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, 366 insertions, 0 deletions

diff --git a/arch/sparc/kernel/wof.S b/arch/sparc/kernel/wof.S
new file mode 100644
index 000000000..96a3a1124
--- /dev/null
+++ b/arch/sparc/kernel/wof.S
@@ -0,0 +1,366 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * wof.S: Sparc window overflow handler.
+ *
+ * Copyright (C) 1995 David S. Miller (davem@caip.rutgers.edu)
+ */
+
+#include <asm/contregs.h>
+#include <asm/page.h>
+#include <asm/ptrace.h>
+#include <asm/psr.h>
+#include <asm/smp.h>
+#include <asm/asi.h>
+#include <asm/winmacro.h>
+#include <asm/asmmacro.h>
+#include <asm/thread_info.h>
+
+/* WARNING: This routine is hairy and _very_ complicated, but it
+ *          must be as fast as possible as it handles the allocation
+ *          of register windows to the user and kernel.  If you touch
+ *          this code be _very_ careful as many other pieces of the
+ *          kernel depend upon how this code behaves.  You have been
+ *          duly warned...
+ */
+
+/* We define macro's for registers which have a fixed
+ * meaning throughout this entire routine.  The 'T' in
+ * the comments mean that the register can only be
+ * accessed when in the 'trap' window, 'G' means
+ * accessible in any window.  Do not change these registers
+ * after they have been set, until you are ready to return
+ * from the trap.
+ */
+#define t_psr       l0 /* %psr at trap time                     T */
+#define t_pc        l1 /* PC for trap return                    T */
+#define t_npc       l2 /* NPC for trap return                   T */
+#define t_wim       l3 /* %wim at trap time                     T */
+#define saved_g5    l5 /* Global save register                  T */
+#define saved_g6    l6 /* Global save register                  T */
+#define curptr      g6 /* Gets set to 'current' then stays      G */
+
+/* Now registers whose values can change within the handler.      */
+#define twin_tmp    l4 /* Temp reg, only usable in trap window  T */
+#define glob_tmp    g5 /* Global temporary reg, usable anywhere G */
+
+	.text
+	.align	4
+	/* BEGINNING OF PATCH INSTRUCTIONS */
+	/* On a 7-window Sparc the boot code patches spnwin_*
+	 * instructions with the following ones.
+	 */
+	.globl	spnwin_patch1_7win, spnwin_patch2_7win, spnwin_patch3_7win
+spnwin_patch1_7win:	sll	%t_wim, 6, %glob_tmp
+spnwin_patch2_7win:	and	%glob_tmp, 0x7f, %glob_tmp
+spnwin_patch3_7win:	and	%twin_tmp, 0x7f, %twin_tmp
+	/* END OF PATCH INSTRUCTIONS */
+
+	/* The trap entry point has done the following:
+	 *
+	 * rd    %psr, %l0
+	 * rd    %wim, %l3
+	 * b     spill_window_entry
+	 * andcc %l0, PSR_PS, %g0
+	 */
+
+	/* Datum current_thread_info->uwinmask contains at all times a bitmask
+	 * where if any user windows are active, at least one bit will
+	 * be set in to mask.  If no user windows are active, the bitmask
+	 * will be all zeroes.
+	 */
+	.globl	spill_window_entry 
+	.globl	spnwin_patch1, spnwin_patch2, spnwin_patch3
+spill_window_entry:
+	/* LOCATION: Trap Window */
+
+	mov	%g5, %saved_g5		! save away global temp register
+	mov	%g6, %saved_g6		! save away 'current' ptr register
+
+	/* Compute what the new %wim will be if we save the
+	 * window properly in this trap handler.
+	 *
+	 * newwim = ((%wim>>1) | (%wim<<(nwindows - 1)));
+	 */
+		srl	%t_wim, 0x1, %twin_tmp
+spnwin_patch1:	sll	%t_wim, 7, %glob_tmp
+		or	%glob_tmp, %twin_tmp, %glob_tmp
+spnwin_patch2:	and	%glob_tmp, 0xff, %glob_tmp
+
+	/* The trap entry point has set the condition codes
+	 * up for us to see if this is from user or kernel.
