struct idpf_dma_mem *bi;
- int num = 1; /* number of idpf_dma_mem to be allocated */
-
- cq->bi.rx_buff[i] = kzalloc_objs(struct idpf_dma_mem, num);
- if (!cq->bi.rx_buff[i])
- goto unwind_alloc_cq_bufs;
-
- bi = cq->bi.rx_buff[i];
-
- bi->va = idpf_alloc_dma_mem(hw, bi, cq->buf_size);
- if (!bi->va) {
- /* unwind will not free the failed entry */
- kfree(cq->bi.rx_buff[i]);
- goto unwind_alloc_cq_bufs;
- }
- }
-
- return 0;
-
-unwind_alloc_cq_bufs:
- /* don't try to free the one that failed... */
- i--;
- for (; i >= 0; i--) {
- idpf_free_dma_mem(hw, cq->bi.rx_buff[i]);
- kfree(cq->bi.rx_buff[i]);
- }
- kfree(cq->bi.rx_buff);
-
- return -ENOMEM;
-}
-
-/**
- * idpf_ctlq_free_desc_ring - Free Control Queue (CQ) rings
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * This assumes the posted send buffers have already been cleaned
- * and de-allocated
- */
-static void idpf_ctlq_free_desc_ring(struct idpf_hw *hw,
- struct idpf_ctlq_info *cq)
-{
- idpf_free_dma_mem(hw, &cq->desc_ring);
-}
-
-/**
- * idpf_ctlq_free_bufs - Free CQ buffer info elements
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * Free the DMA buffers for RX queues, and DMA buffer header for both RX and TX
- * queues. The upper layers are expected to manage freeing of TX DMA buffers
- */
-static void idpf_ctlq_free_bufs(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- void *bi;
-
- if (cq->cq_type == IDPF_CTLQ_TYPE_MAILBOX_RX) {
- int i;
-
- /* free DMA buffers for rx queues*/
- for (i = 0; i < cq->ring_size; i++) {
- if (cq->bi.rx_buff[i]) {
- idpf_free_dma_mem(hw, cq->bi.rx_buff[i]);
- kfree(cq->bi.rx_buff[i]);
- }
- }
-
- bi = (void *)cq->bi.rx_buff;
- } else {
- bi = (void *)cq->bi.tx_msg;
- }
-
- /* free the buffer header */
- kfree(bi);
-}
-
-/**
- * idpf_ctlq_dealloc_ring_res - Free memory allocated for control queue
- * @hw: pointer to hw struct
- * @cq: pointer to the specific Control queue
- *
- * Free the memory used by the ring, buffers and other related structures
- */
-void idpf_ctlq_dealloc_ring_res(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- /* free ring buffers and the ring itself */
- idpf_ctlq_free_bufs(hw, cq);
- idpf_ctlq_free_desc_ring(hw, cq);
-}
-
-/**
- * idpf_ctlq_alloc_ring_res - allocate memory for descriptor ring and bufs
- * @hw: pointer to hw struct
- * @cq: pointer to control queue struct
- *
- * Do *NOT* hold cq_lock when calling this as the memory allocation routines
- * called are not going to be atomic context safe
- */
-int idpf_ctlq_alloc_ring_res(struct idpf_hw *hw, struct idpf_ctlq_info *cq)
-{
- int err;
-
- /* allocate the ring memory */
- err = idpf_ctlq_alloc_desc_ring(hw, cq);
- if (err)
- return err;
-
- /* allocate buffers in the rings */
- err = idpf_ctlq_alloc_bufs(hw, cq);
- if (err)
- goto idpf_init_cq_free_ring;
-
- /* success! */
- return 0;
-
-idpf_init_cq_free_ring:
- idpf_free_dma_mem(hw, &cq->desc_ring);
-
- return err;
-}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_dev.c b/drivers/net/ethernet/intel/idpf/idpf_dev.c
index a9e170eb24b7..92e707ff2b9c 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_dev.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_dev.c
@@ -10,44 +10,32 @@
/**
* idpf_ctlq_reg_init - initialize default mailbox registers
- * @adapter: adapter structure
- * @cq: pointer to the array of create control queues
+ * @mmio: struct that contains MMIO region info
+ * @cci: struct where the register offset pointer to be copied to
*/
-static void idpf_ctlq_reg_init(struct idpf_adapter *adapter,
- struct idpf_ctlq_create_info *cq)
+static void idpf_ctlq_reg_init(struct libie_mmio_info *mmio,
+ struct libie_ctlq_create_info *cci)
{
- int i;
-
- for (i = 0; i < IDPF_NUM_DFLT_MBX_Q; i++) {
- struct idpf_ctlq_create_info *ccq = cq + i;
-
- switch (ccq->type) {
- case IDPF_CTLQ_TYPE_MAILBOX_TX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = PF_FW_ATQH;
- ccq->reg.tail = PF_FW_ATQT;
- ccq->reg.len = PF_FW_ATQLEN;
- ccq->reg.bah = PF_FW_ATQBAH;
- ccq->reg.bal = PF_FW_ATQBAL;
- ccq->reg.len_mask = PF_FW_ATQLEN_ATQLEN_M;
- ccq->reg.len_ena_mask = PF_FW_ATQLEN_ATQENABLE_M;
- ccq->reg.head_mask = PF_FW_ATQH_ATQH_M;
- break;
- case IDPF_CTLQ_TYPE_MAILBOX_RX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = PF_FW_ARQH;
- ccq->reg.tail = PF_FW_ARQT;
- ccq->reg.len = PF_FW_ARQLEN;
- ccq->reg.bah = PF_FW_ARQBAH;
- ccq->reg.bal = PF_FW_ARQBAL;
- ccq->reg.len_mask = PF_FW_ARQLEN_ARQLEN_M;
- ccq->reg.len_ena_mask = PF_FW_ARQLEN_ARQENABLE_M;
- ccq->reg.head_mask = PF_FW_ARQH_ARQH_M;
- break;
- default:
- break;
- }
- }
+ struct libie_ctlq_reg *tx_reg = &cci[LIBIE_CTLQ_TYPE_TX].reg;
+ struct libie_ctlq_reg *rx_reg = &cci[LIBIE_CTLQ_TYPE_RX].reg;
+
+ tx_reg->head = libie_pci_get_mmio_addr(mmio, PF_FW_ATQH);
+ tx_reg->tail = libie_pci_get_mmio_addr(mmio, PF_FW_ATQT);
+ tx_reg->len = libie_pci_get_mmio_addr(mmio, PF_FW_ATQLEN);
+ tx_reg->addr_high = libie_pci_get_mmio_addr(mmio, PF_FW_ATQBAH);
+ tx_reg->addr_low = libie_pci_get_mmio_addr(mmio, PF_FW_ATQBAL);
+ tx_reg->len_mask = PF_FW_ATQLEN_ATQLEN_M;
+ tx_reg->len_ena_mask = PF_FW_ATQLEN_ATQENABLE_M;
+ tx_reg->head_mask = PF_FW_ATQH_ATQH_M;
+
+ rx_reg->head = libie_pci_get_mmio_addr(mmio, PF_FW_ARQH);
+ rx_reg->tail = libie_pci_get_mmio_addr(mmio, PF_FW_ARQT);
+ rx_reg->len = libie_pci_get_mmio_addr(mmio, PF_FW_ARQLEN);
+ rx_reg->addr_high = libie_pci_get_mmio_addr(mmio, PF_FW_ARQBAH);
+ rx_reg->addr_low = libie_pci_get_mmio_addr(mmio, PF_FW_ARQBAL);
+ rx_reg->len_mask = PF_FW_ARQLEN_ARQLEN_M;
+ rx_reg->len_ena_mask = PF_FW_ARQLEN_ARQENABLE_M;
+ rx_reg->head_mask = PF_FW_ARQH_ARQH_M;
}
/**
diff --git a/drivers/net/ethernet/intel/idpf/idpf_ethtool.c b/drivers/net/ethernet/intel/idpf/idpf_ethtool.c
index bb99d9e7c65d..95c45f12b0f9 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_ethtool.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_ethtool.c
@@ -225,7 +225,7 @@ static int idpf_add_flow_steer(struct net_device *netdev,
spin_unlock_bh(&vport_config->flow_steer_list_lock);
if (err)
- goto out;
+ goto out_free_fltr;
rule->vport_id = cpu_to_le32(vport->vport_id);
rule->count = cpu_to_le32(1);
@@ -252,17 +252,15 @@ static int idpf_add_flow_steer(struct net_device *netdev,
break;
default:
err = -EINVAL;
- goto out;
+ goto out_free_fltr;
}
err = idpf_add_del_fsteer_filters(vport->adapter, rule,
VIRTCHNL2_OP_ADD_FLOW_RULE);
- if (err)
- goto out;
-
- if (info->status != cpu_to_le32(VIRTCHNL2_FLOW_RULE_SUCCESS)) {
- err = -EIO;
- goto out;
+ if (err) {
+ /* virtchnl2 rule is already consumed */
+ kfree(fltr);
+ return err;
}
/* Save a copy of the user's flow spec so ethtool can later retrieve it */
@@ -274,9 +272,10 @@ static int idpf_add_flow_steer(struct net_device *netdev,
user_config->num_fsteer_fltrs++;
spin_unlock_bh(&vport_config->flow_steer_list_lock);
- goto out_free_rule;
-out:
+ return 0;
+
+out_free_fltr:
kfree(fltr);
out_free_rule:
kfree(rule);
@@ -319,12 +318,7 @@ static int idpf_del_flow_steer(struct net_device *netdev,
err = idpf_add_del_fsteer_filters(vport->adapter, rule,
VIRTCHNL2_OP_DEL_FLOW_RULE);
if (err)
- goto out;
-
- if (info->status != cpu_to_le32(VIRTCHNL2_FLOW_RULE_SUCCESS)) {
- err = -EIO;
- goto out;
- }
+ return err;
spin_lock_bh(&vport_config->flow_steer_list_lock);
list_for_each_entry_safe(f, iter,
@@ -340,8 +334,6 @@ static int idpf_del_flow_steer(struct net_device *netdev,
out_unlock:
spin_unlock_bh(&vport_config->flow_steer_list_lock);
-out:
- kfree(rule);
return err;
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_lib.c b/drivers/net/ethernet/intel/idpf/idpf_lib.c
index 875472ae77fd..0d131bf0993e 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_lib.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_lib.c
@@ -1371,6 +1371,7 @@ void idpf_statistics_task(struct work_struct *work)
*/
void idpf_mbx_task(struct work_struct *work)
{
+ struct libie_ctlq_xn_recv_params xn_params;
struct idpf_adapter *adapter;
adapter = container_of(work, struct idpf_adapter, mbx_task.work);
@@ -1381,7 +1382,14 @@ void idpf_mbx_task(struct work_struct *work)
queue_delayed_work(adapter->mbx_wq, &adapter->mbx_task,
usecs_to_jiffies(300));
- idpf_recv_mb_msg(adapter, adapter->hw.arq);
+ xn_params = (struct libie_ctlq_xn_recv_params) {
+ .xnm = adapter->xnm,
+ .ctlq = adapter->arq,
+ .ctlq_msg_handler = idpf_recv_event_msg,
+ .budget = LIBIE_CTLQ_MAX_XN_ENTRIES,
+ };
+
+ libie_ctlq_xn_recv(&xn_params);
}
/**
@@ -1909,7 +1917,6 @@ static void idpf_init_hard_reset(struct idpf_adapter *adapter)
idpf_vc_core_deinit(adapter);
if (!is_reset)
reg_ops->trigger_reset(adapter, IDPF_HR_FUNC_RESET);
- idpf_deinit_dflt_mbx(adapter);
} else {
dev_err(dev, "Unhandled hard reset cause\n");
err = -EBADRQC;
@@ -1984,7 +1991,7 @@ void idpf_vc_event_task(struct work_struct *work)
return;
func_reset:
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ libie_ctlq_xn_shutdown(adapter->xnm);
drv_load:
set_bit(IDPF_HR_RESET_IN_PROG, adapter->flags);
idpf_init_hard_reset(adapter);
@@ -2567,44 +2574,6 @@ static int idpf_set_mac(struct net_device *netdev, void *p)
return err;
}
-/**
- * idpf_alloc_dma_mem - Allocate dma memory
- * @hw: pointer to hw struct
- * @mem: pointer to dma_mem struct
- * @size: size of the memory to allocate
- */
-void *idpf_alloc_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem, u64 size)
-{
- struct idpf_adapter *adapter = hw->back;
- size_t sz = ALIGN(size, 4096);
-
- /* The control queue resources are freed under a spinlock, contiguous
- * pages will avoid IOMMU remapping and the use vmap (and vunmap in
- * dma_free_*() path.
- */
- mem->va = dma_alloc_attrs(&adapter->pdev->dev, sz, &mem->pa,
- GFP_KERNEL, DMA_ATTR_FORCE_CONTIGUOUS);
- mem->size = sz;
-
- return mem->va;
-}
-
-/**
- * idpf_free_dma_mem - Free the allocated dma memory
- * @hw: pointer to hw struct
- * @mem: pointer to dma_mem struct
- */
-void idpf_free_dma_mem(struct idpf_hw *hw, struct idpf_dma_mem *mem)
-{
- struct idpf_adapter *adapter = hw->back;
-
- dma_free_attrs(&adapter->pdev->dev, mem->size,
- mem->va, mem->pa, DMA_ATTR_FORCE_CONTIGUOUS);
- mem->size = 0;
- mem->va = NULL;
- mem->pa = 0;
-}
-
static int idpf_hwtstamp_set(struct net_device *netdev,
struct kernel_hwtstamp_config *config,
struct netlink_ext_ack *extack)
diff --git a/drivers/net/ethernet/intel/idpf/idpf_main.c b/drivers/net/ethernet/intel/idpf/idpf_main.c
index d8e19d5f1647..d99f759c55e1 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_main.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_main.c
@@ -133,7 +133,6 @@ static void idpf_remove(struct pci_dev *pdev)
/* Be a good citizen and leave the device clean on exit */
adapter->dev_ops.reg_ops.trigger_reset(adapter, IDPF_HR_FUNC_RESET);
- idpf_deinit_dflt_mbx(adapter);
if (!adapter->netdevs)
goto destroy_wqs;
@@ -170,8 +169,6 @@ static void idpf_remove(struct pci_dev *pdev)
adapter->vport_config = NULL;
kfree(adapter->netdevs);
adapter->netdevs = NULL;
- kfree(adapter->vcxn_mngr);
- adapter->vcxn_mngr = NULL;
mutex_destroy(&adapter->vport_ctrl_lock);
mutex_destroy(&adapter->vector_lock);
@@ -194,7 +191,6 @@ static void idpf_shutdown(struct pci_dev *pdev)
cancel_delayed_work_sync(&adapter->serv_task);
cancel_delayed_work_sync(&adapter->vc_event_task);
idpf_vc_core_deinit(adapter);
- idpf_deinit_dflt_mbx(adapter);
if (system_state == SYSTEM_POWER_OFF)
pci_set_power_state(pdev, PCI_D3hot);
@@ -239,7 +235,6 @@ static int idpf_cfg_device(struct idpf_adapter *adapter)
pci_dbg(pdev, "PCIe PTM is not supported by PCIe bus/controller\n");
pci_set_drvdata(pdev, adapter);
- adapter->hw.back = adapter;
return 0;
}
diff --git a/drivers/net/ethernet/intel/idpf/idpf_mem.h b/drivers/net/ethernet/intel/idpf/idpf_mem.h
deleted file mode 100644
index 2aaabdc02dd2..000000000000
--- a/drivers/net/ethernet/intel/idpf/idpf_mem.h
+++ /dev/null
@@ -1,20 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-only */
-/* Copyright (C) 2023 Intel Corporation */
-
-#ifndef _IDPF_MEM_H_
-#define _IDPF_MEM_H_
-
-#include <linux/io.h>
-
-struct idpf_dma_mem {
- void *va;
- dma_addr_t pa;
- size_t size;
-};
-
-#define idpf_mbx_wr32(a, reg, value) writel((value), ((a)->mbx.vaddr + (reg)))
-#define idpf_mbx_rd32(a, reg) readl((a)->mbx.vaddr + (reg))
-#define idpf_mbx_wr64(a, reg, value) writeq((value), ((a)->mbx.vaddr + (reg)))
-#define idpf_mbx_rd64(a, reg) readq((a)->mbx.vaddr + (reg))
-
-#endif /* _IDPF_MEM_H_ */
diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.h b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
index e101ffb20ae0..a82794c8db3b 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_txrx.h
+++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.h
@@ -236,7 +236,7 @@ enum idpf_tx_ctx_desc_eipt_offload {
(sizeof(u16) * IDPF_RX_MAX_PTYPE_PROTO_IDS))
#define IDPF_RX_PTYPE_HDR_SZ sizeof(struct virtchnl2_get_ptype_info)
#define IDPF_RX_MAX_PTYPES_PER_BUF \
- DIV_ROUND_DOWN_ULL((IDPF_CTLQ_MAX_BUF_LEN - IDPF_RX_PTYPE_HDR_SZ), \
+ DIV_ROUND_DOWN_ULL(LIBIE_CTLQ_MAX_BUF_LEN - IDPF_RX_PTYPE_HDR_SZ, \
IDPF_RX_MAX_PTYPE_SZ)
#define IDPF_GET_PTYPE_SIZE(p) struct_size((p), proto_id, (p)->proto_id_count)
diff --git a/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c b/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
index bf3f6a8bac2d..54535cd8dfaa 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_vf_dev.c
@@ -9,42 +9,32 @@
/**
* idpf_vf_ctlq_reg_init - initialize default mailbox registers
- * @adapter: adapter structure
- * @cq: pointer to the array of create control queues
+ * @mmio: struct that contains MMIO region info
+ * @cci: struct where the register offset pointer to be copied to
*/
-static void idpf_vf_ctlq_reg_init(struct idpf_adapter *adapter,
- struct idpf_ctlq_create_info *cq)
+static void idpf_vf_ctlq_reg_init(struct libie_mmio_info *mmio,
+ struct libie_ctlq_create_info *cci)
{
- for (int i = 0; i < IDPF_NUM_DFLT_MBX_Q; i++) {
- struct idpf_ctlq_create_info *ccq = cq + i;
-
- switch (ccq->type) {
- case IDPF_CTLQ_TYPE_MAILBOX_TX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = VF_ATQH;
- ccq->reg.tail = VF_ATQT;
- ccq->reg.len = VF_ATQLEN;
- ccq->reg.bah = VF_ATQBAH;
- ccq->reg.bal = VF_ATQBAL;
- ccq->reg.len_mask = VF_ATQLEN_ATQLEN_M;
- ccq->reg.len_ena_mask = VF_ATQLEN_ATQENABLE_M;
- ccq->reg.head_mask = VF_ATQH_ATQH_M;
- break;
- case IDPF_CTLQ_TYPE_MAILBOX_RX:
- /* set head and tail registers in our local struct */
- ccq->reg.head = VF_ARQH;
- ccq->reg.tail = VF_ARQT;
- ccq->reg.len = VF_ARQLEN;
- ccq->reg.bah = VF_ARQBAH;
- ccq->reg.bal = VF_ARQBAL;
- ccq->reg.len_mask = VF_ARQLEN_ARQLEN_M;
- ccq->reg.len_ena_mask = VF_ARQLEN_ARQENABLE_M;
- ccq->reg.head_mask = VF_ARQH_ARQH_M;
- break;
- default:
- break;
- }
- }
+ struct libie_ctlq_reg *tx_reg = &cci[LIBIE_CTLQ_TYPE_TX].reg;
+ struct libie_ctlq_reg *rx_reg = &cci[LIBIE_CTLQ_TYPE_RX].reg;
+
+ tx_reg->head = libie_pci_get_mmio_addr(mmio, VF_ATQH);
+ tx_reg->tail = libie_pci_get_mmio_addr(mmio, VF_ATQT);
+ tx_reg->len = libie_pci_get_mmio_addr(mmio, VF_ATQLEN);
+ tx_reg->addr_high = libie_pci_get_mmio_addr(mmio, VF_ATQBAH);
+ tx_reg->addr_low = libie_pci_get_mmio_addr(mmio, VF_ATQBAL);
+ tx_reg->len_mask = VF_ATQLEN_ATQLEN_M;
+ tx_reg->len_ena_mask = VF_ATQLEN_ATQENABLE_M;
+ tx_reg->head_mask = VF_ATQH_ATQH_M;
+
+ rx_reg->head = libie_pci_get_mmio_addr(mmio, VF_ARQH);
+ rx_reg->tail = libie_pci_get_mmio_addr(mmio, VF_ARQT);
+ rx_reg->len = libie_pci_get_mmio_addr(mmio, VF_ARQLEN);
+ rx_reg->addr_high = libie_pci_get_mmio_addr(mmio, VF_ARQBAH);
+ rx_reg->addr_low = libie_pci_get_mmio_addr(mmio, VF_ARQBAL);
+ rx_reg->len_mask = VF_ARQLEN_ARQLEN_M;
+ rx_reg->len_ena_mask = VF_ARQLEN_ARQENABLE_M;
+ rx_reg->head_mask = VF_ARQH_ARQH_M;
}
/**
@@ -157,11 +147,13 @@ static void idpf_vf_reset_reg_init(struct idpf_adapter *adapter)
static void idpf_vf_trigger_reset(struct idpf_adapter *adapter,
enum idpf_flags trig_cause)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_RESET_VF,
+ };
/* Do not send VIRTCHNL2_OP_RESET_VF message on driver unload */
if (trig_cause == IDPF_HR_FUNC_RESET &&
!test_bit(IDPF_REMOVE_IN_PROG, adapter->flags))
- idpf_send_mb_msg(adapter, adapter->hw.asq,
- VIRTCHNL2_OP_RESET_VF, 0, NULL, 0);
+ idpf_send_mb_msg(adapter, &xn_params, NULL, 0);
}
/**
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
index 3e6411a07e4d..13c8505d126f 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.c
@@ -9,20 +9,6 @@
#include "idpf_virtchnl.h"
#include "idpf_ptp.h"
-/**
- * struct idpf_vc_xn_manager - Manager for tracking transactions
- * @ring: backing and lookup for transactions
- * @free_xn_bm: bitmap for free transactions
- * @xn_bm_lock: make bitmap access synchronous where necessary
- * @salt: used to make cookie unique every message
- */
-struct idpf_vc_xn_manager {
- struct idpf_vc_xn ring[IDPF_VC_XN_RING_LEN];
- DECLARE_BITMAP(free_xn_bm, IDPF_VC_XN_RING_LEN);
- spinlock_t xn_bm_lock;
- u8 salt;
-};
-
/**
* idpf_vid_to_vport - Translate vport id to vport pointer
* @adapter: private data struct
@@ -83,79 +69,65 @@ static void idpf_handle_event_link(struct idpf_adapter *adapter,
/**
* idpf_recv_event_msg - Receive virtchnl event message
- * @adapter: Driver specific private structure
+ * @ctx: control queue context
* @ctlq_msg: message to copy from
*
* Receive virtchnl event message
*/
-static void idpf_recv_event_msg(struct idpf_adapter *adapter,
- struct idpf_ctlq_msg *ctlq_msg)
+void idpf_recv_event_msg(struct libie_ctlq_ctx *ctx,
+ struct libie_ctlq_msg *ctlq_msg)
{
- int payload_size = ctlq_msg->ctx.indirect.payload->size;
+ struct kvec *buff = &ctlq_msg->recv_mem;
+ int payload_size = buff->iov_len;
+ struct idpf_adapter *adapter;
struct virtchnl2_event *v2e;
u32 event;
+ adapter = container_of(ctx, struct idpf_adapter, ctlq_ctx);
if (payload_size < sizeof(*v2e)) {
dev_err_ratelimited(&adapter->pdev->dev, "Failed to receive valid payload for event msg (op %d len %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode,
+ ctlq_msg->chnl_opcode,
payload_size);
- return;
+ goto free_rx_buf;
}
- v2e = (struct virtchnl2_event *)ctlq_msg->ctx.indirect.payload->va;
+ v2e = (struct virtchnl2_event *)buff->iov_base;
event = le32_to_cpu(v2e->event);
switch (event) {
case VIRTCHNL2_EVENT_LINK_CHANGE:
idpf_handle_event_link(adapter, v2e);
- return;
+ break;
default:
dev_err(&adapter->pdev->dev,
"Unknown event %d from PF\n", event);
break;
}
+
+free_rx_buf:
+ libie_ctlq_release_rx_buf(buff);
}
/**
* idpf_mb_clean - Reclaim the send mailbox queue entries
* @adapter: driver specific private structure
* @asq: send control queue info
+ * @deinit: release all buffers before destroying the queue
*
- * Reclaim the send mailbox queue entries to be used to send further messages
- *
- * Return: 0 on success, negative on failure
+ * This is a helper function to clean the send mailbox queue entries.
*/
-static int idpf_mb_clean(struct idpf_adapter *adapter,
- struct idpf_ctlq_info *asq)
+static void idpf_mb_clean(struct idpf_adapter *adapter,
+ struct libie_ctlq_info *asq, bool deinit)
{
- u16 i, num_q_msg = IDPF_DFLT_MBX_Q_LEN;
- struct idpf_ctlq_msg **q_msg;
- struct idpf_dma_mem *dma_mem;
- int err;
-
- q_msg = kzalloc_objs(struct idpf_ctlq_msg *, num_q_msg, GFP_ATOMIC);
- if (!q_msg)
- return -ENOMEM;
-
- err = idpf_ctlq_clean_sq(asq, &num_q_msg, q_msg);
- if (err)
- goto err_kfree;
-
- for (i = 0; i < num_q_msg; i++) {
- if (!q_msg[i])
- continue;
- dma_mem = q_msg[i]->ctx.indirect.payload;
- if (dma_mem)
- dma_free_coherent(&adapter->pdev->dev, dma_mem->size,
- dma_mem->va, dma_mem->pa);
- kfree(q_msg[i]);
- kfree(dma_mem);
- }
-
-err_kfree:
- kfree(q_msg);
+ struct libie_ctlq_xn_clean_params clean_params = {
+ .ctx = &adapter->ctlq_ctx,
+ .ctlq = asq,
+ .rel_tx_buf = kfree,
+ .num_msgs = IDPF_DFLT_MBX_Q_LEN,
+ .force = deinit,
+ };
- return err;
+ libie_ctlq_xn_send_clean(&clean_params);
}
#if IS_ENABLED(CONFIG_PTP_1588_CLOCK)
@@ -189,7 +161,7 @@ static bool idpf_ptp_is_mb_msg(u32 op)
* @ctlq_msg: Corresponding control queue message
*/
static void idpf_prepare_ptp_mb_msg(struct idpf_adapter *adapter, u32 op,
- struct idpf_ctlq_msg *ctlq_msg)
+ struct libie_ctlq_msg *ctlq_msg)
{
/* If the message is PTP-related and the secondary mailbox is available,
* send the message through the secondary mailbox.
@@ -197,528 +169,83 @@ static void idpf_prepare_ptp_mb_msg(struct idpf_adapter *adapter, u32 op,
if (!idpf_ptp_is_mb_msg(op) || !adapter->ptp->secondary_mbx.valid)
return;
- ctlq_msg->opcode = idpf_mbq_opc_send_msg_to_peer_drv;
+ ctlq_msg->opcode = LIBIE_CTLQ_SEND_MSG_TO_PEER;
ctlq_msg->func_id = adapter->ptp->secondary_mbx.peer_mbx_q_id;
- ctlq_msg->host_id = adapter->ptp->secondary_mbx.peer_id;
+ ctlq_msg->flags = FIELD_PREP(LIBIE_CTLQ_DESC_FLAG_HOST_ID,
+ adapter->ptp->secondary_mbx.peer_id);
}
#else /* !CONFIG_PTP_1588_CLOCK */
static void idpf_prepare_ptp_mb_msg(struct idpf_adapter *adapter, u32 op,
- struct idpf_ctlq_msg *ctlq_msg)
+ struct libie_ctlq_msg *ctlq_msg)
{ }
#endif /* CONFIG_PTP_1588_CLOCK */
/**
- * idpf_send_mb_msg - Send message over mailbox
+ * idpf_send_mb_msg - send mailbox message to the device control plane
* @adapter: driver specific private structure
- * @asq: control queue to send message to
- * @op: virtchnl opcode
- * @msg_size: size of the payload
- * @msg: pointer to buffer holding the payload
- * @cookie: unique SW generated cookie per message
- *
- * Will prepare the control queue message and initiates the send api
- *
- * Return: 0 on success, negative on failure
- */
-int idpf_send_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *asq,
- u32 op, u16 msg_size, u8 *msg, u16 cookie)
-{
- struct idpf_ctlq_msg *ctlq_msg;
- struct idpf_dma_mem *dma_mem;
- int err;
-
- /* If we are here and a reset is detected nothing much can be
- * done. This thread should silently abort and expected to
- * be corrected with a new run either by user or driver
- * flows after reset
- */
- if (idpf_is_reset_detected(adapter))
- return 0;
-
- err = idpf_mb_clean(adapter, asq);
- if (err)
- return err;
-
- ctlq_msg = kzalloc_obj(*ctlq_msg, GFP_ATOMIC);
- if (!ctlq_msg)
- return -ENOMEM;
-
- dma_mem = kzalloc_obj(*dma_mem, GFP_ATOMIC);
- if (!dma_mem) {
- err = -ENOMEM;
- goto dma_mem_error;
- }
-
- ctlq_msg->opcode = idpf_mbq_opc_send_msg_to_cp;
- ctlq_msg->func_id = 0;
-
- idpf_prepare_ptp_mb_msg(adapter, op, ctlq_msg);
-
- ctlq_msg->data_len = msg_size;
- ctlq_msg->cookie.mbx.chnl_opcode = op;
- ctlq_msg->cookie.mbx.chnl_retval = 0;
- dma_mem->size = IDPF_CTLQ_MAX_BUF_LEN;
- dma_mem->va = dma_alloc_coherent(&adapter->pdev->dev, dma_mem->size,
- &dma_mem->pa, GFP_ATOMIC);
- if (!dma_mem->va) {
- err = -ENOMEM;
- goto dma_alloc_error;
- }
-
- /* It's possible we're just sending an opcode but no buffer */
- if (msg && msg_size)
- memcpy(dma_mem->va, msg, msg_size);
- ctlq_msg->ctx.indirect.payload = dma_mem;
- ctlq_msg->ctx.sw_cookie.data = cookie;
-
- err = idpf_ctlq_send(&adapter->hw, asq, 1, ctlq_msg);
- if (err)
- goto send_error;
-
- return 0;
-
-send_error:
- dma_free_coherent(&adapter->pdev->dev, dma_mem->size, dma_mem->va,
- dma_mem->pa);
-dma_alloc_error:
- kfree(dma_mem);
-dma_mem_error:
- kfree(ctlq_msg);
-
- return err;
-}
-
-/* API for virtchnl "transaction" support ("xn" for short). */
-
-/**
- * idpf_vc_xn_lock - Request exclusive access to vc transaction
- * @xn: struct idpf_vc_xn* to access
- */
-#define idpf_vc_xn_lock(xn) \
- spin_lock(&(xn)->lock)
-
-/**
- * idpf_vc_xn_unlock - Release exclusive access to vc transaction
- * @xn: struct idpf_vc_xn* to access
- */
-#define idpf_vc_xn_unlock(xn) \
- spin_unlock(&(xn)->lock)
-
-/**
- * idpf_vc_xn_release_bufs - Release reference to reply buffer(s) and
- * reset the transaction state.
