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/sh/kernel/cpu/sh2a/fpu.c | 572 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 572 insertions(+) create mode 100644 arch/sh/kernel/cpu/sh2a/fpu.c (limited to 'arch/sh/kernel/cpu/sh2a/fpu.c') diff --git a/arch/sh/kernel/cpu/sh2a/fpu.c b/arch/sh/kernel/cpu/sh2a/fpu.c new file mode 100644 index 000000000..0bcff11a4 --- /dev/null +++ b/arch/sh/kernel/cpu/sh2a/fpu.c @@ -0,0 +1,572 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Save/restore floating point context for signal handlers. + * + * Copyright (C) 1999, 2000 Kaz Kojima & Niibe Yutaka + * + * FIXME! These routines can be optimized in big endian case. + */ +#include +#include +#include +#include +#include +#include + +/* The PR (precision) bit in the FP Status Register must be clear when + * an frchg instruction is executed, otherwise the instruction is undefined. + * Executing frchg with PR set causes a trap on some SH4 implementations. + */ + +#define FPSCR_RCHG 0x00000000 + + +/* + * Save FPU registers onto task structure. + */ +void save_fpu(struct task_struct *tsk) +{ + unsigned long dummy; + + enable_fpu(); + asm volatile("sts.l fpul, @-%0\n\t" + "sts.l fpscr, @-%0\n\t" + "fmov.s fr15, @-%0\n\t" + "fmov.s fr14, @-%0\n\t" + "fmov.s fr13, @-%0\n\t" + "fmov.s fr12, @-%0\n\t" + "fmov.s fr11, @-%0\n\t" + "fmov.s fr10, @-%0\n\t" + "fmov.s fr9, @-%0\n\t" + "fmov.s fr8, @-%0\n\t" + "fmov.s fr7, @-%0\n\t" + "fmov.s fr6, @-%0\n\t" + "fmov.s fr5, @-%0\n\t" + "fmov.s fr4, @-%0\n\t" + "fmov.s fr3, @-%0\n\t" + "fmov.s fr2, @-%0\n\t" + "fmov.s fr1, @-%0\n\t" + "fmov.s fr0, @-%0\n\t" + "lds %3, fpscr\n\t" + : "=r" (dummy) + : "0" ((char *)(&tsk->thread.xstate->hardfpu.status)), + "r" (FPSCR_RCHG), + "r" (FPSCR_INIT) + : "memory"); + + disable_fpu(); +} + +void restore_fpu(struct task_struct *tsk) +{ + unsigned long dummy; + + enable_fpu(); + asm volatile("fmov.s @%0+, fr0\n\t" + "fmov.s @%0+, fr1\n\t" + "fmov.s @%0+, fr2\n\t" + "fmov.s @%0+, fr3\n\t" + "fmov.s @%0+, fr4\n\t" + "fmov.s @%0+, fr5\n\t" + "fmov.s @%0+, fr6\n\t" + "fmov.s @%0+, fr7\n\t" + "fmov.s @%0+, fr8\n\t" + "fmov.s @%0+, fr9\n\t" + "fmov.s @%0+, fr10\n\t" + "fmov.s @%0+, fr11\n\t" + "fmov.s @%0+, fr12\n\t" + "fmov.s @%0+, fr13\n\t" + "fmov.s @%0+, fr14\n\t" + "fmov.s @%0+, fr15\n\t" + "lds.l @%0+, fpscr\n\t" + "lds.l @%0+, fpul\n\t" + : "=r" (dummy) + : "0" (tsk->thread.xstate), "r" (FPSCR_RCHG) + : "memory"); + disable_fpu(); +} + +/* + * Emulate arithmetic ops on denormalized number for some FPU insns. + */ + +/* denormalized float * float */ +static int denormal_mulf(int hx, int hy) +{ + unsigned int ix, iy; + unsigned long long m, n; + int exp, w; + + ix = hx & 0x7fffffff; + iy = hy & 0x7fffffff; + if (iy < 0x00800000 || ix == 0) + return ((hx ^ hy) & 0x80000000); + + exp = (iy & 