Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-nextgrafted

Pull networking updates from Jakub Kicinski:
 "Core:

   - Add dedicated kmem_cache for typical/small skb->head, avoid having
     to access struct page at kfree time, and improve memory use.

   - Introduce sysctl to set default RPS configuration for new netdevs.

   - Define Netlink protocol specification format which can be used to
     describe messages used by each family and auto-generate parsers.
     Add tools for generating kernel data structures and uAPI headers.

   - Expose all net/core sysctls inside netns.

   - Remove 4s sleep in netpoll if carrier is instantly detected on
     boot.

   - Add configurable limit of MDB entries per port, and port-vlan.

   - Continue populating drop reasons throughout the stack.

   - Retire a handful of legacy Qdiscs and classifiers.

  Protocols:

   - Support IPv4 big TCP (TSO frames larger than 64kB).

   - Add IP_LOCAL_PORT_RANGE socket option, to control local port range
     on socket by socket basis.

   - Track and report in procfs number of MPTCP sockets used.

   - Support mixing IPv4 and IPv6 flows in the in-kernel MPTCP path
     manager.

   - IPv6: don't check net.ipv6.route.max_size and rely on garbage
     collection to free memory (similarly to IPv4).

   - Support Penultimate Segment Pop (PSP) flavor in SRv6 (RFC8986).

   - ICMP: add per-rate limit counters.

   - Add support for user scanning requests in ieee802154.

   - Remove static WEP support.

   - Support minimal Wi-Fi 7 Extremely High Throughput (EHT) rate
     reporting.

   - WiFi 7 EHT channel puncturing support (client & AP).

  BPF:

   - Add a rbtree data structure following the "next-gen data structure"
     precedent set by recently added linked list, that is, by using
     kfunc + kptr instead of adding a new BPF map type.

   - Expose XDP hints via kfuncs with initial support for RX hash and
     timestamp metadata.

   - Add BPF_F_NO_TUNNEL_KEY extension to bpf_skb_set_tunnel_key to
     better support decap on GRE tunnel devices not operating in collect
     metadata.

   - Improve x86 JIT's codegen for PROBE_MEM runtime error checks.

   - Remove the need for trace_printk_lock for bpf_trace_printk and
     bpf_trace_vprintk helpers.

   - Extend libbpf's bpf_tracing.h support for tracing arguments of
     kprobes/uprobes and syscall as a special case.

   - Significantly reduce the search time for module symbols by
     livepatch and BPF.

   - Enable cpumasks to be used as kptrs, which is useful for tracing
     programs tracking which tasks end up running on which CPUs in
     different time intervals.

   - Add support for BPF trampoline on s390x and riscv64.

   - Add capability to export the XDP features supported by the NIC.

   - Add __bpf_kfunc tag for marking kernel functions as kfuncs.

   - Add cgroup.memory=nobpf kernel parameter option to disable BPF
     memory accounting for container environments.

  Netfilter:

   - Remove the CLUSTERIP target. It has been marked as obsolete for
     years, and we still have WARN splats wrt races of the out-of-band
     /proc interface installed by this target.

   - Add 'destroy' commands to nf_tables. They are identical to the
     existing 'delete' commands, but do not return an error if the
     referenced object (set, chain, rule...) did not exist.

  Driver API:

   - Improve cpumask_local_spread() locality to help NICs set the right
     IRQ affinity on AMD platforms.

   - Separate C22 and C45 MDIO bus transactions more clearly.

   - Introduce new DCB table to control DSCP rewrite on egress.

   - Support configuration of Physical Layer Collision Avoidance (PLCA)
     Reconciliation Sublayer (RS) (802.3cg-2019). Modern version of
     shared medium Ethernet.

   - Support for MAC Merge layer (IEEE 802.3-2018 clause 99). Allowing
     preemption of low priority frames by high priority frames.

   - Add support for controlling MACSec offload using netlink SET.

   - Rework devlink instance refcounts to allow registration and
     de-registration under the instance lock. Split the code into
     multiple files, drop some of the unnecessarily granular locks and
     factor out common parts of netlink operation handling.

   - Add TX frame aggregation parameters (for USB drivers).

   - Add a new attr TCA_EXT_WARN_MSG to report TC (offload) warning
     messages with notifications for debug.

   - Allow offloading of UDP NEW connections via act_ct.

   - Add support for per action HW stats in TC.

   - Support hardware miss to TC action (continue processing in SW from
     a specific point in the action chain).

   - Warn if old Wireless Extension user space interface is used with
     modern cfg80211/mac80211 drivers. Do not support Wireless
     Extensions for Wi-Fi 7 devices at all. Everyone should switch to
     using nl80211 interface instead.

   - Improve the CAN bit timing configuration. Use extack to return
     error messages directly to user space, update the SJW handling,
     including the definition of a new default value that will benefit
     CAN-FD controllers, by increasing their oscillator tolerance.

  New hardware / drivers:

   - Ethernet:
      - nVidia BlueField-3 support (control traffic driver)
      - Ethernet support for imx93 SoCs
      - Motorcomm yt8531 gigabit Ethernet PHY
      - onsemi NCN26000 10BASE-T1S PHY (with support for PLCA)
      - Microchip LAN8841 PHY (incl. cable diagnostics and PTP)
      - Amlogic gxl MDIO mux

   - WiFi:
      - RealTek RTL8188EU (rtl8xxxu)
      - Qualcomm Wi-Fi 7 devices (ath12k)

   - CAN:
      - Renesas R-Car V4H

  Drivers:

   - Bluetooth:
      - Set Per Platform Antenna Gain (PPAG) for Intel controllers.

