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, 365 insertions, 0 deletions

diff --git a/arch/ia64/lib/memset.S b/arch/ia64/lib/memset.S
new file mode 100644
index 000000000..07a8b92c6
--- /dev/null
+++ b/arch/ia64/lib/memset.S
@@ -0,0 +1,365 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Optimized version of the standard memset() function.
+
+   Copyright (c) 2002 Hewlett-Packard Co/CERN
+	Sverre Jarp <Sverre.Jarp@cern.ch>
+
+   Return: dest
+
+   Inputs:
+        in0:    dest
+        in1:    value
+        in2:    count
+
+   The algorithm is fairly straightforward: set byte by byte until we
+   we get to a 16B-aligned address, then loop on 128 B chunks using an
+   early store as prefetching, then loop on 32B chucks, then clear remaining
+   words, finally clear remaining bytes.
+   Since a stf.spill f0 can store 16B in one go, we use this instruction
+   to get peak speed when value = 0.  */
+
+#include <asm/asmmacro.h>
+#include <asm/export.h>
+#undef ret
+
+#define dest		in0
+#define value		in1
+#define	cnt		in2
+
+#define tmp		r31
+#define save_lc		r30
+#define ptr0		r29
+#define ptr1		r28
+#define ptr2		r27
+#define ptr3		r26
+#define ptr9 		r24
+#define	loopcnt		r23
+#define linecnt		r22
+#define bytecnt		r21
+
+#define fvalue		f6
+
+// This routine uses only scratch predicate registers (p6 - p15)
+#define p_scr		p6			// default register for same-cycle branches
+#define p_nz		p7
+#define p_zr		p8
+#define p_unalgn	p9
+#define p_y		p11
+#define p_n		p12
+#define p_yy		p13
+#define p_nn		p14
+
+#define MIN1		15
+#define MIN1P1HALF	8
+#define LINE_SIZE	128
+#define LSIZE_SH        7			// shift amount
+#define PREF_AHEAD	8
+
+GLOBAL_ENTRY(memset)
+{ .mmi
+	.prologue
+	alloc	tmp = ar.pfs, 3, 0, 0, 0
+	lfetch.nt1 [dest]			//
+	.save   ar.lc, save_lc
+	mov.i	save_lc = ar.lc
+	.body
+} { .mmi
+	mov	ret0 = dest			// return value
+	cmp.ne	p_nz, p_zr = value, r0		// use stf.spill if value is zero
+	cmp.eq	p_scr, p0 = cnt, r0
+;; }
+{ .mmi
+	and	ptr2 = -(MIN1+1), dest		// aligned address
+	and	tmp = MIN1, dest		// prepare to check for correct alignment
+	tbit.nz p_y, p_n = dest, 0		// Do we have an odd address? (M_B_U)
+} { .mib
+	mov	ptr1 = dest
+	mux1	value = value, @brcst		// create 8 identical bytes in word
+(p_scr)	br.ret.dpnt.many rp			// return immediately if count = 0
+;; }
+{ .mib
+	cmp.ne	p_unalgn, p0 = tmp, r0		//
+} { .mib
+	sub	bytecnt = (MIN1+1), tmp		// NB: # of bytes to move is 1 higher than loopcnt
+	cmp.gt	p_scr, p0 = 16, cnt		// is it a minimalistic task?
+(p_scr)	br.cond.dptk.many .move_bytes_unaligned	// go move just a few (M_B_U)
+;; }
+{ .mmi
+(p_unalgn) add	ptr1 = (MIN1+1), ptr2		// after alignment
+(p_unalgn) add	ptr2 = MIN1P1HALF, ptr2		// after alignment
+(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 3	// should we do a st8 ?
+;; }
+{ .mib
+(p_y)	add	cnt = -8, cnt			//
+(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 2	// should we do a st4 ?