- * @xn: struct idpf_vc_xn to update
- */
-static void idpf_vc_xn_release_bufs(struct idpf_vc_xn *xn)
-{
- xn->reply.iov_base = NULL;
- xn->reply.iov_len = 0;
-
- if (xn->state != IDPF_VC_XN_SHUTDOWN)
- xn->state = IDPF_VC_XN_IDLE;
-}
-
-/**
- * idpf_vc_xn_init - Initialize virtchnl transaction object
- * @vcxn_mngr: pointer to vc transaction manager struct
- */
-static void idpf_vc_xn_init(struct idpf_vc_xn_manager *vcxn_mngr)
-{
- int i;
-
- spin_lock_init(&vcxn_mngr->xn_bm_lock);
-
- for (i = 0; i < ARRAY_SIZE(vcxn_mngr->ring); i++) {
- struct idpf_vc_xn *xn = &vcxn_mngr->ring[i];
-
- xn->state = IDPF_VC_XN_IDLE;
- xn->idx = i;
- idpf_vc_xn_release_bufs(xn);
- spin_lock_init(&xn->lock);
- init_completion(&xn->completed);
- }
-
- bitmap_fill(vcxn_mngr->free_xn_bm, IDPF_VC_XN_RING_LEN);
-}
-
-/**
- * idpf_vc_xn_shutdown - Uninitialize virtchnl transaction object
- * @vcxn_mngr: pointer to vc transaction manager struct
+ * @xn_params: Xn send parameters to fill
+ * @send_buf: buffer to send
+ * @send_buf_size: size of the send buffer
*
- * All waiting threads will be woken-up and their transaction aborted. Further
- * operations on that object will fail.
- */
-void idpf_vc_xn_shutdown(struct idpf_vc_xn_manager *vcxn_mngr)
-{
- int i;
-
- spin_lock_bh(&vcxn_mngr->xn_bm_lock);
- bitmap_zero(vcxn_mngr->free_xn_bm, IDPF_VC_XN_RING_LEN);
- spin_unlock_bh(&vcxn_mngr->xn_bm_lock);
-
- for (i = 0; i < ARRAY_SIZE(vcxn_mngr->ring); i++) {
- struct idpf_vc_xn *xn = &vcxn_mngr->ring[i];
-
- idpf_vc_xn_lock(xn);
- xn->state = IDPF_VC_XN_SHUTDOWN;
- idpf_vc_xn_release_bufs(xn);
- idpf_vc_xn_unlock(xn);
- complete_all(&xn->completed);
- }
-}
-
-/**
- * idpf_vc_xn_pop_free - Pop a free transaction from free list
- * @vcxn_mngr: transaction manager to pop from
+ * Fill the Xn parameters with the required info to send a virtchnl message.
+ * The send buffer is DMA mapped in the libie to avoid memcpy.
*
- * Returns NULL if no free transactions
- */
-static
-struct idpf_vc_xn *idpf_vc_xn_pop_free(struct idpf_vc_xn_manager *vcxn_mngr)
-{
- struct idpf_vc_xn *xn = NULL;
- unsigned long free_idx;
-
- spin_lock_bh(&vcxn_mngr->xn_bm_lock);
- free_idx = find_first_bit(vcxn_mngr->free_xn_bm, IDPF_VC_XN_RING_LEN);
- if (free_idx == IDPF_VC_XN_RING_LEN)
- goto do_unlock;
-
- clear_bit(free_idx, vcxn_mngr->free_xn_bm);
- xn = &vcxn_mngr->ring[free_idx];
- xn->salt = vcxn_mngr->salt++;
-
-do_unlock:
- spin_unlock_bh(&vcxn_mngr->xn_bm_lock);
-
- return xn;
-}
-
-/**
- * idpf_vc_xn_push_free - Push a free transaction to free list
- * @vcxn_mngr: transaction manager to push to
- * @xn: transaction to push
- */
-static void idpf_vc_xn_push_free(struct idpf_vc_xn_manager *vcxn_mngr,
- struct idpf_vc_xn *xn)
-{
- idpf_vc_xn_release_bufs(xn);
- spin_lock_bh(&vcxn_mngr->xn_bm_lock);
- set_bit(xn->idx, vcxn_mngr->free_xn_bm);
- spin_unlock_bh(&vcxn_mngr->xn_bm_lock);
-}
-
-/**
- * idpf_vc_xn_exec - Perform a send/recv virtchnl transaction
- * @adapter: driver specific private structure with vcxn_mngr
- * @params: parameters for this particular transaction including
- * -vc_op: virtchannel operation to send
- * -send_buf: kvec iov for send buf and len
- * -recv_buf: kvec iov for recv buf and len (ignored if NULL)
- * -timeout_ms: timeout waiting for a reply (milliseconds)
- * -async: don't wait for message reply, will lose caller context
- * -async_handler: callback to handle async replies
- *
- * @returns >= 0 for success, the size of the initial reply (may or may not be
- * >= @recv_buf.iov_len, but we never overflow @@recv_buf_iov_base). < 0 for
- * error.
- */
-ssize_t idpf_vc_xn_exec(struct idpf_adapter *adapter,
- const struct idpf_vc_xn_params *params)
-{
- const struct kvec *send_buf = ¶ms->send_buf;
- struct idpf_vc_xn *xn;
- ssize_t retval;
- u16 cookie;
-
- xn = idpf_vc_xn_pop_free(adapter->vcxn_mngr);
- /* no free transactions available */
- if (!xn)
- return -ENOSPC;
-
- idpf_vc_xn_lock(xn);
- if (xn->state == IDPF_VC_XN_SHUTDOWN) {
- retval = -ENXIO;
- goto only_unlock;
- } else if (xn->state != IDPF_VC_XN_IDLE) {
- /* We're just going to clobber this transaction even though
- * it's not IDLE. If we don't reuse it we could theoretically
- * eventually leak all the free transactions and not be able to
- * send any messages. At least this way we make an attempt to
- * remain functional even though something really bad is
- * happening that's corrupting what was supposed to be free
- * transactions.
- */
- WARN_ONCE(1, "There should only be idle transactions in free list (idx %d op %d)\n",
- xn->idx, xn->vc_op);
- }
-
- xn->reply = params->recv_buf;
- xn->reply_sz = 0;
- xn->state = params->async ? IDPF_VC_XN_ASYNC : IDPF_VC_XN_WAITING;
- xn->vc_op = params->vc_op;
- xn->async_handler = params->async_handler;
- idpf_vc_xn_unlock(xn);
-
- if (!params->async)
- reinit_completion(&xn->completed);
- cookie = FIELD_PREP(IDPF_VC_XN_SALT_M, xn->salt) |
- FIELD_PREP(IDPF_VC_XN_IDX_M, xn->idx);
-
- retval = idpf_send_mb_msg(adapter, adapter->hw.asq, params->vc_op,
- send_buf->iov_len, send_buf->iov_base,
- cookie);
- if (retval) {
- idpf_vc_xn_lock(xn);
- goto release_and_unlock;
- }
-
- if (params->async)
- return 0;
-
- wait_for_completion_timeout(&xn->completed,
- msecs_to_jiffies(params->timeout_ms));
-
- /* No need to check the return value; we check the final state of the
- * transaction below. It's possible the transaction actually gets more
- * timeout than specified if we get preempted here but after
- * wait_for_completion_timeout returns. This should be non-issue
- * however.
- */
- idpf_vc_xn_lock(xn);
- switch (xn->state) {
- case IDPF_VC_XN_SHUTDOWN:
- retval = -ENXIO;
- goto only_unlock;
- case IDPF_VC_XN_WAITING:
- dev_notice_ratelimited(&adapter->pdev->dev,
- "Transaction timed-out (op:%d cookie:%04x vc_op:%d salt:%02x timeout:%dms)\n",
- params->vc_op, cookie, xn->vc_op,
- xn->salt, params->timeout_ms);
- retval = -ETIME;
- break;
- case IDPF_VC_XN_COMPLETED_SUCCESS:
- retval = xn->reply_sz;
- break;
- case IDPF_VC_XN_COMPLETED_FAILED:
- dev_notice_ratelimited(&adapter->pdev->dev, "Transaction failed (op %d)\n",
- params->vc_op);
- retval = -EIO;
- break;
- default:
- /* Invalid state. */
- WARN_ON_ONCE(1);
- retval = -EIO;
- break;
- }
-
-release_and_unlock:
- idpf_vc_xn_push_free(adapter->vcxn_mngr, xn);
- /* If we receive a VC reply after here, it will be dropped. */
-only_unlock:
- idpf_vc_xn_unlock(xn);
-
- return retval;
-}
-
-/**
- * idpf_vc_xn_forward_async - Handle async reply receives
- * @adapter: private data struct
- * @xn: transaction to handle
- * @ctlq_msg: corresponding ctlq_msg
+ * Cleanup the mailbox queue entries of the previously sent message to
+ * unmap and release the buffer.
*
- * For async sends we're going to lose the caller's context so, if an
- * async_handler was provided, it can deal with the reply, otherwise we'll just
- * check and report if there is an error.
+ * Return: 0 if the request was successful, -%EBUSY if reset is detected
+ * or Tx control queue is full, other negative error code on failure.
*/
-static int
-idpf_vc_xn_forward_async(struct idpf_adapter *adapter, struct idpf_vc_xn *xn,
- const struct idpf_ctlq_msg *ctlq_msg)
+int idpf_send_mb_msg(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size)
{
- int err = 0;
-
- if (ctlq_msg->cookie.mbx.chnl_opcode != xn->vc_op) {
- dev_err_ratelimited(&adapter->pdev->dev, "Async message opcode does not match transaction opcode (msg: %d) (xn: %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode, xn->vc_op);
- xn->reply_sz = 0;
- err = -EINVAL;
- goto release_bufs;
- }
-
- if (xn->async_handler) {
- err = xn->async_handler(adapter, xn, ctlq_msg);
- goto release_bufs;
- }
-
- if (ctlq_msg->cookie.mbx.chnl_retval) {
- xn->reply_sz = 0;
- dev_err_ratelimited(&adapter->pdev->dev, "Async message failure (op %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode);
- err = -EINVAL;
- }
-
-release_bufs:
- idpf_vc_xn_push_free(adapter->vcxn_mngr, xn);
-
- return err;
-}
-
-/**
- * idpf_vc_xn_forward_reply - copy a reply back to receiving thread
- * @adapter: driver specific private structure with vcxn_mngr
- * @ctlq_msg: controlq message to send back to receiving thread
- */
-static int
-idpf_vc_xn_forward_reply(struct idpf_adapter *adapter,
- const struct idpf_ctlq_msg *ctlq_msg)
-{
- const void *payload = NULL;
- size_t payload_size = 0;
- struct idpf_vc_xn *xn;
- u16 msg_info;
- int err = 0;
- u16 xn_idx;
- u16 salt;
-
- msg_info = ctlq_msg->ctx.sw_cookie.data;
- xn_idx = FIELD_GET(IDPF_VC_XN_IDX_M, msg_info);
- if (xn_idx >= ARRAY_SIZE(adapter->vcxn_mngr->ring)) {
- dev_err_ratelimited(&adapter->pdev->dev, "Out of bounds cookie received: %02x\n",
- xn_idx);
- return -EINVAL;
- }
- xn = &adapter->vcxn_mngr->ring[xn_idx];
- idpf_vc_xn_lock(xn);
- salt = FIELD_GET(IDPF_VC_XN_SALT_M, msg_info);
- if (xn->salt != salt) {
- dev_err_ratelimited(&adapter->pdev->dev, "Transaction salt does not match (exp:%d@%02x(%d) != got:%d@%02x)\n",
- xn->vc_op, xn->salt, xn->state,
- ctlq_msg->cookie.mbx.chnl_opcode, salt);
- idpf_vc_xn_unlock(xn);
- return -EINVAL;
- }
-
- switch (xn->state) {
- case IDPF_VC_XN_WAITING:
- /* success */
- break;
- case IDPF_VC_XN_IDLE:
- dev_err_ratelimited(&adapter->pdev->dev, "Unexpected or belated VC reply (op %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode);
- err = -EINVAL;
- goto out_unlock;
- case IDPF_VC_XN_SHUTDOWN:
- /* ENXIO is a bit special here as the recv msg loop uses that
- * know if it should stop trying to clean the ring if we lost
- * the virtchnl. We need to stop playing with registers and
- * yield.
- */
- err = -ENXIO;
- goto out_unlock;
- case IDPF_VC_XN_ASYNC:
- /* Set reply_sz from the actual payload so that async_handler
- * can evaluate the response.
- */
- xn->reply_sz = ctlq_msg->data_len;
- err = idpf_vc_xn_forward_async(adapter, xn, ctlq_msg);
- idpf_vc_xn_unlock(xn);
- return err;
- default:
- dev_err_ratelimited(&adapter->pdev->dev, "Overwriting VC reply (op %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode);
- err = -EBUSY;
- goto out_unlock;
- }
-
- if (ctlq_msg->cookie.mbx.chnl_opcode != xn->vc_op) {
- dev_err_ratelimited(&adapter->pdev->dev, "Message opcode does not match transaction opcode (msg: %d) (xn: %d)\n",
- ctlq_msg->cookie.mbx.chnl_opcode, xn->vc_op);
- xn->reply_sz = 0;
- xn->state = IDPF_VC_XN_COMPLETED_FAILED;
- err = -EINVAL;
- goto out_unlock;
- }
+ struct libie_ctlq_msg ctlq_msg = {};
- if (ctlq_msg->cookie.mbx.chnl_retval) {
- xn->reply_sz = 0;
- xn->state = IDPF_VC_XN_COMPLETED_FAILED;
- err = -EINVAL;
- goto out_unlock;
- }
+ if (idpf_is_reset_detected(adapter)) {
+ if (!libie_cp_can_send_onstack(send_buf_size))
+ kfree(send_buf);
- if (ctlq_msg->data_len) {
- payload = ctlq_msg->ctx.indirect.payload->va;
- payload_size = ctlq_msg->data_len;
+ return -EBUSY;
}
- xn->reply_sz = payload_size;
- xn->state = IDPF_VC_XN_COMPLETED_SUCCESS;
+ idpf_prepare_ptp_mb_msg(adapter, xn_params->chnl_opcode, &ctlq_msg);
+ xn_params->ctlq_msg = ctlq_msg.opcode ? &ctlq_msg : NULL;
- if (xn->reply.iov_base && xn->reply.iov_len && payload_size)
- memcpy(xn->reply.iov_base, payload,
- min_t(size_t, xn->reply.iov_len, payload_size));
+ xn_params->send_buf.iov_base = send_buf;
+ xn_params->send_buf.iov_len = send_buf_size;
+ xn_params->xnm = adapter->xnm;
+ xn_params->ctlq = xn_params->ctlq ? xn_params->ctlq : adapter->asq;
+ xn_params->rel_tx_buf = kfree;
-out_unlock:
- idpf_vc_xn_unlock(xn);
- /* we _cannot_ hold lock while calling complete */
- complete(&xn->completed);
+ idpf_mb_clean(adapter, xn_params->ctlq, false);
- return err;
+ return libie_ctlq_xn_send(xn_params);
}
/**
- * idpf_recv_mb_msg - Receive message over mailbox
+ * idpf_send_mb_msg_kfree - send mailbox message and free the send buffer
* @adapter: driver specific private structure
- * @arq: control queue to receive message from
+ * @xn_params: Xn send parameters to fill
+ * @send_buf: buffer to send, can be released with kfree()
+ * @send_buf_size: size of the send buffer
*
- * Will receive control queue message and posts the receive buffer.
+ * libie_cp functions consume only buffers above certain size,
+ * smaller buffers are assumed to be on the stack. However, for some
+ * commands with variable message size it makes sense to always use kzalloc(),
+ * which means we have to free smaller buffers ourselves.
*
- * Return: 0 on success and negative on failure.
+ * Return: 0 if no unexpected errors were encountered,
+ * negative error code otherwise.
*/
-int idpf_recv_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *arq)
+int idpf_send_mb_msg_kfree(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size)
{
- struct idpf_ctlq_msg ctlq_msg;
- struct idpf_dma_mem *dma_mem;
- int post_err, err;
- u16 num_recv;
-
- while (1) {
- /* This will get <= num_recv messages and output how many
- * actually received on num_recv.
- */
- num_recv = 1;
- err = idpf_ctlq_recv(arq, &num_recv, &ctlq_msg);
- if (err || !num_recv)
- break;
-
- if (ctlq_msg.data_len) {
- dma_mem = ctlq_msg.ctx.indirect.payload;
- } else {
- dma_mem = NULL;
- num_recv = 0;
- }
+ int err = idpf_send_mb_msg(adapter, xn_params, send_buf, send_buf_size);
- if (ctlq_msg.cookie.mbx.chnl_opcode == VIRTCHNL2_OP_EVENT)
- idpf_recv_event_msg(adapter, &ctlq_msg);
- else
- err = idpf_vc_xn_forward_reply(adapter, &ctlq_msg);
-
- post_err = idpf_ctlq_post_rx_buffs(&adapter->hw, arq,
- &num_recv, &dma_mem);
-
- /* If post failed clear the only buffer we supplied */
- if (post_err) {
- if (dma_mem)
- dma_free_coherent(&adapter->pdev->dev,
- dma_mem->size, dma_mem->va,
- dma_mem->pa);
- break;
- }
-
- /* virtchnl trying to shutdown, stop cleaning */
- if (err == -ENXIO)
- break;
- }
+ if (libie_cp_can_send_onstack(send_buf_size))
+ kfree(send_buf);
return err;
}
@@ -768,45 +295,43 @@ struct idpf_queue_set *idpf_alloc_queue_set(struct idpf_adapter *adapter,
static int idpf_send_chunked_msg(struct idpf_adapter *adapter,
const struct idpf_chunked_msg_params *params)
{
- struct idpf_vc_xn_params xn_params = {
- .vc_op = params->vc_op,
+ struct libie_ctlq_xn_send_params xn_params = {
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = params->vc_op,
};
const void *pos = params->chunks;
- u32 num_chunks, num_msgs, buf_sz;
- void *buf __free(kfree) = NULL;
u32 totqs = params->num_chunks;
u32 vid = params->vport_id;
+ u32 num_chunks, num_msgs;
- num_chunks = min(IDPF_NUM_CHUNKS_PER_MSG(params->config_sz,
- params->chunk_sz), totqs);
+ num_chunks = IDPF_NUM_CHUNKS_PER_MSG(params->config_sz,
+ params->chunk_sz);
num_msgs = DIV_ROUND_UP(totqs, num_chunks);
- buf_sz = params->config_sz + num_chunks * params->chunk_sz;
- buf = kzalloc(buf_sz, GFP_KERNEL);
- if (!buf)
- return -ENOMEM;
-
- xn_params.send_buf.iov_base = buf;
-
for (u32 i = 0; i < num_msgs; i++) {
- ssize_t reply_sz;
+ u32 buf_sz;
+ void *buf;
+ int err;
- memset(buf, 0, buf_sz);
- xn_params.send_buf.iov_len = buf_sz;
+ num_chunks = min(num_chunks, totqs);
+ buf_sz = params->config_sz + num_chunks * params->chunk_sz;
+ buf = kzalloc(buf_sz, GFP_KERNEL);
+ if (!buf)
+ return -ENOMEM;
- if (params->prepare_msg(vid, buf, pos, num_chunks) != buf_sz)
+ if (params->prepare_msg(vid, buf, pos, num_chunks) != buf_sz) {
+ kfree(buf);
return -EINVAL;
+ }
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, buf, buf_sz);
+ if (err)
+ return err;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ xn_params.recv_mem = (struct kvec) {};
pos += num_chunks * params->chunk_sz;
totqs -= num_chunks;
-
- num_chunks = min(num_chunks, totqs);
- buf_sz = params->config_sz + num_chunks * params->chunk_sz;
}
return 0;
@@ -881,11 +406,14 @@ static int idpf_wait_for_marker_event(struct idpf_vport *vport)
*/
static int idpf_send_ver_msg(struct idpf_adapter *adapter)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_VERSION,
+ };
+ struct virtchnl2_version_info *vvi_recv;
struct virtchnl2_version_info vvi;
- ssize_t reply_sz;
u32 major, minor;
- int err = 0;
+ int err;
if (adapter->virt_ver_maj) {
vvi.major = cpu_to_le32(adapter->virt_ver_maj);
@@ -895,24 +423,23 @@ static int idpf_send_ver_msg(struct idpf_adapter *adapter)
vvi.minor = cpu_to_le32(IDPF_VIRTCHNL_VERSION_MINOR);
}
- xn_params.vc_op = VIRTCHNL2_OP_VERSION;
- xn_params.send_buf.iov_base = &vvi;
- xn_params.send_buf.iov_len = sizeof(vvi);
- xn_params.recv_buf = xn_params.send_buf;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ err = idpf_send_mb_msg(adapter, &xn_params, &vvi, sizeof(vvi));
+ if (err)
+ return err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz < sizeof(vvi))
- return -EIO;
+ if (xn_params.recv_mem.iov_len < sizeof(*vvi_recv)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
- major = le32_to_cpu(vvi.major);
- minor = le32_to_cpu(vvi.minor);
+ vvi_recv = xn_params.recv_mem.iov_base;
+ major = le32_to_cpu(vvi_recv->major);
+ minor = le32_to_cpu(vvi_recv->minor);
if (major > IDPF_VIRTCHNL_VERSION_MAJOR) {
dev_warn(&adapter->pdev->dev, "Virtchnl major version greater than supported\n");
- return -EINVAL;
+ err = -EINVAL;
+ goto free_rx_buf;
}
if (major == IDPF_VIRTCHNL_VERSION_MAJOR &&
@@ -930,6 +457,9 @@ static int idpf_send_ver_msg(struct idpf_adapter *adapter)
adapter->virt_ver_maj = major;
adapter->virt_ver_min = minor;
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
return err;
}
@@ -942,9 +472,12 @@ static int idpf_send_ver_msg(struct idpf_adapter *adapter)
*/
static int idpf_send_get_caps_msg(struct idpf_adapter *adapter)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_GET_CAPS,
+ };
struct virtchnl2_get_capabilities caps = {};
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ int err;
caps.csum_caps =
cpu_to_le32(VIRTCHNL2_CAP_TX_CSUM_L3_IPV4 |
@@ -1004,20 +537,22 @@ static int idpf_send_get_caps_msg(struct idpf_adapter *adapter)
VIRTCHNL2_CAP_LOOPBACK |
VIRTCHNL2_CAP_PTP);
- xn_params.vc_op = VIRTCHNL2_OP_GET_CAPS;
- xn_params.send_buf.iov_base = ∩︀
- xn_params.send_buf.iov_len = sizeof(caps);
- xn_params.recv_buf.iov_base = &adapter->caps;
- xn_params.recv_buf.iov_len = sizeof(adapter->caps);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ err = idpf_send_mb_msg(adapter, &xn_params, &caps, sizeof(caps));
+ if (err)
+ return err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz < sizeof(adapter->caps))
- return -EIO;
+ if (xn_params.recv_mem.iov_len < sizeof(adapter->caps)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
- return 0;
+ memcpy(&adapter->caps, xn_params.recv_mem.iov_base,
+ sizeof(adapter->caps));
+
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return err;
}
/**
@@ -1062,37 +597,39 @@ static void idpf_decfg_lan_memory_regions(struct idpf_adapter *adapter)
*/
static int idpf_cfg_lan_memory_regions(struct idpf_adapter *adapter)
{
- struct virtchnl2_get_lan_memory_regions *rcvd_regions __free(kfree);
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_GET_LAN_MEMORY_REGIONS,
- .recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN,
- .send_buf.iov_len =
- sizeof(struct virtchnl2_get_lan_memory_regions) +
- sizeof(struct virtchnl2_mem_region),
+ struct virtchnl2_get_lan_memory_regions *send_regions, *rcvd_regions;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_GET_LAN_MEMORY_REGIONS,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int num_regions, size;
- ssize_t reply_sz;
+ size_t send_sz, reply_sz, size;
+ int num_regions;
int err = 0;
- rcvd_regions = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!rcvd_regions)
+ send_sz = sizeof(struct virtchnl2_get_lan_memory_regions) +
+ sizeof(struct virtchnl2_mem_region);
+ send_regions = kzalloc(send_sz, GFP_KERNEL);
+ if (!send_regions)
return -ENOMEM;
- xn_params.recv_buf.iov_base = rcvd_regions;
- rcvd_regions->num_memory_regions = cpu_to_le16(1);
- xn_params.send_buf.iov_base = rcvd_regions;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ send_regions->num_memory_regions = cpu_to_le16(1);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, send_regions,
+ send_sz);
+ if (err)
+ return err;
+ rcvd_regions = xn_params.recv_mem.iov_base;
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz < sizeof(*rcvd_regions)) {
+ err = -EIO;
+ goto rel_rx_buf;
+ }
num_regions = le16_to_cpu(rcvd_regions->num_memory_regions);
size = struct_size(rcvd_regions, mem_reg, num_regions);
- if (reply_sz < size)
- return -EIO;
-
- if (size > IDPF_CTLQ_MAX_BUF_LEN)
- return -EINVAL;
+ if (reply_sz < size) {
+ err = -EIO;
+ goto rel_rx_buf;
+ }
for (int i = 0; i < num_regions; i++) {
struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
@@ -1102,10 +639,14 @@ static int idpf_cfg_lan_memory_regions(struct idpf_adapter *adapter)
len = le64_to_cpu(rcvd_regions->mem_reg[i].size);
if (len && !libie_pci_map_mmio_region(mmio, offset, len)) {
idpf_decfg_lan_memory_regions(adapter);
- return -EIO;
+ err = -EIO;
+ goto rel_rx_buf;
}
}
+rel_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
return err;
}
@@ -1164,24 +705,43 @@ int idpf_add_del_fsteer_filters(struct idpf_adapter *adapter,
struct virtchnl2_flow_rule_add_del *rule,
enum virtchnl2_op opcode)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = opcode,
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ };
+ struct virtchnl2_flow_rule_add_del *rx_rule;
int rule_count = le32_to_cpu(rule->count);
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ size_t send_sz;
+ int err;
if (opcode != VIRTCHNL2_OP_ADD_FLOW_RULE &&
- opcode != VIRTCHNL2_OP_DEL_FLOW_RULE)
+ opcode != VIRTCHNL2_OP_DEL_FLOW_RULE) {
+ kfree(rule);
return -EINVAL;
+ }
+
+ send_sz = struct_size(rule, rule_info, rule_count);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, rule, send_sz);
+ if (err)
+ return err;
+
+ if (xn_params.recv_mem.iov_len < send_sz) {
+ err = -EIO;
+ goto rel_rx;
+ }
- xn_params.vc_op = opcode;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.async = false;
- xn_params.send_buf.iov_base = rule;
- xn_params.send_buf.iov_len = struct_size(rule, rule_info, rule_count);
- xn_params.recv_buf.iov_base = rule;
- xn_params.recv_buf.iov_len = struct_size(rule, rule_info, rule_count);
+ rx_rule = xn_params.recv_mem.iov_base;
+ for (int i = 0; i < rule_count; i++) {
+ if (rx_rule->rule_info[i].status !=
+ cpu_to_le32(VIRTCHNL2_FLOW_RULE_SUCCESS)) {
+ err = -EIO;
+ goto rel_rx;
+ }
+ }
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- return reply_sz < 0 ? reply_sz : 0;
+rel_rx:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -1555,11 +1115,13 @@ int idpf_queue_reg_init(struct idpf_vport *vport,
int idpf_send_create_vport_msg(struct idpf_adapter *adapter,
struct idpf_vport_max_q *max_q)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_CREATE_VPORT,
+ };
struct virtchnl2_create_vport *vport_msg;
- struct idpf_vc_xn_params xn_params = {};
u16 idx = adapter->next_vport;
int err, buf_size;
- ssize_t reply_sz;
buf_size = sizeof(struct virtchnl2_create_vport);
vport_msg = kzalloc(buf_size, GFP_KERNEL);
@@ -1586,33 +1148,29 @@ int idpf_send_create_vport_msg(struct idpf_adapter *adapter,
}
if (!adapter->vport_params_recvd[idx]) {
- adapter->vport_params_recvd[idx] = kzalloc(IDPF_CTLQ_MAX_BUF_LEN,
- GFP_KERNEL);
+ adapter->vport_params_recvd[idx] =
+ kzalloc(LIBIE_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
if (!adapter->vport_params_recvd[idx]) {
err = -ENOMEM;
goto rel_buf;
}
}
- xn_params.vc_op = VIRTCHNL2_OP_CREATE_VPORT;
- xn_params.send_buf.iov_base = vport_msg;
- xn_params.send_buf.iov_len = buf_size;
- xn_params.recv_buf.iov_base = adapter->vport_params_recvd[idx];
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0) {
- err = reply_sz;
- goto free_vport_params;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, vport_msg,
+ sizeof(*vport_msg));
+ if (err) {
+ kfree(adapter->vport_params_recvd[idx]);
+ adapter->vport_params_recvd[idx] = NULL;
+ return err;
}
- kfree(vport_msg);
+ memcpy(adapter->vport_params_recvd[idx], xn_params.recv_mem.iov_base,
+ xn_params.recv_mem.iov_len);
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
return 0;
-free_vport_params:
- kfree(adapter->vport_params_recvd[idx]);
- adapter->vport_params_recvd[idx] = NULL;
rel_buf:
kfree(vport_msg);
@@ -1674,19 +1232,22 @@ int idpf_check_supported_desc_ids(struct idpf_vport *vport)
*/
int idpf_send_destroy_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DESTROY_VPORT,
+ };
struct virtchnl2_vport v_id;
- ssize_t reply_sz;
+ int err;
v_id.vport_id = cpu_to_le32(vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_DESTROY_VPORT;
- xn_params.send_buf.iov_base = &v_id;
- xn_params.send_buf.iov_len = sizeof(v_id);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg(adapter, &xn_params, &v_id, sizeof(v_id));
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return reply_sz < 0 ? reply_sz : 0;
+ return 0;
}
/**
@@ -1698,19 +1259,22 @@ int idpf_send_destroy_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
*/
int idpf_send_enable_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_ENABLE_VPORT,
