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().
  ...

1 files changed, 843 insertions, 0 deletions

diff --git a/arch/sparc/kernel/prom_irqtrans.c b/arch/sparc/kernel/prom_irqtrans.c
new file mode 100644
index 000000000..28aff1c52
--- /dev/null
+++ b/arch/sparc/kernel/prom_irqtrans.c
@@ -0,0 +1,843 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/init.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+
+#include <asm/oplib.h>
+#include <asm/prom.h>
+#include <asm/irq.h>
+#include <asm/upa.h>
+
+#include "prom.h"
+
+#ifdef CONFIG_PCI
+/* PSYCHO interrupt mapping support. */
+#define PSYCHO_IMAP_A_SLOT0	0x0c00UL
+#define PSYCHO_IMAP_B_SLOT0	0x0c20UL
+static unsigned long psycho_pcislot_imap_offset(unsigned long ino)
+{
+	unsigned int bus =  (ino & 0x10) >> 4;
+	unsigned int slot = (ino & 0x0c) >> 2;
+
+	if (bus == 0)
+		return PSYCHO_IMAP_A_SLOT0 + (slot * 8);
+	else
+		return PSYCHO_IMAP_B_SLOT0 + (slot * 8);
+}
+
+#define PSYCHO_OBIO_IMAP_BASE	0x1000UL
+
+#define PSYCHO_ONBOARD_IRQ_BASE		0x20
+#define psycho_onboard_imap_offset(__ino) \
+	(PSYCHO_OBIO_IMAP_BASE + (((__ino) & 0x1f) << 3))
+
+#define PSYCHO_ICLR_A_SLOT0	0x1400UL
+#define PSYCHO_ICLR_SCSI	0x1800UL
+
+#define psycho_iclr_offset(ino)					      \
+	((ino & 0x20) ? (PSYCHO_ICLR_SCSI + (((ino) & 0x1f) << 3)) :  \
+			(PSYCHO_ICLR_A_SLOT0 + (((ino) & 0x1f)<<3)))
+
+static unsigned int psycho_irq_build(struct device_node *dp,
+				     unsigned int ino,
+				     void *_data)
+{
+	unsigned long controller_regs = (unsigned long) _data;
+	unsigned long imap, iclr;
+	unsigned long imap_off, iclr_off;
+	int inofixup = 0;
+
+	ino &= 0x3f;
+	if (ino < PSYCHO_ONBOARD_IRQ_BASE) {
+		/* PCI slot */
+		imap_off = psycho_pcislot_imap_offset(ino);
+	} else {
+		/* Onboard device */
+		imap_off = psycho_onboard_imap_offset(ino);
+	}
+
+	/* Now build the IRQ bucket. */
+	imap = controller_regs + imap_off;
+
+	iclr_off = psycho_iclr_offset(ino);
+	iclr = controller_regs + iclr_off;
+
+	if ((ino & 0x20) == 0)
+		inofixup = ino & 0x03;
+
+	return build_irq(inofixup, iclr, imap);
+}
+
+static void __init psycho_irq_trans_init(struct device_node *dp)
+{
+	const struct linux_prom64_registers *regs;
+
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = psycho_irq_build;
+
+	regs = of_get_property(dp, "reg", NULL);
+	dp->irq_trans->data = (void *) regs[2].phys_addr;
+}
+
+#define sabre_read(__reg) \
+({	u64 __ret; \
+	__asm__ __volatile__("ldxa [%1] %2, %0" \
+			     : "=r" (__ret) \
+			     : "r" (__reg), "i" (ASI_PHYS_BYPASS_EC_E) \
+			     : "memory"); \
+	__ret; \
+})
+
+struct sabre_irq_data {
+	unsigned long controller_regs;
+	unsigned int pci_first_busno;
+};
+#define SABRE_CONFIGSPACE	0x001000000UL
+#define SABRE_WRSYNC		0x1c20UL
+
+#define SABRE_CONFIG_BASE(CONFIG_SPACE)	\
+	(CONFIG_SPACE | (1UL << 24))
+#define SABRE_CONFIG_ENCODE(BUS, DEVFN, REG)	\
+	(((unsigned long)(BUS)   << 16) |	\
+	 ((unsigned long)(DEVFN) << 8)  |	\
+	 ((unsigned long)(REG)))
+
+/* When a device lives behind a bridge deeper in the PCI bus topology
+ * than APB, a special sequence must run to make sure all pending DMA
+ * transfers at the time of IRQ delivery are visible in the coherency
+ * domain by the cpu.  This sequence is to perform a read on the far
+ * side of the non-APB bridge, then perform a read of Sabre's DMA
+ * write-sync register.