+} { .mib
+(p_y)	st8	[ptr2] = value,-4		//
+(p_n)	add	ptr2 = 4, ptr2			//
+;; }
+{ .mib
+(p_yy)	add	cnt = -4, cnt			//
+(p_unalgn) tbit.nz.unc p_y, p_n = bytecnt, 1	// should we do a st2 ?
+} { .mib
+(p_yy)	st4	[ptr2] = value,-2		//
+(p_nn)	add	ptr2 = 2, ptr2			//
+;; }
+{ .mmi
+	mov	tmp = LINE_SIZE+1		// for compare
+(p_y)	add	cnt = -2, cnt			//
+(p_unalgn) tbit.nz.unc p_yy, p_nn = bytecnt, 0	// should we do a st1 ?
+} { .mmi
+	setf.sig fvalue=value			// transfer value to FLP side
+(p_y)	st2	[ptr2] = value,-1		//
+(p_n)	add	ptr2 = 1, ptr2			//
+;; }
+
+{ .mmi
+(p_yy)	st1	[ptr2] = value 			//
+  	cmp.gt	p_scr, p0 = tmp, cnt		// is it a minimalistic task?
+} { .mbb
+(p_yy)	add	cnt = -1, cnt			//
+(p_scr)	br.cond.dpnt.many .fraction_of_line	// go move just a few
+;; }
+
+{ .mib
+	nop.m 0
+	shr.u	linecnt = cnt, LSIZE_SH
+(p_zr)	br.cond.dptk.many .l1b			// Jump to use stf.spill
+;; }
+
+	TEXT_ALIGN(32) // --------------------- //  L1A: store ahead into cache lines; fill later
+{ .mmi
+	and	tmp = -(LINE_SIZE), cnt		// compute end of range
+	mov	ptr9 = ptr1			// used for prefetching
+	and	cnt = (LINE_SIZE-1), cnt	// remainder
+} { .mmi
+	mov	loopcnt = PREF_AHEAD-1		// default prefetch loop
+	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	// check against actual value
+;; }
+{ .mmi
+(p_scr)	add	loopcnt = -1, linecnt		//
+	add	ptr2 = 8, ptr1			// start of stores (beyond prefetch stores)
+	add	ptr1 = tmp, ptr1		// first address beyond total range
+;; }
+{ .mmi
+	add	tmp = -1, linecnt		// next loop count
+	mov.i	ar.lc = loopcnt			//
+;; }
+.pref_l1a:
+{ .mib
+	stf8 [ptr9] = fvalue, 128		// Do stores one cache line apart
+	nop.i	0
+	br.cloop.dptk.few .pref_l1a
+;; }
+{ .mmi
+	add	ptr0 = 16, ptr2			// Two stores in parallel
+	mov.i	ar.lc = tmp			//
+;; }
+.l1ax:
+ { .mmi
+	stf8 [ptr2] = fvalue, 8
+	stf8 [ptr0] = fvalue, 8
+ ;; }
+ { .mmi
+	stf8 [ptr2] = fvalue, 24
+	stf8 [ptr0] = fvalue, 24
+ ;; }
+ { .mmi
+	stf8 [ptr2] = fvalue, 8
+	stf8 [ptr0] = fvalue, 8
+ ;; }
+ { .mmi
+	stf8 [ptr2] = fvalue, 24
+	stf8 [ptr0] = fvalue, 24
+ ;; }
+ { .mmi
+	stf8 [ptr2] = fvalue, 8
+	stf8 [ptr0] = fvalue, 8
+ ;; }
+ { .mmi
+	stf8 [ptr2] = fvalue, 24
+	stf8 [ptr0] = fvalue, 24
+ ;; }
+ { .mmi
+	stf8 [ptr2] = fvalue, 8
+	stf8 [ptr0] = fvalue, 32
+ 	cmp.lt	p_scr, p0 = ptr9, ptr1		// do we need more prefetching?