+	 * Get the load of 'curptr' out of the way.
+	 */
+	LOAD_CURRENT(curptr, twin_tmp)
+
+	andcc	%t_psr, PSR_PS, %g0
+	be,a	spwin_fromuser				! all user wins, branch
+	 save	%g0, %g0, %g0				! Go where saving will occur
+	
+	/* See if any user windows are active in the set. */
+	ld	[%curptr + TI_UWINMASK], %twin_tmp	! grab win mask
+	orcc	%g0, %twin_tmp, %g0			! check for set bits
+	bne	spwin_exist_uwins			! yep, there are some
+	 andn	%twin_tmp, %glob_tmp, %twin_tmp		! compute new uwinmask
+
+	/* Save into the window which must be saved and do it.
+	 * Basically if we are here, this means that we trapped
+	 * from kernel mode with only kernel windows in the register
+	 * file.
+	 */
+	save	%g0, %g0, %g0		! save into the window to stash away
+	wr	%glob_tmp, 0x0, %wim	! set new %wim, this is safe now
+
+spwin_no_userwins_from_kernel:
+	/* LOCATION: Window to be saved */
+
+	STORE_WINDOW(sp)		! stash the window
+	restore	%g0, %g0, %g0		! go back into trap window
+
+	/* LOCATION: Trap window */
+	mov	%saved_g5, %g5		! restore %glob_tmp
+	mov	%saved_g6, %g6		! restore %curptr
+	wr	%t_psr, 0x0, %psr	! restore condition codes in %psr
+	WRITE_PAUSE			! waste some time
+	jmp	%t_pc			! Return from trap
+	rett	%t_npc			! we are done
+
+spwin_exist_uwins:
+	/* LOCATION: Trap window */
+
+	/* Wow, user windows have to be dealt with, this is dirty
+	 * and messy as all hell.  And difficult to follow if you
+	 * are approaching the infamous register window trap handling
+	 * problem for the first time. DON'T LOOK!
+	 *
+	 * Note that how the execution path works out, the new %wim
+	 * will be left for us in the global temporary register,
+	 * %glob_tmp.  We cannot set the new %wim first because we
+	 * need to save into the appropriate window without inducing
+	 * a trap (traps are off, we'd get a watchdog wheee)...
+	 * But first, store the new user window mask calculated
+	 * above.
+	 */
+	st	%twin_tmp, [%curptr + TI_UWINMASK]
+	save	%g0, %g0, %g0		! Go to where the saving will occur
+
+spwin_fromuser:
+	/* LOCATION: Window to be saved */
+	wr	%glob_tmp, 0x0, %wim	! Now it is safe to set new %wim
+
+	/* LOCATION: Window to be saved */
+
+	/* This instruction branches to a routine which will check
+	 * to validity of the users stack pointer by whatever means
+	 * are necessary.  This means that this is architecture
+	 * specific and thus this branch instruction will need to
+	 * be patched at boot time once the machine type is known.
+	 * This routine _shall not_ touch %curptr under any
+	 * circumstances whatsoever!  It will branch back to the
+	 * label 'spwin_good_ustack' if the stack is ok but still
+	 * needs to be dumped (SRMMU for instance will not need to
+	 * do this) or 'spwin_finish_up' if the stack is ok and the
+	 * registers have already been saved.  If the stack is found
+	 * to be bogus for some reason the routine shall branch to
+	 * the label 'spwin_user_stack_is_bolixed' which will take
+	 * care of things at that point.
+	 */
+	b	spwin_srmmu_stackchk
+	 andcc	%sp, 0x7, %g0
+
+spwin_good_ustack:
+	/* LOCATION: Window to be saved */
+
+	/* The users stack is ok and we can safely save it at
+	 * %sp.
+	 */
+	STORE_WINDOW(sp)
+
+spwin_finish_up:
+	restore	%g0, %g0, %g0		/* Back to trap window. */
+
+	/* LOCATION: Trap window */
+
+	/* We have spilled successfully, and we have properly stored
+	 * the appropriate window onto the stack.