+ */
+static void sabre_wsync_handler(unsigned int ino, void *_arg1, void *_arg2)
+{
+	unsigned int phys_hi = (unsigned int) (unsigned long) _arg1;
+	struct sabre_irq_data *irq_data = _arg2;
+	unsigned long controller_regs = irq_data->controller_regs;
+	unsigned long sync_reg = controller_regs + SABRE_WRSYNC;
+	unsigned long config_space = controller_regs + SABRE_CONFIGSPACE;
+	unsigned int bus, devfn;
+	u16 _unused;
+
+	config_space = SABRE_CONFIG_BASE(config_space);
+
+	bus = (phys_hi >> 16) & 0xff;
+	devfn = (phys_hi >> 8) & 0xff;
+
+	config_space |= SABRE_CONFIG_ENCODE(bus, devfn, 0x00);
+
+	__asm__ __volatile__("membar #Sync\n\t"
+			     "lduha [%1] %2, %0\n\t"
+			     "membar #Sync"
+			     : "=r" (_unused)
+			     : "r" ((u16 *) config_space),
+			       "i" (ASI_PHYS_BYPASS_EC_E_L)
+			     : "memory");
+
+	sabre_read(sync_reg);
+}
+
+#define SABRE_IMAP_A_SLOT0	0x0c00UL
+#define SABRE_IMAP_B_SLOT0	0x0c20UL
+#define SABRE_ICLR_A_SLOT0	0x1400UL
+#define SABRE_ICLR_B_SLOT0	0x1480UL
+#define SABRE_ICLR_SCSI		0x1800UL
+#define SABRE_ICLR_ETH		0x1808UL
+#define SABRE_ICLR_BPP		0x1810UL
+#define SABRE_ICLR_AU_REC	0x1818UL
+#define SABRE_ICLR_AU_PLAY	0x1820UL
+#define SABRE_ICLR_PFAIL	0x1828UL
+#define SABRE_ICLR_KMS		0x1830UL
+#define SABRE_ICLR_FLPY		0x1838UL
+#define SABRE_ICLR_SHW		0x1840UL
+#define SABRE_ICLR_KBD		0x1848UL
+#define SABRE_ICLR_MS		0x1850UL
+#define SABRE_ICLR_SER		0x1858UL
+#define SABRE_ICLR_UE		0x1870UL
+#define SABRE_ICLR_CE		0x1878UL
+#define SABRE_ICLR_PCIERR	0x1880UL
+
+static unsigned long sabre_pcislot_imap_offset(unsigned long ino)
+{
+	unsigned int bus =  (ino & 0x10) >> 4;
+	unsigned int slot = (ino & 0x0c) >> 2;
+
+	if (bus == 0)
+		return SABRE_IMAP_A_SLOT0 + (slot * 8);
+	else
+		return SABRE_IMAP_B_SLOT0 + (slot * 8);
+}
+
+#define SABRE_OBIO_IMAP_BASE	0x1000UL
+#define SABRE_ONBOARD_IRQ_BASE	0x20
+#define sabre_onboard_imap_offset(__ino) \
+	(SABRE_OBIO_IMAP_BASE + (((__ino) & 0x1f) << 3))
+
+#define sabre_iclr_offset(ino)					      \
+	((ino & 0x20) ? (SABRE_ICLR_SCSI + (((ino) & 0x1f) << 3)) :  \
+			(SABRE_ICLR_A_SLOT0 + (((ino) & 0x1f)<<3)))
+
+static int sabre_device_needs_wsync(struct device_node *dp)
+{
+	struct device_node *parent = dp->parent;
+	const char *parent_model, *parent_compat;
+
+	/* This traversal up towards the root is meant to
+	 * handle two cases:
+	 *
+	 * 1) non-PCI bus sitting under PCI, such as 'ebus'
+	 * 2) the PCI controller interrupts themselves, which
+	 *    will use the sabre_irq_build but do not need