+ };
struct virtchnl2_vport v_id;
- ssize_t reply_sz;
+ int err;
v_id.vport_id = cpu_to_le32(vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_ENABLE_VPORT;
- xn_params.send_buf.iov_base = &v_id;
- xn_params.send_buf.iov_len = sizeof(v_id);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg(adapter, &xn_params, &v_id, sizeof(v_id));
+ if (err)
+ return err;
- return reply_sz < 0 ? reply_sz : 0;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return 0;
}
/**
@@ -1722,19 +1286,22 @@ int idpf_send_enable_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
*/
int idpf_send_disable_vport_msg(struct idpf_adapter *adapter, u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DISABLE_VPORT,
+ };
struct virtchnl2_vport v_id;
- ssize_t reply_sz;
+ int err;
v_id.vport_id = cpu_to_le32(vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_DISABLE_VPORT;
- xn_params.send_buf.iov_base = &v_id;
- xn_params.send_buf.iov_len = sizeof(v_id);
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg(adapter, &xn_params, &v_id, sizeof(v_id));
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return reply_sz < 0 ? reply_sz : 0;
+ return 0;
}
/**
@@ -2573,11 +2140,14 @@ int idpf_send_delete_queues_msg(struct idpf_adapter *adapter,
struct idpf_queue_id_reg_info *chunks,
u32 vport_id)
{
- struct virtchnl2_del_ena_dis_queues *eq __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DEL_QUEUES,
+ };
+ struct virtchnl2_del_ena_dis_queues *eq;
+ ssize_t buf_size;
u16 num_chunks;
- int buf_size;
+ int err;
num_chunks = chunks->num_chunks;
buf_size = struct_size(eq, chunks.chunks, num_chunks);
@@ -2592,13 +2162,13 @@ int idpf_send_delete_queues_msg(struct idpf_adapter *adapter,
idpf_convert_reg_to_queue_chunks(eq->chunks.chunks, chunks->queue_chunks,
num_chunks);
- xn_params.vc_op = VIRTCHNL2_OP_DEL_QUEUES;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = eq;
- xn_params.send_buf.iov_len = buf_size;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, eq, buf_size);
+ if (err)
+ return err;
- return reply_sz < 0 ? reply_sz : 0;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return 0;
}
/**
@@ -2636,15 +2206,14 @@ int idpf_send_add_queues_msg(struct idpf_adapter *adapter,
struct idpf_q_vec_rsrc *rsrc,
u32 vport_id)
{
- struct virtchnl2_add_queues *vc_msg __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_ADD_QUEUES,
+ };
+ struct virtchnl2_add_queues *vc_msg;
struct virtchnl2_add_queues aq = {};
- ssize_t reply_sz;
- int size;
-
- vc_msg = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!vc_msg)
- return -ENOMEM;
+ size_t size;
+ int err;
aq.vport_id = cpu_to_le32(vport_id);
aq.num_tx_q = cpu_to_le16(rsrc->num_txq);
@@ -2652,29 +2221,38 @@ int idpf_send_add_queues_msg(struct idpf_adapter *adapter,
aq.num_rx_q = cpu_to_le16(rsrc->num_rxq);
aq.num_rx_bufq = cpu_to_le16(rsrc->num_bufq);
- xn_params.vc_op = VIRTCHNL2_OP_ADD_QUEUES;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &aq;
- xn_params.send_buf.iov_len = sizeof(aq);
- xn_params.recv_buf.iov_base = vc_msg;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg(adapter, &xn_params, &aq, sizeof(aq));
+ if (err)
+ return err;
+
+ vc_msg = xn_params.recv_mem.iov_base;
+ if (xn_params.recv_mem.iov_len < sizeof(*vc_msg)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
/* compare vc_msg num queues with vport num queues */
if (le16_to_cpu(vc_msg->num_tx_q) != rsrc->num_txq ||
le16_to_cpu(vc_msg->num_rx_q) != rsrc->num_rxq ||
le16_to_cpu(vc_msg->num_tx_complq) != rsrc->num_complq ||
- le16_to_cpu(vc_msg->num_rx_bufq) != rsrc->num_bufq)
- return -EINVAL;
+ le16_to_cpu(vc_msg->num_rx_bufq) != rsrc->num_bufq) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
size = struct_size(vc_msg, chunks.chunks,
le16_to_cpu(vc_msg->chunks.num_chunks));
- if (reply_sz < size)
- return -EIO;
+ if (xn_params.recv_mem.iov_len < size) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
+
+ err = idpf_vport_init_queue_reg_chunks(vport_config, &vc_msg->chunks);
+
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return idpf_vport_init_queue_reg_chunks(vport_config, &vc_msg->chunks);
+ return err;
}
/**
@@ -2686,49 +2264,51 @@ int idpf_send_add_queues_msg(struct idpf_adapter *adapter,
*/
int idpf_send_alloc_vectors_msg(struct idpf_adapter *adapter, u16 num_vectors)
{
- struct virtchnl2_alloc_vectors *rcvd_vec __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_ALLOC_VECTORS,
+ };
+ struct virtchnl2_alloc_vectors *rcvd_vec;
struct virtchnl2_alloc_vectors ac = {};
- ssize_t reply_sz;
u16 num_vchunks;
- int size;
+ int size, err;
ac.num_vectors = cpu_to_le16(num_vectors);
- rcvd_vec = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!rcvd_vec)
- return -ENOMEM;
+ err = idpf_send_mb_msg(adapter, &xn_params, &ac, sizeof(ac));
+ if (err)
+ return err;
- xn_params.vc_op = VIRTCHNL2_OP_ALLOC_VECTORS;
- xn_params.send_buf.iov_base = ∾
- xn_params.send_buf.iov_len = sizeof(ac);
- xn_params.recv_buf.iov_base = rcvd_vec;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ rcvd_vec = xn_params.recv_mem.iov_base;
+ if (xn_params.recv_mem.iov_len < sizeof(*rcvd_vec)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
num_vchunks = le16_to_cpu(rcvd_vec->vchunks.num_vchunks);
size = struct_size(rcvd_vec, vchunks.vchunks, num_vchunks);
- if (reply_sz < size)
- return -EIO;
-
- if (size > IDPF_CTLQ_MAX_BUF_LEN)
- return -EINVAL;
+ if (xn_params.recv_mem.iov_len < size) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
kfree(adapter->req_vec_chunks);
adapter->req_vec_chunks = kmemdup(rcvd_vec, size, GFP_KERNEL);
- if (!adapter->req_vec_chunks)
- return -ENOMEM;
+ if (!adapter->req_vec_chunks) {
+ err = -ENOMEM;
+ goto free_rx_buf;
+ }
if (le16_to_cpu(adapter->req_vec_chunks->num_vectors) < num_vectors) {
kfree(adapter->req_vec_chunks);
adapter->req_vec_chunks = NULL;
- return -EINVAL;
+ err = -EINVAL;
}
- return 0;
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return err;
}
/**
@@ -2740,24 +2320,28 @@ int idpf_send_alloc_vectors_msg(struct idpf_adapter *adapter, u16 num_vectors)
int idpf_send_dealloc_vectors_msg(struct idpf_adapter *adapter)
{
struct virtchnl2_alloc_vectors *ac = adapter->req_vec_chunks;
- struct virtchnl2_vector_chunks *vcs = &ac->vchunks;
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
- int buf_size;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_DEALLOC_VECTORS,
+ };
+ struct virtchnl2_vector_chunks *vcs;
+ int buf_size, err;
- buf_size = struct_size(vcs, vchunks, le16_to_cpu(vcs->num_vchunks));
+ buf_size = struct_size(&ac->vchunks, vchunks,
+ le16_to_cpu(ac->vchunks.num_vchunks));
+ vcs = kmemdup(&ac->vchunks, buf_size, GFP_KERNEL);
+ if (!vcs)
+ return -ENOMEM;
- xn_params.vc_op = VIRTCHNL2_OP_DEALLOC_VECTORS;
- xn_params.send_buf.iov_base = vcs;
- xn_params.send_buf.iov_len = buf_size;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, vcs, buf_size);
+ if (err)
+ return err;
kfree(adapter->req_vec_chunks);
adapter->req_vec_chunks = NULL;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
return 0;
}
@@ -2781,18 +2365,22 @@ static int idpf_get_max_vfs(struct idpf_adapter *adapter)
*/
int idpf_send_set_sriov_vfs_msg(struct idpf_adapter *adapter, u16 num_vfs)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_SET_SRIOV_VFS,
+ };
struct virtchnl2_sriov_vfs_info svi = {};
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ int err;
svi.num_vfs = cpu_to_le16(num_vfs);
- xn_params.vc_op = VIRTCHNL2_OP_SET_SRIOV_VFS;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &svi;
- xn_params.send_buf.iov_len = sizeof(svi);
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- return reply_sz < 0 ? reply_sz : 0;
+ err = idpf_send_mb_msg(adapter, &xn_params, &svi, sizeof(svi));
+ if (err)
+ return err;
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return 0;
}
/**
@@ -2805,10 +2393,14 @@ int idpf_send_set_sriov_vfs_msg(struct idpf_adapter *adapter, u16 num_vfs)
int idpf_send_get_stats_msg(struct idpf_netdev_priv *np,
struct idpf_port_stats *port_stats)
{
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_GET_STATS,
+ };
struct rtnl_link_stats64 *netstats = &np->netstats;
+ struct virtchnl2_vport_stats *stats_recv;
struct virtchnl2_vport_stats stats_msg = {};
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
+ int err;
/* Don't send get_stats message if the link is down */
@@ -2817,38 +2409,41 @@ int idpf_send_get_stats_msg(struct idpf_netdev_priv *np,
stats_msg.vport_id = cpu_to_le32(np->vport_id);
- xn_params.vc_op = VIRTCHNL2_OP_GET_STATS;
- xn_params.send_buf.iov_base = &stats_msg;
- xn_params.send_buf.iov_len = sizeof(stats_msg);
- xn_params.recv_buf = xn_params.send_buf;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ err = idpf_send_mb_msg(np->adapter, &xn_params, &stats_msg,
+ sizeof(stats_msg));
+ if (err)
+ return err;
- reply_sz = idpf_vc_xn_exec(np->adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz < sizeof(stats_msg))
- return -EIO;
+ if (xn_params.recv_mem.iov_len < sizeof(*stats_recv)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
+
+ stats_recv = xn_params.recv_mem.iov_base;
spin_lock_bh(&np->stats_lock);
- netstats->rx_packets = le64_to_cpu(stats_msg.rx_unicast) +
- le64_to_cpu(stats_msg.rx_multicast) +
- le64_to_cpu(stats_msg.rx_broadcast);
- netstats->tx_packets = le64_to_cpu(stats_msg.tx_unicast) +
- le64_to_cpu(stats_msg.tx_multicast) +
- le64_to_cpu(stats_msg.tx_broadcast);
- netstats->rx_bytes = le64_to_cpu(stats_msg.rx_bytes);
- netstats->tx_bytes = le64_to_cpu(stats_msg.tx_bytes);
- netstats->rx_errors = le64_to_cpu(stats_msg.rx_errors);
- netstats->tx_errors = le64_to_cpu(stats_msg.tx_errors);
- netstats->rx_dropped = le64_to_cpu(stats_msg.rx_discards);
- netstats->tx_dropped = le64_to_cpu(stats_msg.tx_discards);
-
- port_stats->vport_stats = stats_msg;
+ netstats->rx_packets = le64_to_cpu(stats_recv->rx_unicast) +
+ le64_to_cpu(stats_recv->rx_multicast) +
+ le64_to_cpu(stats_recv->rx_broadcast);
+ netstats->tx_packets = le64_to_cpu(stats_recv->tx_unicast) +
+ le64_to_cpu(stats_recv->tx_multicast) +
+ le64_to_cpu(stats_recv->tx_broadcast);
+ netstats->rx_bytes = le64_to_cpu(stats_recv->rx_bytes);
+ netstats->tx_bytes = le64_to_cpu(stats_recv->tx_bytes);
+ netstats->rx_errors = le64_to_cpu(stats_recv->rx_errors);
+ netstats->tx_errors = le64_to_cpu(stats_recv->tx_errors);
+ netstats->rx_dropped = le64_to_cpu(stats_recv->rx_discards);
+ netstats->tx_dropped = le64_to_cpu(stats_recv->tx_discards);
+
+ port_stats->vport_stats = *stats_recv;
spin_unlock_bh(&np->stats_lock);
- return 0;
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+
+ return err;
}
/**
@@ -2868,14 +2463,16 @@ int idpf_send_get_set_rss_lut_msg(struct idpf_adapter *adapter,
struct idpf_rss_data *rss_data,
u32 vport_id, bool get)
{
- struct virtchnl2_rss_lut *recv_rl __free(kfree) = NULL;
- struct virtchnl2_rss_lut *rl __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = get ? VIRTCHNL2_OP_GET_RSS_LUT :
+ VIRTCHNL2_OP_SET_RSS_LUT,
+ };
+ struct virtchnl2_rss_lut *rl, *recv_rl;
int buf_size, lut_buf_size;
struct idpf_vport *vport;
- ssize_t reply_sz;
bool rxhash_ena;
- int i;
+ int i, err;
vport = idpf_vid_to_vport(adapter, vport_id);
if (!vport)
@@ -2889,37 +2486,31 @@ int idpf_send_get_set_rss_lut_msg(struct idpf_adapter *adapter,
return -ENOMEM;
rl->vport_id = cpu_to_le32(vport_id);
-
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = rl;
- xn_params.send_buf.iov_len = buf_size;
-
- if (get) {
- recv_rl = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!recv_rl)
- return -ENOMEM;
- xn_params.vc_op = VIRTCHNL2_OP_GET_RSS_LUT;
- xn_params.recv_buf.iov_base = recv_rl;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- } else {
+ if (!get) {
rl->lut_entries = cpu_to_le16(rss_data->rss_lut_size);
for (i = 0; i < rss_data->rss_lut_size; i++)
rl->lut[i] = rxhash_ena ?
cpu_to_le32(rss_data->rss_lut[i]) : 0;
-
- xn_params.vc_op = VIRTCHNL2_OP_SET_RSS_LUT;
}
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, rl, buf_size);
+ if (err)
+ return err;
+
if (!get)
- return 0;
- if (reply_sz < sizeof(struct virtchnl2_rss_lut))
- return -EIO;
+ goto free_rx_buf;
+ if (xn_params.recv_mem.iov_len < sizeof(*recv_rl)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
+
+ recv_rl = xn_params.recv_mem.iov_base;
lut_buf_size = le16_to_cpu(recv_rl->lut_entries) * sizeof(u32);
- if (reply_sz < lut_buf_size)
- return -EIO;
+ if (xn_params.recv_mem.iov_len < lut_buf_size + sizeof(*recv_rl)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
/* size didn't change, we can reuse existing lut buf */
if (rss_data->rss_lut_size == le16_to_cpu(recv_rl->lut_entries))
@@ -2931,13 +2522,16 @@ int idpf_send_get_set_rss_lut_msg(struct idpf_adapter *adapter,
rss_data->rss_lut = kzalloc(lut_buf_size, GFP_KERNEL);
if (!rss_data->rss_lut) {
rss_data->rss_lut_size = 0;
- return -ENOMEM;
+ err = -ENOMEM;
+ goto free_rx_buf;
}
do_memcpy:
- memcpy(rss_data->rss_lut, recv_rl->lut, rss_data->rss_lut_size);
+ memcpy(rss_data->rss_lut, recv_rl->lut, lut_buf_size);
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -2953,12 +2547,14 @@ int idpf_send_get_set_rss_key_msg(struct idpf_adapter *adapter,
struct idpf_rss_data *rss_data,
u32 vport_id, bool get)
{
- struct virtchnl2_rss_key *recv_rk __free(kfree) = NULL;
- struct virtchnl2_rss_key *rk __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
- ssize_t reply_sz;
- int i, buf_size;
- u16 key_size;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = get ? VIRTCHNL2_OP_GET_RSS_KEY :
+ VIRTCHNL2_OP_SET_RSS_KEY,
+ };
+ struct virtchnl2_rss_key *rk, *recv_rk;
+ u16 key_size, recv_len;
+ int i, buf_size, err;
buf_size = struct_size(rk, key_flex, rss_data->rss_key_size);
rk = kzalloc(buf_size, GFP_KERNEL);
@@ -2966,37 +2562,32 @@ int idpf_send_get_set_rss_key_msg(struct idpf_adapter *adapter,
return -ENOMEM;
rk->vport_id = cpu_to_le32(vport_id);
- xn_params.send_buf.iov_base = rk;
- xn_params.send_buf.iov_len = buf_size;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- if (get) {
- recv_rk = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!recv_rk)
- return -ENOMEM;
-
- xn_params.vc_op = VIRTCHNL2_OP_GET_RSS_KEY;
- xn_params.recv_buf.iov_base = recv_rk;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- } else {
+ if (!get) {
rk->key_len = cpu_to_le16(rss_data->rss_key_size);
for (i = 0; i < rss_data->rss_key_size; i++)
rk->key_flex[i] = rss_data->rss_key[i];
-
- xn_params.vc_op = VIRTCHNL2_OP_SET_RSS_KEY;
}
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, rk, buf_size);
+ if (err)
+ return err;
+
if (!get)
- return 0;
- if (reply_sz < sizeof(struct virtchnl2_rss_key))
- return -EIO;
+ goto free_rx_buf;
+ recv_len = xn_params.recv_mem.iov_len;
+ if (recv_len < sizeof(struct virtchnl2_rss_key)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
+
+ recv_rk = xn_params.recv_mem.iov_base;
key_size = min_t(u16, NETDEV_RSS_KEY_LEN,
le16_to_cpu(recv_rk->key_len));
- if (reply_sz < key_size)
- return -EIO;
+ if (recv_len < key_size) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
/* key len didn't change, reuse existing buf */
if (rss_data->rss_key_size == key_size)
@@ -3007,13 +2598,16 @@ int idpf_send_get_set_rss_key_msg(struct idpf_adapter *adapter,
rss_data->rss_key = kzalloc(key_size, GFP_KERNEL);
if (!rss_data->rss_key) {
rss_data->rss_key_size = 0;
- return -ENOMEM;
+ err = -ENOMEM;
+ goto free_rx_buf;
}
do_memcpy:
memcpy(rss_data->rss_key, recv_rk->key_flex, rss_data->rss_key_size);
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -3190,15 +2784,18 @@ static void idpf_parse_protocol_ids(struct virtchnl2_ptype *ptype,
*/
static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
{
- struct virtchnl2_get_ptype_info *get_ptype_info __free(kfree) = NULL;
- struct virtchnl2_get_ptype_info *ptype_info __free(kfree) = NULL;
struct libeth_rx_pt *singleq_pt_lkup __free(kfree) = NULL;
struct libeth_rx_pt *splitq_pt_lkup __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_GET_PTYPE_INFO,
+ };
+ struct virtchnl2_get_ptype_info *get_ptype_info;
+ struct virtchnl2_get_ptype_info *ptype_info;
+ int err = 0, max_ptype = IDPF_RX_MAX_PTYPE;
+ int buf_size = sizeof(*get_ptype_info);
int ptypes_recvd = 0, ptype_offset;
- u32 max_ptype = IDPF_RX_MAX_PTYPE;
u16 next_ptype_id = 0;
- ssize_t reply_sz;
singleq_pt_lkup = kzalloc_objs(*singleq_pt_lkup, IDPF_RX_MAX_BASE_PTYPE);
if (!singleq_pt_lkup)
@@ -3208,42 +2805,38 @@ static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
if (!splitq_pt_lkup)
return -ENOMEM;
- get_ptype_info = kzalloc_obj(*get_ptype_info);
- if (!get_ptype_info)
- return -ENOMEM;
-
- ptype_info = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!ptype_info)
- return -ENOMEM;
+ while (next_ptype_id < max_ptype) {
+ u16 num_ptypes;
- xn_params.vc_op = VIRTCHNL2_OP_GET_PTYPE_INFO;
- xn_params.send_buf.iov_base = get_ptype_info;
- xn_params.send_buf.iov_len = sizeof(*get_ptype_info);
- xn_params.recv_buf.iov_base = ptype_info;
- xn_params.recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
+ get_ptype_info = kzalloc(buf_size, GFP_KERNEL);
+ if (!get_ptype_info)
+ return -ENOMEM;
- while (next_ptype_id < max_ptype) {
get_ptype_info->start_ptype_id = cpu_to_le16(next_ptype_id);
if ((next_ptype_id + IDPF_RX_MAX_PTYPES_PER_BUF) > max_ptype)
- get_ptype_info->num_ptypes =
- cpu_to_le16(max_ptype - next_ptype_id);
+ num_ptypes = max_ptype - next_ptype_id;
else
- get_ptype_info->num_ptypes =
- cpu_to_le16(IDPF_RX_MAX_PTYPES_PER_BUF);
+ num_ptypes = IDPF_RX_MAX_PTYPES_PER_BUF;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ get_ptype_info->num_ptypes = cpu_to_le16(num_ptypes);
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params,
+ get_ptype_info, buf_size);
+ if (err)
+ return err;
+ ptype_info = xn_params.recv_mem.iov_base;
+ if (xn_params.recv_mem.iov_len < sizeof(*ptype_info)) {
+ err = -EIO;
+ goto free_rx_buf;
+ }
ptypes_recvd += le16_to_cpu(ptype_info->num_ptypes);
- if (ptypes_recvd > max_ptype)
- return -EINVAL;
-
- next_ptype_id = le16_to_cpu(get_ptype_info->start_ptype_id) +
- le16_to_cpu(get_ptype_info->num_ptypes);
+ if (ptypes_recvd > max_ptype) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
+ next_ptype_id = next_ptype_id + num_ptypes;
ptype_offset = IDPF_RX_PTYPE_HDR_SZ;
for (u16 i = 0; i < le16_to_cpu(ptype_info->num_ptypes); i++) {
@@ -3258,14 +2851,18 @@ static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
pt_8 = ptype->ptype_id_8;
ptype_offset += IDPF_GET_PTYPE_SIZE(ptype);
- if (ptype_offset > IDPF_CTLQ_MAX_BUF_LEN)
- return -EINVAL;
+ if (ptype_offset > LIBIE_CTLQ_MAX_BUF_LEN) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
/* 0xFFFF indicates end of ptypes */
if (pt_10 == IDPF_INVALID_PTYPE_ID)
goto out;
- if (pt_10 >= max_ptype)
- return -EINVAL;
+ if (pt_10 >= max_ptype) {
+ err = -EINVAL;
+ goto free_rx_buf;
+ }
idpf_parse_protocol_ids(ptype, &rx_pt);
idpf_finalize_ptype_lookup(&rx_pt);
@@ -3279,13 +2876,18 @@ static int idpf_send_get_rx_ptype_msg(struct idpf_adapter *adapter)
if (!singleq_pt_lkup[pt_8].outer_ip)
singleq_pt_lkup[pt_8] = rx_pt;
}
+
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ xn_params.recv_mem = (struct kvec) {};
}
out:
adapter->splitq_pt_lkup = no_free_ptr(splitq_pt_lkup);
adapter->singleq_pt_lkup = no_free_ptr(singleq_pt_lkup);
+free_rx_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -3313,40 +2915,24 @@ static void idpf_rel_rx_pt_lkup(struct idpf_adapter *adapter)
int idpf_send_ena_dis_loopback_msg(struct idpf_adapter *adapter, u32 vport_id,
bool loopback_ena)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_LOOPBACK,
+ };
struct virtchnl2_loopback loopback;
- ssize_t reply_sz;
+ int err;
loopback.vport_id = cpu_to_le32(vport_id);
loopback.enable = loopback_ena;
- xn_params.vc_op = VIRTCHNL2_OP_LOOPBACK;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &loopback;
- xn_params.send_buf.iov_len = sizeof(loopback);
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
-
- return reply_sz < 0 ? reply_sz : 0;
-}
-
-/**
- * idpf_find_ctlq - Given a type and id, find ctlq info
- * @hw: hardware struct
- * @type: type of ctrlq to find
- * @id: ctlq id to find
- *
- * Returns pointer to found ctlq info struct, NULL otherwise.
- */
-static struct idpf_ctlq_info *idpf_find_ctlq(struct idpf_hw *hw,
- enum idpf_ctlq_type type, int id)
-{
- struct idpf_ctlq_info *cq, *tmp;
+ err = idpf_send_mb_msg(adapter, &xn_params, &loopback,
+ sizeof(loopback));
+ if (err)
+ return err;
- list_for_each_entry_safe(cq, tmp, &hw->cq_list_head, cq_list)
- if (cq->q_id == id && cq->cq_type == type)
- return cq;
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return NULL;
+ return 0;
}
/**
@@ -3357,40 +2943,43 @@ static struct idpf_ctlq_info *idpf_find_ctlq(struct idpf_hw *hw,
*/
int idpf_init_dflt_mbx(struct idpf_adapter *adapter)
{
- struct idpf_ctlq_create_info ctlq_info[] = {
+ struct libie_ctlq_ctx *ctx = &adapter->ctlq_ctx;
+ struct libie_ctlq_create_info ctlq_info[] = {
{
- .type = IDPF_CTLQ_TYPE_MAILBOX_TX,
- .id = IDPF_DFLT_MBX_ID,
+ .type = LIBIE_CTLQ_TYPE_TX,
+ .id = LIBIE_CTLQ_MBX_ID,
.len = IDPF_DFLT_MBX_Q_LEN,
- .buf_size = IDPF_CTLQ_MAX_BUF_LEN
},
{
- .type = IDPF_CTLQ_TYPE_MAILBOX_RX,
- .id = IDPF_DFLT_MBX_ID,
+ .type = LIBIE_CTLQ_TYPE_RX,
+ .id = LIBIE_CTLQ_MBX_ID,
.len = IDPF_DFLT_MBX_Q_LEN,
- .buf_size = IDPF_CTLQ_MAX_BUF_LEN
}
};
- struct idpf_hw *hw = &adapter->hw;
+ struct libie_ctlq_xn_init_params params = {
+ .num_qs = IDPF_NUM_DFLT_MBX_Q,
+ .cctlq_info = ctlq_info,
+ .ctx = ctx,
+ };
int err;
- adapter->dev_ops.reg_ops.ctlq_reg_init(adapter, ctlq_info);
+ adapter->dev_ops.reg_ops.ctlq_reg_init(&ctx->mmio_info,
+ params.cctlq_info);
- err = idpf_ctlq_init(hw, IDPF_NUM_DFLT_MBX_Q, ctlq_info);
+ err = libie_ctlq_xn_init(¶ms);
if (err)
return err;
- hw->asq = idpf_find_ctlq(hw, IDPF_CTLQ_TYPE_MAILBOX_TX,
- IDPF_DFLT_MBX_ID);
- hw->arq = idpf_find_ctlq(hw, IDPF_CTLQ_TYPE_MAILBOX_RX,
- IDPF_DFLT_MBX_ID);
-
- if (!hw->asq || !hw->arq) {
- idpf_ctlq_deinit(hw);
-
+ adapter->asq = libie_find_ctlq(ctx, LIBIE_CTLQ_TYPE_TX,
+ LIBIE_CTLQ_MBX_ID);
+ adapter->arq = libie_find_ctlq(ctx, LIBIE_CTLQ_TYPE_RX,
+ LIBIE_CTLQ_MBX_ID);
+ if (!adapter->asq || !adapter->arq) {
+ libie_ctlq_xn_deinit(params.xnm, ctx);
return -ENOENT;
}
+ adapter->xnm = params.xnm;
adapter->state = __IDPF_VER_CHECK;
return 0;
@@ -3402,12 +2991,13 @@ int idpf_init_dflt_mbx(struct idpf_adapter *adapter)
*/
void idpf_deinit_dflt_mbx(struct idpf_adapter *adapter)
{
- if (adapter->hw.arq && adapter->hw.asq) {
- idpf_mb_clean(adapter, adapter->hw.asq);
- idpf_ctlq_deinit(&adapter->hw);
+ if (adapter->arq && adapter->asq) {
+ idpf_mb_clean(adapter, adapter->asq, true);
+ libie_ctlq_xn_deinit(adapter->xnm, &adapter->ctlq_ctx);
}
- adapter->hw.arq = NULL;
- adapter->hw.asq = NULL;
+
+ adapter->arq = NULL;
+ adapter->asq = NULL;
}
/**
@@ -3478,15 +3068,6 @@ int idpf_vc_core_init(struct idpf_adapter *adapter)
u16 num_max_vports;
int err = 0;
- if (!adapter->vcxn_mngr) {
- adapter->vcxn_mngr = kzalloc_obj(*adapter->vcxn_mngr);
- if (!adapter->vcxn_mngr) {
- err = -ENOMEM;
- goto init_failed;
- }
- }
- idpf_vc_xn_init(adapter->vcxn_mngr);
-
while (adapter->state != __IDPF_INIT_SW) {
switch (adapter->state) {
case __IDPF_VER_CHECK:
@@ -3633,8 +3214,7 @@ int idpf_vc_core_init(struct idpf_adapter *adapter)
* the mailbox again
*/
adapter->state = __IDPF_VER_CHECK;
- if (adapter->vcxn_mngr)
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ idpf_deinit_dflt_mbx(adapter);
set_bit(IDPF_HR_DRV_LOAD, adapter->flags);
queue_delayed_work(adapter->vc_event_wq, &adapter->vc_event_task,
msecs_to_jiffies(task_delay));
@@ -3657,7 +3237,7 @@ void idpf_vc_core_deinit(struct idpf_adapter *adapter)
/* Avoid transaction timeouts when called during reset */
remove_in_prog = test_bit(IDPF_REMOVE_IN_PROG, adapter->flags);
if (!remove_in_prog)
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ idpf_deinit_dflt_mbx(adapter);
idpf_ptp_release(adapter);
idpf_deinit_task(adapter);
@@ -3666,7 +3246,7 @@ void idpf_vc_core_deinit(struct idpf_adapter *adapter)
idpf_intr_rel(adapter);
if (remove_in_prog)
- idpf_vc_xn_shutdown(adapter->vcxn_mngr);
+ idpf_deinit_dflt_mbx(adapter);
cancel_delayed_work_sync(&adapter->serv_task);
cancel_delayed_work_sync(&adapter->mbx_task);
@@ -4203,9 +3783,9 @@ static void idpf_set_mac_type(const u8 *default_mac_addr,
/**
* idpf_mac_filter_async_handler - Async callback for mac filters
- * @adapter: private data struct
- * @xn: transaction for message
- * @ctlq_msg: received message
+ * @ctx: controlq context structure
+ * @buff: response buffer pointer and size
+ * @status: async call return value
*
* In some scenarios driver can't sleep and wait for a reply (e.g.: stack is
* holding rtnl_lock) when adding a new mac filter. It puts us in a difficult
@@ -4213,13 +3793,14 @@ static void idpf_set_mac_type(const u8 *default_mac_addr,
* ultimately do is remove it from our list of mac filters and report the
* error.