0x7f800000) >> 23; + ix &= 0x007fffff; + iy = (iy & 0x007fffff) | 0x00800000; + m = (unsigned long long)ix * iy; + n = m; + w = -1; + while (n) { n >>= 1; w++; } + + /* FIXME: use guard bits */ + exp += w - 126 - 46; + if (exp > 0) + ix = ((int) (m >> (w - 23)) & 0x007fffff) | (exp << 23); + else if (exp + 22 >= 0) + ix = (int) (m >> (w - 22 - exp)) & 0x007fffff; + else + ix = 0; + + ix |= (hx ^ hy) & 0x80000000; + return ix; +} + +/* denormalized double * double */ +static void mult64(unsigned long long x, unsigned long long y, + unsigned long long *highp, unsigned long long *lowp) +{ + unsigned long long sub0, sub1, sub2, sub3; + unsigned long long high, low; + + sub0 = (x >> 32) * (unsigned long) (y >> 32); + sub1 = (x & 0xffffffffLL) * (unsigned long) (y >> 32); + sub2 = (x >> 32) * (unsigned long) (y & 0xffffffffLL); + sub3 = (x & 0xffffffffLL) * (unsigned long) (y & 0xffffffffLL); + low = sub3; + high = 0LL; + sub3 += (sub1 << 32); + if (low > sub3) + high++; + low = sub3; + sub3 += (sub2 << 32); + if (low > sub3) + high++; + low = sub3; + high += (sub1 >> 32) + (sub2 >> 32); + high += sub0; + *lowp = low; + *highp = high; +} + +static inline long long rshift64(unsigned long long mh, + unsigned long long ml, int n) +{ + if (n >= 64) + return mh >> (n - 64); + return (mh << (64 - n)) | (ml >> n); +} + +static long long denormal_muld(long long hx, long long hy) +{ + unsigned long long ix, iy; + unsigned long long mh, ml, nh, nl; + int exp, w; + + ix = hx & 0x7fffffffffffffffLL; + iy = hy & 0x7fffffffffffffffLL; + if (iy < 0x0010000000000000LL || ix == 0) + return ((hx ^ hy) & 0x8000000000000000LL); + + exp = (iy & 0x7ff0000000000000LL) >> 52; + ix &= 0x000fffffffffffffLL; + iy = (iy & 0x000fffffffffffffLL) | 0x0010000000000000LL; + mult64(ix, iy, &mh, &ml); + nh = mh; + nl = ml; + w = -1; + if (nh) { + while (nh) { nh >>= 1; w++;} + w += 64; + } else + while (nl) { nl >>= 1; w++;} + + /* FIXME: use guard bits */ + exp += w - 1022 - 52 * 2; + if (exp > 0) + ix = (rshift64(mh, ml, w - 52) & 0x000fffffffffffffLL) + | ((long long)exp << 52); + else if (exp + 51 >= 0) + ix = rshift64(mh, ml, w - 51 - exp) & 0x000fffffffffffffLL; + else + ix = 0; + + ix |= (hx ^ hy) & 0x8000000000000000LL; + return ix; +} + +/* ix - iy where iy: denormal and ix, iy >= 0 */ +static int denormal_subf1(unsigned int ix, unsigned int iy) +{ + int frac; + int exp; + + if (ix < 0x00800000) + return ix - iy; + + exp = (ix & 0x7f800000) >> 23; + if (exp - 1 > 31) + return ix; + iy >>= exp - 1; + if (iy == 0) + return ix; + + frac = (ix & 0x007fffff) | 0x00800000; + frac -= iy; + while (frac < 0x00800000) { + if (--exp == 0) + return frac; + frac <<= 1; + } + + return (exp << 23) | (frac & 0x007fffff); +} + +/* ix + iy where iy: denormal and ix, iy >= 0 */ +static int denormal_addf1(unsigned