+	 *    the DMA synchronization handling
+	 */
+	while (parent) {
+		if (of_node_is_type(parent, "pci"))
+			break;
+		parent = parent->parent;
+	}
+
+	if (!parent)
+		return 0;
+
+	parent_model = of_get_property(parent,
+				       "model", NULL);
+	if (parent_model &&
+	    (!strcmp(parent_model, "SUNW,sabre") ||
+	     !strcmp(parent_model, "SUNW,simba")))
+		return 0;
+
+	parent_compat = of_get_property(parent,
+					"compatible", NULL);
+	if (parent_compat &&
+	    (!strcmp(parent_compat, "pci108e,a000") ||
+	     !strcmp(parent_compat, "pci108e,a001")))
+		return 0;
+
+	return 1;
+}
+
+static unsigned int sabre_irq_build(struct device_node *dp,
+				    unsigned int ino,
+				    void *_data)
+{
+	struct sabre_irq_data *irq_data = _data;
+	unsigned long controller_regs = irq_data->controller_regs;
+	const struct linux_prom_pci_registers *regs;
+	unsigned long imap, iclr;
+	unsigned long imap_off, iclr_off;
+	int inofixup = 0;
+	int irq;
+
+	ino &= 0x3f;
+	if (ino < SABRE_ONBOARD_IRQ_BASE) {
+		/* PCI slot */
+		imap_off = sabre_pcislot_imap_offset(ino);
+	} else {
+		/* onboard device */
+		imap_off = sabre_onboard_imap_offset(ino);
+	}
+
+	/* Now build the IRQ bucket. */
+	imap = controller_regs + imap_off;
+
+	iclr_off = sabre_iclr_offset(ino);
+	iclr = controller_regs + iclr_off;
+
+	if ((ino & 0x20) == 0)
+		inofixup = ino & 0x03;
+
+	irq = build_irq(inofixup, iclr, imap);
+
+	/* If the parent device is a PCI<->PCI bridge other than
+	 * APB, we have to install a pre-handler to ensure that
+	 * all pending DMA is drained before the interrupt handler
+	 * is run.
+	 */
+	regs = of_get_property(dp, "reg", NULL);
+	if (regs && sabre_device_needs_wsync(dp)) {
+		irq_install_pre_handler(irq,
+					sabre_wsync_handler,
+					(void *) (long) regs->phys_hi,
+					(void *) irq_data);
+	}
+
+	return irq;
+}
+
+static void __init sabre_irq_trans_init(struct device_node *dp)
+{
+	const struct linux_prom64_registers *regs;
+	struct sabre_irq_data *irq_data;
+	const u32 *busrange;
+
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = sabre_irq_build;
+
+	irq_data = prom_early_alloc(sizeof(struct sabre_irq_data));
+
+	regs = of_get_property(dp, "reg", NULL);
+	irq_data->controller_regs = regs[0].phys_addr;
+
+	busrange = of_get_property(dp, "bus-range", NULL);
+	irq_data->pci_first_busno = busrange[0];
+
+	dp->irq_trans->data = irq_data;
+}
+
+/* SCHIZO interrupt mapping support.  Unlike Psycho, for this controller the
+ * imap/iclr registers are per-PBM.