   - Ethernet NICs:
      - Intel (1G, igc):
         - support TSN / Qbv / packet scheduling features of i226 model
      - Intel (100G, ice):
         - use GNSS subsystem instead of TTY
         - multi-buffer XDP support
         - extend support for GPIO pins to E823 devices
      - nVidia/Mellanox:
         - update the shared buffer configuration on PFC commands
         - implement PTP adjphase function for HW offset control
         - TC support for Geneve and GRE with VF tunnel offload
         - more efficient crypto key management method
         - multi-port eswitch support
      - Netronome/Corigine:
         - add DCB IEEE support
         - support IPsec offloading for NFP3800
      - Freescale/NXP (enetc):
         - support XDP_REDIRECT for XDP non-linear buffers
         - improve reconfig, avoid link flap and waiting for idle
         - support MAC Merge layer
      - Other NICs:
         - sfc/ef100: add basic devlink support for ef100
         - ionic: rx_push mode operation (writing descriptors via MMIO)
         - bnxt: use the auxiliary bus abstraction for RDMA
         - r8169: disable ASPM and reset bus in case of tx timeout
         - cpsw: support QSGMII mode for J721e CPSW9G
         - cpts: support pulse-per-second output
         - ngbe: add an mdio bus driver
         - usbnet: optimize usbnet_bh() by avoiding unnecessary queuing
         - r8152: handle devices with FW with NCM support
         - amd-xgbe: support 10Mbps, 2.5GbE speeds and rx-adaptation
         - virtio-net: support multi buffer XDP
         - virtio/vsock: replace virtio_vsock_pkt with sk_buff
         - tsnep: XDP support

   - Ethernet high-speed switches:
      - nVidia/Mellanox (mlxsw):
         - add support for latency TLV (in FW control messages)
      - Microchip (sparx5):
         - separate explicit and implicit traffic forwarding rules, make
           the implicit rules always active
         - add support for egress DSCP rewrite
         - IS0 VCAP support (Ingress Classification)
         - IS2 VCAP filters (protos, L3 addrs, L4 ports, flags, ToS
           etc.)
         - ES2 VCAP support (Egress Access Control)
         - support for Per-Stream Filtering and Policing (802.1Q,
           8.6.5.1)

   - Ethernet embedded switches:
      - Marvell (mv88e6xxx):
         - add MAB (port auth) offload support
         - enable PTP receive for mv88e6390
      - NXP (ocelot):
         - support MAC Merge layer
         - support for the the vsc7512 internal copper phys
      - Microchip:
         - lan9303: convert to PHYLINK
         - lan966x: support TC flower filter statistics
         - lan937x: PTP support for KSZ9563/KSZ8563 and LAN937x
         - lan937x: support Credit Based Shaper configuration
         - ksz9477: support Energy Efficient Ethernet
      - other:
         - qca8k: convert to regmap read/write API, use bulk operations
         - rswitch: Improve TX timestamp accuracy

   - Intel WiFi (iwlwifi):
      - EHT (Wi-Fi 7) rate reporting
      - STEP equalizer support: transfer some STEP (connection to radio
        on platforms with integrated wifi) related parameters from the
        BIOS to the firmware.

   - Qualcomm 802.11ax WiFi (ath11k):
      - IPQ5018 support
      - Fine Timing Measurement (FTM) responder role support
      - channel 177 support

   - MediaTek WiFi (mt76):
      - per-PHY LED support
      - mt7996: EHT (Wi-Fi 7) support
      - Wireless Ethernet Dispatch (WED) reset support
      - switch to using page pool allocator

   - RealTek WiFi (rtw89):
      - support new version of Bluetooth co-existance

   - Mobile:
      - rmnet: support TX aggregation"

* tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1872 commits)
  page_pool: add a comment explaining the fragment counter usage
  net: ethtool: fix __ethtool_dev_mm_supported() implementation
  ethtool: pse-pd: Fix double word in comments
  xsk: add linux/vmalloc.h to xsk.c
  sefltests: netdevsim: wait for devlink instance after netns removal
  selftest: fib_tests: Always cleanup before exit
  net/mlx5e: Align IPsec ASO result memory to be as required by hardware
  net/mlx5e: TC, Set CT miss to the specific ct action instance
  net/mlx5e: Rename CHAIN_TO_REG to MAPPED_OBJ_TO_REG
  net/mlx5: Refactor tc miss handling to a single function
  net/mlx5: Kconfig: Make tc offload depend on tc skb extension
  net/sched: flower: Support hardware miss to tc action
  net/sched: flower: Move filter handle initialization earlier
  net/sched: cls_api: Support hardware miss to tc action
  net/sched: Rename user cookie and act cookie
  sfc: fix builds without CONFIG_RTC_LIB
  sfc: clean up some inconsistent indentings
  net/mlx4_en: Introduce flexible array to silence overflow warning
  net: lan966x: Fix possible deadlock inside PTP
  net/ulp: Remove redundant ->clone() test in inet_clone_ulp().
  ...

