From 5b7c4cabbb65f5c469464da6c5f614cbd7f730f2 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 21 Feb 2023 18:24:12 -0800 Subject: Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next 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(). ... --- arch/sparc/lib/NGmemcpy.S | 509 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 509 insertions(+) create mode 100644 arch/sparc/lib/NGmemcpy.S (limited to 'arch/sparc/lib/NGmemcpy.S') diff --git a/arch/sparc/lib/NGmemcpy.S b/arch/sparc/lib/NGmemcpy.S new file mode 100644 index 000000000..ee51c1230 --- /dev/null +++ b/arch/sparc/lib/NGmemcpy.S @@ -0,0 +1,509 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* NGmemcpy.S: Niagara optimized memcpy. + * + * Copyright (C) 2006, 2007 David S. Miller (davem@davemloft.net) + */ + +#ifdef __KERNEL__ +#include +#include +#include +#define GLOBAL_SPARE %g7 +#define RESTORE_ASI(TMP) \ + wr %g0, ASI_AIUS, %asi +#else +#define GLOBAL_SPARE %g5 +#define RESTORE_ASI(TMP) \ + wr %g0, ASI_PNF, %asi +#endif + +#ifdef __sparc_v9__ +#define SAVE_AMOUNT 128 +#else +#define SAVE_AMOUNT 64 +#endif + +#ifndef STORE_ASI +#define STORE_ASI ASI_BLK_INIT_QUAD_LDD_P +#endif + +#ifndef EX_LD +#define EX_LD(x,y) x +#endif + +#ifndef EX_ST +#define EX_ST(x,y) x +#endif + +#ifndef LOAD +#ifndef MEMCPY_DEBUG +#define LOAD(type,addr,dest) type [addr], dest +#else +#define LOAD(type,addr,dest) type##a [addr] 0x80, dest +#endif +#endif + +#ifndef LOAD_TWIN +#define LOAD_TWIN(addr_reg,dest0,dest1) \ + ldda [addr_reg] ASI_BLK_INIT_QUAD_LDD_P, dest0 +#endif + +#ifndef STORE +#define STORE(type,src,addr) type src, [addr] +#endif + +#ifndef STORE_INIT +#ifndef SIMULATE_NIAGARA_ON_NON_NIAGARA +#define STORE_INIT(src,addr) stxa src, [addr] %asi +#else +#define STORE_INIT(src,addr) stx src, [addr + 0x00] +#endif +#endif + +#ifndef FUNC_NAME +#define FUNC_NAME NGmemcpy +#endif + +#ifndef PREAMBLE +#define PREAMBLE +#endif + +#ifndef XCC +#define XCC xcc +#endif + + .register %g2,#scratch + .register %g3,#scratch + + .text +#ifndef EX_RETVAL +#define EX_RETVAL(x) x +__restore_asi: + ret + wr %g0, ASI_AIUS, %asi + restore +ENTRY(NG_ret_i2_plus_i4_plus_1) + ba,pt %xcc, __restore_asi + add %i2, %i5, %i0 +ENDPROC(NG_ret_i2_plus_i4_plus_1) +ENTRY(NG_ret_i2_plus_g1) + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1) +ENTRY(NG_ret_i2_plus_g1_minus_8) + sub %g1, 8, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_minus_8) +ENTRY(NG_ret_i2_plus_g1_minus_16) + sub %g1, 16, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_minus_16) +ENTRY(NG_ret_i2_plus_g1_minus_24) + sub %g1, 24, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_minus_24) +ENTRY(NG_ret_i2_plus_g1_minus_32) + sub %g1, 32, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_minus_32) +ENTRY(NG_ret_i2_plus_g1_minus_40) + sub %g1, 40, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_minus_40) +ENTRY(NG_ret_i2_plus_g1_minus_48) + sub %g1, 48, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_minus_48) +ENTRY(NG_ret_i2_plus_g1_minus_56) + sub %g1, 56, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_minus_56) +ENTRY(NG_ret_i2_plus_i4) + ba,pt %xcc, __restore_asi + add %i2, %i4, %i0 +ENDPROC(NG_ret_i2_plus_i4) +ENTRY(NG_ret_i2_plus_i4_minus_8) + sub %i4, 8, %i4 + ba,pt %xcc, __restore_asi + add %i2, %i4, %i0 +ENDPROC(NG_ret_i2_plus_i4_minus_8) +ENTRY(NG_ret_i2_plus_8) + ba,pt %xcc, __restore_asi + add %i2, 8, %i0 +ENDPROC(NG_ret_i2_plus_8) +ENTRY(NG_ret_i2_plus_4) + ba,pt %xcc, __restore_asi + add %i2, 4, %i0 +ENDPROC(NG_ret_i2_plus_4) +ENTRY(NG_ret_i2_plus_1) + ba,pt %xcc, __restore_asi + add %i2, 1, %i0 +ENDPROC(NG_ret_i2_plus_1) +ENTRY(NG_ret_i2_plus_g1_plus_1) + add %g1, 1, %g1 + ba,pt %xcc, __restore_asi + add %i2, %g1, %i0 +ENDPROC(NG_ret_i2_plus_g1_plus_1) +ENTRY(NG_ret_i2) + ba,pt %xcc, __restore_asi + mov %i2, %i0 +ENDPROC(NG_ret_i2) +ENTRY(NG_ret_i2_and_7_plus_i4) + and %i2, 7, %i2 + ba,pt %xcc, __restore_asi + add %i2, %i4, %i0 +ENDPROC(NG_ret_i2_and_7_plus_i4) +#endif + + .align 64 + + .globl FUNC_NAME + .type FUNC_NAME,#function +FUNC_NAME: /* %i0=dst, %i1=src, %i2=len */ + PREAMBLE + save %sp, -SAVE_AMOUNT, %sp + srlx %i2, 31, %g2 + cmp %g2, 0 + tne %xcc, 5 + mov %i0, %o0 + cmp %i2, 0 + be,pn %XCC, 85f + or %o0, %i1, %i3 + cmp %i2, 16 + blu,a,pn %XCC, 80f + or %i3, %i2, %i3 + + /* 2 blocks (128 bytes) is the minimum we can do the block + * copy with. We need to ensure that we'll iterate at least + * once in the block copy loop. At worst we'll need to align + * the destination to a 64-byte boundary which can chew up + * to (64 - 1) bytes from the length before we perform the + * block copy loop. + */ + cmp %i2, (2 * 64) + blu,pt %XCC, 70f + andcc %i3, 0x7, %g0 + + /* %o0: dst + * %i1: src + * %i2: len (known to be >= 128) + * + * The block copy loops will use %i4/%i5,%g2/%g3 as + * temporaries while copying the data. + */ + + LOAD(prefetch, %i1, #one_read) + wr %g0, STORE_ASI, %asi + + /* Align destination on 64-byte boundary. */ + andcc %o0, (64 - 1), %i4 + be,pt %XCC, 2f + sub %i4, 64, %i4 + sub %g0, %i4, %i4 ! bytes to align dst + sub %i2, %i4, %i2 +1: subcc %i4, 1, %i4 + EX_LD(LOAD(ldub, %i1, %g1), NG_ret_i2_plus_i4_plus_1) + EX_ST(STORE(stb, %g1, %o0), NG_ret_i2_plus_i4_plus_1) + add %i1, 1, %i1 + bne,pt %XCC, 1b + add %o0, 1, %o0 + + /* If the source is on a 16-byte boundary we can do + * the direct block copy loop. If it is 8-byte aligned + * we can do the 16-byte loads offset by -8 bytes and the + * init stores offset by one register. + * + * If the source is not even 8-byte aligned, we need to do + * shifting and masking (basically integer faligndata). + * + * The careful bit with init stores is that if we store + * to any part of the cache line we have to store the whole + * cacheline else we can end up with corrupt L2 cache line + * contents. Since the loop works on 64-bytes of 64-byte + * aligned store data at a time, this is easy to ensure. + */ +2: + andcc %i1, (16 - 1), %i4 + andn %i2, (64 - 1), %g1 ! block copy loop iterator + be,pt %XCC, 50f + sub %i2, %g1, %i2 ! final sub-block copy bytes + + cmp %i4, 8 + be,pt %XCC, 10f + sub %i1, %i4, %i1 + + /* Neither 8-byte nor 16-byte aligned, shift and mask. */ + and %i4, 0x7, GLOBAL_SPARE + sll GLOBAL_SPARE, 3, GLOBAL_SPARE + mov 64, %i5 + EX_LD(LOAD_TWIN(%i1, %g2, %g3), NG_ret_i2_plus_g1) + sub %i5, GLOBAL_SPARE, %i5 + mov 16, %o4 + mov 32, %o5 + mov 48, %o7 + mov 64, %i3 + + bg,pn %XCC, 9f + nop + +#define MIX_THREE_WORDS(WORD1, WORD2, WORD3, PRE_SHIFT, POST_SHIFT, TMP) \ + sllx WORD1, POST_SHIFT, WORD1; \ + srlx WORD2, PRE_SHIFT, TMP; \ + sllx WORD2, POST_SHIFT, WORD2; \ + or WORD1, TMP, WORD1; \ + srlx WORD3, PRE_SHIFT, TMP; \ + or WORD2, TMP, WORD2; + +8: EX_LD(LOAD_TWIN(%i1 + %o4, %o2, %o3), NG_ret_i2_plus_g1) + MIX_THREE_WORDS(%g2, %g3, %o2, %i5, GLOBAL_SPARE, %o1) + LOAD(prefetch, %i1 + %i3, #one_read) + + EX_ST(STORE_INIT(%g2, %o0 + 0x00), NG_ret_i2_plus_g1) + EX_ST(STORE_INIT(%g3, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8) + + EX_LD(LOAD_TWIN(%i1 + %o5, %g2, %g3), NG_ret_i2_plus_g1_minus_16) + MIX_THREE_WORDS(%o2, %o3, %g2, %i5, GLOBAL_SPARE, %o1) + + EX_ST(STORE_INIT(%o2, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16) + EX_ST(STORE_INIT(%o3, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24) + + EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1_minus_32) + MIX_THREE_WORDS(%g2, %g3, %o2, %i5, GLOBAL_SPARE, %o1) + + EX_ST(STORE_INIT(%g2, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32) + EX_ST(STORE_INIT(%g3, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40) + + EX_LD(LOAD_TWIN(%i1 + %i3, %g2, %g3), NG_ret_i2_plus_g1_minus_48) + add %i1, 64, %i1 + MIX_THREE_WORDS(%o2, %o3, %g2, %i5, GLOBAL_SPARE, %o1) + + EX_ST(STORE_INIT(%o2, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48) + EX_ST(STORE_INIT(%o3, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56) + + subcc %g1, 64, %g1 + bne,pt %XCC, 8b + add %o0, 64, %o0 + + ba,pt %XCC, 60f + add %i1, %i4, %i1 + +9: EX_LD(LOAD_TWIN(%i1 + %o4, %o2, %o3), NG_ret_i2_plus_g1) + MIX_THREE_WORDS(%g3, %o2, %o3, %i5, GLOBAL_SPARE, %o1) + LOAD(prefetch, %i1 + %i3, #one_read) + + EX_ST(STORE_INIT(%g3, %o0 + 0x00), NG_ret_i2_plus_g1) + EX_ST(STORE_INIT(%o2, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8) + + EX_LD(LOAD_TWIN(%i1 + %o5, %g2, %g3), NG_ret_i2_plus_g1_minus_16) + MIX_THREE_WORDS(%o3, %g2, %g3, %i5, GLOBAL_SPARE, %o1) + + EX_ST(STORE_INIT(%o3, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16) + EX_ST(STORE_INIT(%g2, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24) + + EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1_minus_32) + MIX_THREE_WORDS(%g3, %o2, %o3, %i5, GLOBAL_SPARE, %o1) + + EX_ST(STORE_INIT(%g3, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32) + EX_ST(STORE_INIT(%o2, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40) + + EX_LD(LOAD_TWIN(%i1 + %i3, %g2, %g3), NG_ret_i2_plus_g1_minus_48) + add %i1, 64, %i1 + MIX_THREE_WORDS(%o3, %g2, %g3, %i5, GLOBAL_SPARE, %o1) + + EX_ST(STORE_INIT(%o3, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48) + EX_ST(STORE_INIT(%g2, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56) + + subcc %g1, 64, %g1 + bne,pt %XCC, 9b + add %o0, 64, %o0 + + ba,pt %XCC, 60f + add %i1, %i4, %i1 + +10: /* Destination is 64-byte aligned, source was only 8-byte + * aligned but it has been subtracted by 8 and we perform + * one twin load ahead, then add 8 back into source when + * we finish the loop. + */ + EX_LD(LOAD_TWIN(%i1, %o4, %o5), NG_ret_i2_plus_g1) + mov 16, %o7 + mov 32, %g2 + mov 48, %g3 + mov 64, %o1 +1: EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1) + LOAD(prefetch, %i1 + %o1, #one_read) + EX_ST(STORE_INIT(%o5, %o0 + 0x00), NG_ret_i2_plus_g1) ! initializes cache line + EX_ST(STORE_INIT(%o2, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8) + EX_LD(LOAD_TWIN(%i1 + %g2, %o4, %o5), NG_ret_i2_plus_g1_minus_16) + EX_ST(STORE_INIT(%o3, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16) + EX_ST(STORE_INIT(%o4, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24) + EX_LD(LOAD_TWIN(%i1 + %g3, %o2, %o3), NG_ret_i2_plus_g1_minus_32) + EX_ST(STORE_INIT(%o5, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32) + EX_ST(STORE_INIT(%o2, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40) + EX_LD(LOAD_TWIN(%i1 + %o1, %o4, %o5), NG_ret_i2_plus_g1_minus_48) + add %i1, 64, %i1 + EX_ST(STORE_INIT(%o3, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48) + EX_ST(STORE_INIT(%o4, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56) + subcc %g1, 64, %g1 + bne,pt %XCC, 1b + add %o0, 64, %o0 + + ba,pt %XCC, 60f + add %i1, 0x8, %i1 + +50: /* Destination is 64-byte aligned, and source is 16-byte + * aligned. + */ + mov 16, %o7 + mov 32, %g2 + mov 48, %g3 + mov 64, %o1 +1: EX_LD(LOAD_TWIN(%i1 + %g0, %o4, %o5), NG_ret_i2_plus_g1) + EX_LD(LOAD_TWIN(%i1 + %o7, %o2, %o3), NG_ret_i2_plus_g1) + LOAD(prefetch, %i1 + %o1, #one_read) + EX_ST(STORE_INIT(%o4, %o0 + 0x00), NG_ret_i2_plus_g1) ! initializes cache line + EX_ST(STORE_INIT(%o5, %o0 + 0x08), NG_ret_i2_plus_g1_minus_8) + EX_LD(LOAD_TWIN(%i1 + %g2, %o4, %o5), NG_ret_i2_plus_g1_minus_16) + EX_ST(STORE_INIT(%o2, %o0 + 0x10), NG_ret_i2_plus_g1_minus_16) + EX_ST(STORE_INIT(%o3, %o0 + 0x18), NG_ret_i2_plus_g1_minus_24) + EX_LD(LOAD_TWIN(%i1 + %g3, %o2, %o3), NG_ret_i2_plus_g1_minus_32) + add %i1, 64, %i1 + EX_ST(STORE_INIT(%o4, %o0 + 0x20), NG_ret_i2_plus_g1_minus_32) + EX_ST(STORE_INIT(%o5, %o0 + 