*/
-static int idpf_mac_filter_async_handler(struct idpf_adapter *adapter,
- struct idpf_vc_xn *xn,
- const struct idpf_ctlq_msg *ctlq_msg)
+static void idpf_mac_filter_async_handler(void *ctx,
+ struct kvec *buff,
+ int status)
{
struct virtchnl2_mac_addr_list *ma_list;
struct idpf_vport_config *vport_config;
struct virtchnl2_mac_addr *mac_addr;
+ struct idpf_adapter *adapter = ctx;
struct idpf_mac_filter *f, *tmp;
struct list_head *ma_list_head;
struct idpf_vport *vport;
@@ -4227,18 +3808,18 @@ static int idpf_mac_filter_async_handler(struct idpf_adapter *adapter,
int i;
/* if success we're done, we're only here if something bad happened */
- if (!ctlq_msg->cookie.mbx.chnl_retval)
- return 0;
+ if (!status || status == -ETIMEDOUT)
+ return;
+ ma_list = buff->iov_base;
/* make sure at least struct is there */
- if (xn->reply_sz < sizeof(*ma_list))
+ if (buff->iov_len < sizeof(*ma_list))
goto invalid_payload;
- ma_list = ctlq_msg->ctx.indirect.payload->va;
mac_addr = ma_list->mac_addr_list;
num_entries = le16_to_cpu(ma_list->num_mac_addr);
/* we should have received a buffer at least this big */
- if (xn->reply_sz < struct_size(ma_list, mac_addr_list, num_entries))
+ if (buff->iov_len < struct_size(ma_list, mac_addr_list, num_entries))
goto invalid_payload;
vport = idpf_vid_to_vport(adapter, le32_to_cpu(ma_list->vport_id));
@@ -4258,16 +3839,13 @@ static int idpf_mac_filter_async_handler(struct idpf_adapter *adapter,
if (ether_addr_equal(mac_addr[i].addr, f->macaddr))
list_del(&f->list);
spin_unlock_bh(&vport_config->mac_filter_list_lock);
- dev_err_ratelimited(&adapter->pdev->dev, "Received error sending MAC filter request (op %d)\n",
- xn->vc_op);
-
- return 0;
+ dev_err_ratelimited(&adapter->pdev->dev, "Received error %d on sending MAC filter request\n",
+ status);
+ return;
invalid_payload:
- dev_err_ratelimited(&adapter->pdev->dev, "Received invalid MAC filter payload (op %d) (len %zd)\n",
- xn->vc_op, xn->reply_sz);
-
- return -EINVAL;
+ dev_err_ratelimited(&adapter->pdev->dev, "Received invalid MAC filter payload (len %zd)\n",
+ buff->iov_len);
}
/**
@@ -4286,19 +3864,21 @@ int idpf_add_del_mac_filters(struct idpf_adapter *adapter,
const u8 *default_mac_addr, u32 vport_id,
bool add, bool async)
{
- struct virtchnl2_mac_addr_list *ma_list __free(kfree) = NULL;
struct virtchnl2_mac_addr *mac_addr __free(kfree) = NULL;
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = add ? VIRTCHNL2_OP_ADD_MAC_ADDR :
+ VIRTCHNL2_OP_DEL_MAC_ADDR,
+ };
+ struct virtchnl2_mac_addr_list *ma_list;
u32 num_msgs, total_filters = 0;
struct idpf_mac_filter *f;
- ssize_t reply_sz;
- int i = 0, k;
+ int i = 0;
- xn_params.vc_op = add ? VIRTCHNL2_OP_ADD_MAC_ADDR :
- VIRTCHNL2_OP_DEL_MAC_ADDR;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.async = async;
- xn_params.async_handler = idpf_mac_filter_async_handler;
+ if (async) {
+ xn_params.resp_cb = idpf_mac_filter_async_handler;
+ xn_params.send_ctx = adapter;
+ }
spin_lock_bh(&vport_config->mac_filter_list_lock);
@@ -4353,32 +3933,31 @@ int idpf_add_del_mac_filters(struct idpf_adapter *adapter,
*/
num_msgs = DIV_ROUND_UP(total_filters, IDPF_NUM_FILTERS_PER_MSG);
- for (i = 0, k = 0; i < num_msgs; i++) {
- u32 entries_size, buf_size, num_entries;
+ for (u32 i = 0, k = 0; i < num_msgs; i++) {
+ u32 entries_size, num_entries;
+ size_t buf_size;
+ int err;
num_entries = min_t(u32, total_filters,
IDPF_NUM_FILTERS_PER_MSG);
entries_size = sizeof(struct virtchnl2_mac_addr) * num_entries;
buf_size = struct_size(ma_list, mac_addr_list, num_entries);
- if (!ma_list || num_entries != IDPF_NUM_FILTERS_PER_MSG) {
- kfree(ma_list);
- ma_list = kzalloc(buf_size, GFP_ATOMIC);
- if (!ma_list)
- return -ENOMEM;
- } else {
- memset(ma_list, 0, buf_size);
- }
+ ma_list = kzalloc(buf_size, GFP_ATOMIC);
+ if (!ma_list)
+ return -ENOMEM;
ma_list->vport_id = cpu_to_le32(vport_id);
ma_list->num_mac_addr = cpu_to_le16(num_entries);
memcpy(ma_list->mac_addr_list, &mac_addr[k], entries_size);
- xn_params.send_buf.iov_base = ma_list;
- xn_params.send_buf.iov_len = buf_size;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
+ err = idpf_send_mb_msg_kfree(adapter, &xn_params, ma_list,
+ buf_size);
+ if (err)
+ return err;
+
+ if (!async)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
k += num_entries;
total_filters -= num_entries;
@@ -4387,6 +3966,26 @@ int idpf_add_del_mac_filters(struct idpf_adapter *adapter,
return 0;
}
+/**
+ * idpf_promiscuous_async_handler - async callback for promiscuous mode
+ * @ctx: controlq context structure
+ * @buff: response buffer pointer and size
+ * @status: async call return value
+ *
+ * Nobody is waiting for the promiscuous virtchnl message response. Print
+ * an error message if something went wrong and return.
+ */
+static void idpf_promiscuous_async_handler(void *ctx,
+ struct kvec *buff,
+ int status)
+{
+ struct idpf_adapter *adapter = ctx;
+
+ if (status)
+ dev_err_ratelimited(&adapter->pdev->dev, "Failed to set promiscuous mode: %d\n",
+ status);
+}
+
/**
* idpf_set_promiscuous - set promiscuous and send message to mailbox
* @adapter: Driver specific private structure
@@ -4401,9 +4000,13 @@ int idpf_set_promiscuous(struct idpf_adapter *adapter,
struct idpf_vport_user_config_data *config_data,
u32 vport_id)
{
- struct idpf_vc_xn_params xn_params = {};
+ struct libie_ctlq_xn_send_params xn_params = {
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ .chnl_opcode = VIRTCHNL2_OP_CONFIG_PROMISCUOUS_MODE,
+ .resp_cb = idpf_promiscuous_async_handler,
+ .send_ctx = adapter,
+ };
struct virtchnl2_promisc_info vpi;
- ssize_t reply_sz;
u16 flags = 0;
if (test_bit(__IDPF_PROMISC_UC, config_data->user_flags))
@@ -4414,15 +4017,7 @@ int idpf_set_promiscuous(struct idpf_adapter *adapter,
vpi.vport_id = cpu_to_le32(vport_id);
vpi.flags = cpu_to_le16(flags);
- xn_params.vc_op = VIRTCHNL2_OP_CONFIG_PROMISCUOUS_MODE;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = &vpi;
- xn_params.send_buf.iov_len = sizeof(vpi);
- /* setting promiscuous is only ever done asynchronously */
- xn_params.async = true;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
-
- return reply_sz < 0 ? reply_sz : 0;
+ return idpf_send_mb_msg(adapter, &xn_params, &vpi, sizeof(vpi));
}
/**
@@ -4440,26 +4035,39 @@ int idpf_idc_rdma_vc_send_sync(struct iidc_rdma_core_dev_info *cdev_info,
u8 *recv_msg, u16 *recv_len)
{
struct idpf_adapter *adapter = pci_get_drvdata(cdev_info->pdev);
- struct idpf_vc_xn_params xn_params = { };
- ssize_t reply_sz;
- u16 recv_size;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_RDMA,
+ .timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
+ };
+ u8 on_stack_buf[LIBIE_CP_TX_COPYBREAK];
+ void *send_buf;
+ int err;
- if (!recv_msg || !recv_len || msg_size > IDPF_CTLQ_MAX_BUF_LEN)
+ if (!recv_msg || !recv_len || msg_size > LIBIE_CTLQ_MAX_BUF_LEN)
return -EINVAL;
- recv_size = min_t(u16, *recv_len, IDPF_CTLQ_MAX_BUF_LEN);
- *recv_len = 0;
- xn_params.vc_op = VIRTCHNL2_OP_RDMA;
- xn_params.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC;
- xn_params.send_buf.iov_base = send_msg;
- xn_params.send_buf.iov_len = msg_size;
- xn_params.recv_buf.iov_base = recv_msg;
- xn_params.recv_buf.iov_len = recv_size;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- *recv_len = reply_sz;
+ if (!libie_cp_can_send_onstack(msg_size)) {
+ send_buf = kzalloc(msg_size, GFP_KERNEL);
+ if (!send_buf)
+ return -ENOMEM;
+ } else {
+ send_buf = on_stack_buf;
+ }
- return 0;
+ memcpy(send_buf, send_msg, msg_size);
+ err = idpf_send_mb_msg(adapter, &xn_params, send_buf, msg_size);
+ if (err)
+ return err;
+
+ if (xn_params.recv_mem.iov_len > *recv_len) {
+ err = -EINVAL;
+ goto rel_buf;
+ }
+
+ *recv_len = xn_params.recv_mem.iov_len;
+ memcpy(recv_msg, xn_params.recv_mem.iov_base, *recv_len);
+rel_buf:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
EXPORT_SYMBOL_GPL(idpf_idc_rdma_vc_send_sync);
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h
index 6992b768cef4..86a44b6e1488 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl.h
@@ -7,86 +7,6 @@
#include <linux/intel/virtchnl2.h>
#define IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC (60 * 1000)
-#define IDPF_VC_XN_IDX_M GENMASK(7, 0)
-#define IDPF_VC_XN_SALT_M GENMASK(15, 8)
-#define IDPF_VC_XN_RING_LEN U8_MAX
-
-/**
- * enum idpf_vc_xn_state - Virtchnl transaction status
- * @IDPF_VC_XN_IDLE: not expecting a reply, ready to be used
- * @IDPF_VC_XN_WAITING: expecting a reply, not yet received
- * @IDPF_VC_XN_COMPLETED_SUCCESS: a reply was expected and received, buffer
- * updated
- * @IDPF_VC_XN_COMPLETED_FAILED: a reply was expected and received, but there
- * was an error, buffer not updated
- * @IDPF_VC_XN_SHUTDOWN: transaction object cannot be used, VC torn down
- * @IDPF_VC_XN_ASYNC: transaction sent asynchronously and doesn't have the
- * return context; a callback may be provided to handle
- * return
- */
-enum idpf_vc_xn_state {
- IDPF_VC_XN_IDLE = 1,
- IDPF_VC_XN_WAITING,
- IDPF_VC_XN_COMPLETED_SUCCESS,
- IDPF_VC_XN_COMPLETED_FAILED,
- IDPF_VC_XN_SHUTDOWN,
- IDPF_VC_XN_ASYNC,
-};
-
-struct idpf_vc_xn;
-/* Callback for asynchronous messages */
-typedef int (*async_vc_cb) (struct idpf_adapter *, struct idpf_vc_xn *,
- const struct idpf_ctlq_msg *);
-
-/**
- * struct idpf_vc_xn - Data structure representing virtchnl transactions
- * @completed: virtchnl event loop uses that to signal when a reply is
- * available, uses kernel completion API
- * @lock: protects the transaction state fields below
- * @state: virtchnl event loop stores the data below, protected by @lock
- * @reply_sz: Original size of reply, may be > reply_buf.iov_len; it will be
- * truncated on its way to the receiver thread according to
- * reply_buf.iov_len.
- * @reply: Reference to the buffer(s) where the reply data should be written
- * to. May be 0-length (then NULL address permitted) if the reply data
- * should be ignored.
- * @async_handler: if sent asynchronously, a callback can be provided to handle
- * the reply when it's received
- * @vc_op: corresponding opcode sent with this transaction
- * @idx: index used as retrieval on reply receive, used for cookie
- * @salt: changed every message to make unique, used for cookie
- */
-struct idpf_vc_xn {
- struct completion completed;
- spinlock_t lock;
- enum idpf_vc_xn_state state;
- size_t reply_sz;
- struct kvec reply;
- async_vc_cb async_handler;
- u32 vc_op;
- u8 idx;
- u8 salt;
-};
-
-/**
- * struct idpf_vc_xn_params - Parameters for executing transaction
- * @send_buf: kvec for send buffer
- * @recv_buf: kvec for recv buffer, may be NULL, must then have zero length
- * @timeout_ms: timeout to wait for reply
- * @async: send message asynchronously, will not wait on completion
- * @async_handler: If sent asynchronously, optional callback handler. The user
- * must be careful when using async handlers as the memory for
- * the recv_buf _cannot_ be on stack if this is async.
- * @vc_op: virtchnl op to send
- */
-struct idpf_vc_xn_params {
- struct kvec send_buf;
- struct kvec recv_buf;
- int timeout_ms;
- bool async;
- async_vc_cb async_handler;
- u32 vc_op;
-};
struct idpf_adapter;
struct idpf_netdev_priv;
@@ -96,8 +16,6 @@ struct idpf_vport_max_q;
struct idpf_vport_config;
struct idpf_vport_user_config_data;
-ssize_t idpf_vc_xn_exec(struct idpf_adapter *adapter,
- const struct idpf_vc_xn_params *params);
int idpf_init_dflt_mbx(struct idpf_adapter *adapter);
void idpf_deinit_dflt_mbx(struct idpf_adapter *adapter);
int idpf_vc_core_init(struct idpf_adapter *adapter);
@@ -124,9 +42,14 @@ bool idpf_sideband_action_ena(struct idpf_vport *vport,
struct ethtool_rx_flow_spec *fsp);
unsigned int idpf_fsteer_max_rules(struct idpf_vport *vport);
-int idpf_recv_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *arq);
-int idpf_send_mb_msg(struct idpf_adapter *adapter, struct idpf_ctlq_info *asq,
- u32 op, u16 msg_size, u8 *msg, u16 cookie);
+void idpf_recv_event_msg(struct libie_ctlq_ctx *ctx,
+ struct libie_ctlq_msg *ctlq_msg);
+int idpf_send_mb_msg(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size);
+int idpf_send_mb_msg_kfree(struct idpf_adapter *adapter,
+ struct libie_ctlq_xn_send_params *xn_params,
+ void *send_buf, size_t send_buf_size);
struct idpf_queue_ptr {
enum virtchnl2_queue_type type;
@@ -214,7 +137,6 @@ int idpf_send_get_set_rss_key_msg(struct idpf_adapter *adapter,
int idpf_send_get_set_rss_lut_msg(struct idpf_adapter *adapter,
struct idpf_rss_data *rss_data,
u32 vport_id, bool get);
-void idpf_vc_xn_shutdown(struct idpf_vc_xn_manager *vcxn_mngr);
int idpf_idc_rdma_vc_send_sync(struct iidc_rdma_core_dev_info *cdev_info,
u8 *send_msg, u16 msg_size,
u8 *recv_msg, u16 *recv_len);
diff --git a/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c b/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
index 8d8fb498e092..6d44021c222b 100644
--- a/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
+++ b/drivers/net/ethernet/intel/idpf/idpf_virtchnl_ptp.c
@@ -15,7 +15,6 @@
*/
int idpf_ptp_get_caps(struct idpf_adapter *adapter)
{
- struct virtchnl2_ptp_get_caps *recv_ptp_caps_msg __free(kfree) = NULL;
struct virtchnl2_ptp_get_caps send_ptp_caps_msg = {
.caps = cpu_to_le32(VIRTCHNL2_CAP_PTP_GET_DEVICE_CLK_TIME |
VIRTCHNL2_CAP_PTP_GET_DEVICE_CLK_TIME_MB |
@@ -24,34 +23,34 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
VIRTCHNL2_CAP_PTP_ADJ_DEVICE_CLK_MB |
VIRTCHNL2_CAP_PTP_TX_TSTAMPS_MB)
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_CAPS,
- .send_buf.iov_base = &send_ptp_caps_msg,
- .send_buf.iov_len = sizeof(send_ptp_caps_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_CAPS,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
struct virtchnl2_ptp_cross_time_reg_offsets cross_tstamp_offsets;
struct libie_mmio_info *mmio = &adapter->ctlq_ctx.mmio_info;
struct virtchnl2_ptp_clk_adj_reg_offsets clk_adj_offsets;
struct virtchnl2_ptp_clk_reg_offsets clock_offsets;
+ struct virtchnl2_ptp_get_caps *recv_ptp_caps_msg;
struct idpf_ptp_secondary_mbx *scnd_mbx;
struct idpf_ptp *ptp = adapter->ptp;
enum idpf_ptp_access access_type;
u32 temp_offset;
- int reply_sz;
+ size_t reply_sz;
+ int err;
- recv_ptp_caps_msg = kzalloc_obj(struct virtchnl2_ptp_get_caps);
- if (!recv_ptp_caps_msg)
- return -ENOMEM;
+ err = idpf_send_mb_msg(adapter, &xn_params, &send_ptp_caps_msg,
+ sizeof(send_ptp_caps_msg));
+ if (err)
+ return err;
- xn_params.recv_buf.iov_base = recv_ptp_caps_msg;
- xn_params.recv_buf.iov_len = sizeof(*recv_ptp_caps_msg);
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz != sizeof(*recv_ptp_caps_msg)) {
+ err = -EIO;
+ goto free_resp;
+ }
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- else if (reply_sz != sizeof(*recv_ptp_caps_msg))
- return -EIO;
+ recv_ptp_caps_msg = xn_params.recv_mem.iov_base;
ptp->caps = le32_to_cpu(recv_ptp_caps_msg->caps);
ptp->base_incval = le64_to_cpu(recv_ptp_caps_msg->base_incval);
@@ -112,7 +111,7 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
discipline_clock:
access_type = ptp->adj_dev_clk_time_access;
if (access_type != IDPF_PTP_DIRECT)
- return 0;
+ goto free_resp;
clk_adj_offsets = recv_ptp_caps_msg->clk_adj_offsets;
@@ -145,7 +144,9 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
ptp->dev_clk_regs.phy_shadj_h =
libie_pci_get_mmio_addr(mmio, temp_offset);
- return 0;
+free_resp:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -160,28 +161,34 @@ int idpf_ptp_get_caps(struct idpf_adapter *adapter)
int idpf_ptp_get_dev_clk_time(struct idpf_adapter *adapter,
struct idpf_ptp_dev_timers *dev_clk_time)
{
+ struct virtchnl2_ptp_get_dev_clk_time *get_dev_clk_time_resp;
struct virtchnl2_ptp_get_dev_clk_time get_dev_clk_time_msg;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_DEV_CLK_TIME,
- .send_buf.iov_base = &get_dev_clk_time_msg,
- .send_buf.iov_len = sizeof(get_dev_clk_time_msg),
- .recv_buf.iov_base = &get_dev_clk_time_msg,
- .recv_buf.iov_len = sizeof(get_dev_clk_time_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_DEV_CLK_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ size_t reply_sz;
u64 dev_time;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(get_dev_clk_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &get_dev_clk_time_msg,
+ sizeof(get_dev_clk_time_msg));
+ if (err)
+ return err;
- dev_time = le64_to_cpu(get_dev_clk_time_msg.dev_time_ns);
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz != sizeof(*get_dev_clk_time_resp)) {
+ err = -EIO;
+ goto free_resp;
+ }
+
+ get_dev_clk_time_resp = xn_params.recv_mem.iov_base;
+ dev_time = le64_to_cpu(get_dev_clk_time_resp->dev_time_ns);
dev_clk_time->dev_clk_time_ns = dev_time;
- return 0;
+free_resp:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -197,27 +204,30 @@ int idpf_ptp_get_dev_clk_time(struct idpf_adapter *adapter,
int idpf_ptp_get_cross_time(struct idpf_adapter *adapter,
struct idpf_ptp_dev_timers *cross_time)
{
- struct virtchnl2_ptp_get_cross_time cross_time_msg;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_CROSS_TIME,
- .send_buf.iov_base = &cross_time_msg,
- .send_buf.iov_len = sizeof(cross_time_msg),
- .recv_buf.iov_base = &cross_time_msg,
- .recv_buf.iov_len = sizeof(cross_time_msg),
+ struct virtchnl2_ptp_get_cross_time cross_time_send, *cross_time_recv;
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_CROSS_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err = 0;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(cross_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &cross_time_send,
+ sizeof(cross_time_send));
+ if (err)
+ return err;
- cross_time->dev_clk_time_ns = le64_to_cpu(cross_time_msg.dev_time_ns);
- cross_time->sys_time_ns = le64_to_cpu(cross_time_msg.sys_time_ns);
+ if (xn_params.recv_mem.iov_len != sizeof(*cross_time_recv)) {
+ err = -EIO;
+ goto free_resp;
+ }
- return 0;
+ cross_time_recv = xn_params.recv_mem.iov_base;
+ cross_time->dev_clk_time_ns = le64_to_cpu(cross_time_recv->dev_time_ns);
+ cross_time->sys_time_ns = le64_to_cpu(cross_time_recv->sys_time_ns);
+
+free_resp:
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
+ return err;
}
/**
@@ -234,23 +244,18 @@ int idpf_ptp_set_dev_clk_time(struct idpf_adapter *adapter, u64 time)
struct virtchnl2_ptp_set_dev_clk_time set_dev_clk_time_msg = {
.dev_time_ns = cpu_to_le64(time),
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_SET_DEV_CLK_TIME,
- .send_buf.iov_base = &set_dev_clk_time_msg,
- .send_buf.iov_len = sizeof(set_dev_clk_time_msg),
- .recv_buf.iov_base = &set_dev_clk_time_msg,
- .recv_buf.iov_len = sizeof(set_dev_clk_time_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_SET_DEV_CLK_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(set_dev_clk_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &set_dev_clk_time_msg,
+ sizeof(set_dev_clk_time_msg));
+ if (!err)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -267,23 +272,18 @@ int idpf_ptp_adj_dev_clk_time(struct idpf_adapter *adapter, s64 delta)
struct virtchnl2_ptp_adj_dev_clk_time adj_dev_clk_time_msg = {
.delta = cpu_to_le64(delta),
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_TIME,
- .send_buf.iov_base = &adj_dev_clk_time_msg,
- .send_buf.iov_len = sizeof(adj_dev_clk_time_msg),
- .recv_buf.iov_base = &adj_dev_clk_time_msg,
- .recv_buf.iov_len = sizeof(adj_dev_clk_time_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_TIME,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(adj_dev_clk_time_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &adj_dev_clk_time_msg,
+ sizeof(adj_dev_clk_time_msg));
+ if (!err)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -301,23 +301,18 @@ int idpf_ptp_adj_dev_clk_fine(struct idpf_adapter *adapter, u64 incval)
struct virtchnl2_ptp_adj_dev_clk_fine adj_dev_clk_fine_msg = {
.incval = cpu_to_le64(incval),
};
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_FINE,
- .send_buf.iov_base = &adj_dev_clk_fine_msg,
- .send_buf.iov_len = sizeof(adj_dev_clk_fine_msg),
- .recv_buf.iov_base = &adj_dev_clk_fine_msg,
- .recv_buf.iov_len = sizeof(adj_dev_clk_fine_msg),
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_ADJ_DEV_CLK_FINE,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
- int reply_sz;
+ int err;
- reply_sz = idpf_vc_xn_exec(adapter, &xn_params);
- if (reply_sz < 0)
- return reply_sz;
- if (reply_sz != sizeof(adj_dev_clk_fine_msg))
- return -EIO;
+ err = idpf_send_mb_msg(adapter, &xn_params, &adj_dev_clk_fine_msg,
+ sizeof(adj_dev_clk_fine_msg));
+ if (!err)
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
- return 0;
+ return err;
}
/**
@@ -336,18 +331,16 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
struct virtchnl2_ptp_tx_tstamp_latch_caps tx_tstamp_latch_caps;
struct idpf_ptp_vport_tx_tstamp_caps *tstamp_caps;
struct idpf_ptp_tx_tstamp *ptp_tx_tstamp, *tmp;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP_CAPS,
- .send_buf.iov_base = &send_tx_tstamp_caps,
- .send_buf.iov_len = sizeof(send_tx_tstamp_caps),
- .recv_buf.iov_len = IDPF_CTLQ_MAX_BUF_LEN,
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP_CAPS,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
};
enum idpf_ptp_access tstamp_access, get_dev_clk_access;
struct idpf_ptp *ptp = vport->adapter->ptp;
struct list_head *head;
- int err = 0, reply_sz;
+ size_t reply_sz;
u16 num_latches;
+ int err = 0;
u32 size;
if (!ptp)
@@ -359,19 +352,19 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
get_dev_clk_access == IDPF_PTP_NONE)
return -EOPNOTSUPP;
- rcv_tx_tstamp_caps = kzalloc(IDPF_CTLQ_MAX_BUF_LEN, GFP_KERNEL);
- if (!rcv_tx_tstamp_caps)
- return -ENOMEM;
-
send_tx_tstamp_caps.vport_id = cpu_to_le32(vport->vport_id);
- xn_params.recv_buf.iov_base = rcv_tx_tstamp_caps;
- reply_sz = idpf_vc_xn_exec(vport->adapter, &xn_params);
- if (reply_sz < 0) {
- err = reply_sz;
+ err = idpf_send_mb_msg(vport->adapter, &xn_params, &send_tx_tstamp_caps,
+ sizeof(send_tx_tstamp_caps));
+ if (err)
+ return err;
+
+ rcv_tx_tstamp_caps = xn_params.recv_mem.iov_base;
+ reply_sz = xn_params.recv_mem.iov_len;
+ if (reply_sz < sizeof(*rcv_tx_tstamp_caps)) {
+ err = -EIO;
goto get_tstamp_caps_out;
}
-
num_latches = le16_to_cpu(rcv_tx_tstamp_caps->num_latches);
size = struct_size(rcv_tx_tstamp_caps, tstamp_latches, num_latches);
if (reply_sz != size) {
@@ -426,7 +419,7 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
}
vport->tx_tstamp_caps = tstamp_caps;
- kfree(rcv_tx_tstamp_caps);
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
return 0;
@@ -439,7 +432,7 @@ int idpf_ptp_get_vport_tstamps_caps(struct idpf_vport *vport)
kfree(tstamp_caps);
get_tstamp_caps_out:
- kfree(rcv_tx_tstamp_caps);
+ libie_ctlq_release_rx_buf(&xn_params.recv_mem);
return err;
}
@@ -536,9 +529,9 @@ idpf_ptp_get_tstamp_value(struct idpf_vport *vport,
/**
* idpf_ptp_get_tx_tstamp_async_handler - Async callback for getting Tx tstamps
- * @adapter: Driver specific private structure
- * @xn: transaction for message
- * @ctlq_msg: received message
+ * @ctx: adapter pointer
+ * @mem: address and size of the response
+ * @status: return value of the request
*
* Read the tstamps Tx tstamp values from a received message and put them
* directly to the skb. The number of timestamps to read is specified by
@@ -546,22 +539,23 @@ idpf_ptp_get_tstamp_value(struct idpf_vport *vport,
*
* Return: 0 on success, -errno otherwise.
*/
-static int
-idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
- struct idpf_vc_xn *xn,
- const struct idpf_ctlq_msg *ctlq_msg)
+static void
+idpf_ptp_get_tx_tstamp_async_handler(void *ctx, struct kvec *mem, int status)
{
struct virtchnl2_ptp_get_vport_tx_tstamp_latches *recv_tx_tstamp_msg;
struct idpf_ptp_vport_tx_tstamp_caps *tx_tstamp_caps;
struct virtchnl2_ptp_tx_tstamp_latch tstamp_latch;
struct idpf_ptp_tx_tstamp *tx_tstamp, *tmp;
struct idpf_vport *tstamp_vport = NULL;
+ struct idpf_adapter *adapter = ctx;
struct list_head *head;
u16 num_latches;
u32 vport_id;
- int err = 0;
- recv_tx_tstamp_msg = ctlq_msg->ctx.indirect.payload->va;
+ if (status)
+ return;
+
+ recv_tx_tstamp_msg = mem->iov_base;
vport_id = le32_to_cpu(recv_tx_tstamp_msg->vport_id);
idpf_for_each_vport(adapter, vport) {
@@ -575,7 +569,7 @@ idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
}
if (!tstamp_vport || !tstamp_vport->tx_tstamp_caps)
- return -EINVAL;
+ return;
tx_tstamp_caps = tstamp_vport->tx_tstamp_caps;
num_latches = le16_to_cpu(recv_tx_tstamp_msg->num_latches);
@@ -589,13 +583,13 @@ idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
if (!tstamp_latch.valid)
continue;
- if (list_empty(head)) {
- err = -ENOBUFS;
+ if (list_empty(head))
goto unlock;
- }
list_for_each_entry_safe(tx_tstamp, tmp, head, list_member) {
if (tstamp_latch.index == tx_tstamp->idx) {
+ int err;
+
list_del(&tx_tstamp->list_member);
err = idpf_ptp_get_tstamp_value(tstamp_vport,
&tstamp_latch,
@@ -610,8 +604,6 @@ idpf_ptp_get_tx_tstamp_async_handler(struct idpf_adapter *adapter,
unlock:
spin_unlock_bh(&tx_tstamp_caps->latches_lock);
-
- return err;
}
/**
@@ -627,15 +619,15 @@ int idpf_ptp_get_tx_tstamp(struct idpf_vport *vport)
{
struct virtchnl2_ptp_get_vport_tx_tstamp_latches *send_tx_tstamp_msg;
struct idpf_ptp_vport_tx_tstamp_caps *tx_tstamp_caps;
- struct idpf_vc_xn_params xn_params = {
- .vc_op = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP,
+ struct libie_ctlq_xn_send_params xn_params = {
+ .chnl_opcode = VIRTCHNL2_OP_PTP_GET_VPORT_TX_TSTAMP,
.timeout_ms = IDPF_VC_XN_DEFAULT_TIMEOUT_MSEC,
- .async = true,
- .async_handler = idpf_ptp_get_tx_tstamp_async_handler,
+ .resp_cb = idpf_ptp_get_tx_tstamp_async_handler,
+ .send_ctx = vport->adapter,
};
struct idpf_ptp_tx_tstamp *ptp_tx_tstamp;
- int reply_sz, size, msg_size;
struct list_head *head;
+ int size, msg_size;
bool state_upd;
u16 id = 0;
@@ -668,11 +660,7 @@ int idpf_ptp_get_tx_tstamp(struct idpf_vport *vport)
msg_size = struct_size(send_tx_tstamp_msg, tstamp_latches, id);
send_tx_tstamp_msg->vport_id = cpu_to_le32(vport->vport_id);
send_tx_tstamp_msg->num_latches = cpu_to_le16(id);
- xn_params.send_buf.iov_base = send_tx_tstamp_msg;
- xn_params.send_buf.iov_len = msg_size;
-
- reply_sz = idpf_vc_xn_exec(vport->adapter, &xn_params);
- kfree(send_tx_tstamp_msg);
- return min(reply_sz, 0);
+ return idpf_send_mb_msg_kfree(vport->adapter, &xn_params,
+ send_tx_tstamp_msg, msg_size);
}
--
2.47.0
^ permalink raw reply related
* [PATCH iwl-next v7 14/14] ixd: add devlink support
From: Larysa Zaremba @ 2026-04-28 8:26 UTC (permalink / raw)
To: intel-wired-lan, Tony Nguyen
Cc: aleksander.lobakin, sridhar.samudrala, Michal Swiatkowski,
Larysa Zaremba, Fijalkowski, Maciej, Emil Tantilov, Madhu Chittim,
Josh Hay, Keller, Jacob E, jayaprakash.shanmugam, Jiri Pirko,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
Simon Horman, Jonathan Corbet, Richard Cochran, Przemek Kitszel,
Andrew Lunn, netdev, linux-doc, linux-kernel, Bharath R
In-Reply-To: <20260428082654.44364-1-larysa.zaremba@intel.com>
From: Amritha Nambiar <amritha.nambiar@intel.com>
Enable initial support for the devlink interface with the ixd driver. The
ixd hardware is a single function PCIe device. So, the PCIe adapter gets
its own devlink instance to manage device-wide resources or configuration.
$ devlink dev show
pci/0000:83:00.6
$ devlink dev info pci/0000:83:00.6
pci/0000:83:00.6:
driver ixd
serial_number 00-a0-c9-ff-ff-23-45-67
versions:
fixed:
device.type MEV
running:
virtchnl 2.0
Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
Tested-by: Bharath R <Bharath.r@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
Documentation/networking/devlink/index.rst | 1 +
Documentation/networking/devlink/ixd.rst | 30 ++++++
drivers/net/ethernet/intel/ixd/Kconfig | 1 +
drivers/net/ethernet/intel/ixd/Makefile | 1 +
drivers/net/ethernet/intel/ixd/ixd_devlink.c | 97 ++++++++++++++++++++
drivers/net/ethernet/intel/ixd/ixd_devlink.h | 44 +++++++++
drivers/net/ethernet/intel/ixd/ixd_main.c | 16 +++-
7 files changed, 187 insertions(+), 3 deletions(-)
create mode 100644 Documentation/networking/devlink/ixd.rst
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_devlink.c
create mode 100644 drivers/net/ethernet/intel/ixd/ixd_devlink.h
diff --git a/Documentation/networking/devlink/index.rst b/Documentation/networking/devlink/index.rst
index f7ba7dcf477d..f0c077843fa7 100644
--- a/Documentation/networking/devlink/index.rst
+++ b/Documentation/networking/devlink/index.rst
@@ -88,6 +88,7 @@ parameters, info versions, and other features it supports.
ionic
iosm
ixgbe
+ ixd
kvaser_pciefd
kvaser_usb
mlx4
diff --git a/Documentation/networking/devlink/ixd.rst b/Documentation/networking/devlink/ixd.rst
new file mode 100644
index 000000000000..17b63c8425aa
--- /dev/null
+++ b/Documentation/networking/devlink/ixd.rst
@@ -0,0 +1,30 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+===================
+ixd devlink support
+===================
+
+This document describes the devlink features implemented by the ``ixd``
+device driver.