10 files changed, 967 insertions, 0 deletions

diff --git a/arch/mips/sgi-ip30/Makefile b/arch/mips/sgi-ip30/Makefile
new file mode 100644
index 000000000..18cf561b3
--- /dev/null
+++ b/arch/mips/sgi-ip30/Makefile
@@ -0,0 +1,9 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Makefile for the IP30 specific kernel interface routines under Linux.
+#
+
+obj-y	:= ip30-irq.o ip30-power.o ip30-setup.o ip30-timer.o ip30-xtalk.o
+
+obj-$(CONFIG_EARLY_PRINTK)	+= ip30-console.o
+obj-$(CONFIG_SMP)		+= ip30-smp.o
diff --git a/arch/mips/sgi-ip30/Platform b/arch/mips/sgi-ip30/Platform
new file mode 100644
index 000000000..f6f11517e
--- /dev/null
+++ b/arch/mips/sgi-ip30/Platform
@@ -0,0 +1,5 @@
+#
+# SGI-IP30 (Octane/Octane2)
+#
+cflags-$(CONFIG_SGI_IP30)		+= -I$(srctree)/arch/mips/include/asm/mach-ip30
+load-$(CONFIG_SGI_IP30)			+= 0xa800000020004000
diff --git a/arch/mips/sgi-ip30/ip30-common.h b/arch/mips/sgi-ip30/ip30-common.h
new file mode 100644
index 000000000..7b5db24b6
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-common.h
@@ -0,0 +1,23 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#ifndef __IP30_COMMON_H
+#define __IP30_COMMON_H
+
+/*
+ * Power Switch is wired via BaseIO BRIDGE slot #6.
+ *
+ * ACFail is wired via BaseIO BRIDGE slot #7.
+ */
+#define IP30_POWER_IRQ		HEART_L2_INT_POWER_BTN
+
+#define IP30_HEART_L0_IRQ	(MIPS_CPU_IRQ_BASE + 2)
+#define IP30_HEART_L1_IRQ	(MIPS_CPU_IRQ_BASE + 3)
+#define IP30_HEART_L2_IRQ	(MIPS_CPU_IRQ_BASE + 4)
+#define IP30_HEART_TIMER_IRQ	(MIPS_CPU_IRQ_BASE + 5)
+#define IP30_HEART_ERR_IRQ	(MIPS_CPU_IRQ_BASE + 6)
+
+extern void __init ip30_install_ipi(void);
+extern struct plat_smp_ops ip30_smp_ops;
+extern void __init ip30_per_cpu_init(void);
+
+#endif /* __IP30_COMMON_H */
diff --git a/arch/mips/sgi-ip30/ip30-console.c b/arch/mips/sgi-ip30/ip30-console.c
new file mode 100644
index 000000000..b91f8c4fd
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-console.c
@@ -0,0 +1,23 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/io.h>
+
+#include <asm/sn/ioc3.h>
+
+static inline struct ioc3_uartregs *console_uart(void)
+{
+	struct ioc3 *ioc3;
+
+	ioc3 = (struct ioc3 *)((void *)(0x900000001f600000));
+	return &ioc3->sregs.uarta;
+}
+
+void prom_putchar(char c)
+{
+	struct ioc3_uartregs *uart = console_uart();
+
+	while ((readb(&uart->iu_lsr) & 0x20) == 0)
+		cpu_relax();
+
+	writeb(c, &uart->iu_thr);
+}
diff --git a/arch/mips/sgi-ip30/ip30-irq.c b/arch/mips/sgi-ip30/ip30-irq.c
new file mode 100644
index 000000000..423c32cb6
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-irq.c
@@ -0,0 +1,330 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * ip30-irq.c: Highlevel interrupt handling for IP30 architecture.
+ */
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqdomain.h>
+#include <linux/percpu.h>
+#include <linux/spinlock.h>
+#include <linux/tick.h>
+#include <linux/types.h>
+
+#include <asm/irq_cpu.h>
+#include <asm/sgi/heart.h>
+
+#include "ip30-common.h"
+
+struct heart_irq_data {
+	u64	*irq_mask;
+	int	cpu;
+};
+
+static DECLARE_BITMAP(heart_irq_map, HEART_NUM_IRQS);
+
+static DEFINE_PER_CPU(unsigned long, irq_enable_mask);
+
+static inline int heart_alloc_int(void)
+{
+	int bit;
+
+again:
+	bit = find_first_zero_bit(heart_irq_map, HEART_NUM_IRQS);
+	if (bit >= HEART_NUM_IRQS)
+		return -ENOSPC;
+
+	if (test_and_set_bit(bit, heart_irq_map))
+		goto again;
+
+	return bit;
+}
+
+static void ip30_error_irq(struct irq_desc *desc)
+{
+	u64 pending, mask, cause, error_irqs, err_reg;
+	int cpu = smp_processor_id();
+	int i;
+
+	pending = heart_read(&heart_regs->isr);
+	mask = heart_read(&heart_regs->imr[cpu]);
+	cause = heart_read(&heart_regs->cause);
+	error_irqs = (pending & HEART_L4_INT_MASK & mask);
+
+	/* Bail if there's nothing to process (how did we get here, then?) */
+	if (unlikely(!error_irqs))
+		return;
+
+	/* Prevent any of the error IRQs from firing again. */
+	heart_write(mask & ~(pending), &heart_regs->imr[cpu]);
+
+	/* Ack all error IRQs. */
+	heart_write(HEART_L4_INT_MASK, &heart_regs->clear_isr);
+
+	/*
+	 * If we also have a cause value, then something happened, so loop
+	 * through the error IRQs and report a "heart attack" for each one
+	 * and print the value of the HEART cause register.  This is really
+	 * primitive right now, but it should hopefully work until a more
+	 * robust error handling routine can be put together.
+	 *
+	 * Refer to heart.h for the HC_* macros to work out the cause
+	 * that got us here.
+	 */
+	if (cause) {
+		pr_alert("IP30: CPU%d: HEART ATTACK! ISR = 0x%.16llx, IMR = 0x%.16llx, CAUSE = 0x%.16llx\n",
+			 cpu, pending, mask, cause);
+
+		if (cause & HC_COR_MEM_ERR) {