int ix, unsigned int iy) +{ + int frac; + int exp; + + if (ix < 0x00800000) + return ix + iy; + + exp = (ix & 0x7f800000) >> 23; + if (exp - 1 > 31) + return ix; + iy >>= exp - 1; + if (iy == 0) + return ix; + + frac = (ix & 0x007fffff) | 0x00800000; + frac += iy; + if (frac >= 0x01000000) { + frac >>= 1; + ++exp; + } + + return (exp << 23) | (frac & 0x007fffff); +} + +static int denormal_addf(int hx, int hy) +{ + unsigned int ix, iy; + int sign; + + if ((hx ^ hy) & 0x80000000) { + sign = hx & 0x80000000; + ix = hx & 0x7fffffff; + iy = hy & 0x7fffffff; + if (iy < 0x00800000) { + ix = denormal_subf1(ix, iy); + if ((int) ix < 0) { + ix = -ix; + sign ^= 0x80000000; + } + } else { + ix = denormal_subf1(iy, ix); + sign ^= 0x80000000; + } + } else { + sign = hx & 0x80000000; + ix = hx & 0x7fffffff; + iy = hy & 0x7fffffff; + if (iy < 0x00800000) + ix = denormal_addf1(ix, iy); + else + ix = denormal_addf1(iy, ix); + } + + return sign | ix; +} + +/* ix - iy where iy: denormal and ix, iy >= 0 */ +static long long denormal_subd1(unsigned long long ix, unsigned long long iy) +{ + long long frac; + int exp; + + if (ix < 0x0010000000000000LL) + return ix - iy; + + exp = (ix & 0x7ff0000000000000LL) >> 52; + if (exp - 1 > 63) + return ix; + iy >>= exp - 1; + if (iy == 0) + return ix; + + frac = (ix & 0x000fffffffffffffLL) | 0x0010000000000000LL; + frac -= iy; + while (frac < 0x0010000000000000LL) { + if (--exp == 0) + return frac; + frac <<= 1; + } + + return ((long long)exp << 52) | (frac & 0x000fffffffffffffLL); +} + +/* ix + iy where iy: denormal and ix, iy >= 0 */ +static long long denormal_addd1(unsigned long long ix, unsigned long long iy) +{ + long long frac; + long long exp; + + if (ix < 0x0010000000000000LL) + return ix + iy; + + exp = (ix & 0x7ff0000000000000LL) >> 52; + if (exp - 1 > 63) + return ix; + iy >>= exp - 1; + if (iy == 0) + return ix; + + frac = (ix & 0x000fffffffffffffLL) | 0x0010000000000000LL; + frac += iy; + if (frac >= 0x0020000000000000LL) { + frac >>= 1; + ++exp; + } + + return (exp << 52) | (frac & 0x000fffffffffffffLL); +} + +static long long denormal_addd(long long hx, long long hy) +{ + unsigned long long ix, iy; + long long sign; + + if ((hx ^ hy) & 0x8000000000000000LL) { + sign = hx & 0x8000000000000000LL; + ix = hx & 0x7fffffffffffffffLL; + iy = hy & 0x7fffffffffffffffLL; + if (iy < 0x0010000000000000LL) { + ix = denormal_subd1(ix, iy); + if ((int) ix < 0) { + ix = -ix; + sign ^= 0x8000000000000000LL; + } + } else { + ix = denormal_subd1(iy, ix); + sign ^= 0x8000000000000000LL; + } + } else { + sign = hx & 0x8000000000000000LL; + ix = hx & 0x7fffffffffffffffLL; + iy = hy & 0x7fffffffffffffffLL; + if (iy < 0x0010000000000000LL) + ix = denormal_addd1(ix, iy); + else + ix = denormal_addd1(iy, ix); + } + + return sign | ix; +} + +/** + * denormal_to_double - Given denormalized float number, + * store double float + * + * @fpu: Pointer to sh_fpu_hard structure + * @n: Index to FP register + */ +static void +denormal_to_double (struct sh_fpu_hard_struct *fpu, int n) +{ + unsigned long du, dl; + unsigned long x = fpu->fpul; + int exp = 1023 - 126; + + if (x != 0 && (x & 0x7f800000) == 0) { + du = (x & 0x80000000); + while ((x & 0x00800000) == 0) { + x <<= 1; + exp--; + } + x &= 0x007fffff; + du |= (exp << 20) | (x >> 3); + dl = x << 29; + + fpu->fp_regs[n] = du; + fpu->fp_regs[n+1] = dl; + } +} + +/** + * ieee_fpe_handler - Handle denormalized number exception + * + * @regs: Pointer to register structure + * + * Returns 1 when it's handled (should not cause exception). + */ +static int +ieee_fpe_handler (struct pt_regs *regs) +{ + unsigned short insn = *(unsigned short *) regs->pc; + unsigned short finsn; + unsigned long nextpc; + int nib[4] = { + (insn >> 12) & 0xf, + (insn >> 8) & 0xf, + (insn >> 4) & 0xf, + insn & 0xf}; + + if (nib[0] == 0xb || + (nib[0] == 0x4 && nib[2] == 0x0 && nib[3] == 0xb)) /* bsr & jsr */ + regs->pr = regs->pc + 4; + if (nib[0] == 0xa || nib[0] == 0xb) { /* bra & bsr */ + nextpc = regs->pc + 4 + ((short) ((insn & 0xfff) << 4) >> 3); + finsn = *(unsigned short *) (regs->pc + 2); + } else if (nib[0] == 0x8 && nib[1] == 0xd) { /* bt/s */ + if (regs->sr & 1) + nextpc = regs->pc + 4 + ((char) (insn & 0xff) << 1); + else + nextpc = regs->pc + 4; + finsn = *(unsigned short *) (regs->pc + 2); + } else if (nib[0] == 0x8 && nib[1] == 0xf) { /* bf/s */ + if (regs->sr & 1) + nextpc = regs->pc + 4; + else + nextpc = regs->pc + 4 + ((char) (insn & 0xff) << 1); + finsn = *(unsigned short *) (regs->pc + 2); + } else if (nib[0] == 0x4 && nib[3] == 0xb && + (nib[2] == 0x0 || nib[2] == 0x2)) { /* jmp & jsr */ + nextpc = regs->regs[nib[1]]; + finsn = *(unsigned short *) (regs->pc + 2); + } else if (nib[0] == 0x0 && nib[3] == 0x3 && + (nib[2] == 0x0 || nib[2] == 0x2)) { /* braf & bsrf */ + nextpc = regs->pc + 4 + regs->regs[nib[1]]; + finsn = *(unsigned short *) (regs->pc + 2); + } else if (insn == 0x000b) { /* rts */ + nextpc = regs->pr; + finsn = *(unsigned short *) (regs->pc + 2); + } else { + nextpc = regs->pc + 2; + finsn = insn; + } + +#define FPSCR_FPU_ERROR (1 << 17) + + if ((finsn & 0xf1ff) == 0xf0ad) { /* fcnvsd */ + struct task_struct *tsk = current; + + if ((tsk->thread.xstate->hardfpu.fpscr & FPSCR_FPU_ERROR)) { + /* FPU error */ + denormal_to_double (&tsk->thread.xstate->hardfpu, + (finsn >> 8) & 0xf); + } else + return 0; + + regs->pc = nextpc; + return 1; + } else if ((finsn & 0xf00f) == 0xf002) { /* fmul */ + struct task_struct *tsk = current; + int fpscr; + int n, m, prec; + unsigned int hx, hy; + + n = (finsn >> 8) & 0xf; + m = (finsn >> 4) & 0xf; + hx = tsk->thread.xstate->hardfpu.fp_regs[n]; + hy = tsk->thread.xstate->hardfpu.fp_regs[m]; + fpscr = tsk->thread.xstate->hardfpu.fpscr; + prec = fpscr & (1 << 19); + + if ((fpscr & FPSCR_FPU_ERROR) + && (prec && ((hx & 0x7fffffff) < 0x00100000 + || (hy & 0x7fffffff) < 0x00100000))) { + long long llx, lly; + + /* FPU error because of denormal */ + llx = ((long long) hx << 32) + | tsk->thread.xstate->hardfpu.fp_regs[n+1]; + lly = ((long long) hy << 32) + | tsk->thread.xstate->hardfpu.fp_regs[m+1]; + if ((hx & 0x7fffffff) >= 0x00100000) + llx = denormal_muld(lly, llx); + else + llx = denormal_muld(llx, lly); + tsk->thread.xstate->hardfpu.fp_regs[n] = llx >> 32; + tsk->thread.xstate->hardfpu.fp_regs[n+1] = llx & 0xffffffff; + } else if ((fpscr & FPSCR_FPU_ERROR) + && (!prec && ((hx & 0x7fffffff) < 0x00800000 + || (hy & 0x7fffffff) < 0x00800000))) { + /* FPU error because of denormal */ + if ((hx & 0x7fffffff) >= 0x00800000) + hx = denormal_mulf(hy, hx); + else + hx = denormal_mulf(hx, hy); + tsk->thread.xstate->hardfpu.fp_regs[n] = hx; + } else + return 0; + + regs->pc = nextpc; + return 1; + } else if ((finsn & 0xf00e) == 0xf000) { /* fadd, fsub */ + struct task_struct *tsk = current; + int fpscr; + int n, m, prec; + unsigned int hx, hy; + + n = (finsn >> 8) & 0xf; + m = (finsn >> 4) & 0xf; + hx = tsk->thread.xstate->hardfpu.fp_regs[n]; + hy = tsk->thread.xstate->hardfpu.fp_regs[m]; + fpscr = tsk->thread.xstate->hardfpu.fpscr; + prec = fpscr & (1 << 19); + + if ((fpscr & FPSCR_FPU_ERROR) + && (prec && ((hx & 0x7fffffff) < 0x00100000 + || (hy & 0x7fffffff) < 0x00100000))) { + long long llx, lly; + + /* FPU error because of denormal */ + llx = ((long long) hx << 32) + | tsk->thread.xstate->hardfpu.fp_regs[n+1]; + lly = ((long long) hy << 32) + | tsk->thread.xstate->hardfpu.fp_regs[m+1]; + if ((finsn & 0xf00f) == 0xf000) + llx = denormal_addd(llx, lly); + else + llx = denormal_addd(llx, lly ^ (1LL << 63)); + tsk->thread.xstate->hardfpu.fp_regs[n] = llx >> 32; + tsk->thread.xstate->hardfpu.fp_regs[n+1] = llx & 0xffffffff; + } else if ((fpscr & FPSCR_FPU_ERROR) + && (!prec && ((hx & 0x7fffffff) < 0x00800000 + || (hy & 0x7fffffff) < 0x00800000))) { + /* FPU error because of denormal */ + if ((finsn & 0xf00f) == 0xf000) + hx = denormal_addf(hx, hy); + else + hx = denormal_addf(hx, hy ^ 0x80000000); + tsk->thread.xstate->hardfpu.fp_regs[n] = hx; + } else + return 0; + + regs->pc = nextpc; + return 1; + } + + return 0; +} + +BUILD_TRAP_HANDLER(fpu_error) +{ + struct task_struct *tsk = current; + TRAP_HANDLER_DECL; + + __unlazy_fpu(tsk, regs); + if (ieee_fpe_handler(regs)) { + tsk->thread.xstate->hardfpu.fpscr &= + ~(FPSCR_CAUSE_MASK | FPSCR_FLAG_MASK); + grab_fpu(regs); + restore_fpu(tsk); + task_thread_info(tsk)->status |= TS_USEDFPU; + return; + } + + force_sig(SIGFPE); +} -- cgit v1.2.3