+ */
+#define SCHIZO_IMAP_BASE	0x1000UL
+#define SCHIZO_ICLR_BASE	0x1400UL
+
+static unsigned long schizo_imap_offset(unsigned long ino)
+{
+	return SCHIZO_IMAP_BASE + (ino * 8UL);
+}
+
+static unsigned long schizo_iclr_offset(unsigned long ino)
+{
+	return SCHIZO_ICLR_BASE + (ino * 8UL);
+}
+
+static unsigned long schizo_ino_to_iclr(unsigned long pbm_regs,
+					unsigned int ino)
+{
+
+	return pbm_regs + schizo_iclr_offset(ino);
+}
+
+static unsigned long schizo_ino_to_imap(unsigned long pbm_regs,
+					unsigned int ino)
+{
+	return pbm_regs + schizo_imap_offset(ino);
+}
+
+#define schizo_read(__reg) \
+({	u64 __ret; \
+	__asm__ __volatile__("ldxa [%1] %2, %0" \
+			     : "=r" (__ret) \
+			     : "r" (__reg), "i" (ASI_PHYS_BYPASS_EC_E) \
+			     : "memory"); \
+	__ret; \
+})
+#define schizo_write(__reg, __val) \
+	__asm__ __volatile__("stxa %0, [%1] %2" \
+			     : /* no outputs */ \
+			     : "r" (__val), "r" (__reg), \
+			       "i" (ASI_PHYS_BYPASS_EC_E) \
+			     : "memory")
+
+static void tomatillo_wsync_handler(unsigned int ino, void *_arg1, void *_arg2)
+{
+	unsigned long sync_reg = (unsigned long) _arg2;
+	u64 mask = 1UL << (ino & IMAP_INO);
+	u64 val;
+	int limit;
+
+	schizo_write(sync_reg, mask);
+
+	limit = 100000;
+	val = 0;
+	while (--limit) {
+		val = schizo_read(sync_reg);
+		if (!(val & mask))
+			break;
+	}
+	if (limit <= 0) {
+		printk("tomatillo_wsync_handler: DMA won't sync [%llx:%llx]\n",
+		       val, mask);
+	}
+
+	if (_arg1) {
+		static unsigned char cacheline[64]
+			__attribute__ ((aligned (64)));
+
+		__asm__ __volatile__("rd %%fprs, %0\n\t"
+				     "or %0, %4, %1\n\t"
+				     "wr %1, 0x0, %%fprs\n\t"
+				     "stda %%f0, [%5] %6\n\t"
+				     "wr %0, 0x0, %%fprs\n\t"
+				     "membar #Sync"
+				     : "=&r" (mask), "=&r" (val)
+				     : "0" (mask), "1" (val),
+				     "i" (FPRS_FEF), "r" (&cacheline[0]),
+				     "i" (ASI_BLK_COMMIT_P));
+	}
+}
+
+struct schizo_irq_data {
+	unsigned long pbm_regs;
+	unsigned long sync_reg;
+	u32 portid;
+	int chip_version;
+};
+
+static unsigned int schizo_irq_build(struct device_node *dp,
+				     unsigned int ino,
+				     void *_data)
+{
+	struct schizo_irq_data *irq_data = _data;
+	unsigned long pbm_regs = irq_data->pbm_regs;
+	unsigned long imap, iclr;
+	int ign_fixup;
+	int irq;
+	int is_tomatillo;
+
+	ino &= 0x3f;
+
+	/* Now build the IRQ bucket. */
+	imap = schizo_ino_to_imap(pbm_regs, ino);
+	iclr = schizo_ino_to_iclr(pbm_regs, ino);
+
+	/* On Schizo, no inofixup occurs.  This is because each
+	 * INO has it's own IMAP register.  On Psycho and Sabre
+	 * there is only one IMAP register for each PCI slot even
+	 * though four different INOs can be generated by each
+	 * PCI slot.
+	 *
+	 * But, for JBUS variants (essentially, Tomatillo), we have
+	 * to fixup the lowest bit of the interrupt group number.
+	 */
+	ign_fixup = 0;
+
+	is_tomatillo = (irq_data->sync_reg != 0UL);
+
+	if (is_tomatillo) {
+		if (irq_data->portid & 1)
+			ign_fixup = (1 << 6);
+	}
+
+	irq = build_irq(ign_fixup, iclr, imap);
+
+	if (is_tomatillo) {
+		irq_install_pre_handler(irq,
+					tomatillo_wsync_handler,
+					((irq_data->chip_version <= 4) ?