+ ;; }
+{ .mmb
+	stf8 [ptr2] = fvalue, 24
+(p_scr)	stf8 [ptr9] = fvalue, 128
+	br.cloop.dptk.few .l1ax
+;; }
+{ .mbb
+	cmp.le  p_scr, p0 = 8, cnt		// just a few bytes left ?
+(p_scr) br.cond.dpnt.many  .fraction_of_line	// Branch no. 2
+	br.cond.dpnt.many  .move_bytes_from_alignment	// Branch no. 3
+;; }
+
+	TEXT_ALIGN(32)
+.l1b:	// ------------------------------------ //  L1B: store ahead into cache lines; fill later
+{ .mmi
+	and	tmp = -(LINE_SIZE), cnt		// compute end of range
+	mov	ptr9 = ptr1			// used for prefetching
+	and	cnt = (LINE_SIZE-1), cnt	// remainder
+} { .mmi
+	mov	loopcnt = PREF_AHEAD-1		// default prefetch loop
+	cmp.gt	p_scr, p0 = PREF_AHEAD, linecnt	// check against actual value
+;; }
+{ .mmi
+(p_scr)	add	loopcnt = -1, linecnt
+	add	ptr2 = 16, ptr1			// start of stores (beyond prefetch stores)
+	add	ptr1 = tmp, ptr1		// first address beyond total range
+;; }
+{ .mmi
+	add	tmp = -1, linecnt		// next loop count
+	mov.i	ar.lc = loopcnt
+;; }
+.pref_l1b:
+{ .mib
+	stf.spill [ptr9] = f0, 128		// Do stores one cache line apart
+	nop.i   0
+	br.cloop.dptk.few .pref_l1b
+;; }
+{ .mmi
+	add	ptr0 = 16, ptr2			// Two stores in parallel
+	mov.i	ar.lc = tmp
+;; }
+.l1bx:
+ { .mmi
+	stf.spill [ptr2] = f0, 32
+	stf.spill [ptr0] = f0, 32
+ ;; }
+ { .mmi
+	stf.spill [ptr2] = f0, 32
+	stf.spill [ptr0] = f0, 32
+ ;; }
+ { .mmi
+	stf.spill [ptr2] = f0, 32
+	stf.spill [ptr0] = f0, 64
+ 	cmp.lt	p_scr, p0 = ptr9, ptr1		// do we need more prefetching?
+ ;; }
+{ .mmb
+	stf.spill [ptr2] = f0, 32
+(p_scr)	stf.spill [ptr9] = f0, 128
+	br.cloop.dptk.few .l1bx
+;; }
+{ .mib
+	cmp.gt  p_scr, p0 = 8, cnt		// just a few bytes left ?
+(p_scr)	br.cond.dpnt.many  .move_bytes_from_alignment	//
+;; }
+
+.fraction_of_line:
+{ .mib
+	add	ptr2 = 16, ptr1
+	shr.u	loopcnt = cnt, 5   		// loopcnt = cnt / 32
+;; }
+{ .mib
+	cmp.eq	p_scr, p0 = loopcnt, r0
+	add	loopcnt = -1, loopcnt
+(p_scr)	br.cond.dpnt.many .store_words
+;; }
+{ .mib
+	and	cnt = 0x1f, cnt			// compute the remaining cnt
+	mov.i   ar.lc = loopcnt
+;; }
+	TEXT_ALIGN(32)
+.l2:	// ------------------------------------ //  L2A:  store 32B in 2 cycles
+{ .mmb
+	stf8	[ptr1] = fvalue, 8
+	stf8	[ptr2] = fvalue, 8
+;; } { .mmb
+	stf8	[ptr1] = fvalue, 24
+	stf8	[ptr2] = fvalue, 24
+	br.cloop.dptk.many .l2
+;; }
+.store_words:
+{ .mib
+	cmp.gt	p_scr, p0 = 8, cnt		// just a few bytes left ?