+			err_reg = heart_read(&heart_regs->mem_err_addr);
+			pr_alert("  HEART_MEMERR_ADDR = 0x%.16llx\n", err_reg);
+		}
+
+		/* i = 63; i >= 51; i-- */
+		for (i = HEART_ERR_MASK_END; i >= HEART_ERR_MASK_START; i--)
+			if ((pending >> i) & 1)
+				pr_alert("  HEART Error IRQ #%d\n", i);
+
+		/* XXX: Seems possible to loop forever here, so panic(). */
+		panic("IP30: Fatal Error !\n");
+	}
+
+	/* Unmask the error IRQs. */
+	heart_write(mask, &heart_regs->imr[cpu]);
+}
+
+static void ip30_normal_irq(struct irq_desc *desc)
+{
+	int cpu = smp_processor_id();
+	struct irq_domain *domain;
+	u64 pend, mask;
+	int ret;
+
+	pend = heart_read(&heart_regs->isr);
+	mask = (heart_read(&heart_regs->imr[cpu]) &
+		(HEART_L0_INT_MASK | HEART_L1_INT_MASK | HEART_L2_INT_MASK));
+
+	pend &= mask;
+	if (unlikely(!pend))
+		return;
+
+#ifdef CONFIG_SMP
+	if (pend & BIT_ULL(HEART_L2_INT_RESCHED_CPU_0)) {
+		heart_write(BIT_ULL(HEART_L2_INT_RESCHED_CPU_0),
+			    &heart_regs->clear_isr);
+		scheduler_ipi();
+	} else if (pend & BIT_ULL(HEART_L2_INT_RESCHED_CPU_1)) {
+		heart_write(BIT_ULL(HEART_L2_INT_RESCHED_CPU_1),
+			    &heart_regs->clear_isr);
+		scheduler_ipi();
+	} else if (pend & BIT_ULL(HEART_L2_INT_CALL_CPU_0)) {
+		heart_write(BIT_ULL(HEART_L2_INT_CALL_CPU_0),
+			    &heart_regs->clear_isr);
+		generic_smp_call_function_interrupt();
+	} else if (pend & BIT_ULL(HEART_L2_INT_CALL_CPU_1)) {
+		heart_write(BIT_ULL(HEART_L2_INT_CALL_CPU_1),
+			    &heart_regs->clear_isr);
+		generic_smp_call_function_interrupt();
+	} else
+#endif
+	{
+		domain = irq_desc_get_handler_data(desc);
+		ret = generic_handle_domain_irq(domain, __ffs(pend));
+		if (ret)
+			spurious_interrupt();
+	}
+}
+
+static void ip30_ack_heart_irq(struct irq_data *d)
+{
+	heart_write(BIT_ULL(d->hwirq), &heart_regs->clear_isr);
+}
+
+static void ip30_mask_heart_irq(struct irq_data *d)
+{
+	struct heart_irq_data *hd = irq_data_get_irq_chip_data(d);
+	unsigned long *mask = &per_cpu(irq_enable_mask, hd->cpu);
+
+	clear_bit(d->hwirq, mask);
+	heart_write(*mask, &heart_regs->imr[hd->cpu]);
+}
+
+static void ip30_mask_and_ack_heart_irq(struct irq_data *d)
+{
+	struct heart_irq_data *hd = irq_data_get_irq_chip_data(d);
+	unsigned long *mask = &per_cpu(irq_enable_mask, hd->cpu);
+
+	clear_bit(d->hwirq, mask);
+	heart_write(*mask, &heart_regs->imr[hd->cpu]);
+	heart_write(BIT_ULL(d->hwirq), &heart_regs->clear_isr);
+}
+
+static void ip30_unmask_heart_irq(struct irq_data *d)
+{
+	struct heart_irq_data *hd = irq_data_get_irq_chip_data(d);
+	unsigned long *mask = &per_cpu(irq_enable_mask, hd->cpu);
+
+	set_bit(d->hwirq, mask);
+	heart_write(*mask, &heart_regs->imr[hd->cpu]);
+}
+
+static int ip30_set_heart_irq_affinity(struct irq_data *d,
+				       const struct cpumask *mask, bool force)
+{
+	struct heart_irq_data *hd = irq_data_get_irq_chip_data(d);
+
+	if (!hd)
+		return -EINVAL;
+
+	if (irqd_is_started(d))
+		ip30_mask_and_ack_heart_irq(d);
+
+	hd->cpu = cpumask_first_and(mask, cpu_online_mask);
+
+	if (irqd_is_started(d))
+		ip30_unmask_heart_irq(d);
+
+	irq_data_update_effective_affinity(d, cpumask_of(hd->cpu));
+
+	return 0;
+}
+
+static struct irq_chip heart_irq_chip = {
+	.name			= "HEART",
+	.irq_ack		= ip30_ack_heart_irq,
+	.irq_mask		= ip30_mask_heart_irq,
+	.irq_mask_ack		= ip30_mask_and_ack_heart_irq,
+	.irq_unmask		= ip30_unmask_heart_irq,
+	.irq_set_affinity	= ip30_set_heart_irq_affinity,
+};
+
+static int heart_domain_alloc(struct irq_domain *domain, unsigned int virq,
+			      unsigned int nr_irqs, void *arg)
+{
+	struct irq_alloc_info *info = arg;
+	struct heart_irq_data *hd;
+	int hwirq;
+
+	if (nr_irqs > 1 || !info)
+		return -EINVAL;
+
+	hd = kzalloc(sizeof(*hd), GFP_KERNEL);
+	if (!hd)
+		return -ENOMEM;
+
+	hwirq = heart_alloc_int();
+	if (hwirq < 0) {
+		kfree(hd);
+		return -EAGAIN;
+	}
+	irq_domain_set_info(domain, virq, hwirq, &heart_irq_chip, hd,
+			    handle_level_irq, NULL, NULL);
+
+	return 0;
+}
+
+static void heart_domain_free(struct irq_domain *domain,
+			      unsigned int virq, unsigned int nr_irqs)
+{
+	struct irq_data *irqd;
+
+	if (nr_irqs > 1)
+		return;
+
+	irqd = irq_domain_get_irq_data(domain, virq);
+	if (irqd) {
+		clear_bit(irqd->hwirq, heart_irq_map);
+		kfree(irqd->chip_data);
+	}
+}
+
+static const struct irq_domain_ops heart_domain_ops = {
+	.alloc = heart_domain_alloc,
+	.free  = heart_domain_free,
+};
+
+void __init ip30_install_ipi(void)
+{
+	int cpu = smp_processor_id();
+	unsigned long *mask = &per_cpu(irq_enable_mask, cpu);
+
+	set_bit(HEART_L2_INT_RESCHED_CPU_0 + cpu, mask);
+	heart_write(BIT_ULL(HEART_L2_INT_RESCHED_CPU_0 + cpu),
+		    &heart_regs->clear_isr);