+					 (void *) 1 : (void *) 0),
+					(void *) irq_data->sync_reg);
+	}
+
+	return irq;
+}
+
+static void __init __schizo_irq_trans_init(struct device_node *dp,
+					   int is_tomatillo)
+{
+	const struct linux_prom64_registers *regs;
+	struct schizo_irq_data *irq_data;
+
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = schizo_irq_build;
+
+	irq_data = prom_early_alloc(sizeof(struct schizo_irq_data));
+
+	regs = of_get_property(dp, "reg", NULL);
+	dp->irq_trans->data = irq_data;
+
+	irq_data->pbm_regs = regs[0].phys_addr;
+	if (is_tomatillo)
+		irq_data->sync_reg = regs[3].phys_addr + 0x1a18UL;
+	else
+		irq_data->sync_reg = 0UL;
+	irq_data->portid = of_getintprop_default(dp, "portid", 0);
+	irq_data->chip_version = of_getintprop_default(dp, "version#", 0);
+}
+
+static void __init schizo_irq_trans_init(struct device_node *dp)
+{
+	__schizo_irq_trans_init(dp, 0);
+}
+
+static void __init tomatillo_irq_trans_init(struct device_node *dp)
+{
+	__schizo_irq_trans_init(dp, 1);
+}
+
+static unsigned int pci_sun4v_irq_build(struct device_node *dp,
+					unsigned int devino,
+					void *_data)
+{
+	u32 devhandle = (u32) (unsigned long) _data;
+
+	return sun4v_build_irq(devhandle, devino);
+}
+
+static void __init pci_sun4v_irq_trans_init(struct device_node *dp)
+{
+	const struct linux_prom64_registers *regs;
+
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = pci_sun4v_irq_build;
+
+	regs = of_get_property(dp, "reg", NULL);
+	dp->irq_trans->data = (void *) (unsigned long)
+		((regs->phys_addr >> 32UL) & 0x0fffffff);
+}
+
+struct fire_irq_data {
+	unsigned long pbm_regs;
+	u32 portid;
+};
+
+#define FIRE_IMAP_BASE	0x001000
+#define FIRE_ICLR_BASE	0x001400
+
+static unsigned long fire_imap_offset(unsigned long ino)
+{
+	return FIRE_IMAP_BASE + (ino * 8UL);
+}
+
+static unsigned long fire_iclr_offset(unsigned long ino)
+{
+	return FIRE_ICLR_BASE + (ino * 8UL);
+}
+
+static unsigned long fire_ino_to_iclr(unsigned long pbm_regs,
+					    unsigned int ino)
+{
+	return pbm_regs + fire_iclr_offset(ino);
+}
+
+static unsigned long fire_ino_to_imap(unsigned long pbm_regs,
+					    unsigned int ino)
+{
+	return pbm_regs + fire_imap_offset(ino);
+}
+
+static unsigned int fire_irq_build(struct device_node *dp,
+					 unsigned int ino,
+					 void *_data)
+{
+	struct fire_irq_data *irq_data = _data;
+	unsigned long pbm_regs = irq_data->pbm_regs;
+	unsigned long imap, iclr;
+	unsigned long int_ctrlr;
+
+	ino &= 0x3f;
+
+	/* Now build the IRQ bucket. */
+	imap = fire_ino_to_imap(pbm_regs, ino);
+	iclr = fire_ino_to_iclr(pbm_regs, ino);
+
+	/* Set the interrupt controller number.  */
+	int_ctrlr = 1 << 6;
+	upa_writeq(int_ctrlr, imap);
+
+	/* The interrupt map registers do not have an INO field
+	 * like other chips do.  They return zero in the INO
+	 * field, and the interrupt controller number is controlled
+	 * in bits 6 to 9.  So in order for build_irq() to get
+	 * the INO right we pass it in as part of the fixup
+	 * which will get added to the map register zero value
+	 * read by build_irq().