+(p_scr)	br.cond.dpnt.many .move_bytes_from_alignment	// Branch
+;; }
+
+{ .mmi
+	stf8	[ptr1] = fvalue, 8		// store
+	cmp.le	p_y, p_n = 16, cnt
+	add	cnt = -8, cnt			// subtract
+;; }
+{ .mmi
+(p_y)	stf8	[ptr1] = fvalue, 8		// store
+(p_y)	cmp.le.unc p_yy, p_nn = 16, cnt
+(p_y)	add	cnt = -8, cnt			// subtract
+;; }
+{ .mmi						// store
+(p_yy)	stf8	[ptr1] = fvalue, 8
+(p_yy)	add	cnt = -8, cnt			// subtract
+;; }
+
+.move_bytes_from_alignment:
+{ .mib
+	cmp.eq	p_scr, p0 = cnt, r0
+	tbit.nz.unc p_y, p0 = cnt, 2		// should we terminate with a st4 ?
+(p_scr)	br.cond.dpnt.few .restore_and_exit
+;; }
+{ .mib
+(p_y)	st4	[ptr1] = value,4
+	tbit.nz.unc p_yy, p0 = cnt, 1		// should we terminate with a st2 ?
+;; }
+{ .mib
+(p_yy)	st2	[ptr1] = value,2
+	tbit.nz.unc p_y, p0 = cnt, 0		// should we terminate with a st1 ?
+;; }
+
+{ .mib
+(p_y)	st1	[ptr1] = value
+;; }
+.restore_and_exit:
+{ .mib
+	nop.m	0
+	mov.i	ar.lc = save_lc
+	br.ret.sptk.many rp
+;; }
+
+.move_bytes_unaligned:
+{ .mmi
+       .pred.rel "mutex",p_y, p_n
+       .pred.rel "mutex",p_yy, p_nn
+(p_n)	cmp.le  p_yy, p_nn = 4, cnt
+(p_y)	cmp.le  p_yy, p_nn = 5, cnt
+(p_n)	add	ptr2 = 2, ptr1
+} { .mmi
+(p_y)	add	ptr2 = 3, ptr1
+(p_y)	st1	[ptr1] = value, 1		// fill 1 (odd-aligned) byte [15, 14 (or less) left]
+(p_y)	add	cnt = -1, cnt
+;; }
+{ .mmi
+(p_yy)	cmp.le.unc p_y, p0 = 8, cnt
+	add	ptr3 = ptr1, cnt		// prepare last store
+	mov.i	ar.lc = save_lc
+} { .mmi
+(p_yy)	st2	[ptr1] = value, 4		// fill 2 (aligned) bytes
+(p_yy)	st2	[ptr2] = value, 4		// fill 2 (aligned) bytes [11, 10 (o less) left]
+(p_yy)	add	cnt = -4, cnt
+;; }
+{ .mmi
+(p_y)	cmp.le.unc p_yy, p0 = 8, cnt
+	add	ptr3 = -1, ptr3			// last store
+	tbit.nz p_scr, p0 = cnt, 1		// will there be a st2 at the end ?
+} { .mmi
+(p_y)	st2	[ptr1] = value, 4		// fill 2 (aligned) bytes
+(p_y)	st2	[ptr2] = value, 4		// fill 2 (aligned) bytes [7, 6 (or less) left]
+(p_y)	add	cnt = -4, cnt
+;; }
+{ .mmi
+(p_yy)	st2	[ptr1] = value, 4		// fill 2 (aligned) bytes
+(p_yy)	st2	[ptr2] = value, 4		// fill 2 (aligned) bytes [3, 2 (or less) left]
+	tbit.nz p_y, p0 = cnt, 0		// will there be a st1 at the end ?
+} { .mmi
+(p_yy)	add	cnt = -4, cnt
+;; }
+{ .mmb
+(p_scr)	st2	[ptr1] = value			// fill 2 (aligned) bytes
+(p_y)	st1	[ptr3] = value			// fill last byte (using ptr3)
+	br.ret.sptk.many rp
+}
+END(memset)
+EXPORT_SYMBOL(memset)