+	 */
+
+	/* Restore saved globals */
+	mov	%saved_g5, %g5
+	mov	%saved_g6, %g6
+
+	wr	%t_psr, 0x0, %psr
+	WRITE_PAUSE
+	jmp	%t_pc
+	rett	%t_npc
+
+spwin_user_stack_is_bolixed:
+	/* LOCATION: Window to be saved */
+
+	/* Wheee, user has trashed his/her stack.  We have to decide
+	 * how to proceed based upon whether we came from kernel mode
+	 * or not.  If we came from kernel mode, toss the window into
+	 * a special buffer and proceed, the kernel _needs_ a window
+	 * and we could be in an interrupt handler so timing is crucial.
+	 * If we came from user land we build a full stack frame and call
+	 * c-code to gun down the process.
+	 */
+	rd	%psr, %glob_tmp
+	andcc	%glob_tmp, PSR_PS, %g0
+	bne	spwin_bad_ustack_from_kernel
+	 nop
+
+	/* Oh well, throw this one window into the per-task window
+	 * buffer, the first one.
+	 */
+	st	%sp, [%curptr + TI_RWIN_SPTRS]
+	STORE_WINDOW(curptr + TI_REG_WINDOW)
+	restore	%g0, %g0, %g0
+
+	/* LOCATION: Trap Window */
+
+	/* Back in the trap window, update winbuffer save count. */
+	mov	1, %twin_tmp
+	st	%twin_tmp, [%curptr + TI_W_SAVED]
+
+		/* Compute new user window mask.  What we are basically
+		 * doing is taking two windows, the invalid one at trap
+		 * time and the one we attempted to throw onto the users
+		 * stack, and saying that everything else is an ok user
+		 * window.  umask = ((~(%t_wim | %wim)) & valid_wim_bits)
+		 */
+		rd	%wim, %twin_tmp
+		or	%twin_tmp, %t_wim, %twin_tmp
+		not	%twin_tmp
+spnwin_patch3:	and	%twin_tmp, 0xff, %twin_tmp	! patched on 7win Sparcs
+		st	%twin_tmp, [%curptr + TI_UWINMASK]
+
+#define STACK_OFFSET (THREAD_SIZE - TRACEREG_SZ - STACKFRAME_SZ)
+
+	sethi	%hi(STACK_OFFSET), %sp
+	or	%sp, %lo(STACK_OFFSET), %sp
+	add	%curptr, %sp, %sp
+
+	/* Restore the saved globals and build a pt_regs frame. */
+	mov	%saved_g5, %g5
+	mov	%saved_g6, %g6
+	STORE_PT_ALL(sp, t_psr, t_pc, t_npc, g1)
+
+	sethi	%hi(STACK_OFFSET), %g6
+	or	%g6, %lo(STACK_OFFSET), %g6
+	sub	%sp, %g6, %g6		! curptr
+
+	/* Turn on traps and call c-code to deal with it. */
+	wr	%t_psr, PSR_ET, %psr
+	nop
+	call	window_overflow_fault
+	 nop
+
+	/* Return from trap if C-code actually fixes things, if it
+	 * doesn't then we never get this far as the process will
+	 * be given the look of death from Commander Peanut.
+	 */
+	b	ret_trap_entry
+	 clr	%l6
+
+spwin_bad_ustack_from_kernel:
+	/* LOCATION: Window to be saved */
+
+	/* The kernel provoked a spill window trap, but the window we
+	 * need to save is a user one and the process has trashed its
+	 * stack pointer.  We need to be quick, so we throw it into
+	 * a per-process window buffer until we can properly handle
+	 * this later on.