0x28), NG_ret_i2_plus_g1_minus_40) + EX_ST(STORE_INIT(%o2, %o0 + 0x30), NG_ret_i2_plus_g1_minus_48) + EX_ST(STORE_INIT(%o3, %o0 + 0x38), NG_ret_i2_plus_g1_minus_56) + subcc %g1, 64, %g1 + bne,pt %XCC, 1b + add %o0, 64, %o0 + /* fall through */ + +60: + membar #Sync + + /* %i2 contains any final bytes still needed to be copied + * over. If anything is left, we copy it one byte at a time. + */ + RESTORE_ASI(%i3) + brz,pt %i2, 85f + sub %o0, %i1, %i3 + ba,a,pt %XCC, 90f + nop + + .align 64 +70: /* 16 < len <= 64 */ + bne,pn %XCC, 75f + sub %o0, %i1, %i3 + +72: + andn %i2, 0xf, %i4 + and %i2, 0xf, %i2 +1: subcc %i4, 0x10, %i4 + EX_LD(LOAD(ldx, %i1, %o4), NG_ret_i2_plus_i4) + add %i1, 0x08, %i1 + EX_LD(LOAD(ldx, %i1, %g1), NG_ret_i2_plus_i4) + sub %i1, 0x08, %i1 + EX_ST(STORE(stx, %o4, %i1 + %i3), NG_ret_i2_plus_i4) + add %i1, 0x8, %i1 + EX_ST(STORE(stx, %g1, %i1 + %i3), NG_ret_i2_plus_i4_minus_8) + bgu,pt %XCC, 1b + add %i1, 0x8, %i1 +73: andcc %i2, 0x8, %g0 + be,pt %XCC, 1f + nop + sub %i2, 0x8, %i2 + EX_LD(LOAD(ldx, %i1, %o4), NG_ret_i2_plus_8) + EX_ST(STORE(stx, %o4, %i1 + %i3), NG_ret_i2_plus_8) + add %i1, 0x8, %i1 +1: andcc %i2, 0x4, %g0 + be,pt %XCC, 1f + nop + sub %i2, 0x4, %i2 + EX_LD(LOAD(lduw, %i1, %i5), NG_ret_i2_plus_4) + EX_ST(STORE(stw, %i5, %i1 + %i3), NG_ret_i2_plus_4) + add %i1, 0x4, %i1 +1: cmp %i2, 0 + be,pt %XCC, 85f + nop + ba,pt %xcc, 90f + nop + +75: + andcc %o0, 0x7, %g1 + sub %g1, 0x8, %g1 + be,pn %icc, 2f + sub %g0, %g1, %g1 + sub %i2, %g1, %i2 + +1: subcc %g1, 1, %g1 + EX_LD(LOAD(ldub, %i1, %i5), NG_ret_i2_plus_g1_plus_1) + EX_ST(STORE(stb, %i5, %i1 + %i3), NG_ret_i2_plus_g1_plus_1) + bgu,pt %icc, 1b + add %i1, 1, %i1 + +2: add %i1, %i3, %o0 + andcc %i1, 0x7, %g1 + bne,pt %icc, 8f + sll %g1, 3, %g1 + + cmp %i2, 16 + bgeu,pt %icc, 72b + nop + ba,a,pt %xcc, 73b + +8: mov 64, %i3 + andn %i1, 0x7, %i1 + EX_LD(LOAD(ldx, %i1, %g2), NG_ret_i2) + sub %i3, %g1, %i3 + andn %i2, 0x7, %i4 + sllx %g2, %g1, %g2 +1: add %i1, 0x8, %i1 + EX_LD(LOAD(ldx, %i1, %g3), NG_ret_i2_and_7_plus_i4) + subcc %i4, 0x8, %i4 + srlx %g3, %i3, %i5 + or %i5, %g2, %i5 + EX_ST(STORE(stx, %i5, %o0), NG_ret_i2_and_7_plus_i4) + add %o0, 0x8, %o0 + bgu,pt %icc, 1b + sllx %g3, %g1, %g2 + + srl %g1, 3, %g1 + andcc %i2, 0x7, %i2 + be,pn %icc, 85f + add %i1, %g1, %i1 + ba,pt %xcc, 90f + sub %o0, %i1, %i3 + + .align 64 +80: /* 0 < len <= 16 */ + andcc %i3, 0x3, %g0 + bne,pn %XCC, 90f + sub %o0, %i1, %i3 + +1: + subcc %i2, 4, %i2 + EX_LD(LOAD(lduw, %i1, %g1), NG_ret_i2_plus_4) + EX_ST(STORE(stw, %g1, %i1 + %i3), NG_ret_i2_plus_4) + bgu,pt %XCC, 1b + add %i1, 4, %i1 + +85: ret + restore EX_RETVAL(%i0), %g0, %o0 + + .align 32 +90: + subcc %i2, 1, %i2 + EX_LD(LOAD(ldub, %i1, %g1), NG_ret_i2_plus_1) + EX_ST(STORE(stb, %g1, %i1 + %i3), NG_ret_i2_plus_1) + bgu,pt %XCC, 90b + add %i1, 1, %i1 + ret + restore EX_RETVAL(%i0), %g0, %o0 + + .size FUNC_NAME, .-FUNC_NAME -- cgit v1.2.3