+
+Info versions
+=============
+
+The ``ixd`` driver reports the following versions
+
+.. list-table:: devlink info versions implemented
+ :widths: 5 5 5 90
+
+ * - Name
+ - Type
+ - Example
+ - Description
+ * - ``device.type``
+ - fixed
+ - MEV
+ - The hardware type for this device
+ * - ``virtchnl``
+ - running
+ - 2.0
+ - 2-digit version number (major.minor) of the communication channel
+ (virtchnl) used by the device.
diff --git a/drivers/net/ethernet/intel/ixd/Kconfig b/drivers/net/ethernet/intel/ixd/Kconfig
index 24510c50070e..34181c59dcdc 100644
--- a/drivers/net/ethernet/intel/ixd/Kconfig
+++ b/drivers/net/ethernet/intel/ixd/Kconfig
@@ -7,6 +7,7 @@ config IXD
select LIBETH
select LIBIE_CP
select LIBIE_PCI
+ select NET_DEVLINK
help
This driver supports Intel(R) Control Plane PCI Function
of Intel E2100 and later IPUs and FNICs.
diff --git a/drivers/net/ethernet/intel/ixd/Makefile b/drivers/net/ethernet/intel/ixd/Makefile
index 90abf231fb16..03760a2580b9 100644
--- a/drivers/net/ethernet/intel/ixd/Makefile
+++ b/drivers/net/ethernet/intel/ixd/Makefile
@@ -8,5 +8,6 @@ obj-$(CONFIG_IXD) += ixd.o
ixd-y := ixd_main.o
ixd-y += ixd_ctlq.o
ixd-y += ixd_dev.o
+ixd-y += ixd_devlink.o
ixd-y += ixd_lib.o
ixd-y += ixd_virtchnl.o
diff --git a/drivers/net/ethernet/intel/ixd/ixd_devlink.c b/drivers/net/ethernet/intel/ixd/ixd_devlink.c
new file mode 100644
index 000000000000..23ab11226978
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_devlink.c
@@ -0,0 +1,97 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2025, Intel Corporation. */
+
+#include "ixd.h"
+#include "ixd_devlink.h"
+
+#define IXD_DEVLINK_INFO_LEN 128
+
+/**
+ * ixd_fill_dsn - Get the serial number for the ixd device
+ * @adapter: adapter to query
+ * @buf: storage buffer for the info request
+ */
+static void ixd_fill_dsn(struct ixd_adapter *adapter, char *buf)
+{
+ u8 dsn[8];
+
+ /* Copy the DSN into an array in Big Endian format */
+ put_unaligned_be64(pci_get_dsn(adapter->cp_ctx.mmio_info.pdev), dsn);
+
+ snprintf(buf, IXD_DEVLINK_INFO_LEN, "%8phD", dsn);
+}
+
+/**
+ * ixd_fill_device_name - Get the name of the underlying hardware
+ * @adapter: adapter to query
+ * @buf: storage buffer for the info request
+ * @buf_size: size of the storage buffer
+ */
+static void ixd_fill_device_name(struct ixd_adapter *adapter, char *buf,
+ size_t buf_size)
+{
+ if (adapter->caps.device_type == cpu_to_le32(VIRTCHNL2_MEV_DEVICE))
+ snprintf(buf, buf_size, "%s", "MEV");
+ else
+ snprintf(buf, buf_size, "%s", "UNKNOWN");
+}
+
+/**
+ * ixd_devlink_info_get - .info_get devlink handler
+ * @devlink: devlink instance structure
+ * @req: the devlink info request
+ * @extack: extended netdev ack structure
+ *
+ * Callback for the devlink .info_get operation. Reports information about the
+ * device.
+ *
+ * Return: zero on success or an error code on failure.
+ */
+static int ixd_devlink_info_get(struct devlink *devlink,
+ struct devlink_info_req *req,
+ struct netlink_ext_ack *extack)
+{
+ struct ixd_adapter *adapter = devlink_priv(devlink);
+ char buf[IXD_DEVLINK_INFO_LEN];
+ int err;
+
+ ixd_fill_dsn(adapter, buf);
+ err = devlink_info_serial_number_put(req, buf);
+ if (err)
+ return err;
+
+ ixd_fill_device_name(adapter, buf, IXD_DEVLINK_INFO_LEN);
+ err = devlink_info_version_fixed_put(req, "device.type", buf);
+ if (err)
+ return err;
+
+ snprintf(buf, sizeof(buf), "%u.%u",
+ adapter->vc_ver.major, adapter->vc_ver.minor);
+
+ return devlink_info_version_running_put(req, "virtchnl", buf);
+}
+
+static const struct devlink_ops ixd_devlink_ops = {
+ .info_get = ixd_devlink_info_get,
+};
+
+/**
+ * ixd_adapter_alloc - Allocate devlink and return adapter pointer
+ * @dev: the device to allocate for
+ *
+ * Allocate a devlink instance for this device and return the private area as
+ * the adapter structure.
+ *
+ * Return: adapter structure on success, NULL on failure
+ */
+struct ixd_adapter *ixd_adapter_alloc(struct device *dev)
+{
+ struct devlink *devlink;
+
+ devlink = devlink_alloc(&ixd_devlink_ops, sizeof(struct ixd_adapter),
+ dev);
+ if (!devlink)
+ return NULL;
+
+ return devlink_priv(devlink);
+}
diff --git a/drivers/net/ethernet/intel/ixd/ixd_devlink.h b/drivers/net/ethernet/intel/ixd/ixd_devlink.h
new file mode 100644
index 000000000000..c43ce0655de2
--- /dev/null
+++ b/drivers/net/ethernet/intel/ixd/ixd_devlink.h
@@ -0,0 +1,44 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (c) 2025, Intel Corporation. */
+
+#ifndef _IXD_DEVLINK_H_
+#define _IXD_DEVLINK_H_
+#include <net/devlink.h>
+
+struct ixd_adapter *ixd_adapter_alloc(struct device *dev);
+
+/**
+ * ixd_devlink_free - teardown the devlink
+ * @adapter: the adapter structure to free
+ *
+ */
+static inline void ixd_devlink_free(struct ixd_adapter *adapter)
+{
+ struct devlink *devlink = priv_to_devlink(adapter);
+
+ devlink_free(devlink);
+}
+
+/**
+ * ixd_devlink_unregister - Unregister devlink resources for this adapter.
+ * @adapter: the adapter structure to cleanup
+ *
+ * Releases resources used by devlink and cleans up associated memory.
+ */
+static inline void ixd_devlink_unregister(struct ixd_adapter *adapter)
+{
+ devlink_unregister(priv_to_devlink(adapter));
+}
+
+/**
+ * ixd_devlink_register - Register devlink interface for this adapter
+ * @adapter: pointer to ixd adapter structure to be associated with devlink
+ *
+ * Register the devlink instance associated with this adapter
+ */
+static inline void ixd_devlink_register(struct ixd_adapter *adapter)
+{
+ devlink_register(priv_to_devlink(adapter));
+}
+
+#endif /* _IXD_DEVLINK_H_ */
diff --git a/drivers/net/ethernet/intel/ixd/ixd_main.c b/drivers/net/ethernet/intel/ixd/ixd_main.c
index 6d5e6aca77df..ea6aa793a6a7 100644
--- a/drivers/net/ethernet/intel/ixd/ixd_main.c
+++ b/drivers/net/ethernet/intel/ixd/ixd_main.c
@@ -4,6 +4,7 @@
#include "ixd.h"
#include "ixd_ctlq.h"
#include "ixd_lan_regs.h"
+#include "ixd_devlink.h"
MODULE_DESCRIPTION("Intel(R) Control Plane Function Device Driver");
MODULE_IMPORT_NS("LIBIE_CP");
@@ -21,11 +22,14 @@ static void ixd_remove(struct pci_dev *pdev)
/* Do not mix removal with (re)initialization */
cancel_delayed_work_sync(&adapter->init_task.init_work);
+ ixd_devlink_unregister(adapter);
+
/* Leave the device clean on exit */
ixd_trigger_reset(adapter);
ixd_deinit_dflt_mbx(adapter);
libie_pci_unmap_all_mmio_regions(&adapter->cp_ctx.mmio_info);
+ ixd_devlink_free(adapter);
}
/**
@@ -93,7 +97,7 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
if (WARN_ON(ent->device != IXD_DEV_ID_CPF))
return -EINVAL;
- adapter = devm_kzalloc(&pdev->dev, sizeof(*adapter), GFP_KERNEL);
+ adapter = ixd_adapter_alloc(&pdev->dev);
if (!adapter)
return -ENOMEM;
@@ -102,13 +106,13 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
err = libie_pci_init_dev(pdev);
if (err)
- return err;
+ goto free_adapter;
pci_set_drvdata(pdev, adapter);
err = ixd_iomap_regions(adapter);
if (err)
- return err;
+ goto free_adapter;
INIT_DELAYED_WORK(&adapter->init_task.init_work,
ixd_init_task);
@@ -118,7 +122,13 @@ static int ixd_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
queue_delayed_work(system_unbound_wq, &adapter->init_task.init_work,
msecs_to_jiffies(500));
+ ixd_devlink_register(adapter);
+
return 0;
+
+free_adapter:
+ ixd_devlink_free(adapter);
+ return err;
}
static const struct pci_device_id ixd_pci_tbl[] = {
--
2.47.0
^ permalink raw reply related
* [PATCH iproute2-next] tc: use ll_init_map() only when needed
From: Eric Dumazet @ 2026-04-28 8:28 UTC (permalink / raw)
To: David Ahern, Stephen Hemminger
Cc: David S . Miller, Jakub Kicinski, Paolo Abeni, netdev,
eric.dumazet, Eric Dumazet
Some setups can have thousands of devices.
ll_init_map() is rather expensive for them.
Only call ll_init_map() in the following cases:
1) tc runs in batch mode.
2) tc runs in monitor mode.
3) tc dumps qdiscs/classes/filters for all netdev.
This greatly reduces RTNL pressure on common operations.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
tc/f_route.c | 1 -
tc/m_mirred.c | 2 --
tc/tc.c | 1 +
tc/tc_class.c | 6 ++----
tc/tc_filter.c | 32 ++++++++++++++++----------------
tc/tc_qdisc.c | 5 ++---
6 files changed, 21 insertions(+), 26 deletions(-)
diff --git a/tc/f_route.c b/tc/f_route.c
index 96b99b06be179c90dfb0d24d9577b5206dd98ef1..c234ddf0cb9309556844c85047d0f9ef53a2fd7b 100644
--- a/tc/f_route.c
+++ b/tc/f_route.c
@@ -76,7 +76,6 @@ static int route_parse_opt(const struct filter_util *qu, char *handle, int argc,
__u32 id;
NEXT_ARG();
- ll_init_map(&rth);
if ((id = ll_name_to_index(*argv)) <= 0) {
fprintf(stderr, "Illegal \"fromif\"\n");
return -1;
diff --git a/tc/m_mirred.c b/tc/m_mirred.c
index f4da3c76284460d7f4bb73c7c3c172732f054632..e9ae5199432e828026f15ce226c5f5f84998c660 100644
--- a/tc/m_mirred.c
+++ b/tc/m_mirred.c
@@ -213,8 +213,6 @@ parse_direction(const struct action_util *a, int *argc_p, char ***argv_p,
if (d[0]) {
int idx;
- ll_init_map(&rth);
-
idx = ll_name_to_index(d);
if (!idx)
return nodev(d);
diff --git a/tc/tc.c b/tc/tc.c
index 0fc658c881f0f553f1f1f8d87b46943d4d0eed8f..7d69e4d533a98b9c071cb954443a049a52252f2c 100644
--- a/tc/tc.c
+++ b/tc/tc.c
@@ -243,6 +243,7 @@ static int batch(const char *name)
return -1;
}
+ ll_init_map(&rth);
ret = do_batch(name, force, tc_batch_cmd, NULL);
rtnl_close(&rth);
diff --git a/tc/tc_class.c b/tc/tc_class.c
index 6d707d8c924f4b5e90201d4de8e9779f82ce17f3..9aace019e6c8b96fb51edb4fe89bea5b7469beed 100644
--- a/tc/tc_class.c
+++ b/tc/tc_class.c
@@ -136,11 +136,11 @@ static int tc_class_modify(int cmd, unsigned int flags, int argc, char **argv)
}
if (d[0]) {
- ll_init_map(&rth);
-
req.t.tcm_ifindex = ll_name_to_index(d);
if (!req.t.tcm_ifindex)
return -nodev(d);
+ } else {
+ ll_init_map(&rth);
}
if (rtnl_talk(&rth, &req.n, NULL) < 0)
@@ -437,8 +437,6 @@ static int tc_class_list(int argc, char **argv)
argc--; argv++;
}
- ll_init_map(&rth);
-
if (d[0]) {
t.tcm_ifindex = ll_name_to_index(d);
if (!t.tcm_ifindex)
diff --git a/tc/tc_filter.c b/tc/tc_filter.c
index 7db850bda11a3408ec66b457c9fc590f3a734f65..be70e360849e5058192640da0f061810a99e2ea8 100644
--- a/tc/tc_filter.c
+++ b/tc/tc_filter.c
@@ -188,16 +188,17 @@ static int tc_filter_modify(int cmd, unsigned int flags, int argc, char **argv)
addattr_l(&req.n, sizeof(req), TCA_KIND, k, strlen(k)+1);
if (d[0]) {
- ll_init_map(&rth);
-
req.t.tcm_ifindex = ll_name_to_index(d);
if (req.t.tcm_ifindex == 0) {
fprintf(stderr, "Cannot find device \"%s\"\n", d);
return 1;
}
- } else if (block_index) {
- req.t.tcm_ifindex = TCM_IFINDEX_MAGIC_BLOCK;
- req.t.tcm_block_index = block_index;
+ } else {
+ ll_init_map(&rth);
+ if (block_index) {
+ req.t.tcm_ifindex = TCM_IFINDEX_MAGIC_BLOCK;
+ req.t.tcm_block_index = block_index;
+ }
}
if (q) {
@@ -539,8 +540,6 @@ static int tc_filter_get(int cmd, unsigned int flags, int argc, char **argv)
}
if (d[0]) {
- ll_init_map(&rth);
-
req.t.tcm_ifindex = ll_name_to_index(d);
if (!req.t.tcm_ifindex)
return -nodev(d);
@@ -704,21 +703,22 @@ static int tc_filter_list(int cmd, int argc, char **argv)
req.t.tcm_info = TC_H_MAKE(prio<<16, protocol);
- ll_init_map(&rth);
-
if (d[0]) {
req.t.tcm_ifindex = ll_name_to_index(d);
if (!req.t.tcm_ifindex)
return -nodev(d);
filter_ifindex = req.t.tcm_ifindex;
- } else if (block_index) {
- if (!tc_qdisc_block_exists(block_index)) {
- fprintf(stderr, "Cannot find block \"%u\"\n", block_index);
- return 1;
+ } else {
+ ll_init_map(&rth);
+ if (block_index) {
+ if (!tc_qdisc_block_exists(block_index)) {
+ fprintf(stderr, "Cannot find block \"%u\"\n", block_index);
+ return 1;
+ }
+ req.t.tcm_ifindex = TCM_IFINDEX_MAGIC_BLOCK;
+ req.t.tcm_block_index = block_index;
+ filter_block_index = block_index;
}
- req.t.tcm_ifindex = TCM_IFINDEX_MAGIC_BLOCK;
- req.t.tcm_block_index = block_index;
- filter_block_index = block_index;
}
if (filter_chain_index_set)
diff --git a/tc/tc_qdisc.c b/tc/tc_qdisc.c
index 7eb9a31baa31fb91a8141c9eca1135e76a52fa8b..252ea861f249a5a71cf782389776e68c3e48c6cc 100644
--- a/tc/tc_qdisc.c
+++ b/tc/tc_qdisc.c
@@ -193,8 +193,6 @@ static int tc_qdisc_modify(int cmd, unsigned int flags, int argc, char **argv)
if (d[0]) {
int idx;
- ll_init_map(&rth);
-
idx = ll_name_to_index(d);
if (!idx)
return -nodev(d);
@@ -411,13 +409,14 @@ static int tc_qdisc_list(int argc, char **argv)
argc--; argv++;
}
- ll_init_map(&rth);
if (d[0]) {
req.t.tcm_ifindex = ll_name_to_index(d);
if (!req.t.tcm_ifindex)
return -nodev(d);
filter_ifindex = req.t.tcm_ifindex;
+ } else {
+ ll_init_map(&rth);
}
if (dump_invisible) {
--
2.54.0.545.g6539524ca2-goog
^ permalink raw reply related
* Re: [PATCH net] net: add net_iov_init() and use it to initialize ->page_type
From: Byungchul Park @ 2026-04-28 8:14 UTC (permalink / raw)
To: Vlastimil Babka (SUSE)
Cc: Jakub Kicinski, davem, netdev, edumazet, pabeni, andrew+netdev,
horms, asml.silence, axboe, almasrymina, sdf, hawk, akpm, rppt,
io-uring, kernel_team
In-Reply-To: <6dcbee42-df9b-4c0a-b153-aad953441fad@kernel.org>
On Tue, Apr 28, 2026 at 09:57:08AM +0200, Vlastimil Babka (SUSE) wrote:
> +Cc Byungchul
>
> On 4/28/26 04:53, Jakub Kicinski wrote:
> > Commit db359fccf212 ("mm: introduce a new page type for page pool in
> > page type") added a page_type field to struct net_iov at the same
> > offset as struct page::page_type, so that page_pool_set_pp_info() can
> > call __SetPageNetpp() uniformly on both pages and net_iovs.
> >
> > The page-type API requires the field to hold the UINT_MAX "no type"
> > sentinel before a type can be set; for real struct page that invariant
> > is established by the page allocator on free. struct net_iov is not
> > allocated through the page allocator, so the field is left as zero
My bad. Overlooked the point. Thanks for the fix.
Acked-by: Byungchul Park <byungchul@sk.com>
Byungchul
> > (io_uring zcrx, which uses __GFP_ZERO) or as slab garbage (devmem,
> > which uses kvmalloc_objs() without zeroing). When the page pool then
> > calls page_pool_set_pp_info() on a freshly-bound niov,
> > __SetPageNetpp()'s VM_BUG_ON_PAGE(page->page_type != UINT_MAX) fires
> > and the kernel BUGs. Triggered in selftests by io_uring zcrx setup
> > through the fbnic queue restart path:
> >
> > kernel BUG at ./include/linux/page-flags.h:1062!
> > RIP: 0010:page_pool_set_pp_info (./include/linux/page-flags.h:1062
> > net/core/page_pool.c:716)
> > Call Trace:
> > <TASK>
> > net_mp_niov_set_page_pool (net/core/page_pool.c:1360)
> > io_pp_zc_alloc_netmems (io_uring/zcrx.c:1089 io_uring/zcrx.c:1110)
> > fbnic_fill_bdq (./include/net/page_pool/helpers.h:160
> > drivers/net/ethernet/meta/fbnic/fbnic_txrx.c:906)
> > __fbnic_nv_restart (drivers/net/ethernet/meta/fbnic/fbnic_txrx.c:2470
> > drivers/net/ethernet/meta/fbnic/fbnic_txrx.c:2874)
> > fbnic_queue_start (drivers/net/ethernet/meta/fbnic/fbnic_txrx.c:2903)
> > netdev_rx_queue_reconfig (net/core/netdev_rx_queue.c:137)
> > __netif_mp_open_rxq (net/core/netdev_rx_queue.c:234)
> > io_register_zcrx (io_uring/zcrx.c:818 io_uring/zcrx.c:903)
> > __io_uring_register (io_uring/register.c:931)
> > __do_sys_io_uring_register (io_uring/register.c:1029)
> > do_syscall_64 (arch/x86/entry/syscall_64.c:63
> > arch/x86/entry/syscall_64.c:94)
> > </TASK>
> >
> > The same path is reachable through devmem dmabuf binding via
> > netdev_nl_bind_rx_doit() -> net_devmem_bind_dmabuf_to_queue().
> >
> > Add a net_iov_init() helper that stamps ->owner, ->type and the
> > ->page_type sentinel, and use it from both the devmem and io_uring
> > zcrx niov init loops.
> >
> > Fixes: db359fccf212 ("mm: introduce a new page type for page pool in page type")
> > Signed-off-by: Jakub Kicinski <kuba@kernel.org>
>
> Acked-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
>
> > ---
> > CC: asml.silence@gmail.com
> > CC: axboe@kernel.dk
> > CC: almasrymina@google.com
> > CC: sdf@fomichev.me
> > CC: hawk@kernel.org
> > CC: akpm@linux-foundation.org
> > CC: rppt@kernel.org
> > CC: vbabka@kernel.org
> > CC: io-uring@vger.kernel.org
> > ---
> > include/net/netmem.h | 15 +++++++++++++++
> > io_uring/zcrx.c | 3 +--
> > net/core/devmem.c | 3 +--
> > 3 files changed, 17 insertions(+), 4 deletions(-)
> >
> > diff --git a/include/net/netmem.h b/include/net/netmem.h
> > index 507b74c9f52d..78fe51e5756b 100644
> > --- a/include/net/netmem.h
> > +++ b/include/net/netmem.h
> > @@ -127,6 +127,21 @@ static inline unsigned int net_iov_idx(const struct net_iov *niov)
> > return niov - net_iov_owner(niov)->niovs;
> > }
> >
> > +/* Initialize a niov: stamp the owning area, the memory provider type,
> > + * and the page_type "no type" sentinel expected by the page-type API
> > + * (see PAGE_TYPE_OPS in <linux/page-flags.h>) so that
> > + * page_pool_set_pp_info() can later call __SetPageNetpp() on a niov
> > + * cast to struct page.
> > + */
> > +static inline void net_iov_init(struct net_iov *niov,
> > + struct net_iov_area *owner,
> > + enum net_iov_type type)
> > +{
> > + niov->owner = owner;
> > + niov->type = type;
> > + niov->page_type = UINT_MAX;
> > +}
> > +
> > /* netmem */
> >
> > /**
> > diff --git a/io_uring/zcrx.c b/io_uring/zcrx.c
> > index 7b93c87b8371..19837e0b5e91 100644
> > --- a/io_uring/zcrx.c
> > +++ b/io_uring/zcrx.c
> > @@ -495,10 +495,9 @@ static int io_zcrx_create_area(struct io_zcrx_ifq *ifq,
> > for (i = 0; i < nr_iovs; i++) {
> > struct net_iov *niov = &area->nia.niovs[i];
> >
> > - niov->owner = &area->nia;
> > + net_iov_init(niov, &area->nia, NET_IOV_IOURING);
> > area->freelist[i] = i;
> > atomic_set(&area->user_refs[i], 0);
> > - niov->type = NET_IOV_IOURING;
> > }
> >
> > if (ifq->dev) {
> > diff --git a/net/core/devmem.c b/net/core/devmem.c
> > index cde4c89bc146..468344739db2 100644
> > --- a/net/core/devmem.c
> > +++ b/net/core/devmem.c
> > @@ -297,8 +297,7 @@ net_devmem_bind_dmabuf(struct net_device *dev,
> >
> > for (i = 0; i < owner->area.num_niovs; i++) {
> > niov = &owner->area.niovs[i];
> > - niov->type = NET_IOV_DMABUF;
> > - niov->owner = &owner->area;
> > + net_iov_init(niov, &owner->area, NET_IOV_DMABUF);
> > page_pool_set_dma_addr_netmem(net_iov_to_netmem(niov),
> > net_devmem_get_dma_addr(niov));
> > if (direction == DMA_TO_DEVICE)
^ permalink raw reply
* RE: [PATCH iwl-net v1 2/2] ice: use NETIF_F_HW_CSUM instead of IP/IPV6
From: Loktionov, Aleksandr @ 2026-04-28 8:34 UTC (permalink / raw)
To: Michal Swiatkowski, intel-wired-lan@lists.osuosl.org
Cc: netdev@vger.kernel.org, jramaseu@redhat.com, Nguyen, Anthony L,
Kitszel, Przemyslaw
In-Reply-To: <20260428070647.777141-3-michal.swiatkowski@linux.intel.com>
> -----Original Message-----
> From: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
> Sent: Tuesday, April 28, 2026 9:07 AM
> To: intel-wired-lan@lists.osuosl.org
> Cc: netdev@vger.kernel.org; jramaseu@redhat.com; Nguyen, Anthony L
> <anthony.l.nguyen@intel.com>; Kitszel, Przemyslaw
> <przemyslaw.kitszel@intel.com>; Loktionov, Aleksandr
> <aleksandr.loktionov@intel.com>; Michal Swiatkowski
> <michal.swiatkowski@linux.intel.com>
> Subject: [PATCH iwl-net v1 2/2] ice: use NETIF_F_HW_CSUM instead of
> IP/IPV6
>
> The hardware is capable of calculating checksum for IPV6 packets with
> extension header. To not drop such packets switch from IP/IPV6
> checksum to HW_CSUM.
I'd recommend "To not drop" -> "To avoid dropping"
>
> HW_CSUM is also used in previous generation (i40e).
>
> Previously HW_CSUM was used to indicate that hardware supports general
> checksum. Drop it assuming that if the hardware supports it, it is
> used.
>
> Disabling offload for E830 in case of TSO isn't needed anymore as the
> check for TSO is done in Tx path just before preparation of the
> special GCS descriptor.
>
> The commit from Fixes didn't introduce a bug, it just shown that the
> driver is doing sth wrong with the checksum features.
>
Except commit message nits
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
> Suggested-by: Jakub Ramaseuski <jramaseu@redhat.com>
> Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
> Fixes: 04c20a9356f2 ("net: skip offload for NETIF_F_IPV6_CSUM if ipv6
> header contains extension")
> Signed-off-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
> ---
> drivers/net/ethernet/intel/ice/ice_main.c | 21 +--------------------
> 1 file changed, 1 insertion(+), 20 deletions(-)
>
> diff --git a/drivers/net/ethernet/intel/ice/ice_main.c
> b/drivers/net/ethernet/intel/ice/ice_main.c
> index 15550216fbf0..0f2f949af536 100644
> --- a/drivers/net/ethernet/intel/ice/ice_main.c
> +++ b/drivers/net/ethernet/intel/ice/ice_main.c
> @@ -3500,9 +3500,8 @@ void ice_set_netdev_features(struct net_device
> *netdev)
> NETIF_F_RXHASH;
>
> csumo_features = NETIF_F_RXCSUM |
> - NETIF_F_IP_CSUM |
> NETIF_F_SCTP_CRC |
> - NETIF_F_IPV6_CSUM;
> + NETIF_F_HW_CSUM;
>
> vlano_features = NETIF_F_HW_VLAN_CTAG_FILTER |
> NETIF_F_HW_VLAN_CTAG_TX |
> @@ -3564,12 +3563,6 @@ void ice_set_netdev_features(struct net_device
> *netdev)
> /* Allow core to manage IRQs affinity */
> netif_set_affinity_auto(netdev);
>
> - /* Mutual exclusivity for TSO and GCS is enforced by the set
> features
> - * ndo callback.
> - */
> - if (ice_is_feature_supported(pf, ICE_F_GCS))
> - netdev->hw_features |= NETIF_F_HW_CSUM;
> -
> netif_set_tso_max_size(netdev, ICE_MAX_TSO_SIZE); }
>
> @@ -6489,18 +6482,6 @@ ice_set_features(struct net_device *netdev,
> netdev_features_t features)
> if (changed & NETIF_F_LOOPBACK)
> ret = ice_set_loopback(vsi, !!(features &
> NETIF_F_LOOPBACK));
>
> - /* Due to E830 hardware limitations, TSO (NETIF_F_ALL_TSO) with
> GCS
> - * (NETIF_F_HW_CSUM) is not supported.
> - */
> - if (ice_is_feature_supported(pf, ICE_F_GCS) &&
> - ((features & NETIF_F_HW_CSUM) && (features &
> NETIF_F_ALL_TSO))) {
> - if (netdev->features & NETIF_F_HW_CSUM)
> - dev_err(ice_pf_to_dev(pf), "To enable TSO, you
> must first disable HW checksum.\n");
> - else
> - dev_err(ice_pf_to_dev(pf), "To enable HW
> checksum, you must first disable TSO.\n");
> - return -EIO;
> - }
> -
> return ret;
> }
>
> --
> 2.49.0
^ permalink raw reply
* [PATCH net] net: airoha: Move entries to queue head in case of DMA mapping failure in airoha_dev_xmit()
From: Lorenzo Bianconi @ 2026-04-28 8:44 UTC (permalink / raw)
To: Andrew Lunn, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, Jacob Keller
Cc: Simon Horman, linux-arm-kernel, linux-mediatek, netdev,
Lorenzo Bianconi
In order to respect the original descriptor order and avoid any
potential IOMMU fault or memory corruption, move pending queue entries
to the head of hw queue tx_list if the DMA mapping of current inflight
packet fails in airoha_dev_xmit routine.
Fixes: 3f47e67dff1f7 ("net: airoha: Add the capability to consume out-of-order DMA tx descriptors")
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
---
drivers/net/ethernet/airoha/airoha_eth.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/airoha/airoha_eth.c b/drivers/net/ethernet/airoha/airoha_eth.c
index 5effb4a4ae84..82018a085e46 100644
--- a/drivers/net/ethernet/airoha/airoha_eth.c
+++ b/drivers/net/ethernet/airoha/airoha_eth.c
@@ -2123,14 +2123,14 @@ static netdev_tx_t airoha_dev_xmit(struct sk_buff *skb,
return NETDEV_TX_OK;
error_unmap:
- while (!list_empty(&tx_list)) {
+ list_for_each_entry(e, &tx_list, list) {
e = list_first_entry(&tx_list, struct airoha_queue_entry,
list);
dma_unmap_single(dev->dev.parent, e->dma_addr, e->dma_len,
DMA_TO_DEVICE);
e->dma_addr = 0;
- list_move_tail(&e->list, &q->tx_list);
}
+ list_splice(&tx_list, &q->tx_list);
spin_unlock_bh(&q->lock);
error:
---
base-commit: 3bc179bc7146c26c9dff75d2943d10528274e301
change-id: 20260428-airoha-xmit-unmap-error-path-8125594efb23
Best regards,
--
Lorenzo Bianconi <lorenzo@kernel.org>
^ permalink raw reply related
* [PATCH 0/5] selftests/xsk: foundational UMEM refactoring
From: Tushar Vyavahare @ 2026-04-28 8:03 UTC (permalink / raw)
To: netdev, magnus.karlsson, maciej.fijalkowski, stfomichev,
kernelxing, davem, kuba, pabeni, ast, daniel, tirthendu.sarkar,
tushar.vyavahare
Cc: bpf
This series refactors foundational UMEM handling in selftests/xsk.
It centralizes UMEM property setup through helpers, removes ifobject-level
UMEM ownership state, and normalizes umem_size/mmap_size usage across the
code paths touched here.
No functional behavior change is intended; the focus is internal
consistency and maintainability.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Tushar Vyavahare <tushar.vyavahare@intel.com>
Tushar Vyavahare (5):
selftests/xsk: Introduce helpers for setting UMEM properties
selftests/xsk: Eliminate umem reference from ifobject
selftests/xsk: Remove umem from pkt_generate parameters
selftests/xsk: Use umem_size() helper consistently
selftests/xsk: Introduce mmap_size in umem struct
.../selftests/bpf/prog_tests/test_xsk.c | 195 +++++++++---------
.../selftests/bpf/prog_tests/test_xsk.h | 33 +--
2 files changed, 119 insertions(+), 109 deletions(-)
--
2.43.0
^ permalink raw reply
* [PATCH 1/5] selftests/xsk: Introduce helpers for setting UMEM properties
From: Tushar Vyavahare @ 2026-04-28 8:03 UTC (permalink / raw)
To: netdev, magnus.karlsson, maciej.fijalkowski, stfomichev,
kernelxing, davem, kuba, pabeni, ast, daniel, tirthendu.sarkar,
tushar.vyavahare
Cc: bpf
In-Reply-To: <20260428080317.702124-1-tushar.vyavahare@intel.com>
UMEM properties are set via open-coded field assignments in multiple test
paths, which makes updates noisy and error-prone.