+	set_bit(HEART_L2_INT_CALL_CPU_0 + cpu, mask);
+	heart_write(BIT_ULL(HEART_L2_INT_CALL_CPU_0 + cpu),
+		    &heart_regs->clear_isr);
+
+	heart_write(*mask, &heart_regs->imr[cpu]);
+}
+
+void __init arch_init_irq(void)
+{
+	struct irq_domain *domain;
+	struct fwnode_handle *fn;
+	unsigned long *mask;
+	int i;
+
+	mips_cpu_irq_init();
+
+	/* Mask all IRQs. */
+	heart_write(HEART_CLR_ALL_MASK, &heart_regs->imr[0]);
+	heart_write(HEART_CLR_ALL_MASK, &heart_regs->imr[1]);
+	heart_write(HEART_CLR_ALL_MASK, &heart_regs->imr[2]);
+	heart_write(HEART_CLR_ALL_MASK, &heart_regs->imr[3]);
+
+	/* Ack everything. */
+	heart_write(HEART_ACK_ALL_MASK, &heart_regs->clear_isr);
+
+	/* Enable specific HEART error IRQs for each CPU. */
+	mask = &per_cpu(irq_enable_mask, 0);
+	*mask |= HEART_CPU0_ERR_MASK;
+	heart_write(*mask, &heart_regs->imr[0]);
+	mask = &per_cpu(irq_enable_mask, 1);
+	*mask |= HEART_CPU1_ERR_MASK;
+	heart_write(*mask, &heart_regs->imr[1]);
+
+	/*
+	 * Some HEART bits are reserved by hardware or by software convention.
+	 * Mark these as reserved right away so they won't be accidentally
+	 * used later.
+	 */
+	set_bit(HEART_L0_INT_GENERIC, heart_irq_map);
+	set_bit(HEART_L0_INT_FLOW_CTRL_HWTR_0, heart_irq_map);
+	set_bit(HEART_L0_INT_FLOW_CTRL_HWTR_1, heart_irq_map);
+	set_bit(HEART_L2_INT_RESCHED_CPU_0, heart_irq_map);
+	set_bit(HEART_L2_INT_RESCHED_CPU_1, heart_irq_map);
+	set_bit(HEART_L2_INT_CALL_CPU_0, heart_irq_map);
+	set_bit(HEART_L2_INT_CALL_CPU_1, heart_irq_map);
+	set_bit(HEART_L3_INT_TIMER, heart_irq_map);
+
+	/* Reserve the error interrupts (#51 to #63). */
+	for (i = HEART_L4_INT_XWID_ERR_9; i <= HEART_L4_INT_HEART_EXCP; i++)
+		set_bit(i, heart_irq_map);
+
+	fn = irq_domain_alloc_named_fwnode("HEART");
+	WARN_ON(fn == NULL);
+	if (!fn)
+		return;
+	domain = irq_domain_create_linear(fn, HEART_NUM_IRQS,
+					  &heart_domain_ops, NULL);
+	WARN_ON(domain == NULL);
+	if (!domain)
+		return;
+
+	irq_set_default_host(domain);
+
+	irq_set_percpu_devid(IP30_HEART_L0_IRQ);
+	irq_set_chained_handler_and_data(IP30_HEART_L0_IRQ, ip30_normal_irq,
+					 domain);
+	irq_set_percpu_devid(IP30_HEART_L1_IRQ);
+	irq_set_chained_handler_and_data(IP30_HEART_L1_IRQ, ip30_normal_irq,
+					 domain);
+	irq_set_percpu_devid(IP30_HEART_L2_IRQ);
+	irq_set_chained_handler_and_data(IP30_HEART_L2_IRQ, ip30_normal_irq,
+					 domain);
+	irq_set_percpu_devid(IP30_HEART_ERR_IRQ);
+	irq_set_chained_handler_and_data(IP30_HEART_ERR_IRQ, ip30_error_irq,
+					 domain);
+}
diff --git a/arch/mips/sgi-ip30/ip30-power.c b/arch/mips/sgi-ip30/ip30-power.c
new file mode 100644
index 000000000..120b3f3d5
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-power.c
@@ -0,0 +1,41 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * ip30-power.c: Software powerdown and reset handling for IP30 architecture.
+ *
+ * Copyright (C) 2004-2007 Stanislaw Skowronek <skylark@unaligned.org>
+ *               2014 Joshua Kinard <kumba@gentoo.org>
+ *               2009 Johannes Dickgreber <tanzy@gmx.de>
+ */
+
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/notifier.h>
+#include <linux/delay.h>
+#include <linux/rtc/ds1685.h>
+#include <linux/interrupt.h>
+#include <linux/pm.h>
+
+#include <asm/reboot.h>
+#include <asm/sgi/heart.h>
+
+static void __noreturn ip30_machine_restart(char *cmd)
+{
+	/*
+	 * Execute HEART cold reset
+	 *   Yes, it's cold-HEARTed!
+	 */
+	heart_write((heart_read(&heart_regs->mode) | HM_COLD_RST),
+		    &heart_regs->mode);
+	unreachable();
+}
+
+static int __init ip30_reboot_setup(void)
+{
+	_machine_restart = ip30_machine_restart;
+
+	return 0;
+}
+
+subsys_initcall(ip30_reboot_setup);
diff --git a/arch/mips/sgi-ip30/ip30-setup.c b/arch/mips/sgi-ip30/ip30-setup.c
new file mode 100644
index 000000000..75a34684e
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-setup.c
@@ -0,0 +1,138 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * SGI IP30 miscellaneous setup bits.
+ *
+ * Copyright (C) 2004-2007 Stanislaw Skowronek <skylark@unaligned.org>
+ *               2007 Joshua Kinard <kumba@gentoo.org>
+ *               2009 Johannes Dickgreber <tanzy@gmx.de>
+ */
+
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/percpu.h>
+#include <linux/memblock.h>
+
+#include <asm/smp-ops.h>
+#include <asm/sgialib.h>
+#include <asm/time.h>
+#include <asm/sgi/heart.h>
+
+#include "ip30-common.h"
+
+/* Structure of accessible HEART registers located in XKPHYS space. */
+struct ip30_heart_regs __iomem *heart_regs = HEART_XKPHYS_BASE;
+
+/*
+ * ARCS will report up to the first 1GB of
+ * memory if queried.  Anything beyond that
+ * is marked as reserved.
+ */
+#define IP30_MAX_PROM_MEMORY	_AC(0x40000000, UL)
+
+/*
+ * Memory in the Octane starts at 512MB
+ */
+#define IP30_MEMORY_BASE	_AC(0x20000000, UL)
+
+/*
+ * If using ARCS to probe for memory, then
+ * remaining memory will start at this offset.
+ */
+#define IP30_REAL_MEMORY_START  (IP30_MEMORY_BASE + IP30_MAX_PROM_MEMORY)
+
+#define MEM_SHIFT(x) ((x) >> 20)
+
+static void __init ip30_mem_init(void)
+{
+	unsigned long total_mem;
+	phys_addr_t addr;
+	phys_addr_t size;
+	u32 memcfg;
+	int i;
+
+	total_mem = 0;
+	for (i = 0; i < HEART_MEMORY_BANKS; i++) {
+		memcfg = __raw_readl(&heart_regs->mem_cfg.l[i]);
+		if (!(memcfg & HEART_MEMCFG_VALID))
+			continue;
+
+		addr = memcfg & HEART_MEMCFG_ADDR_MASK;
+		addr <<= HEART_MEMCFG_UNIT_SHIFT;
+		addr += IP30_MEMORY_BASE;
+		size = memcfg & HEART_MEMCFG_SIZE_MASK;
+		size >>= HEART_MEMCFG_SIZE_SHIFT;
+		size += 1;
+		size <<= HEART_MEMCFG_UNIT_SHIFT;
+
+		total_mem += size;
+
+		if (addr >= IP30_REAL_MEMORY_START)
+			memblock_phys_free(addr, size);
+		else if ((addr + size) > IP30_REAL_MEMORY_START)
+			memblock_phys_free(IP30_REAL_MEMORY_START,
+					   size - IP30_MAX_PROM_MEMORY);
+	}
+	pr_info("Detected %luMB of physical memory.\n", MEM_SHIFT(total_mem));
+}
+
+/**
+ * ip30_cpu_time_init - platform time initialization.
+ */
+static void __init ip30_cpu_time_init(void)
+{
+	int cpu = smp_processor_id();
+	u64 heart_compare;
+	unsigned int start, end;
+	int time_diff;
+
+	heart_compare = (heart_read(&heart_regs->count) +
+			 (HEART_CYCLES_PER_SEC / 10));
+	start = read_c0_count();
+	while ((heart_read(&heart_regs->count) - heart_compare) & 0x800000)
+		cpu_relax();
+
+	end = read_c0_count();
+	time_diff = (int)end - (int)start;
+	mips_hpt_frequency = time_diff * 10;
+	pr_info("IP30: CPU%d: %d MHz CPU detected.\n", cpu,
+		(mips_hpt_frequency * 2) / 1000000);
+}
+
+void __init ip30_per_cpu_init(void)
+{
+	/* Disable all interrupts. */
+	clear_c0_status(ST0_IM);
+
+	ip30_cpu_time_init();
+#ifdef CONFIG_SMP
+	ip30_install_ipi();
+#endif
+
+	enable_percpu_irq(IP30_HEART_L0_IRQ, IRQ_TYPE_NONE);
+	enable_percpu_irq(IP30_HEART_L1_IRQ, IRQ_TYPE_NONE);
+	enable_percpu_irq(IP30_HEART_L2_IRQ, IRQ_TYPE_NONE);
+	enable_percpu_irq(IP30_HEART_ERR_IRQ, IRQ_TYPE_NONE);
+}
+
+/**
+ * plat_mem_setup - despite the name, misc setup happens here.
+ */
+void __init plat_mem_setup(void)
+{
+	ip30_mem_init();
+
+	/* XXX: Hard lock on /sbin/init if this flag isn't specified. */
+	prom_flags |= PROM_FLAG_DONT_FREE_TEMP;
+
+#ifdef CONFIG_SMP
+	register_smp_ops(&ip30_smp_ops);
+#else
+	ip30_per_cpu_init();
+#endif
+
+	ioport_resource.start = 0;
+	ioport_resource.end = ~0UL;
+	set_io_port_base(IO_BASE);
+}
diff --git a/arch/mips/sgi-ip30/ip30-smp.c b/arch/mips/sgi-ip30/ip30-smp.c
new file mode 100644
index 000000000..4bfe65460
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-smp.c
@@ -0,0 +1,149 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * ip30-smp.c: SMP on IP30 architecture.
+ * Based off of the original IP30 SMP code, with inspiration from ip27-smp.c
+ * and smp-bmips.c.
+ *
+ * Copyright (C) 2005-2007 Stanislaw Skowronek <skylark@unaligned.org>
+ *               2006-2007, 2014-2015 Joshua Kinard <kumba@gentoo.org>
+ *               2009 Johannes Dickgreber <tanzy@gmx.de>
+ */
+
+#include <linux/init.h>
+#include <linux/sched.h>
+#include <linux/sched/task_stack.h>
+
+#include <asm/time.h>
+#include <asm/sgi/heart.h>
+
+#include "ip30-common.h"
+
+#define MPCONF_MAGIC	0xbaddeed2
+#define	MPCONF_ADDR	0xa800000000000600L
+#define MPCONF_SIZE	0x80
+#define MPCONF(x)	(MPCONF_ADDR + (x) * MPCONF_SIZE)
+
+/* HEART can theoretically do 4 CPUs, but only 2 are physically possible */
+#define MP_NCPU		2
+
+struct mpconf {
+	u32 magic;
+	u32 prid;
+	u32 physid;
+	u32 virtid;
+	u32 scachesz;
+	u16 fanloads;
+	u16 res;
+	void *launch;
+	void *rendezvous;
+	u64 res2[3];
+	void *stackaddr;
+	void *lnch_parm;
+	void *rndv_parm;
+	u32 idleflag;
+};
+
+static void ip30_smp_send_ipi_single(int cpu, u32 action)
+{
+	int irq;
+
+	switch (action) {
+	case SMP_RESCHEDULE_YOURSELF:
+		irq = HEART_L2_INT_RESCHED_CPU_0;
+		break;
+	case SMP_CALL_FUNCTION:
+		irq = HEART_L2_INT_CALL_CPU_0;
+		break;
+	default:
+		panic("IP30: Unknown action value in %s!\n", __func__);
+	}
+
+	irq += cpu;
+
+	/* Poke the other CPU -- it's got mail! */
+	heart_write(BIT_ULL(irq), &heart_regs->set_isr);
+}
+
+static void ip30_smp_send_ipi_mask(const struct cpumask *mask, u32 action)
+{
+	u32 i;
+
+	for_each_cpu(i, mask)
+		ip30_smp_send_ipi_single(i, action);
+}
+
+static void __init ip30_smp_setup(void)
+{
+	int i;
+	int ncpu = 0;
+	struct mpconf *mpc;
+
+	init_cpu_possible(cpumask_of(0));
+