+	 */
+	ino |= (irq_data->portid << 6);
+	ino -= int_ctrlr;
+	return build_irq(ino, iclr, imap);
+}
+
+static void __init fire_irq_trans_init(struct device_node *dp)
+{
+	const struct linux_prom64_registers *regs;
+	struct fire_irq_data *irq_data;
+
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = fire_irq_build;
+
+	irq_data = prom_early_alloc(sizeof(struct fire_irq_data));
+
+	regs = of_get_property(dp, "reg", NULL);
+	dp->irq_trans->data = irq_data;
+
+	irq_data->pbm_regs = regs[0].phys_addr;
+	irq_data->portid = of_getintprop_default(dp, "portid", 0);
+}
+#endif /* CONFIG_PCI */
+
+#ifdef CONFIG_SBUS
+/* INO number to IMAP register offset for SYSIO external IRQ's.
+ * This should conform to both Sunfire/Wildfire server and Fusion
+ * desktop designs.
+ */
+#define SYSIO_IMAP_SLOT0	0x2c00UL
+#define SYSIO_IMAP_SLOT1	0x2c08UL
+#define SYSIO_IMAP_SLOT2	0x2c10UL
+#define SYSIO_IMAP_SLOT3	0x2c18UL
+#define SYSIO_IMAP_SCSI		0x3000UL
+#define SYSIO_IMAP_ETH		0x3008UL
+#define SYSIO_IMAP_BPP		0x3010UL
+#define SYSIO_IMAP_AUDIO	0x3018UL
+#define SYSIO_IMAP_PFAIL	0x3020UL
+#define SYSIO_IMAP_KMS		0x3028UL
+#define SYSIO_IMAP_FLPY		0x3030UL
+#define SYSIO_IMAP_SHW		0x3038UL
+#define SYSIO_IMAP_KBD		0x3040UL
+#define SYSIO_IMAP_MS		0x3048UL
+#define SYSIO_IMAP_SER		0x3050UL
+#define SYSIO_IMAP_TIM0		0x3060UL
+#define SYSIO_IMAP_TIM1		0x3068UL
+#define SYSIO_IMAP_UE		0x3070UL
+#define SYSIO_IMAP_CE		0x3078UL
+#define SYSIO_IMAP_SBERR	0x3080UL
+#define SYSIO_IMAP_PMGMT	0x3088UL
+#define SYSIO_IMAP_GFX		0x3090UL
+#define SYSIO_IMAP_EUPA		0x3098UL
+
+#define bogon     ((unsigned long) -1)
+static unsigned long sysio_irq_offsets[] = {
+	/* SBUS Slot 0 --> 3, level 1 --> 7 */
+	SYSIO_IMAP_SLOT0, SYSIO_IMAP_SLOT0, SYSIO_IMAP_SLOT0, SYSIO_IMAP_SLOT0,
+	SYSIO_IMAP_SLOT0, SYSIO_IMAP_SLOT0, SYSIO_IMAP_SLOT0, SYSIO_IMAP_SLOT0,
+	SYSIO_IMAP_SLOT1, SYSIO_IMAP_SLOT1, SYSIO_IMAP_SLOT1, SYSIO_IMAP_SLOT1,
+	SYSIO_IMAP_SLOT1, SYSIO_IMAP_SLOT1, SYSIO_IMAP_SLOT1, SYSIO_IMAP_SLOT1,
+	SYSIO_IMAP_SLOT2, SYSIO_IMAP_SLOT2, SYSIO_IMAP_SLOT2, SYSIO_IMAP_SLOT2,
+	SYSIO_IMAP_SLOT2, SYSIO_IMAP_SLOT2, SYSIO_IMAP_SLOT2, SYSIO_IMAP_SLOT2,
+	SYSIO_IMAP_SLOT3, SYSIO_IMAP_SLOT3, SYSIO_IMAP_SLOT3, SYSIO_IMAP_SLOT3,
+	SYSIO_IMAP_SLOT3, SYSIO_IMAP_SLOT3, SYSIO_IMAP_SLOT3, SYSIO_IMAP_SLOT3,
+
+	/* Onboard devices (not relevant/used on SunFire). */
+	SYSIO_IMAP_SCSI,
+	SYSIO_IMAP_ETH,
+	SYSIO_IMAP_BPP,
+	bogon,
+	SYSIO_IMAP_AUDIO,
+	SYSIO_IMAP_PFAIL,
+	bogon,
+	bogon,
+	SYSIO_IMAP_KMS,
+	SYSIO_IMAP_FLPY,
+	SYSIO_IMAP_SHW,
+	SYSIO_IMAP_KBD,
+	SYSIO_IMAP_MS,
+	SYSIO_IMAP_SER,
+	bogon,
+	bogon,
+	SYSIO_IMAP_TIM0,
+	SYSIO_IMAP_TIM1,
+	bogon,
+	bogon,
+	SYSIO_IMAP_UE,
+	SYSIO_IMAP_CE,
+	SYSIO_IMAP_SBERR,
+	SYSIO_IMAP_PMGMT,
+	SYSIO_IMAP_GFX,
+	SYSIO_IMAP_EUPA,
+};
+
+#undef bogon
+
+#define NUM_SYSIO_OFFSETS ARRAY_SIZE(sysio_irq_offsets)
+
+/* Convert Interrupt Mapping register pointer to associated
+ * Interrupt Clear register pointer, SYSIO specific version.
+ */
+#define SYSIO_ICLR_UNUSED0	0x3400UL
+#define SYSIO_ICLR_SLOT0	0x3408UL
+#define SYSIO_ICLR_SLOT1	0x3448UL
+#define SYSIO_ICLR_SLOT2	0x3488UL
+#define SYSIO_ICLR_SLOT3	0x34c8UL
+static unsigned long sysio_imap_to_iclr(unsigned long imap)
+{
+	unsigned long diff = SYSIO_ICLR_UNUSED0 - SYSIO_IMAP_SLOT0;
+	return imap + diff;
+}
+
+static unsigned int sbus_of_build_irq(struct device_node *dp,
+				      unsigned int ino,
+				      void *_data)
+{
+	unsigned long reg_base = (unsigned long) _data;
+	const struct linux_prom_registers *regs;
+	unsigned long imap, iclr;
+	int sbus_slot = 0;
+	int sbus_level = 0;
+
+	ino &= 0x3f;
+
+	regs = of_get_property(dp, "reg", NULL);
+	if (regs)
+		sbus_slot = regs->which_io;
+
+	if (ino < 0x20)
+		ino += (sbus_slot * 8);
+
+	imap = sysio_irq_offsets[ino];
+	if (imap == ((unsigned long)-1)) {
+		prom_printf("get_irq_translations: Bad SYSIO INO[%x]\n",
+			    ino);
+		prom_halt();
+	}
+	imap += reg_base;
+
+	/* SYSIO inconsistency.  For external SLOTS, we have to select
+	 * the right ICLR register based upon the lower SBUS irq level
+	 * bits.
+	 */
+	if (ino >= 0x20) {
+		iclr = sysio_imap_to_iclr(imap);
+	} else {
+		sbus_level = ino & 0x7;
+
+		switch(sbus_slot) {
+		case 0:
+			iclr = reg_base + SYSIO_ICLR_SLOT0;
+			break;
+		case 1:
+			iclr = reg_base + SYSIO_ICLR_SLOT1;
+			break;
+		case 2:
+			iclr = reg_base + SYSIO_ICLR_SLOT2;
+			break;
+		default:
+		case 3:
+			iclr = reg_base + SYSIO_ICLR_SLOT3;
+			break;
+		}
+
+		iclr += ((unsigned long)sbus_level - 1UL) * 8UL;
+	}
+	return build_irq(sbus_level, iclr, imap);
+}
+
+static void __init sbus_irq_trans_init(struct device_node *dp)
+{
+	const struct linux_prom64_registers *regs;
+
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = sbus_of_build_irq;
+
+	regs = of_get_property(dp, "reg", NULL);
+	dp->irq_trans->data = (void *) (unsigned long) regs->phys_addr;
+}
+#endif /* CONFIG_SBUS */
+
+
+static unsigned int central_build_irq(struct device_node *dp,
+				      unsigned int ino,
+				      void *_data)
+{
+	struct device_node *central_dp = _data;
+	struct platform_device *central_op = of_find_device_by_node(central_dp);
+	struct resource *res;
+	unsigned long imap, iclr;
+	u32 tmp;
+
+	if (of_node_name_eq(dp, "eeprom")) {
+		res = &central_op->resource[5];
+	} else if (of_node_name_eq(dp, "zs")) {
+		res = &central_op->resource[4];
+	} else if (of_node_name_eq(dp, "clock-board")) {
+		res = &central_op->resource[3];
+	} else {
+		return ino;
+	}
+
+	imap = res->start + 0x00UL;
+	iclr = res->start + 0x10UL;
+
+	/* Set the INO state to idle, and disable.  */