Introduce three helpers to set UMEM properties through a single interface.
This keeps setup logic consistent across tests and makes future refactoring
simpler.
No functional behavior change is intended.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Tushar Vyavahare <tushar.vyavahare@intel.com>
---
.../selftests/bpf/prog_tests/test_xsk.c | 48 +++++++++++--------
1 file changed, 28 insertions(+), 20 deletions(-)
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xsk.c b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
index 7950c504ed28..d5cd8f590b25 100644
--- a/tools/testing/selftests/bpf/prog_tests/test_xsk.c
+++ b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
@@ -303,6 +303,22 @@ static void test_spec_reset(struct test_spec *test)
__test_spec_init(test, test->ifobj_tx, test->ifobj_rx);
}
+static void test_spec_set_unaligned(struct test_spec *test)
+{
+ test_spec_set_unaligned(test);
+}
+
+static void test_spec_set_frame_size(struct test_spec *test, u32 size)
+{
+ test->ifobj_tx->umem->frame_size = size;
+ test->ifobj_rx->umem->frame_size = size;
+}
+
+static void test_spec_set_frame_headroom(struct test_spec *test, u32 size)
+{
+ test->ifobj_rx->umem->frame_headroom = size;
+}
+
static void test_spec_set_xdp_prog(struct test_spec *test, struct bpf_program *xdp_prog_rx,
struct bpf_program *xdp_prog_tx, struct bpf_map *xskmap_rx,
struct bpf_map *xskmap_tx)
@@ -1953,7 +1969,7 @@ int testapp_xdp_prog_cleanup(struct test_spec *test)
int testapp_headroom(struct test_spec *test)
{
- test->ifobj_rx->umem->frame_headroom = UMEM_HEADROOM_TEST_SIZE;
+ test_spec_set_frame_headroom(test, UMEM_HEADROOM_TEST_SIZE);
return testapp_validate_traffic(test);
}
@@ -1968,8 +1984,8 @@ int testapp_stats_rx_dropped(struct test_spec *test)
if (pkt_stream_replace_half(test, (MIN_PKT_SIZE * 3) + umem_tr, 0))
return TEST_FAILURE;
- test->ifobj_rx->umem->frame_headroom = test->ifobj_rx->umem->frame_size -
- XDP_PACKET_HEADROOM - (MIN_PKT_SIZE * 2) - umem_tr;
+ test_spec_set_frame_headroom(test, test->ifobj_rx->umem->frame_size -
+ XDP_PACKET_HEADROOM - (MIN_PKT_SIZE * 2) - umem_tr);
if (pkt_stream_receive_half(test))
return TEST_FAILURE;
test->ifobj_rx->validation_func = validate_rx_dropped;
@@ -2025,8 +2041,7 @@ int testapp_stats_fill_empty(struct test_spec *test)
int testapp_send_receive_unaligned(struct test_spec *test)
{
- test->ifobj_tx->umem->unaligned_mode = true;
- test->ifobj_rx->umem->unaligned_mode = true;
+ test_spec_set_unaligned(test);
/* Let half of the packets straddle a 4K buffer boundary */
if (pkt_stream_replace_half(test, MIN_PKT_SIZE, -MIN_PKT_SIZE / 2))
return TEST_FAILURE;
@@ -2037,8 +2052,7 @@ int testapp_send_receive_unaligned(struct test_spec *test)
int testapp_send_receive_unaligned_mb(struct test_spec *test)
{
test->mtu = MAX_ETH_JUMBO_SIZE;
- test->ifobj_tx->umem->unaligned_mode = true;
- test->ifobj_rx->umem->unaligned_mode = true;
+ test_spec_set_unaligned(test);
if (pkt_stream_replace(test, DEFAULT_PKT_CNT, MAX_ETH_JUMBO_SIZE))
return TEST_FAILURE;
return testapp_validate_traffic(test);
@@ -2204,7 +2218,7 @@ int testapp_poll_txq_tmout(struct test_spec *test)
{
test->ifobj_tx->use_poll = true;
/* create invalid frame by set umem frame_size and pkt length equal to 2048 */
- test->ifobj_tx->umem->frame_size = 2048;
+ test_spec_set_frame_size(test, 2048);
if (pkt_stream_replace(test, 2 * DEFAULT_PKT_CNT, 2048))
return TEST_FAILURE;
return testapp_validate_traffic_single_thread(test, test->ifobj_tx);
@@ -2337,8 +2351,7 @@ int testapp_send_receive(struct test_spec *test)
int testapp_send_receive_2k_frame(struct test_spec *test)
{
- test->ifobj_tx->umem->frame_size = 2048;
- test->ifobj_rx->umem->frame_size = 2048;
+ test_spec_set_frame_size(test, 2048);
if (pkt_stream_replace(test, DEFAULT_PKT_CNT, MIN_PKT_SIZE))
return TEST_FAILURE;
return testapp_validate_traffic(test);
@@ -2363,15 +2376,13 @@ int testapp_aligned_inv_desc(struct test_spec *test)
int testapp_aligned_inv_desc_2k_frame(struct test_spec *test)
{
- test->ifobj_tx->umem->frame_size = 2048;
- test->ifobj_rx->umem->frame_size = 2048;
+ test_spec_set_frame_size(test, 2048);
return testapp_invalid_desc(test);
}
int testapp_unaligned_inv_desc(struct test_spec *test)
{
- test->ifobj_tx->umem->unaligned_mode = true;
- test->ifobj_rx->umem->unaligned_mode = true;
+ test_spec_set_unaligned(test);
return testapp_invalid_desc(test);
}
@@ -2380,10 +2391,8 @@ int testapp_unaligned_inv_desc_4001_frame(struct test_spec *test)
u64 page_size, umem_size;
/* Odd frame size so the UMEM doesn't end near a page boundary. */
- test->ifobj_tx->umem->frame_size = 4001;
- test->ifobj_rx->umem->frame_size = 4001;
- test->ifobj_tx->umem->unaligned_mode = true;
- test->ifobj_rx->umem->unaligned_mode = true;
+ test_spec_set_frame_size(test, 4001);
+ test_spec_set_unaligned(test);
/* This test exists to test descriptors that staddle the end of
* the UMEM but not a page.
*/
@@ -2402,8 +2411,7 @@ int testapp_aligned_inv_desc_mb(struct test_spec *test)
int testapp_unaligned_inv_desc_mb(struct test_spec *test)
{
- test->ifobj_tx->umem->unaligned_mode = true;
- test->ifobj_rx->umem->unaligned_mode = true;
+ test_spec_set_unaligned(test);
return testapp_invalid_desc_mb(test);
}
--
2.43.0
^ permalink raw reply related
* [PATCH 2/5] selftests/xsk: Eliminate umem reference from ifobject
From: Tushar Vyavahare @ 2026-04-28 8:03 UTC (permalink / raw)
To: netdev, magnus.karlsson, maciej.fijalkowski, stfomichev,
kernelxing, davem, kuba, pabeni, ast, daniel, tirthendu.sarkar,
tushar.vyavahare
Cc: bpf
In-Reply-To: <20260428080317.702124-1-tushar.vyavahare@intel.com>
Each AF_XDP socket is associated with a UMEM, but ifobject currently
keeps only a single ifobject-level UMEM reference. That makes it awkward
to model cases where multiple sockets on the same interface eventually
need distinct UMEM instances.
Move UMEM storage into xsk_socket_info and keep xsk_socket_info::umem as
the active pointer used by the code paths. For now, all sockets on an
interface still point at the first socket's UMEM, which preserves
existing behaviour while removing the ifobject-level ownership.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Tushar Vyavahare <tushar.vyavahare@intel.com>
---
.../selftests/bpf/prog_tests/test_xsk.c | 124 +++++++++---------
.../selftests/bpf/prog_tests/test_xsk.h | 32 ++---
2 files changed, 80 insertions(+), 76 deletions(-)
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xsk.c b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
index d5cd8f590b25..ad5ada9216fc 100644
--- a/tools/testing/selftests/bpf/prog_tests/test_xsk.c
+++ b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
@@ -65,9 +65,9 @@ static void gen_eth_hdr(struct xsk_socket_info *xsk, struct ethhdr *eth_hdr)
eth_hdr->h_proto = htons(ETH_P_LOOPBACK);
}
-static bool is_umem_valid(struct ifobject *ifobj)
+static bool is_umem_valid(struct xsk_umem_info *umem)
{
- return !!ifobj->umem->umem;
+ return !!umem->umem;
}
static u32 mode_to_xdp_flags(enum test_mode mode)
@@ -229,24 +229,26 @@ static void __test_spec_init(struct test_spec *test, struct ifobject *ifobj_tx,
ifobj->tx_on = false;
}
- memset(ifobj->umem, 0, sizeof(*ifobj->umem));
- ifobj->umem->num_frames = DEFAULT_UMEM_BUFFERS;
- ifobj->umem->frame_size = XSK_UMEM__DEFAULT_FRAME_SIZE;
-
for (j = 0; j < MAX_SOCKETS; j++) {
- memset(&ifobj->xsk_arr[j], 0, sizeof(ifobj->xsk_arr[j]));
- ifobj->xsk_arr[j].rxqsize = XSK_RING_CONS__DEFAULT_NUM_DESCS;
+ struct xsk_socket_info *xsk = &ifobj->xsk_arr[j];
+
+ memset(xsk, 0, sizeof(*xsk));
+ xsk->rxqsize = XSK_RING_CONS__DEFAULT_NUM_DESCS;
+ xsk->umem = &ifobj->xsk_arr[0].umem_real;
ifobj->xsk_arr[j].batch_size = DEFAULT_BATCH_SIZE;
if (i == 0)
- ifobj->xsk_arr[j].pkt_stream = test->tx_pkt_stream_default;
+ xsk->pkt_stream = test->tx_pkt_stream_default;
else
- ifobj->xsk_arr[j].pkt_stream = test->rx_pkt_stream_default;
+ xsk->pkt_stream = test->rx_pkt_stream_default;
- memcpy(ifobj->xsk_arr[j].src_mac, g_mac, ETH_ALEN);
- memcpy(ifobj->xsk_arr[j].dst_mac, g_mac, ETH_ALEN);
- ifobj->xsk_arr[j].src_mac[5] += ((j * 2) + 0);
- ifobj->xsk_arr[j].dst_mac[5] += ((j * 2) + 1);
+ memcpy(xsk->src_mac, g_mac, ETH_ALEN);
+ memcpy(xsk->dst_mac, g_mac, ETH_ALEN);
+ xsk->src_mac[5] += ((j * 2) + 0);
+ xsk->dst_mac[5] += ((j * 2) + 1);
}
+
+ ifobj->xsk->umem->num_frames = DEFAULT_UMEM_BUFFERS;
+ ifobj->xsk->umem->frame_size = XSK_UMEM__DEFAULT_FRAME_SIZE;
}
if (ifobj_tx->hw_ring_size_supp)
@@ -305,18 +307,19 @@ static void test_spec_reset(struct test_spec *test)
static void test_spec_set_unaligned(struct test_spec *test)
{
- test_spec_set_unaligned(test);
+ test->ifobj_tx->xsk->umem->unaligned_mode = true;
+ test->ifobj_rx->xsk->umem->unaligned_mode = true;
}
static void test_spec_set_frame_size(struct test_spec *test, u32 size)
{
- test->ifobj_tx->umem->frame_size = size;
- test->ifobj_rx->umem->frame_size = size;
+ test->ifobj_tx->xsk->umem->frame_size = size;
+ test->ifobj_rx->xsk->umem->frame_size = size;
}
static void test_spec_set_frame_headroom(struct test_spec *test, u32 size)
{
- test->ifobj_rx->umem->frame_headroom = size;
+ test->ifobj_rx->xsk->umem->frame_headroom = size;
}
static void test_spec_set_xdp_prog(struct test_spec *test, struct bpf_program *xdp_prog_rx,
@@ -1007,7 +1010,7 @@ static int __receive_pkts(struct test_spec *test, struct xsk_socket_info *xsk)
return TEST_FAILURE;
if (!ret) {
- if (!is_umem_valid(test->ifobj_tx))
+ if (!is_umem_valid(umem))
return TEST_PASS;
ksft_print_msg("ERROR: [%s] Poll timed out\n", __func__);
@@ -1167,7 +1170,7 @@ static int __send_pkts(struct ifobject *ifobject, struct xsk_socket_info *xsk, b
{
u32 i, idx = 0, valid_pkts = 0, valid_frags = 0, buffer_len;
struct pkt_stream *pkt_stream = xsk->pkt_stream;
- struct xsk_umem_info *umem = ifobject->umem;
+ struct xsk_umem_info *umem = xsk->umem;
bool use_poll = ifobject->use_poll;
struct pollfd fds = { };
int ret;
@@ -1226,7 +1229,7 @@ static int __send_pkts(struct ifobject *ifobject, struct xsk_socket_info *xsk, b
while (nb_frags_left--) {
struct xdp_desc *tx_desc = xsk_ring_prod__tx_desc(&xsk->tx, idx + i);
- tx_desc->addr = pkt_get_addr(pkt, ifobject->umem);
+ tx_desc->addr = pkt_get_addr(pkt, xsk->umem);
if (pkt_stream->verbatim) {
tx_desc->len = pkt->len;
tx_desc->options = pkt->options;
@@ -1319,7 +1322,7 @@ bool all_packets_sent(struct test_spec *test, unsigned long *bitmap)
static int send_pkts(struct test_spec *test, struct ifobject *ifobject)
{
- bool timeout = !is_umem_valid(test->ifobj_rx);
+ bool timeout = !is_umem_valid(test->ifobj_rx->xsk->umem);
DECLARE_BITMAP(bitmap, test->nb_sockets);
u32 i, ret;
@@ -1504,14 +1507,20 @@ static int xsk_configure(struct test_spec *test, struct ifobject *ifobject,
static int thread_common_ops_tx(struct test_spec *test, struct ifobject *ifobject)
{
- int ret = xsk_configure(test, ifobject, test->ifobj_rx->umem, true);
+ struct xsk_umem_info *umem_rx;
+ int ret;
+ if (!test->ifobj_rx || !test->ifobj_rx->xsk_arr[0].umem ||
+ !test->ifobj_rx->xsk_arr[0].umem->umem) {
+ ksft_print_msg("Error: RX UMEM is not initialized before shared-UMEM TX setup\n");
+ return -EINVAL;
+ }
+
+ umem_rx = test->ifobj_rx->xsk_arr[0].umem;
+ ret = xsk_configure(test, ifobject, umem_rx, true);
if (ret)
return ret;
- ifobject->xsk = &ifobject->xsk_arr[0];
ifobject->xskmap = test->ifobj_rx->xskmap;
- memcpy(ifobject->umem, test->ifobj_rx->umem, sizeof(struct xsk_umem_info));
- ifobject->umem->base_addr = 0;
return 0;
}
@@ -1564,6 +1573,7 @@ static int xsk_populate_fill_ring(struct xsk_umem_info *umem, struct pkt_stream
static int thread_common_ops(struct test_spec *test, struct ifobject *ifobject)
{
+ struct xsk_umem_info *umem = ifobject->xsk->umem;
LIBBPF_OPTS(bpf_xdp_query_opts, opts);
int mmap_flags;
u64 umem_sz;
@@ -1571,10 +1581,10 @@ static int thread_common_ops(struct test_spec *test, struct ifobject *ifobject)
int ret;
u32 i;
- umem_sz = ifobject->umem->num_frames * ifobject->umem->frame_size;
+ umem_sz = umem->num_frames * umem->frame_size;
mmap_flags = MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE;
- if (ifobject->umem->unaligned_mode)
+ if (umem->unaligned_mode)
mmap_flags |= MAP_HUGETLB | MAP_HUGE_2MB;
if (ifobject->shared_umem)
@@ -1584,11 +1594,11 @@ static int thread_common_ops(struct test_spec *test, struct ifobject *ifobject)
if (bufs == MAP_FAILED)
return -errno;
- ret = xsk_configure_umem(ifobject, ifobject->umem, bufs, umem_sz);
+ ret = xsk_configure_umem(ifobject, umem, bufs, umem_sz);
if (ret)
return ret;
- ret = xsk_configure(test, ifobject, ifobject->umem, false);
+ ret = xsk_configure(test, ifobject, umem, false);
if (ret)
return ret;
@@ -1597,21 +1607,19 @@ static int thread_common_ops(struct test_spec *test, struct ifobject *ifobject)
if (!ifobject->rx_on)
return 0;
- ret = xsk_populate_fill_ring(ifobject->umem, ifobject->xsk->pkt_stream,
+ ret = xsk_populate_fill_ring(umem, ifobject->xsk->pkt_stream,
ifobject->use_fill_ring);
if (ret)
return ret;
for (i = 0; i < test->nb_sockets; i++) {
- ifobject->xsk = &ifobject->xsk_arr[i];
- ret = xsk_update_xskmap(ifobject->xskmap, ifobject->xsk->xsk, i);
+ ret = xsk_update_xskmap(ifobject->xskmap, ifobject->xsk_arr[i].xsk, i);
if (ret)
return ret;
}
return 0;
}
-
void *worker_testapp_validate_tx(void *arg)
{
struct test_spec *test = (struct test_spec *)arg;
@@ -1691,14 +1699,15 @@ void *worker_testapp_validate_rx(void *arg)
static void testapp_clean_xsk_umem(struct ifobject *ifobj)
{
- u64 umem_sz = ifobj->umem->num_frames * ifobj->umem->frame_size;
+ struct xsk_umem_info *umem = ifobj->xsk->umem;
+ u64 umem_sz = umem->num_frames * umem->frame_size;
if (ifobj->shared_umem)
umem_sz *= 2;
umem_sz = ceil_u64(umem_sz, HUGEPAGE_SIZE) * HUGEPAGE_SIZE;
- xsk_umem__delete(ifobj->umem->umem);
- munmap(ifobj->umem->buffer, umem_sz);
+ xsk_umem__delete(umem->umem);
+ munmap(umem->buffer, umem_sz);
}
static void handler(int signum)
@@ -1860,9 +1869,11 @@ static int testapp_validate_traffic(struct test_spec *test)
{
struct ifobject *ifobj_rx = test->ifobj_rx;
struct ifobject *ifobj_tx = test->ifobj_tx;
+ struct xsk_umem_info *umem_rx = ifobj_rx->xsk->umem;
+ struct xsk_umem_info *umem_tx = ifobj_tx->xsk->umem;
- if ((ifobj_rx->umem->unaligned_mode && !ifobj_rx->unaligned_supp) ||
- (ifobj_tx->umem->unaligned_mode && !ifobj_tx->unaligned_supp)) {
+ if ((umem_rx->unaligned_mode && !ifobj_rx->unaligned_supp) ||
+ (umem_tx->unaligned_mode && !ifobj_tx->unaligned_supp)) {
ksft_print_msg("No huge pages present.\n");
return TEST_SKIP;
}
@@ -1975,6 +1986,7 @@ int testapp_headroom(struct test_spec *test)
int testapp_stats_rx_dropped(struct test_spec *test)
{
+ struct xsk_umem_info *umem = test->ifobj_rx->xsk->umem;
u32 umem_tr = test->ifobj_tx->umem_tailroom;
if (test->mode == TEST_MODE_ZC) {
@@ -1984,8 +1996,8 @@ int testapp_stats_rx_dropped(struct test_spec *test)
if (pkt_stream_replace_half(test, (MIN_PKT_SIZE * 3) + umem_tr, 0))
return TEST_FAILURE;
- test_spec_set_frame_headroom(test, test->ifobj_rx->umem->frame_size -
- XDP_PACKET_HEADROOM - (MIN_PKT_SIZE * 2) - umem_tr);
+ test_spec_set_frame_headroom(test, umem->frame_size - XDP_PACKET_HEADROOM -
+ (MIN_PKT_SIZE * 2) - umem_tr);
if (pkt_stream_receive_half(test))
return TEST_FAILURE;
test->ifobj_rx->validation_func = validate_rx_dropped;
@@ -2078,7 +2090,7 @@ int testapp_send_receive_mb(struct test_spec *test)
int testapp_invalid_desc_mb(struct test_spec *test)
{
- struct xsk_umem_info *umem = test->ifobj_tx->umem;
+ struct xsk_umem_info *umem = test->ifobj_tx->xsk->umem;
u64 umem_size = umem->num_frames * umem->frame_size;
struct pkt pkts[] = {
/* Valid packet for synch to start with */
@@ -2119,7 +2131,7 @@ int testapp_invalid_desc_mb(struct test_spec *test)
int testapp_invalid_desc(struct test_spec *test)
{
- struct xsk_umem_info *umem = test->ifobj_tx->umem;
+ struct xsk_umem_info *umem = test->ifobj_tx->xsk->umem;
u64 umem_size = umem->num_frames * umem->frame_size;
struct pkt pkts[] = {
/* Zero packet address allowed */
@@ -2397,7 +2409,8 @@ int testapp_unaligned_inv_desc_4001_frame(struct test_spec *test)
* the UMEM but not a page.
*/
page_size = sysconf(_SC_PAGESIZE);
- umem_size = test->ifobj_tx->umem->num_frames * test->ifobj_tx->umem->frame_size;
+ umem_size = test->ifobj_tx->xsk->umem->num_frames *
+ test->ifobj_tx->xsk->umem->frame_size;
assert(umem_size % page_size > MIN_PKT_SIZE);
assert(umem_size % page_size < page_size - MIN_PKT_SIZE);
@@ -2455,9 +2468,9 @@ int testapp_hw_sw_max_ring_size(struct test_spec *test)
test->total_steps = 2;
test->ifobj_tx->ring.tx_pending = test->ifobj_tx->ring.tx_max_pending;
test->ifobj_tx->ring.rx_pending = test->ifobj_tx->ring.rx_max_pending;
- test->ifobj_rx->umem->num_frames = max_descs;
- test->ifobj_rx->umem->fill_size = max_descs;
- test->ifobj_rx->umem->comp_size = max_descs;
+ test->ifobj_rx->xsk->umem->num_frames = max_descs;
+ test->ifobj_rx->xsk->umem->fill_size = max_descs;
+ test->ifobj_rx->xsk->umem->comp_size = max_descs;
test->ifobj_tx->xsk->batch_size = XSK_RING_PROD__DEFAULT_NUM_DESCS;
test->ifobj_rx->xsk->batch_size = XSK_RING_PROD__DEFAULT_NUM_DESCS;
@@ -2595,25 +2608,16 @@ struct ifobject *ifobject_create(void)
return NULL;
ifobj->xsk_arr = calloc(MAX_SOCKETS, sizeof(*ifobj->xsk_arr));
- if (!ifobj->xsk_arr)
- goto out_xsk_arr;
-
- ifobj->umem = calloc(1, sizeof(*ifobj->umem));
- if (!ifobj->umem)
- goto out_umem;
+ if (!ifobj->xsk_arr) {
+ free(ifobj);
+ return NULL;
+ }
return ifobj;
-
-out_umem:
- free(ifobj->xsk_arr);
-out_xsk_arr:
- free(ifobj);
- return NULL;
}
void ifobject_delete(struct ifobject *ifobj)
{
- free(ifobj->umem);
free(ifobj->xsk_arr);
free(ifobj);
}
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xsk.h b/tools/testing/selftests/bpf/prog_tests/test_xsk.h
index 1ab8aee4ce56..accda2456fe6 100644
--- a/tools/testing/selftests/bpf/prog_tests/test_xsk.h
+++ b/tools/testing/selftests/bpf/prog_tests/test_xsk.h
@@ -80,9 +80,25 @@ typedef int (*validation_func_t)(struct ifobject *ifobj);
typedef void *(*thread_func_t)(void *arg);
typedef int (*test_func_t)(struct test_spec *test);
+struct xsk_umem_info {
+ struct xsk_ring_prod fq;
+ struct xsk_ring_cons cq;
+ struct xsk_umem *umem;
+ u64 next_buffer;
+ u32 num_frames;
+ u32 frame_headroom;
+ void *buffer;
+ u32 frame_size;
+ u32 base_addr;
+ u32 fill_size;
+ u32 comp_size;
+ bool unaligned_mode;
+};
+
struct xsk_socket_info {
struct xsk_ring_cons rx;
struct xsk_ring_prod tx;
+ struct xsk_umem_info umem_real;
struct xsk_umem_info *umem;
struct xsk_socket *xsk;
struct pkt_stream *pkt_stream;
@@ -97,21 +113,6 @@ struct xsk_socket_info {
int kick_rx(struct xsk_socket_info *xsk);
int kick_tx(struct xsk_socket_info *xsk);
-struct xsk_umem_info {
- struct xsk_ring_prod fq;
- struct xsk_ring_cons cq;
- struct xsk_umem *umem;
- u64 next_buffer;
- u32 num_frames;
- u32 frame_headroom;
- void *buffer;
- u32 frame_size;
- u32 base_addr;
- u32 fill_size;
- u32 comp_size;
- bool unaligned_mode;
-};
-
struct set_hw_ring {
u32 default_tx;
u32 default_rx;
@@ -123,7 +124,6 @@ struct ifobject {
char ifname[MAX_INTERFACE_NAME_CHARS];
struct xsk_socket_info *xsk;
struct xsk_socket_info *xsk_arr;
- struct xsk_umem_info *umem;
thread_func_t func_ptr;
validation_func_t validation_func;
struct xsk_xdp_progs *xdp_progs;
--
2.43.0
^ permalink raw reply related
* [PATCH 3/5] selftests/xsk: Remove umem from pkt_generate parameters
From: Tushar Vyavahare @ 2026-04-28 8:03 UTC (permalink / raw)
To: netdev, magnus.karlsson, maciej.fijalkowski, stfomichev,
kernelxing, davem, kuba, pabeni, ast, daniel, tirthendu.sarkar,
tushar.vyavahare
Cc: bpf
In-Reply-To: <20260428080317.702124-1-tushar.vyavahare@intel.com>
After moving UMEM ownership to socket-level state, passing UMEM
explicitly into pkt_generate() is redundant and can drift from the
active socket state.
Drop the UMEM parameter and use the socket-attached UMEM directly. This
simplifies call sites and keeps packet generation aligned with the
configured socket/UMEM pairing.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Tushar Vyavahare <tushar.vyavahare@intel.com>
---
tools/testing/selftests/bpf/prog_tests/test_xsk.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xsk.c b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
index ad5ada9216fc..ea229ea446ca 100644
--- a/tools/testing/selftests/bpf/prog_tests/test_xsk.c
+++ b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
@@ -630,10 +630,10 @@ static void pkt_stream_cancel(struct pkt_stream *pkt_stream)
pkt_stream->current_pkt_nb--;
}
-static void pkt_generate(struct xsk_socket_info *xsk, struct xsk_umem_info *umem, u64 addr, u32 len,
- u32 pkt_nb, u32 bytes_written)
+static void pkt_generate(struct xsk_socket_info *xsk, u64 addr, u32 len, u32 pkt_nb,
+ u32 bytes_written)
{
- void *data = xsk_umem__get_data(umem->buffer, addr);
+ void *data = xsk_umem__get_data(xsk->umem->buffer, addr);
if (len < MIN_PKT_SIZE)
return;
@@ -1241,7 +1241,7 @@ static int __send_pkts(struct ifobject *ifobject, struct xsk_socket_info *xsk, b
tx_desc->options = 0;
}
if (pkt->valid)
- pkt_generate(xsk, umem, tx_desc->addr, tx_desc->len, pkt->pkt_nb,
+ pkt_generate(xsk, tx_desc->addr, tx_desc->len, pkt->pkt_nb,
bytes_written);
bytes_written += tx_desc->len;
--
2.43.0
^ permalink raw reply related
* [PATCH 4/5] selftests/xsk: Use umem_size() helper consistently
From: Tushar Vyavahare @ 2026-04-28 8:03 UTC (permalink / raw)
To: netdev, magnus.karlsson, maciej.fijalkowski, stfomichev,
kernelxing, davem, kuba, pabeni, ast, daniel, tirthendu.sarkar,
tushar.vyavahare
Cc: bpf
In-Reply-To: <20260428080317.702124-1-tushar.vyavahare@intel.com>
Replace remaining open-coded `umem->num_frames * umem->frame_size`
calculations in test_xsk.c with the existing `umem_size()` helper.
This keeps UMEM size computation centralized, avoids duplicated arithmetic,
and improves readability with no intended behavior change.
Signed-off-by: Tushar Vyavahare <tushar.vyavahare@intel.com>
---
.../selftests/bpf/prog_tests/test_xsk.c | 26 +++++++++----------
1 file changed, 13 insertions(+), 13 deletions(-)
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xsk.c b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
index ea229ea446ca..ba4768f83eb6 100644
--- a/tools/testing/selftests/bpf/prog_tests/test_xsk.c
+++ b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
@@ -828,12 +828,12 @@ static bool is_frag_valid(struct xsk_umem_info *umem, u64 addr, u32 len, u32 exp
u32 bytes_processed)
{
u32 seqnum, pkt_nb, *pkt_data, words_to_end, expected_seqnum;
+ u64 umem_sz = umem_size(umem);
void *data = xsk_umem__get_data(umem->buffer, addr);
addr -= umem->base_addr;
- if (addr >= umem->num_frames * umem->frame_size ||
- addr + len > umem->num_frames * umem->frame_size) {
+ if (addr >= umem_sz || addr + len > umem_sz) {
ksft_print_msg("Frag invalid addr: %llx len: %u\n",
(unsigned long long)addr, len);
return false;
@@ -1581,7 +1581,7 @@ static int thread_common_ops(struct test_spec *test, struct ifobject *ifobject)
int ret;
u32 i;
- umem_sz = umem->num_frames * umem->frame_size;
+ umem_sz = umem_size(umem);
mmap_flags = MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE;
if (umem->unaligned_mode)
@@ -1700,7 +1700,7 @@ void *worker_testapp_validate_rx(void *arg)
static void testapp_clean_xsk_umem(struct ifobject *ifobj)
{
struct xsk_umem_info *umem = ifobj->xsk->umem;
- u64 umem_sz = umem->num_frames * umem->frame_size;
+ u64 umem_sz = umem_size(umem);
if (ifobj->shared_umem)
umem_sz *= 2;
@@ -2091,7 +2091,7 @@ int testapp_send_receive_mb(struct test_spec *test)
int testapp_invalid_desc_mb(struct test_spec *test)
{
struct xsk_umem_info *umem = test->ifobj_tx->xsk->umem;
- u64 umem_size = umem->num_frames * umem->frame_size;
+ u64 umem_sz = umem_size(umem);
struct pkt pkts[] = {
/* Valid packet for synch to start with */
{0, MIN_PKT_SIZE, 0, true, 0},
@@ -2101,7 +2101,7 @@ int testapp_invalid_desc_mb(struct test_spec *test)
{0, 0, 0, false, 0},
/* Invalid address in the second frame */
{0, XSK_UMEM__LARGE_FRAME_SIZE, 0, false, XDP_PKT_CONTD},
- {umem_size, XSK_UMEM__LARGE_FRAME_SIZE, 0, false, XDP_PKT_CONTD},
+ {umem_sz, XSK_UMEM__LARGE_FRAME_SIZE, 0, false, XDP_PKT_CONTD},
/* Invalid len in the middle */
{0, XSK_UMEM__LARGE_FRAME_SIZE, 0, false, XDP_PKT_CONTD},
{0, XSK_UMEM__INVALID_FRAME_SIZE, 0, false, XDP_PKT_CONTD},
@@ -2132,7 +2132,7 @@ int testapp_invalid_desc_mb(struct test_spec *test)
int testapp_invalid_desc(struct test_spec *test)
{
struct xsk_umem_info *umem = test->ifobj_tx->xsk->umem;
- u64 umem_size = umem->num_frames * umem->frame_size;
+ u64 umem_sz = umem_size(umem);
struct pkt pkts[] = {
/* Zero packet address allowed */
{0, MIN_PKT_SIZE, 0, true},
@@ -2143,11 +2143,11 @@ int testapp_invalid_desc(struct test_spec *test)
/* Packet too large */
{0, XSK_UMEM__INVALID_FRAME_SIZE, 0, false},
/* Up to end of umem allowed */
- {umem_size - MIN_PKT_SIZE - 2 * umem->frame_size, MIN_PKT_SIZE, 0, true},
+ {umem_sz - MIN_PKT_SIZE - 2 * umem->frame_size, MIN_PKT_SIZE, 0, true},
/* After umem ends */
- {umem_size, MIN_PKT_SIZE, 0, false},
+ {umem_sz, MIN_PKT_SIZE, 0, false},
/* Straddle the end of umem */
- {umem_size - MIN_PKT_SIZE / 2, MIN_PKT_SIZE, 0, false},
+ {umem_sz - MIN_PKT_SIZE / 2, MIN_PKT_SIZE, 0, false},
/* Straddle a 4K boundary */
{0x1000 - MIN_PKT_SIZE / 2, MIN_PKT_SIZE, 0, false},
/* Straddle a 2K boundary */
@@ -2165,9 +2165,9 @@ int testapp_invalid_desc(struct test_spec *test)
}
if (test->ifobj_tx->shared_umem) {
- pkts[4].offset += umem_size;
- pkts[5].offset += umem_size;
- pkts[6].offset += umem_size;
+ pkts[4].offset += umem_sz;
+ pkts[5].offset += umem_sz;
+ pkts[6].offset += umem_sz;
}
if (pkt_stream_generate_custom(test, pkts, ARRAY_SIZE(pkts)))
--
2.43.0
^ permalink raw reply related
* [PATCH 5/5] selftests/xsk: Introduce mmap_size in umem struct
From: Tushar Vyavahare @ 2026-04-28 8:03 UTC (permalink / raw)
To: netdev, magnus.karlsson, maciej.fijalkowski, stfomichev,
kernelxing, davem, kuba, pabeni, ast, daniel, tirthendu.sarkar,
tushar.vyavahare
Cc: bpf
In-Reply-To: <20260428080317.702124-1-tushar.vyavahare@intel.com>
UMEM teardown currently recomputes the `munmap()` length
from frame geometry (`num_frames * frame_size`, shared-UMEM
multiplier, and hugepage rounding).