+	/* Scan the MPCONF structure and enumerate available CPUs. */
+	for (i = 0; i < MP_NCPU; i++) {
+		mpc = (struct mpconf *)MPCONF(i);
+		if (mpc->magic == MPCONF_MAGIC) {
+			set_cpu_possible(i, true);
+			__cpu_number_map[i] = ++ncpu;
+			__cpu_logical_map[ncpu] = i;
+			pr_info("IP30: Slot: %d, PrID: %.8x, PhyID: %d, VirtID: %d\n",
+				i, mpc->prid, mpc->physid, mpc->virtid);
+		}
+	}
+	pr_info("IP30: Detected %d CPU(s) present.\n", ncpu);
+
+	/*
+	 * Set the coherency algorithm to '5' (cacheable coherent
+	 * exclusive on write).  This is needed on IP30 SMP, especially
+	 * for R14000 CPUs, otherwise, instruction bus errors will
+	 * occur upon reaching userland.
+	 */
+	change_c0_config(CONF_CM_CMASK, CONF_CM_CACHABLE_COW);
+}
+
+static void __init ip30_smp_prepare_cpus(unsigned int max_cpus)
+{
+	/* nothing to do here */
+}
+
+static int __init ip30_smp_boot_secondary(int cpu, struct task_struct *idle)
+{
+	struct mpconf *mpc = (struct mpconf *)MPCONF(cpu);
+
+	/* Stack pointer (sp). */
+	mpc->stackaddr = (void *)__KSTK_TOS(idle);
+
+	/* Global pointer (gp). */
+	mpc->lnch_parm = task_thread_info(idle);
+
+	mb(); /* make sure stack and lparm are written */
+
+	/* Boot CPUx. */
+	mpc->launch = smp_bootstrap;
+
+	/* CPUx now executes smp_bootstrap, then ip30_smp_finish */
+	return 0;
+}
+
+static void __init ip30_smp_init_cpu(void)
+{
+	ip30_per_cpu_init();
+}
+
+static void __init ip30_smp_finish(void)
+{
+	enable_percpu_irq(get_c0_compare_int(), IRQ_TYPE_NONE);
+	local_irq_enable();
+}
+
+struct plat_smp_ops __read_mostly ip30_smp_ops = {
+	.send_ipi_single	= ip30_smp_send_ipi_single,
+	.send_ipi_mask		= ip30_smp_send_ipi_mask,
+	.smp_setup		= ip30_smp_setup,
+	.prepare_cpus		= ip30_smp_prepare_cpus,
+	.boot_secondary		= ip30_smp_boot_secondary,
+	.init_secondary		= ip30_smp_init_cpu,
+	.smp_finish		= ip30_smp_finish,
+	.prepare_boot_cpu	= ip30_smp_init_cpu,
+};
diff --git a/arch/mips/sgi-ip30/ip30-timer.c b/arch/mips/sgi-ip30/ip30-timer.c
new file mode 100644
index 000000000..d13e10547
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-timer.c
@@ -0,0 +1,63 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * ip30-timer.c: Clocksource/clockevent support for the
+ *               HEART chip in SGI Octane (IP30) systems.
+ *
+ * Copyright (C) 2004-2007 Stanislaw Skowronek <skylark@unaligned.org>
+ * Copyright (C) 2009 Johannes Dickgreber <tanzy@gmx.de>
+ * Copyright (C) 2011 Joshua Kinard <kumba@gentoo.org>
+ */
+
+#include <linux/clocksource.h>
+#include <linux/cpumask.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/percpu.h>
+#include <linux/sched_clock.h>
+
+#include <asm/time.h>
+#include <asm/cevt-r4k.h>
+#include <asm/sgi/heart.h>
+
+static u64 ip30_heart_counter_read(struct clocksource *cs)
+{
+	return heart_read(&heart_regs->count);
+}
+
+struct clocksource ip30_heart_clocksource = {
+	.name	= "HEART",
+	.rating	= 400,
+	.read	= ip30_heart_counter_read,
+	.mask	= CLOCKSOURCE_MASK(52),
+	.flags	= (CLOCK_SOURCE_IS_CONTINUOUS | CLOCK_SOURCE_VALID_FOR_HRES),
+};
+
+static u64 notrace ip30_heart_read_sched_clock(void)
+{
+	return heart_read(&heart_regs->count);
+}
+
+static void __init ip30_heart_clocksource_init(void)
+{
+	struct clocksource *cs = &ip30_heart_clocksource;
+
+	clocksource_register_hz(cs, HEART_CYCLES_PER_SEC);
+
+	sched_clock_register(ip30_heart_read_sched_clock, 52,
+			     HEART_CYCLES_PER_SEC);
+}
+
+void __init plat_time_init(void)
+{
+	int irq = get_c0_compare_int();
+
+	cp0_timer_irq_installed = 1;
+	c0_compare_irqaction.percpu_dev_id = &mips_clockevent_device;
+	c0_compare_irqaction.flags &= ~IRQF_SHARED;
+	irq_set_handler(irq, handle_percpu_devid_irq);
+	irq_set_percpu_devid(irq);
+	setup_percpu_irq(irq, &c0_compare_irqaction);
+	enable_percpu_irq(irq, IRQ_TYPE_NONE);
+
+	ip30_heart_clocksource_init();
+}
diff --git a/arch/mips/sgi-ip30/ip30-xtalk.c b/arch/mips/sgi-ip30/ip30-xtalk.c
new file mode 100644
index 000000000..7ceb2b23e
--- /dev/null
+++ b/arch/mips/sgi-ip30/ip30-xtalk.c
@@ -0,0 +1,186 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * ip30-xtalk.c - Very basic Crosstalk (XIO) detection support.
+ *   Copyright (C) 2004-2007 Stanislaw Skowronek <skylark@unaligned.org>
+ *   Copyright (C) 2009 Johannes Dickgreber <tanzy@gmx.de>
+ *   Copyright (C) 2007, 2014-2016 Joshua Kinard <kumba@gentoo.org>
+ */
+
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/platform_data/sgi-w1.h>
+#include <linux/platform_data/xtalk-bridge.h>
+
+#include <asm/xtalk/xwidget.h>
+#include <asm/pci/bridge.h>
+
+#define IP30_SWIN_BASE(widget) \
+		(0x0000000010000000 | (((unsigned long)(widget)) << 24))
+
+#define IP30_RAW_SWIN_BASE(widget)	(IO_BASE + IP30_SWIN_BASE(widget))
+
+#define IP30_SWIN_SIZE		(1 << 24)
+
+#define IP30_WIDGET_XBOW        _AC(0x0, UL)    /* XBow is always 0 */
+#define IP30_WIDGET_HEART       _AC(0x8, UL)    /* HEART is always 8 */
+#define IP30_WIDGET_PCI_BASE    _AC(0xf, UL)    /* BaseIO PCI is always 15 */
+
+#define XTALK_NODEV             0xffffffff
+
+#define XBOW_REG_LINK_STAT_0    0x114
+#define XBOW_REG_LINK_BLK_SIZE  0x40
+#define XBOW_REG_LINK_ALIVE     0x80000000
+
+#define HEART_INTR_ADDR		0x00000080
+
+#define xtalk_read	__raw_readl
+
+static void bridge_platform_create(int widget, int masterwid)
+{
+	struct xtalk_bridge_platform_data *bd;
+	struct sgi_w1_platform_data *wd;
+	struct platform_device *pdev_wd;
+	struct platform_device *pdev_bd;
+	struct resource w1_res;
+
+	wd = kzalloc(sizeof(*wd), GFP_KERNEL);
+	if (!wd) {
+		pr_warn("xtalk:%x bridge create out of memory\n", widget);
+		return;
+	}
+
+	snprintf(wd->dev_id, sizeof(wd->dev_id), "bridge-%012lx",
+		 IP30_SWIN_BASE(widget));
+
+	memset(&w1_res, 0, sizeof(w1_res));
+	w1_res.start = IP30_SWIN_BASE(widget) +
+				offsetof(struct bridge_regs, b_nic);
+	w1_res.end = w1_res.start + 3;
+	w1_res.flags = IORESOURCE_MEM;
+
+	pdev_wd = platform_device_alloc("sgi_w1", PLATFORM_DEVID_AUTO);
+	if (!pdev_wd) {
+		pr_warn("xtalk:%x bridge create out of memory\n", widget);
+		goto err_kfree_wd;
+	}
+	if (platform_device_add_resources(pdev_wd, &w1_res, 1)) {
+		pr_warn("xtalk:%x bridge failed to add platform resources.\n", widget);
+		goto err_put_pdev_wd;
+	}
+	if (platform_device_add_data(pdev_wd, wd, sizeof(*wd))) {
+		pr_warn("xtalk:%x bridge failed to add platform data.\n", widget);
+		goto err_put_pdev_wd;
+	}
+	if (platform_device_add(pdev_wd)) {
+		pr_warn("xtalk:%x bridge failed to add platform device.\n", widget);
+		goto err_put_pdev_wd;
+	}
+	/* platform_device_add_data() duplicates the data */
+	kfree(wd);
+
+	bd = kzalloc(sizeof(*bd), GFP_KERNEL);
+	if (!bd) {
+		pr_warn("xtalk:%x bridge create out of memory\n", widget);
+		goto err_unregister_pdev_wd;
+	}
+	pdev_bd = platform_device_alloc("xtalk-bridge", PLATFORM_DEVID_AUTO);
+	if (!pdev_bd) {
+		pr_warn("xtalk:%x bridge create out of memory\n", widget);
+		goto err_kfree_bd;
+	}
+
+	bd->bridge_addr	= IP30_RAW_SWIN_BASE(widget);
+	bd->intr_addr	= HEART_INTR_ADDR;
+	bd->nasid	= 0;
+	bd->masterwid	= masterwid;
+
+	bd->mem.name	= "Bridge PCI MEM";
+	bd->mem.start	= IP30_SWIN_BASE(widget) + BRIDGE_DEVIO0;
+	bd->mem.end	= IP30_SWIN_BASE(widget) + IP30_SWIN_SIZE - 1;
+	bd->mem.flags	= IORESOURCE_MEM;
+	bd->mem_offset	= IP30_SWIN_BASE(widget);
+
+	bd->io.name	= "Bridge PCI IO";
+	bd->io.start	= IP30_SWIN_BASE(widget) + BRIDGE_DEVIO0;
+	bd->io.end	= IP30_SWIN_BASE(widget) + IP30_SWIN_SIZE - 1;
+	bd->io.flags	= IORESOURCE_IO;
+	bd->io_offset	= IP30_SWIN_BASE(widget);
+
+	if (platform_device_add_data(pdev_bd, bd, sizeof(*bd))) {
+		pr_warn("xtalk:%x bridge failed to add platform data.\n", widget);
+		goto err_put_pdev_bd;
+	}
+	if (platform_device_add(pdev_bd)) {
+		pr_warn("xtalk:%x bridge failed to add platform device.\n", widget);
+		goto err_put_pdev_bd;
+	}
+	/* platform_device_add_data() duplicates the data */
+	kfree(bd);
+	pr_info("xtalk:%x bridge widget\n", widget);
+	return;
+
+err_put_pdev_bd:
+	platform_device_put(pdev_bd);
+err_kfree_bd:
+	kfree(bd);
+err_unregister_pdev_wd:
+	platform_device_unregister(pdev_wd);
+	return;
+err_put_pdev_wd:
+	platform_device_put(pdev_wd);
+err_kfree_wd:
+	kfree(wd);
+	return;
+}
+
+static unsigned int __init xbow_widget_active(s8 wid)
+{
+	unsigned int link_stat;
+
+	link_stat = xtalk_read((void *)(IP30_RAW_SWIN_BASE(IP30_WIDGET_XBOW) +
+					XBOW_REG_LINK_STAT_0 +
+					XBOW_REG_LINK_BLK_SIZE *
+					(wid - 8)));
+
+	return (link_stat & XBOW_REG_LINK_ALIVE) ? 1 : 0;
+}
+
+static void __init xtalk_init_widget(s8 wid, s8 masterwid)
+{
+	xwidget_part_num_t partnum;
+	widgetreg_t widget_id;
+
+	if (!xbow_widget_active(wid))
+		return;
+
+	widget_id = xtalk_read((void *)(IP30_RAW_SWIN_BASE(wid) + WIDGET_ID));
+
+	partnum = XWIDGET_PART_NUM(widget_id);
+
+	switch (partnum) {
+	case BRIDGE_WIDGET_PART_NUM:
+	case XBRIDGE_WIDGET_PART_NUM:
+		bridge_platform_create(wid, masterwid);
+		break;
+	default:
+		pr_info("xtalk:%x unknown widget (0x%x)\n", wid, partnum);
+		break;
+	}
+}
+
+static int __init ip30_xtalk_init(void)
+{
+	int i;
+
+	/*
+	 * Walk widget IDs backwards so that BaseIO is probed first.  This
+	 * ensures that the BaseIO IOC3 is always detected as eth0.
+	 */
+	for (i = IP30_WIDGET_PCI_BASE; i > IP30_WIDGET_HEART; i--)
+		xtalk_init_widget(i, IP30_WIDGET_HEART);
+
+	return 0;
+}
+
+arch_initcall(ip30_xtalk_init);