+	upa_writel(0, iclr);
+	upa_readl(iclr);
+
+	tmp = upa_readl(imap);
+	tmp &= ~0x80000000;
+	upa_writel(tmp, imap);
+
+	return build_irq(0, iclr, imap);
+}
+
+static void __init central_irq_trans_init(struct device_node *dp)
+{
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = central_build_irq;
+
+	dp->irq_trans->data = dp;
+}
+
+struct irq_trans {
+	const char *name;
+	void (*init)(struct device_node *);
+};
+
+#ifdef CONFIG_PCI
+static struct irq_trans __initdata pci_irq_trans_table[] = {
+	{ "SUNW,sabre", sabre_irq_trans_init },
+	{ "pci108e,a000", sabre_irq_trans_init },
+	{ "pci108e,a001", sabre_irq_trans_init },
+	{ "SUNW,psycho", psycho_irq_trans_init },
+	{ "pci108e,8000", psycho_irq_trans_init },
+	{ "SUNW,schizo", schizo_irq_trans_init },
+	{ "pci108e,8001", schizo_irq_trans_init },
+	{ "SUNW,schizo+", schizo_irq_trans_init },
+	{ "pci108e,8002", schizo_irq_trans_init },
+	{ "SUNW,tomatillo", tomatillo_irq_trans_init },
+	{ "pci108e,a801", tomatillo_irq_trans_init },
+	{ "SUNW,sun4v-pci", pci_sun4v_irq_trans_init },
+	{ "pciex108e,80f0", fire_irq_trans_init },
+};
+#endif
+
+static unsigned int sun4v_vdev_irq_build(struct device_node *dp,
+					 unsigned int devino,
+					 void *_data)
+{
+	u32 devhandle = (u32) (unsigned long) _data;
+
+	return sun4v_build_irq(devhandle, devino);
+}
+
+static void __init sun4v_vdev_irq_trans_init(struct device_node *dp)
+{
+	const struct linux_prom64_registers *regs;
+
+	dp->irq_trans = prom_early_alloc(sizeof(struct of_irq_controller));
+	dp->irq_trans->irq_build = sun4v_vdev_irq_build;
+
+	regs = of_get_property(dp, "reg", NULL);
+	dp->irq_trans->data = (void *) (unsigned long)
+		((regs->phys_addr >> 32UL) & 0x0fffffff);
+}
+
+void __init irq_trans_init(struct device_node *dp)
+{
+#ifdef CONFIG_PCI
+	const char *model;
+	int i;
+#endif
+
+#ifdef CONFIG_PCI
+	model = of_get_property(dp, "model", NULL);
+	if (!model)
+		model = of_get_property(dp, "compatible", NULL);
+	if (model) {
+		for (i = 0; i < ARRAY_SIZE(pci_irq_trans_table); i++) {
+			struct irq_trans *t = &pci_irq_trans_table[i];
+
+			if (!strcmp(model, t->name)) {
+				t->init(dp);
+				return;
+			}
+		}
+	}
+#endif
+#ifdef CONFIG_SBUS
+	if (of_node_name_eq(dp, "sbus") ||
+	    of_node_name_eq(dp, "sbi")) {
+		sbus_irq_trans_init(dp);
+		return;
+	}
+#endif
+	if (of_node_name_eq(dp, "fhc") &&
+	    of_node_name_eq(dp->parent, "central")) {
+		central_irq_trans_init(dp);
+		return;
+	}
+	if (of_node_name_eq(dp, "virtual-devices") ||
+	    of_node_name_eq(dp, "niu")) {
+		sun4v_vdev_irq_trans_init(dp);
+		return;
+	}
+}