That duplicates setup logic in cleanup and makes teardown depend on
re-deriving values instead of using the exact mapping metadata.
Store the finalized mapping length in `xsk_umem_info` as `mmap_size` when
the UMEM mapping is created, and use that value during teardown.
This removes duplicated size arithmetic in cleanup and guarantees that
`munmap()` uses the same canonical size that was established at setup time.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Tushar Vyavahare <tushar.vyavahare@intel.com>
---
tools/testing/selftests/bpf/prog_tests/test_xsk.c | 9 +++------
tools/testing/selftests/bpf/prog_tests/test_xsk.h | 1 +
2 files changed, 4 insertions(+), 6 deletions(-)
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xsk.c b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
index ba4768f83eb6..ba1bcfc6b74b 100644
--- a/tools/testing/selftests/bpf/prog_tests/test_xsk.c
+++ b/tools/testing/selftests/bpf/prog_tests/test_xsk.c
@@ -1594,6 +1594,8 @@ static int thread_common_ops(struct test_spec *test, struct ifobject *ifobject)
if (bufs == MAP_FAILED)
return -errno;
+ umem->mmap_size = ceil_u64(umem_sz, HUGEPAGE_SIZE) * HUGEPAGE_SIZE;
+
ret = xsk_configure_umem(ifobject, umem, bufs, umem_sz);
if (ret)
return ret;
@@ -1700,14 +1702,9 @@ void *worker_testapp_validate_rx(void *arg)
static void testapp_clean_xsk_umem(struct ifobject *ifobj)
{
struct xsk_umem_info *umem = ifobj->xsk->umem;
- u64 umem_sz = umem_size(umem);
-
- if (ifobj->shared_umem)
- umem_sz *= 2;
- umem_sz = ceil_u64(umem_sz, HUGEPAGE_SIZE) * HUGEPAGE_SIZE;
xsk_umem__delete(umem->umem);
- munmap(umem->buffer, umem_sz);
+ munmap(umem->buffer, umem->mmap_size);
}
static void handler(int signum)
diff --git a/tools/testing/selftests/bpf/prog_tests/test_xsk.h b/tools/testing/selftests/bpf/prog_tests/test_xsk.h
index accda2456fe6..0355a9d06b4e 100644
--- a/tools/testing/selftests/bpf/prog_tests/test_xsk.h
+++ b/tools/testing/selftests/bpf/prog_tests/test_xsk.h
@@ -85,6 +85,7 @@ struct xsk_umem_info {
struct xsk_ring_cons cq;
struct xsk_umem *umem;
u64 next_buffer;
+ u64 mmap_size;
u32 num_frames;
u32 frame_headroom;
void *buffer;
--
2.43.0
^ permalink raw reply related
* RE: [Intel-wired-lan] [PATCH iwl-next v7 07/14] idpf: refactor idpf to use libie_pci APIs
From: Loktionov, Aleksandr @ 2026-04-28 8:47 UTC (permalink / raw)
To: Zaremba, Larysa, intel-wired-lan@lists.osuosl.org,
Nguyen, Anthony L
Cc: Lobakin, Aleksander, Samudrala, Sridhar, Michal Swiatkowski,
Zaremba, Larysa, Fijalkowski, Maciej, Tantilov, Emil S,
Chittim, Madhu, Hay, Joshua A, Keller, Jacob E,
Shanmugam, Jayaprakash, Jiri Pirko, David S. Miller, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Simon Horman, Jonathan Corbet,
Richard Cochran, Kitszel, Przemyslaw, Andrew Lunn,
netdev@vger.kernel.org, linux-doc@vger.kernel.org,
linux-kernel@vger.kernel.org, Salin, Samuel
In-Reply-To: <20260428082654.44364-8-larysa.zaremba@intel.com>
> -----Original Message-----
> From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf
> Of Larysa Zaremba
> Sent: Tuesday, April 28, 2026 10:27 AM
> To: intel-wired-lan@lists.osuosl.org; Nguyen, Anthony L
> <anthony.l.nguyen@intel.com>
> Cc: Lobakin, Aleksander <aleksander.lobakin@intel.com>; Samudrala,
> Sridhar <sridhar.samudrala@intel.com>; Michal Swiatkowski
> <michal.swiatkowski@linux.intel.com>; Zaremba, Larysa
> <larysa.zaremba@intel.com>; Fijalkowski, Maciej
> <maciej.fijalkowski@intel.com>; Tantilov, Emil S
> <emil.s.tantilov@intel.com>; Chittim, Madhu <madhu.chittim@intel.com>;
> Hay, Joshua A <joshua.a.hay@intel.com>; Keller, Jacob E
> <jacob.e.keller@intel.com>; Shanmugam, Jayaprakash
> <jayaprakash.shanmugam@intel.com>; Jiri Pirko <jiri@resnulli.us>;
> David S. Miller <davem@davemloft.net>; Eric Dumazet
> <edumazet@google.com>; Jakub Kicinski <kuba@kernel.org>; Paolo Abeni
> <pabeni@redhat.com>; Simon Horman <horms@kernel.org>; Jonathan Corbet
> <corbet@lwn.net>; Richard Cochran <richardcochran@gmail.com>; Kitszel,
> Przemyslaw <przemyslaw.kitszel@intel.com>; Andrew Lunn
> <andrew+netdev@lunn.ch>; netdev@vger.kernel.org; linux-
> doc@vger.kernel.org; linux-kernel@vger.kernel.org; Salin, Samuel
> <samuel.salin@intel.com>
> Subject: [Intel-wired-lan] [PATCH iwl-next v7 07/14] idpf: refactor
> idpf to use libie_pci APIs
>
> From: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
>
> Use libie_pci init and MMIO APIs where possible, struct idpf_hw cannot
> be deleted for now as it also houses control queues that will be
> refactored later. Use libie_cp header for libie_ctlq_ctx that contains
> mmio info from the start in order to not increase the diff later.
>
> Reviewed-by: Madhu Chittim <madhu.chittim@intel.com>
> Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
> Signed-off-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
> Co-developed-by: Larysa Zaremba <larysa.zaremba@intel.com>
> Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
> Tested-by: Samuel Salin <Samuel.salin@intel.com>
> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
> ---
> drivers/net/ethernet/intel/idpf/Kconfig | 1 +
> drivers/net/ethernet/intel/idpf/idpf.h | 70 +-------
> .../net/ethernet/intel/idpf/idpf_controlq.c | 26 ++-
> .../net/ethernet/intel/idpf/idpf_controlq.h | 2 -
> drivers/net/ethernet/intel/idpf/idpf_dev.c | 61 ++++---
> drivers/net/ethernet/intel/idpf/idpf_idc.c | 38 ++--
> drivers/net/ethernet/intel/idpf/idpf_lib.c | 7 +-
> drivers/net/ethernet/intel/idpf/idpf_main.c | 114 ++++++------
> drivers/net/ethernet/intel/idpf/idpf_vf_dev.c | 57 +++---
> .../net/ethernet/intel/idpf/idpf_virtchnl.c | 169 +++++++++--------
> -
> .../ethernet/intel/idpf/idpf_virtchnl_ptp.c | 58 +++---
> 11 files changed, 288 insertions(+), 315 deletions(-)
>
> diff --git a/drivers/net/ethernet/intel/idpf/Kconfig
> b/drivers/net/ethernet/intel/idpf/Kconfig
> index adab2154125b..586df3a4afe9 100644
> --- a/drivers/net/ethernet/intel/idpf/Kconfig
> +++ b/drivers/net/ethernet/intel/idpf/Kconfig
> @@ -6,6 +6,7 @@ config IDPF
> depends on PCI_MSI
> depends on PTP_1588_CLOCK_OPTIONAL
> select DIMLIB
> + select LIBIE_CP
> select LIBETH_XDP
> help
> This driver supports Intel(R) Infrastructure Data Path
> Function diff --git a/drivers/net/ethernet/intel/idpf/idpf.h
> b/drivers/net/ethernet/intel/idpf/idpf.h
> index 0d08f51be7e3..efdb58990a8b 100644
> --- a/drivers/net/ethernet/intel/idpf/idpf.h
> +++ b/drivers/net/ethernet/intel/idpf/idpf.h
> @@ -23,6 +23,7 @@ struct idpf_rss_data;
>
> #include <linux/intel/iidc_rdma.h>
> #include <linux/intel/iidc_rdma_idpf.h>
> +#include <linux/intel/libie/controlq.h>
> #include <linux/intel/virtchnl2.h>
>
> #include "idpf_txrx.h"
> @@ -625,6 +626,7 @@ struct idpf_vc_xn_manager;
> * @flags: See enum idpf_flags
> * @reset_reg: See struct idpf_reset_reg
> * @hw: Device access data
...
> for (i = 0; i < num_vecs; i++) {
> struct idpf_q_vector *q_vector = &rsrc->q_vectors[i];
> u16 vec_id = rsrc->q_vector_idxs[i] - IDPF_MBX_Q_VEC;
> struct idpf_intr_reg *intr = &q_vector->intr_reg;
> + struct idpf_vec_regs *reg = ®_vals[vec_id];
> u32 spacing;
>
> - intr->dyn_ctl = idpf_get_reg_addr(adapter,
> -
> reg_vals[vec_id].dyn_ctl_reg);
> + intr->dyn_ctl = libie_pci_get_mmio_addr(mmio,
> + reg->dyn_ctl_reg);
Stray TAB after =
> intr->dyn_ctl_intena_m = PF_GLINT_DYN_CTL_INTENA_M;
> intr->dyn_ctl_intena_msk_m =
> PF_GLINT_DYN_CTL_INTENA_MSK_M;
> intr->dyn_ctl_itridx_s = PF_GLINT_DYN_CTL_ITR_INDX_S; @@
...
>
> return 0;
> }
> --
> 2.47.0
^ permalink raw reply
* Re: [PATCH net] tcp: do not shrink window clamp when SO_RCVBUF is locked
From: Ankit Jain @ 2026-04-28 8:47 UTC (permalink / raw)
To: Eric Dumazet
Cc: netdev, davem, dsahern, ncardwell, kuniyu, kuba, pabeni, horms,
quic_stranche, quic_subashab, linux-kernel, karen.badiryan,
ajay.kaher, alexey.makhalov, vamsi-krishna.brahmajosyula,
yin.ding, tapas.kundu, stable
In-Reply-To: <CANn89iJOTxeF30wO7+0GoLmMAZGpCq+JUM5EQe5siNfYzEtZkw@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 1969 bytes --]
On Mon, Apr 27, 2026 at 9:08 PM Eric Dumazet <edumazet@google.com> wrote:
>
> On Mon, Apr 27, 2026 at 8:32 AM Ankit Jain <ankit-aj.jain@broadcom.com> wrote:
> >
> > When an application explicitly sets SO_RCVBUF, the window clamp should
> > not be dynamically recalculated based on the memory scaling_ratio.
> >
> > Currently, tcp_measure_rcv_mss() aggressively crushes the window clamp
> > down when it sees a poor skb->len to skb->truesize ratio. If the
> > application explicitly locked the buffer via SO_RCVBUF, this
> > recalculation causes the advertised window to drop severely.
> >
> > If the window drops below the interface MSS, it triggers Silly Window
> > Syndrome (SWS) avoidance on the sender. The sender defers transmission
> > and drops the connection into a perpetual 200ms PROBE0 timer loop,
> > drastically reducing throughput.
> >
> > This is highly reproducible on loopback interfaces (MTU 65536) using
> > Java-based workloads (like Tomcat/GemFire) where the JVM sets SO_RCVBUF
> > to 32K or 64K. The bloated loopback truesize forces the scaling ratio
> > to drop, crushing the window clamp to ~26K, instantly triggering SWS
> > stalls and causing gigabyte transfers to take minutes instead of
> > milliseconds.
> >
> > Since the application locked the buffer, the kernel should respect the
> > clamp boundary and not dynamically crush it based on runtime ratios.
> >
> > Fixes: a2cbb1603943 ("tcp: Update window clamping condition")
> > Cc: stable@vger.kernel.org
> > Reported-by: Karen Badiryan <karen.badiryan@broadcom.com>
> > Signed-off-by: Ankit Jain <ankit-aj.jain@broadcom.com>
>
> Make sure to add a selftests (in ./tools/testing/selftests/net/packetdrill/ )
>
> Thanks.
Hi Eric,
Thank you for the review.
I will formalize the packetdrill sequence from the commit notes into
a proper runnable .pkt selftest. I will include it as the second patch
in the upcoming v2 series.
Thanks,
Ankit
[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 5420 bytes --]
^ permalink raw reply
* Re: [PATCH] net: phonet: fix BUG_ON() in pn_socket_autobind()
From: Deepanshu Kartikey @ 2026-04-28 8:48 UTC (permalink / raw)
To: Jakub Kicinski
Cc: pabeni, netdev, syzbot+706f5eb79044e686c794, horms, linux-kernel,
edumazet, davem, courmisch
In-Reply-To: <20260428005433.716529-1-kuba@kernel.org>
On Tue, Apr 28, 2026 at 6:24 AM Jakub Kicinski <kuba@kernel.org> wrote:
>
> This is an AI-generated review of your patch. The human sending this
> email has considered the AI review valid, or at least plausible.
>
> For local reproduction steps see https://netdev-ai.bots.linux.dev/ai-local.html
> ---
> net: phonet: fix BUG_ON() in pn_socket_autobind()
>
> This commit fixes a kernel crash in pn_socket_autobind() by checking
> the socket's bound state directly before calling pn_socket_bind(),
> eliminating ambiguous -EINVAL interpretation.
>
> Closes: https://syzkaller.appspot.com/bug?extid=706f5eb79044e686c794
>
> > pn_socket_autobind() calls pn_socket_bind() and treats
> > -EINVAL as a signal that the socket was already bound,
> > then uses BUG_ON() to verify it:
> >
> > if (err != -EINVAL)
> > return err;
> > BUG_ON(!pn_port(pn_sk(sock->sk)->sobject));
> >
> > However, pn_socket_bind() returns -EINVAL in multiple
> > cases:
> >
> > 1. address length too short
>
> The commit message lists "address length too short" as a cause of the
> BUG_ON crash, but is this accurate?
>
> Looking at pn_socket_autobind(), it always passes
> len=sizeof(struct sockaddr_pn) to pn_socket_bind():
>
> err = pn_socket_bind(sock, (struct sockaddr_unsized *)&sa,
> sizeof(struct sockaddr_pn));
>
> While pn_socket_bind() does have this check:
>
> if (len < sizeof(struct sockaddr_pn))
> return -EINVAL;
>
> This case cannot be triggered through pn_socket_autobind() since it
> always passes the correct length. The BUG_ON is only reachable through
> pn_socket_autobind(), so listing this as a cause seems misleading.
>
> Should the commit message be updated to remove case 1, or clarify that
> it's describing pn_socket_bind() in general rather than paths that can
> trigger the BUG_ON?
>
> > 2. socket not in TCP_CLOSE state
>
> Can case 2 (socket not in TCP_CLOSE state) actually trigger the BUG_ON
> crash with sobject still zero?
>
> The combined check in pn_socket_bind() is:
>
> if (sk->sk_state != TCP_CLOSE || pn_port(pn->sobject)) {
> err = -EINVAL;
> goto out;
> }
>
> For the BUG_ON to fire, we need sk_state != TCP_CLOSE while sobject
> remains zero (unbound). Looking at phonet state transitions:
>
> - pn_socket_connect() binds the socket before changing state
> - pn_socket_listen() binds the socket before changing state
> - pep_sock_accept() creates sockets with state=TCP_SYN_RECV but also
> sets sobject
> - pep_ioctl(SIOCPNENABLEPIPE) checks sobject before changing state
>
> Is there a code path that changes sk_state to non-TCP_CLOSE without
> first binding the socket? The syzbot crash is real, but the mechanism
> described couldn't be verified. Could this involve a race condition or
> a different code path?
>
> > 3. socket already bound <- only intended case
> >
> > When -EINVAL comes from cases 1 or 2, sobject is still
> > zero (never assigned), causing BUG_ON to fire and crash
> > the kernel.
>
> Given the issues with cases 1 and 2 above, is the explanation of the
> crash mechanism accurate? Case 1 cannot happen from pn_socket_autobind(),
> and case 2's ability to leave sobject at zero while triggering -EINVAL
> could not be verified.
>
> The syzbot crash is confirmed (RIP at pn_socket_autobind
> net/phonet/socket.c:213 via pn_socket_sendmsg), but could the commit
> message more accurately describe how the crash actually occurs?
>
> > diff --git a/net/phonet/socket.c b/net/phonet/socket.c
> > index c4af26357144..5a55e7d14e85 100644
> > --- a/net/phonet/socket.c
> > +++ b/net/phonet/socket.c
> > @@ -204,14 +204,14 @@ static int pn_socket_autobind(struct socket *sock)
> > struct sockaddr_pn sa;
> > int err;
> >
> > + if (pn_port(pn_sk(sock->sk)->sobject))
> > + return 0; /* socket was already bound */
> > +
> > memset(&sa, 0, sizeof(sa));
> > sa.spn_family = AF_PHONET;
> > err = pn_socket_bind(sock, (struct sockaddr_unsized *)&sa,
> > sizeof(struct sockaddr_pn));
> > - if (err != -EINVAL)
> > - return err;
> > - BUG_ON(!pn_port(pn_sk(sock->sk)->sobject));
> > - return 0; /* socket was already bound */
> > + return err;
> > }
>
> The fix itself is correct and removes the dangerous BUG_ON(). However,
> the pn_port(sobject) check is performed without holding any lock. Could
> a concurrent pn_socket_bind() race with this check? This is likely
> benign since pn_socket_bind() re-checks under lock_sock(), but is this
> the intended behavior?
> --
> pw-bot: cr
Hi Jakub,
Thank you for the detailed review!
You are correct on all three points. Let me address
each one:
1. Case 1 ("address length too short") is indeed
impossible from pn_socket_autobind() since it
always passes sizeof(struct sockaddr_pn). I will
remove this from the commit message in v2.
2. Regarding case 2 (sk_state != TCP_CLOSE with
sobject == 0), I verified this using the syzbot
C reproducer. The exact path is:
r0 = socket$kcm(AF_KCM, SOCK_SEQPACKET, 0)
listen(r0, ...)
→ sk_state = TCP_LISTEN, sobject = 0
r3 = accept4(r0, ...)
→ creates NEW socket r3
→ sk_state = TCP_SYN_RECV (not TCP_CLOSE!)
→ sobject = 0 (never bound!)
sendmsg(r3)
→ pn_socket_autobind(r3)
→ pn_socket_bind() sees sk_state != TCP_CLOSE
while sobject is still 0
→ returns -EINVAL
→ BUG_ON fires!
sobject is only set in two places:
a. successful pn_socket_bind() → get_port() assigns port
b. pep_sock_accept() → copied from request
Since r3 comes from accept4() on a KCM socket
(not a PEP phonet socket), pep_sock_accept() is
never called and sobject stays zero. So sk_state
!= TCP_CLOSE and sobject == 0 can indeed be true
simultaneously.
3. Regarding the missing lock: the check:
if (pn_port(pn_sk(sock->sk)->sobject))
return 0;
is performed without lock_sock(). A concurrent
pn_socket_bind() could race with this read.
However since pn_socket_bind() re-checks
pn_port(sobject) under lock_sock(), any racing
bind would be caught there. So the race is
benign. I can add a comment clarifying this
if desired.
I will send a v2 with an updated commit message
reflecting the above.
Thanks,
Deepanshu Kartikey
^ permalink raw reply
* RE: [Intel-wired-lan] [PATCH iwl-next v7 13/14] ixd: add the core initialization
From: Loktionov, Aleksandr @ 2026-04-28 8:50 UTC (permalink / raw)
To: Zaremba, Larysa, intel-wired-lan@lists.osuosl.org,
Nguyen, Anthony L
Cc: Lobakin, Aleksander, Samudrala, Sridhar, Michal Swiatkowski,
Zaremba, Larysa, Fijalkowski, Maciej, Tantilov, Emil S,
Chittim, Madhu, Hay, Joshua A, Keller, Jacob E,
Shanmugam, Jayaprakash, Jiri Pirko, David S. Miller, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Simon Horman, Jonathan Corbet,
Richard Cochran, Kitszel, Przemyslaw, Andrew Lunn,
netdev@vger.kernel.org, linux-doc@vger.kernel.org,
linux-kernel@vger.kernel.org, R, Bharath
In-Reply-To: <20260428082654.44364-14-larysa.zaremba@intel.com>
> -----Original Message-----
> From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf
> Of Larysa Zaremba
> Sent: Tuesday, April 28, 2026 10:27 AM
> To: intel-wired-lan@lists.osuosl.org; Nguyen, Anthony L
> <anthony.l.nguyen@intel.com>
> Cc: Lobakin, Aleksander <aleksander.lobakin@intel.com>; Samudrala,
> Sridhar <sridhar.samudrala@intel.com>; Michal Swiatkowski
> <michal.swiatkowski@linux.intel.com>; Zaremba, Larysa
> <larysa.zaremba@intel.com>; Fijalkowski, Maciej
> <maciej.fijalkowski@intel.com>; Tantilov, Emil S
> <emil.s.tantilov@intel.com>; Chittim, Madhu <madhu.chittim@intel.com>;
> Hay, Joshua A <joshua.a.hay@intel.com>; Keller, Jacob E
> <jacob.e.keller@intel.com>; Shanmugam, Jayaprakash
> <jayaprakash.shanmugam@intel.com>; Jiri Pirko <jiri@resnulli.us>;
> David S. Miller <davem@davemloft.net>; Eric Dumazet
> <edumazet@google.com>; Jakub Kicinski <kuba@kernel.org>; Paolo Abeni
> <pabeni@redhat.com>; Simon Horman <horms@kernel.org>; Jonathan Corbet
> <corbet@lwn.net>; Richard Cochran <richardcochran@gmail.com>; Kitszel,
> Przemyslaw <przemyslaw.kitszel@intel.com>; Andrew Lunn
> <andrew+netdev@lunn.ch>; netdev@vger.kernel.org; linux-
> doc@vger.kernel.org; linux-kernel@vger.kernel.org; R, Bharath
> <bharath.r@intel.com>
> Subject: [Intel-wired-lan] [PATCH iwl-next v7 13/14] ixd: add the core
> initialization
>
> As the mailbox is setup, initialize the core. This makes use of the
> send and receive mailbox message framework for virtchnl communication
> between the driver and device Control Plane (CP).
>
> To start with, driver confirms the virtchnl version with the CP. Once
> that is done, it requests and gets the required capabilities and
> resources needed such as max vectors, queues, vports etc.
>
> Use a unified way of handling the virtchnl messages, where a single
> function handles all related memory management and the caller only
> provides the callbacks to fill the send buffer and to handle the
> response.
>
> Place generic control queue message handling separately to facilitate
> the addition of protocols other than virtchannel in the future.
>
> Co-developed-by: Amritha Nambiar <amritha.nambiar@intel.com>
> Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
> Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
> Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
> Tested-by: Bharath R <Bharath.r@intel.com>
> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
> ---
> drivers/net/ethernet/intel/ixd/Makefile | 2 +
> drivers/net/ethernet/intel/ixd/ixd.h | 10 +
> drivers/net/ethernet/intel/ixd/ixd_ctlq.c | 149 +++++++++++++++
> drivers/net/ethernet/intel/ixd/ixd_ctlq.h | 33 ++++
> drivers/net/ethernet/intel/ixd/ixd_lib.c | 25 ++-
> drivers/net/ethernet/intel/ixd/ixd_main.c | 3 +
> drivers/net/ethernet/intel/ixd/ixd_virtchnl.c | 178
> ++++++++++++++++++ drivers/net/ethernet/intel/ixd/ixd_virtchnl.h |
> 12 ++
> 8 files changed, 411 insertions(+), 1 deletion(-) create mode 100644
> drivers/net/ethernet/intel/ixd/ixd_ctlq.c
> create mode 100644 drivers/net/ethernet/intel/ixd/ixd_ctlq.h
> create mode 100644 drivers/net/ethernet/intel/ixd/ixd_virtchnl.c
> create mode 100644 drivers/net/ethernet/intel/ixd/ixd_virtchnl.h
>
> diff --git a/drivers/net/ethernet/intel/ixd/Makefile
> b/drivers/net/ethernet/intel/ixd/Makefile
> index 164b2c86952f..90abf231fb16 100644
> --- a/drivers/net/ethernet/intel/ixd/Makefile
> +++ b/drivers/net/ethernet/intel/ixd/Makefile
> @@ -6,5 +6,7 @@
> obj-$(CONFIG_IXD) += ixd.o
>
> ixd-y := ixd_main.o
> +ixd-y += ixd_ctlq.o
> ixd-y += ixd_dev.o
> ixd-y += ixd_lib.o
> +ixd-y += ixd_virtchnl.o
> diff --git a/drivers/net/ethernet/intel/ixd/ixd.h
> b/drivers/net/ethernet/intel/ixd/ixd.h
> index 99c44f2aa659..98d1f22534b5 100644
> --- a/drivers/net/ethernet/intel/ixd/ixd.h
> +++ b/drivers/net/ethernet/intel/ixd/ixd.h
> @@ -10,19 +10,29 @@
> * struct ixd_adapter - Data structure representing a CPF
> * @cp_ctx: Control plane communication context
> * @init_task: Delayed initialization after reset
...
> diff --git a/drivers/net/ethernet/intel/ixd/ixd_ctlq.h
> b/drivers/net/ethernet/intel/ixd/ixd_ctlq.h
> new file mode 100644
> index 000000000000..f450a3a0828f
> --- /dev/null
> +++ b/drivers/net/ethernet/intel/ixd/ixd_ctlq.h
> @@ -0,0 +1,33 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/* Copyright (C) 2025 Intel Corporation */
> +
> +#ifndef _IXD_CTLQ_H_
> +#define _IXD_CTLQ_H_
> +
> +#include "linux/intel/virtchnl2.h"
Every other file use #include <linux/intel/virtchnl2.h> why do you need #include "linux/intel/virtchnl2.h" here?
Please explain
> +
> +#define IXD_CTLQ_TIMEOUT 2000
> +
...
> --
> 2.47.0
^ permalink raw reply
* RE: [Intel-wired-lan] [PATCH iwl-next v7 14/14] ixd: add devlink support
From: Loktionov, Aleksandr @ 2026-04-28 8:53 UTC (permalink / raw)
To: Zaremba, Larysa, intel-wired-lan@lists.osuosl.org,
Nguyen, Anthony L
Cc: Lobakin, Aleksander, Samudrala, Sridhar, Michal Swiatkowski,
Zaremba, Larysa, Fijalkowski, Maciej, Tantilov, Emil S,
Chittim, Madhu, Hay, Joshua A, Keller, Jacob E,
Shanmugam, Jayaprakash, Jiri Pirko, David S. Miller, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Simon Horman, Jonathan Corbet,
Richard Cochran, Kitszel, Przemyslaw, Andrew Lunn,
netdev@vger.kernel.org, linux-doc@vger.kernel.org,
linux-kernel@vger.kernel.org, R, Bharath
In-Reply-To: <20260428082654.44364-15-larysa.zaremba@intel.com>
> -----Original Message-----
> From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf
> Of Larysa Zaremba
> Sent: Tuesday, April 28, 2026 10:27 AM
> To: intel-wired-lan@lists.osuosl.org; Nguyen, Anthony L
> <anthony.l.nguyen@intel.com>
> Cc: Lobakin, Aleksander <aleksander.lobakin@intel.com>; Samudrala,
> Sridhar <sridhar.samudrala@intel.com>; Michal Swiatkowski
> <michal.swiatkowski@linux.intel.com>; Zaremba, Larysa
> <larysa.zaremba@intel.com>; Fijalkowski, Maciej
> <maciej.fijalkowski@intel.com>; Tantilov, Emil S
> <emil.s.tantilov@intel.com>; Chittim, Madhu <madhu.chittim@intel.com>;
> Hay, Joshua A <joshua.a.hay@intel.com>; Keller, Jacob E
> <jacob.e.keller@intel.com>; Shanmugam, Jayaprakash
> <jayaprakash.shanmugam@intel.com>; Jiri Pirko <jiri@resnulli.us>;
> David S. Miller <davem@davemloft.net>; Eric Dumazet
> <edumazet@google.com>; Jakub Kicinski <kuba@kernel.org>; Paolo Abeni
> <pabeni@redhat.com>; Simon Horman <horms@kernel.org>; Jonathan Corbet
> <corbet@lwn.net>; Richard Cochran <richardcochran@gmail.com>; Kitszel,
> Przemyslaw <przemyslaw.kitszel@intel.com>; Andrew Lunn
> <andrew+netdev@lunn.ch>; netdev@vger.kernel.org; linux-
> doc@vger.kernel.org; linux-kernel@vger.kernel.org; R, Bharath
> <bharath.r@intel.com>
> Subject: [Intel-wired-lan] [PATCH iwl-next v7 14/14] ixd: add devlink
> support
>
> From: Amritha Nambiar <amritha.nambiar@intel.com>
>
> Enable initial support for the devlink interface with the ixd driver.