+	 */
+	SAVE_BOLIXED_USER_STACK(curptr, glob_tmp)
+	restore	%g0, %g0, %g0
+
+	/* LOCATION: Trap window */
+
+	/* Restore globals, condition codes in the %psr and
+	 * return from trap.  Note, restoring %g6 when returning
+	 * to kernel mode is not necessarily these days. ;-)
+	 */
+	mov	%saved_g5, %g5
+	mov	%saved_g6, %g6
+
+	wr	%t_psr, 0x0, %psr
+	WRITE_PAUSE
+
+	jmp	%t_pc
+	rett	%t_npc
+
+/* Undefine the register macros which would only cause trouble
+ * if used below.  This helps find 'stupid' coding errors that
+ * produce 'odd' behavior.  The routines below are allowed to
+ * make usage of glob_tmp and t_psr so we leave them defined.
+ */
+#undef twin_tmp
+#undef curptr
+#undef t_pc
+#undef t_npc
+#undef t_wim
+#undef saved_g5
+#undef saved_g6
+
+/* Now come the per-architecture window overflow stack checking routines.
+ * As noted above %curptr cannot be touched by this routine at all.
+ */
+
+	/* This is a generic SRMMU routine.  As far as I know this
+	 * works for all current v8/srmmu implementations, we'll
+	 * see...
+	 */
+	.globl	spwin_srmmu_stackchk
+spwin_srmmu_stackchk:
+	/* LOCATION: Window to be saved on the stack */
+
+	/* Because of SMP concerns and speed we play a trick.
+	 * We disable fault traps in the MMU control register,
+	 * Execute the stores, then check the fault registers
+	 * to see what happens.  I can hear Linus now
+	 * "disgusting... broken hardware...".
+	 *
+	 * But first, check to see if the users stack has ended
+	 * up in kernel vma, then we would succeed for the 'wrong'
+	 * reason... ;(  Note that the 'sethi' below assumes the
+	 * kernel is page aligned, which should always be the case.
+	 */
+	/* Check results of callers andcc %sp, 0x7, %g0 */
+	bne	spwin_user_stack_is_bolixed
+	 sethi   %hi(PAGE_OFFSET), %glob_tmp
+	cmp	%glob_tmp, %sp
+	bleu	spwin_user_stack_is_bolixed
+	 mov	AC_M_SFSR, %glob_tmp
+
+	/* Clear the fault status and turn on the no_fault bit. */
+LEON_PI(lda	[%glob_tmp] ASI_LEON_MMUREGS, %g0)	! eat SFSR
+SUN_PI_(lda	[%glob_tmp] ASI_M_MMUREGS, %g0)		! eat SFSR
+
+LEON_PI(lda	[%g0] ASI_LEON_MMUREGS, %glob_tmp)	! read MMU control
+SUN_PI_(lda	[%g0] ASI_M_MMUREGS, %glob_tmp)		! read MMU control
+	or	%glob_tmp, 0x2, %glob_tmp		! or in no_fault bit
+LEON_PI(sta	%glob_tmp, [%g0] ASI_LEON_MMUREGS)	! set it
+SUN_PI_(sta	%glob_tmp, [%g0] ASI_M_MMUREGS)		! set it
+
+	/* Dump the registers and cross fingers. */
+	STORE_WINDOW(sp)
+
+	/* Clear the no_fault bit and check the status. */
+	andn	%glob_tmp, 0x2, %glob_tmp
+LEON_PI(sta	%glob_tmp, [%g0] ASI_LEON_MMUREGS)
+SUN_PI_(sta	%glob_tmp, [%g0] ASI_M_MMUREGS)
+
+	mov	AC_M_SFAR, %glob_tmp
+LEON_PI(lda	[%glob_tmp] ASI_LEON_MMUREGS, %g0)
+SUN_PI_(lda	[%glob_tmp] ASI_M_MMUREGS, %g0)
+
+	mov	AC_M_SFSR, %glob_tmp
+LEON_PI(lda	[%glob_tmp] ASI_LEON_MMUREGS, %glob_tmp)
+SUN_PI_(lda	[%glob_tmp] ASI_M_MMUREGS, %glob_tmp)
+	andcc	%glob_tmp, 0x2, %g0			! did we fault?
+	be,a	spwin_finish_up + 0x4			! cool beans, success
+	 restore %g0, %g0, %g0
+
+	rd	%psr, %glob_tmp
+	b	spwin_user_stack_is_bolixed + 0x4	! we faulted, ugh
+	 nop