> The ixd hardware is a single function PCIe device. So, the PCIe
> adapter gets its own devlink instance to manage device-wide resources
> or configuration.
>
> $ devlink dev show
> pci/0000:83:00.6
>
> $ devlink dev info pci/0000:83:00.6
> pci/0000:83:00.6:
> driver ixd
> serial_number 00-a0-c9-ff-ff-23-45-67
> versions:
> fixed:
> device.type MEV
> running:
> virtchnl 2.0
>
> Signed-off-by: Amritha Nambiar <amritha.nambiar@intel.com>
> Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
> Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
> Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
> Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
> Tested-by: Bharath R <Bharath.r@intel.com>
> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
> ---
> Documentation/networking/devlink/index.rst | 1 +
> Documentation/networking/devlink/ixd.rst | 30 ++++++
> drivers/net/ethernet/intel/ixd/Kconfig | 1 +
> drivers/net/ethernet/intel/ixd/Makefile | 1 +
> drivers/net/ethernet/intel/ixd/ixd_devlink.c | 97
> ++++++++++++++++++++ drivers/net/ethernet/intel/ixd/ixd_devlink.h |
> 44 +++++++++
> drivers/net/ethernet/intel/ixd/ixd_main.c | 16 +++-
> 7 files changed, 187 insertions(+), 3 deletions(-) create mode
> 100644 Documentation/networking/devlink/ixd.rst
> create mode 100644 drivers/net/ethernet/intel/ixd/ixd_devlink.c
> create mode 100644 drivers/net/ethernet/intel/ixd/ixd_devlink.h
>
> diff --git a/Documentation/networking/devlink/index.rst
> b/Documentation/networking/devlink/index.rst
> index f7ba7dcf477d..f0c077843fa7 100644
> --- a/Documentation/networking/devlink/index.rst
> +++ b/Documentation/networking/devlink/index.rst
> @@ -88,6 +88,7 @@ parameters, info versions, and other features it
> supports.
> ionic
> iosm
> ixgbe
...
> --- /dev/null
> +++ b/drivers/net/ethernet/intel/ixd/ixd_devlink.h
> @@ -0,0 +1,44 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +/* Copyright (c) 2025, Intel Corporation. */
> +
devlink.c has:
> +// SPDX-License-Identifier: GPL-2.0
But devlink.h has:
> +/* SPDX-License-Identifier: GPL-2.0 */
Why?
> +#ifndef _IXD_DEVLINK_H_
> +#define _IXD_DEVLINK_H_
> +#include <net/devlink.h>
...
> }
>
> static const struct pci_device_id ixd_pci_tbl[] = {
> --
> 2.47.0
^ permalink raw reply
* Re: [PATCH] vsock/virtio: fix memory leak in virtio_transport_recv_listen()
From: Stefano Garzarella @ 2026-04-28 8:57 UTC (permalink / raw)
To: Deepanshu Kartikey
Cc: mst, jasowang, xuanzhuo, eperezma, stefanha, davem, edumazet,
kuba, pabeni, horms, virtualization, kvm, netdev, linux-kernel,
syzbot+1b2c9c4a0f8708082678
In-Reply-To: <CADhLXY6APdmMQLSxN9b66pV5mp5sDwJRR7_BJAh13BaK-qoa=w@mail.gmail.com>
On Tue, Apr 28, 2026 at 01:54:36PM +0530, Deepanshu Kartikey wrote:
>On Mon, Apr 27, 2026 at 7:15 PM Stefano Garzarella <sgarzare@redhat.com> wrote:
>>
>> On Fri, Apr 24, 2026 at 08:33:10PM +0530, Deepanshu Kartikey wrote:
>> >Two bugs exist in virtio_transport_recv_listen():
>>
>> Two bugs, two fixes, two patches usually.
>>
>> >
>> >1. On the transport assignment error path, sk_acceptq_added() is called
>> > but sk_acceptq_removed() is never called when vsock_assign_transport()
>> > fails or assigns a different transport than expected. This causes the
>> > parent listener's accept backlog counter to be permanently inflated,
>> > eventually causing sk_acceptq_is_full() to reject legitimate incoming
>> > connections.
>>
>> Wait, I can't see this issue. sk_acceptq_added() is called after the
>> vsock_assign_transport(), so why we should call sk_acceptq_removed()
>> in the error path of vsock_assign_transport()?
>>
>> Maybe I'm missing something.
>>
>> >
>> >2. There is a race between __vsock_release() and vsock_enqueue_accept().
>> > __vsock_release() sets sk->sk_shutdown to SHUTDOWN_MASK and flushes
>> > the accept queue under the parent socket lock. However,
>> > virtio_transport_recv_listen() checks sk_shutdown and subsequently
>> > calls vsock_enqueue_accept() without holding the parent socket lock.
>>
>> Are you sure about this?
>>
>> virtio_transport_recv_listen() is called only by
>> virtio_transport_recv_pkt() after calling lock_sock(sk), so I'm really
>> confused.
>>
>> > This means a child socket can be enqueued after __vsock_release() has
>> > already flushed the queue, causing the child socket and its associated
>> > resources to leak
>> > permanently. The existing comment in the code hints at this race but
>> > the fix was never implemented.
>>
>> Are you referring to:
>> /* __vsock_release() might have already flushed accept_queue.
>> * Subsequent enqueues would lead to a memory leak.
>> */
>> if (sk->sk_shutdown == SHUTDOWN_MASK) {
>> virtio_transport_reset_no_sock(t, skb, sock_net(sk));
>> return -ESHUTDOWN;
>> }
>>
>> In this case I think we are saying that we are doing this check exactly
>> to avoid that issue.
>>
>> >
>> >Fix both issues: add sk_acceptq_removed() on the transport error path,
>>
>> Again, better to fix the 2 issues with 2 patches (same series is fine).
>>
>> >and re-check sk->sk_shutdown under the parent socket lock before calling
>> >vsock_enqueue_accept() to close the race window. The child socket lock
>> >is released before acquiring the parent socket lock to maintain correct
>> >lock ordering (parent before child).
>> >
>>
>> We are missing the Fixes tag, and I think we can target the `net` tree
>> with this patch (i.e. [PATCH net]), see:
>> https://www.kernel.org/doc/html/next/process/maintainer-netdev.html
>>
>> >Reported-by: syzbot+1b2c9c4a0f8708082678@syzkaller.appspotmail.com
>> >Closes: https://syzkaller.appspot.com/bug?extid=1b2c9c4a0f8708082678
>> >Tested-by: syzbot+1b2c9c4a0f8708082678@syzkaller.appspotmail.com
>> >Signed-off-by: Deepanshu Kartikey <kartikey406@gmail.com>
>> >---
>> > net/vmw_vsock/virtio_transport_common.c | 13 +++++++++++--
>> > 1 file changed, 11 insertions(+), 2 deletions(-)
>> >
>> >diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
>> >index 416d533f493d..fad5fa4a4296 100644
>> >--- a/net/vmw_vsock/virtio_transport_common.c
>> >+++ b/net/vmw_vsock/virtio_transport_common.c
>> >@@ -1578,6 +1578,7 @@ virtio_transport_recv_listen(struct sock *sk, struct sk_buff *skb,
>> > */
>> > if (ret || vchild->transport != &t->transport) {
>> > release_sock(child);
>> >+ sk_acceptq_removed(sk);
>>
>> Ditto, are we sure about this?
>>
>> > virtio_transport_reset_no_sock(t, skb, sock_net(sk));
>> > sock_put(child);
>> > return ret;
>> >@@ -1588,11 +1589,19 @@ virtio_transport_recv_listen(struct sock *sk, struct sk_buff *skb,
>> > child->sk_write_space(child);
>> >
>> > vsock_insert_connected(vchild);
>> >+ release_sock(child);
>> >+ lock_sock(sk);
>>
>> IMO this is a deadlock with the lock_sock(sk) called by the caller.
>>
>> Also a comment would be helpful here to explain why we're doing this.
>>
>> >+ if (sk->sk_shutdown == SHUTDOWN_MASK) {
>> >+ release_sock(sk);
>> >+ sk_acceptq_removed(sk);
>> >+ virtio_transport_reset_no_sock(t, skb, sock_net(sk));
>> >+ sock_put(child);
>> >+ return -ESHUTDOWN;
>>
>> Since this is very similar to the error path of
>> vsock_assign_transport(), I think it would be better to start by
>> defining a common error path for the function and use labels to exit, so
>> we can avoid duplicating the code multiple times.
>>
>> >+ }
>> > vsock_enqueue_accept(sk, child);
>> >+ release_sock(sk);
>> > virtio_transport_send_response(vchild, skb);
>> >
>> >- release_sock(child);
>> >-
>>
>> TBH I'm really worried about this patch since both fixes are completely
>> wrong IMO.
>>
>> Thanks,
>> Stefano
>>
>> > sk->sk_data_ready(sk);
>> > return 0;
>> > }
>> >--
>> >2.43.0
>> >
>>
>
>
>Hi Stefano,
>
>Thank you for the detailed review!
>
>You are correct on both points. I apologize for the confusion — I was
>looking at an older version of the code where sk_acceptq_added() was
>called BEFORE vsock_assign_transport(), which made the
>sk_acceptq_removed() fix appear necessary. In the current kernel,
>sk_acceptq_added() is already moved to after vsock_assign_transport(),
>so that issue no longer exists.
>
>Regarding the lock_sock(sk) fix — you are also correct that
>virtio_transport_recv_pkt() already holds lock_sock(sk) before calling
>virtio_transport_recv_listen(), so our second fix would indeed cause a
>deadlock. I missed that completely.
>
>I am still investigating the root cause of the memory leak reported by
>syzbot. The backtrace points to the vsock loopback path
>(vsock_loopback_work), so I am looking there next. I will send a v2
>once I have a correct analysis and fix.
Okay, thanks for looking into that issue, feel free to chat here, or in
reply to the syzbot report if you have some new findings.
Thanks,
Stefano
^ permalink raw reply
* RE: [Intel-wired-lan] [PATCH iwl-next v7 07/14] idpf: refactor idpf to use libie_pci APIs
From: Loktionov, Aleksandr @ 2026-04-28 9:00 UTC (permalink / raw)
To: Zaremba, Larysa, intel-wired-lan@lists.osuosl.org,
Nguyen, Anthony L
Cc: Lobakin, Aleksander, Samudrala, Sridhar, Michal Swiatkowski,
Zaremba, Larysa, Fijalkowski, Maciej, Tantilov, Emil S,
Chittim, Madhu, Hay, Joshua A, Keller, Jacob E,
Shanmugam, Jayaprakash, Jiri Pirko, David S. Miller, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Simon Horman, Jonathan Corbet,
Richard Cochran, Kitszel, Przemyslaw, Andrew Lunn,
netdev@vger.kernel.org, linux-doc@vger.kernel.org,
linux-kernel@vger.kernel.org, Salin, Samuel
In-Reply-To: <20260428082654.44364-8-larysa.zaremba@intel.com>
> -----Original Message-----
> From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf
> Of Larysa Zaremba
> Sent: Tuesday, April 28, 2026 10:27 AM
> To: intel-wired-lan@lists.osuosl.org; Nguyen, Anthony L
> <anthony.l.nguyen@intel.com>
> Cc: Lobakin, Aleksander <aleksander.lobakin@intel.com>; Samudrala,
> Sridhar <sridhar.samudrala@intel.com>; Michal Swiatkowski
> <michal.swiatkowski@linux.intel.com>; Zaremba, Larysa
> <larysa.zaremba@intel.com>; Fijalkowski, Maciej
> <maciej.fijalkowski@intel.com>; Tantilov, Emil S
> <emil.s.tantilov@intel.com>; Chittim, Madhu <madhu.chittim@intel.com>;
> Hay, Joshua A <joshua.a.hay@intel.com>; Keller, Jacob E
> <jacob.e.keller@intel.com>; Shanmugam, Jayaprakash
> <jayaprakash.shanmugam@intel.com>; Jiri Pirko <jiri@resnulli.us>;
> David S. Miller <davem@davemloft.net>; Eric Dumazet
> <edumazet@google.com>; Jakub Kicinski <kuba@kernel.org>; Paolo Abeni
> <pabeni@redhat.com>; Simon Horman <horms@kernel.org>; Jonathan Corbet
> <corbet@lwn.net>; Richard Cochran <richardcochran@gmail.com>; Kitszel,
> Przemyslaw <przemyslaw.kitszel@intel.com>; Andrew Lunn
> <andrew+netdev@lunn.ch>; netdev@vger.kernel.org; linux-
> doc@vger.kernel.org; linux-kernel@vger.kernel.org; Salin, Samuel
> <samuel.salin@intel.com>
> Subject: [Intel-wired-lan] [PATCH iwl-next v7 07/14] idpf: refactor
> idpf to use libie_pci APIs
>
> From: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
>
> Use libie_pci init and MMIO APIs where possible, struct idpf_hw cannot
> be deleted for now as it also houses control queues that will be
> refactored later. Use libie_cp header for libie_ctlq_ctx that contains
> mmio info from the start in order to not increase the diff later.
>
> Reviewed-by: Madhu Chittim <madhu.chittim@intel.com>
> Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
> Signed-off-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com>
> Co-developed-by: Larysa Zaremba <larysa.zaremba@intel.com>
> Signed-off-by: Larysa Zaremba <larysa.zaremba@intel.com>
> Tested-by: Samuel Salin <Samuel.salin@intel.com>
> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
> ---
> drivers/net/ethernet/intel/idpf/Kconfig | 1 +
> drivers/net/ethernet/intel/idpf/idpf.h | 70 +-------
> .../net/ethernet/intel/idpf/idpf_controlq.c | 26 ++-
> .../net/ethernet/intel/idpf/idpf_controlq.h | 2 -
> drivers/net/ethernet/intel/idpf/idpf_dev.c | 61 ++++---
> drivers/net/ethernet/intel/idpf/idpf_idc.c | 38 ++--
> drivers/net/ethernet/intel/idpf/idpf_lib.c | 7 +-
> drivers/net/ethernet/intel/idpf/idpf_main.c | 114 ++++++------
> drivers/net/ethernet/intel/idpf/idpf_vf_dev.c | 57 +++---
> .../net/ethernet/intel/idpf/idpf_virtchnl.c | 169 +++++++++--------
> -
> .../ethernet/intel/idpf/idpf_virtchnl_ptp.c | 58 +++---
> 11 files changed, 288 insertions(+), 315 deletions(-)
>
> diff --git a/drivers/net/ethernet/intel/idpf/Kconfig
> b/drivers/net/ethernet/intel/idpf/Kconfig
> index adab2154125b..586df3a4afe9 100644
> --- a/drivers/net/ethernet/intel/idpf/Kconfig
> +++ b/drivers/net/ethernet/intel/idpf/Kconfig
> @@ -6,6 +6,7 @@ config IDPF
> depends on PCI_MSI
> depends on PTP_1588_CLOCK_OPTIONAL
> select DIMLIB
...
> >dev_ops.static_reg_info;
> + bool is_static = false;
> +
> + for (uint j = 0; j < IDPF_MMIO_REG_NUM_STATIC; j++)
I think you need to use here unsigned int, didn't checkpach.pl complain about it?
> + if (mr->offset == static_regs[j].start)
> + is_static = true;
...
>
> return 0;
> }
> --
> 2.47.0
^ permalink raw reply
* [PATCH net] netfilter: nf_nat: avoid invalid nat_net pointer use on failed nf_nat_init()
From: Mathias Krause @ 2026-04-28 9:09 UTC (permalink / raw)
To: netfilter-devel
Cc: Pablo Neira Ayuso, Florian Westphal, netdev, Mathias Krause
We ran into below KASAN splat, which is mostly uninteresting, beside
for having nf_nat_register_fn() in the call chain as a cause for the
offending access:
==================================================================
BUG: KASAN: slab-out-of-bounds in nf_nat_register_fn+0x5f9/0x640
Read of size 8 at addr ffff890031e54c20 by task iptables/9510
CPU: 0 UID: 0 PID: 9510 Comm: iptables Not tainted 6.18.18-grsec-full-20260320181326 #1 PREEMPT(voluntary)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
Call Trace:
<TASK>
[…] dump_stack_lvl+0xee/0x160 ffff88004117eeb8
[…] print_report+0x6e/0x640 ffff88004117eee0
[…] ? __phys_addr+0x8e/0x140 ffff88004117eef0
[…] ? kasan_addr_to_slab+0x51/0xe0 ffff88004117ef08
[…] ? complete_report_info+0xec/0x1c0 ffff88004117ef20
[…] ? nf_nat_register_fn+0x5f9/0x640 ffff88004117ef48
[…] kasan_report+0xbc/0x140 ffff88004117ef50
[…] ? nf_nat_register_fn+0x5f9/0x640 ffff88004117ef90
[…] nf_nat_register_fn+0x5f9/0x640 ffff88004117eff8
[…] ? nf_nat_icmp_reply_translation+0x6e0/0x6e0 ffff88004117f070
[…] nf_tables_register_hook.part.0+0xa0/0x220 ffff88004117f080
[…] nf_tables_addchain.constprop.0+0x1054/0x1fc0 ffff88004117f0b8
[…] ? nft_chain_lookup.part.0+0x4ce/0xac0 ffff88004117f130
[…] ? nf_tables_abort+0x3d80/0x3d80 ffff88004117f190
[…] ? nf_tables_dumpreset_obj+0x100/0x100 ffff88004117f1c8
[…] ? nft_table_lookup.part.0+0x255/0x300 ffff88004117f310
[…] ? nf_tables_newchain+0x21a4/0x2fa0 ffff88004117f358
[…] nf_tables_newchain+0x21a4/0x2fa0 ffff88004117f360
[…] ? nf_tables_addchain.constprop.0+0x1fc0/0x1fc0 ffff88004117f458
[…] ? nla_get_range_signed+0x4a0/0x4a0 ffff88004117f488
[…] ? lock_acquire+0x16f/0x320 ffff88004117f490
[…] ? find_held_lock+0x3b/0xe0 ffff88004117f4b0
[…] ? __nla_parse+0x45/0x80 ffff88004117f500
[…] nfnetlink_rcv_batch+0xbca/0x19a0 ffff88004117f550
[…] ? nfnetlink_net_exit_batch+0x120/0x120 ffff88004117f618
[…] ? __sanitizer_cov_trace_switch+0x63/0xe0 ffff88004117f720
[…] ? gr_acl_handle_mmap+0x1c4/0x320 ffff88004117f7c0
[…] ? nla_get_range_signed+0x4a0/0x4a0 ffff88004117f7e8
[…] ? gr_is_capable+0x6f/0xe0 ffff88004117f830
[…] ? __nla_parse+0x45/0x80 ffff88004117f860
[…] ? skb_pull+0x103/0x1a0 ffff88004117f880
[…] nfnetlink_rcv+0x3db/0x4a0 ffff88004117f8b0
[…] ? nfnetlink_rcv_batch+0x19a0/0x19a0 ffff88004117f8d8
[…] ? netlink_lookup+0xe2/0x240 ffff88004117f900
[…] netlink_unicast+0x74b/0xb00 ffff88004117f930
[…] ? netlink_attachskb+0xb20/0xb20 ffff88004117f980
[…] ? __check_object_size+0x3e/0xaa0 ffff88004117f998
[…] ? security_netlink_send+0x51/0x160 ffff88004117f9c8
[…] netlink_sendmsg+0xa03/0x1200 ffff88004117f9f8
[…] ? netlink_unicast+0xb00/0xb00 ffff88004117fa70
[…] ? netlink_unicast+0xb00/0xb00 ffff88004117fac8
[…] ? ____sys_sendmsg+0xe2a/0x1040 ffff88004117faf8
[…] ____sys_sendmsg+0xe2a/0x1040 ffff88004117fb00
[…] ? kernel_recvmsg+0x300/0x300 ffff88004117fb60
[…] ? reacquire_held_locks+0xe9/0x260 ffff88004117fbc8
[…] ___sys_sendmsg+0x138/0x200 ffff88004117fbf8
[…] ? do_recvmmsg+0x7e0/0x7e0 ffff88004117fc30
[…] ? lockdep_hardirqs_on_prepare+0x101/0x1e0 ffff88004117fc50
[…] ? lock_acquire+0x16f/0x320 ffff88004117fd20
[…] ? lock_acquire+0x16f/0x320 ffff88004117fd58
[…] ? find_held_lock+0x3b/0xe0 ffff88004117fd70
[…] __sys_sendmsg+0x17a/0x260 ffff88004117fdc8
[…] ? __sys_sendmsg_sock+0x80/0x80 ffff88004117fdf0
[…] ? syscall_trace_enter+0x15e/0x2c0 ffff88004117fe98
[…] do_syscall_64+0x7d/0x400 ffff88004117fec8
[…] entry_SYSCALL_64_safe_stack+0x4a/0x60 ffff88004117fef8
</TASK>
==================================================================
The out-of-bounds report, though, is a red herring as it is for an
access that shouldn't have happened in the first place.
When nf_nat_init() fails to register its BPF kfuncs, it'll unwind and,
among others, call unregister_pernet_subsys() to deregister its per-net
ops. This makes the previously allocated net id available for reuse by
the next caller of register_pernet_subsys(), in our case, synproxy.
However, 'nat_net_id' will still hold the previously allocated value.
If nf_nat.o gets build as a module, all this doesn't matter. A failed
initialization routine makes the module fail to load and any dependent
module won't be able to load either. However, if nf_nat.o is built-in,
a failing init won't /completely/ make its functionality unavailable to
dependent modules, namely the code and static data is still there, free
to be called by modules like nft_chain_nat.ko.
Case in point, nft_chain_nat registers hooks that'll call into nf_nat
which, in our case, failed to initialize and therefore won't have a
valid net id nor related net_nat object any more.
Code in nf_nat, namely nf_nat_register_fn() and nf_nat_unregister_fn(),
still making use of the reallocated net id, lead to a type confusion as
the call to net_generic() will no longer return memory belonging to an
object suited to fit 'struct nat_net' but 'struct synproxy_net' instead.
The latter is only 24 bytes on 64-bit systems, much smaller than struct
nat_net which is 176 bytes, perfectly explaining the OOB KASAN report.
Detect and handle a failed nf_nat_init() by testing the 'nf_nat_hook'
pointer which will be reset to NULL on initialization errors to prevent
the usage of an invalid nat_net pointer.
As this check is only needed when nf_nat.o is built-in, guard it by
'#ifndef MODULE...'.
Fixes: cbc1dd5b659f ("netfilter: nf_nat: Fix possible memory leak in nf_nat_init()")
Signed-off-by: Mathias Krause <minipli@grsecurity.net>
---
net/netfilter/nf_nat_core.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c
index 3b5434e4ec9c..76a150b9d418 100644
--- a/net/netfilter/nf_nat_core.c
+++ b/net/netfilter/nf_nat_core.c
@@ -1187,6 +1187,16 @@ int nf_nat_register_fn(struct net *net, u8 pf, const struct nf_hook_ops *ops,
struct nf_hook_ops *nat_ops;
int i, ret;
+#ifndef MODULE
+ /* If nf_nat_core is built-in and nf_nat_init() fails, dependent
+ * modules like nft_chain_nat.ko may still call this function.
+ * However, nat_net would be invalid, likely pointing to some other
+ * per-net structure.
+ */
+ if (WARN_ON_ONCE(!nf_nat_hook))
+ return -EOPNOTSUPP;
+#endif
+
if (WARN_ON_ONCE(pf >= ARRAY_SIZE(nat_net->nat_proto_net)))
return -EINVAL;
--
2.47.3
^ permalink raw reply related
* Re: [PATCH net] net: ipv6: fix NOREF dst use in seg6 and rpl lwtunnels
From: Paolo Abeni @ 2026-04-28 9:14 UTC (permalink / raw)
To: Andrea Mayer, Sebastian Andrzej Siewior
Cc: davem, dsahern, edumazet, kuba, horms, clrkwllms, rostedt,
david.lebrun, alex.aring, Justin Iurman, stefano.salsano, netdev,
linux-rt-devel, linux-kernel, stable
In-Reply-To: <20260425160856.8cebade5eae1dcaec7af8bfe@uniroma2.it>
On 4/25/26 4:08 PM, Andrea Mayer wrote:
> On Thu, 23 Apr 2026 10:00:56 +0200
> Sebastian Andrzej Siewior <bigeasy@linutronix.de> wrote:
>
> Hi Sebastian,
>
> thanks for the review, and to Simon and Justin as well.
>
>
>> On 2026-04-21 11:47:35 [+0200], Andrea Mayer wrote:
>>>
>>> [snip]
>>
>> So the dst passed to skb_dst_set_noref() has no reference count. The fix
>> is to use skb_dst_force() to increment the refcount on it. But this
>> requires that we are in the same RCU section. And I guess we are since
>> none of the warnings are visible.
>
> Yes. lwtunnel_input() holds rcu_read_lock() around ops->input(), which is
> where seg6_input_core()/rpl_input() execute. The skb_dst_force() is called
> within that RCU section.
>
>
>> Doesn't this make ip6_route_input() on RT fragile in general due to the
>> RT6_LOOKUP_F_DST_NOREF usage or here something special about the two
>> files that are patched?
>> Based on your explanation it all makes sense, I am just not sure if this
>> race is limited to those two are if there is more to it.
>
> seg6_input_core() and rpl_input() cache the dst via dst_cache_set_ip6(), which
> invokes dst_hold(). The dst_hold() calls rcuref_get(), failing on a zero
> refcount and triggering a WARN, but the pointer is still stored in the cache.
> After the RCU grace period completes the dst is freed, and a subsequent
> dst_cache_get() returns a dangling pointer.
>
> The other callers of ip6_route_input() (e.g., ipv6_srh_rcv, ipv6_rpl_srh_rcv,
> ip6_rcv_finish_core) consume the NOREF dst without caching it. Even if the
> pcpu_rt's refcount is concurrently dropped to zero, the dst memory remains
> valid because dst_release() defers the actual free via call_rcu_hurry() and the
> caller is still inside the RCU read-side critical section.
>
>
>>> [snip]
>>>
>>> Fixes: af4a2209b134 ("ipv6: sr: use dst_cache in seg6_input")
>>> Fixes: a7a29f9c361f ("net: ipv6: add rpl sr tunnel")
>>
>> If having PREEMPT_RT_NEEDS_BH_LOCK unset is the requirement then the
>> right fixes: would be
>> Fixes: 3253cb49cbad4 ("softirq: Allow to drop the softirq-BKL lock on PREEMPT_RT")
>>
>> as prior this commit the race is not possible, right?
>
> I built and tested kernels at 3253cb49cbad and its parent fd4e876f59b7 (both
> CONFIG_PREEMPT_RT=y, without the fix): no issues at fd4e876f59b7.
> At 3253cb49cbad, a pcpu_rt cmpxchg contention in rt6_make_pcpu_route() shows
> up, which was addressed in 1adaea51c61b. I also tested at 1adaea51c61b, and at
> that point the dst_hold() race described in this patch appears.
>
> The seg6/rpl code obtains a NOREF dst from ip6_route_input(), does not promote
> it via skb_dst_force(), and passes it to dst_cache_set_ip6() which calls
> dst_hold(). This pattern has been present since af4a2209b134 and a7a29f9c361f,
> and the current Fixes: tags point to the commits where it was introduced.
> Does that seem reasonable?
I think the above is correct, but also pointing to 3253cb49cbad4 would
be correct, since the latter is required to exploit the problem. I also
think cases like this one it's better to avoid the repost (since the
constant ML flood) and to err on the conservative side (older hash). So
I'm applying the patch as-is.
Thanks,
Paolo
^ permalink raw reply
* Re: [Intel-wired-lan] [PATCH v2] ice: wait for reset completion in ice_resume()
From: Paul Menzel @ 2026-04-28 9:17 UTC (permalink / raw)
To: Aaron Ma
Cc: Tony Nguyen, Przemek Kitszel, Andrew Lunn, David S. Miller,
Eric Dumazet, Jakub Kicinski, Paolo Abeni, netdev, linux-kernel,
Akeem G Abodunrin, Jesse Brandeburg, intel-wired-lan, Kohei Enju
In-Reply-To: <CAJ6xRxUEcbadApMg0i7ngcqYMUacrGNvCrUZ96sqkW22TsC7iA@mail.gmail.com>
Dear Aaron,
Thank you for your reply.
Am 28.04.26 um 09:53 schrieb Aaron Ma:
> On Mon, Apr 27, 2026 at 6:13 PM Paul Menzel wrote:
>> Am 24.04.26 um 05:03 schrieb Aaron Ma via Intel-wired-lan:
>>> ice_resume() schedules an asynchronous PF reset and returns
>>> immediately. The reset runs later in ice_service_task(). If
>>> userspace tries to bring up the net device before the reset
>>> finishes, ice_open() fails with -EBUSY:
>>>
>>> ice_resume()
>>> ice_schedule_reset() # sets ICE_PFR_REQ, returns
>>> ...
>>> ice_open()
>>> ice_is_reset_in_progress() # ICE_PFR_REQ still set, -EBUSY
>>> ...
>>> ice_service_task()
>>> ice_do_reset()
>>> ice_rebuild() # clears ICE_PFR_REQ, too late
>>>
>>> Reproduced on E800 series NICs during suspend/resume with irdma
>>> enabled, where the aux device probe widens the race window.
>>
>> Please document, how you reproduced it, and also paste possible messages
>> by Linux or NetworkManager, so that people can easily search for the commit.
>
> The error message is "can't open net device while reset is in progress"
> I can add it in v3 if you like.
Yes, that’d be great.
> > > Wait for the reset to complete before returning from ice_resume().
>>
>> Please mention the delay length in the commit message.
>
> The timeout is 10 * HZ (10 seconds), matching the existing usage in
> ice_devlink_info_get() for the same ice_wait_for_reset() call. In
> practice the wait completes in ~300ms.
I often wonder, where the delay values come from. Maybe mention, that
you copied it.
>>> Fixes: 769c500dcc1e ("ice: Add advanced power mgmt for WoL")
>>> Cc: stable@vger.kernel.org
>>> Signed-off-by: Aaron Ma <aaron.ma@canonical.com>
>>> ---
>>> v2: reword comment to clarify best-effort semantics (Kohei Enju)
>>>
>>> drivers/net/ethernet/intel/ice/ice_main.c | 9 +++++++++
>>> 1 file changed, 9 insertions(+)
>>>
>>> diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
>>> index 5f92377d4dfc2..a81eb21ea87c1 100644
>>> --- a/drivers/net/ethernet/intel/ice/ice_main.c
>>> +++ b/drivers/net/ethernet/intel/ice/ice_main.c
>>> @@ -5635,6 +5635,15 @@ static int ice_resume(struct device *dev)
>>> /* Restart the service task */
>>> mod_timer(&pf->serv_tmr, round_jiffies(jiffies + pf->serv_tmr_period));
>>>
>>> + /* Best-effort wait for the scheduled reset to finish so that the
>>> + * device is operational before returning. Without this, userspace
>>> + * (e.g. NetworkManager) may try to open the net device while the
>>> + * asynchronous reset is still in progress, hitting -EBUSY.
>>> + */
>>> + ret = ice_wait_for_reset(pf, 10 * HZ);
>>
>> Why not pass a delay in micro/milliseconds?
>
> ice_wait_for_reset() takes jiffies — that's the existing API.
It’s recommended to use `msecs_to_jiffies()` to make it HZ invariant.
>>> + if (ret)
>>> + dev_err(dev, "Wait for reset failed during resume: %d\n", ret);
>>
>> Mention the delay?
>
> Good point. I'll include the timeout in the error message in v3.
Awesome.
[…]
Thanks,
Paul
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox