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/m68k/math-emu/fp_arith.c | 699 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 699 insertions(+) create mode 100644 arch/m68k/math-emu/fp_arith.c (limited to 'arch/m68k/math-emu/fp_arith.c') diff --git a/arch/m68k/math-emu/fp_arith.c b/arch/m68k/math-emu/fp_arith.c new file mode 100644 index 000000000..f4a06492c --- /dev/null +++ b/arch/m68k/math-emu/fp_arith.c @@ -0,0 +1,699 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + + fp_arith.c: floating-point math routines for the Linux-m68k + floating point emulator. + + Copyright (c) 1998-1999 David Huggins-Daines. + + Somewhat based on the AlphaLinux floating point emulator, by David + Mosberger-Tang. + + */ + +#include "fp_emu.h" +#include "multi_arith.h" +#include "fp_arith.h" + +const struct fp_ext fp_QNaN = +{ + .exp = 0x7fff, + .mant = { .m64 = ~0 } +}; + +const struct fp_ext fp_Inf = +{ + .exp = 0x7fff, +}; + +/* let's start with the easy ones */ + +struct fp_ext * +fp_fabs(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "fabs\n"); + + fp_monadic_check(dest, src); + + dest->sign = 0; + + return dest; +} + +struct fp_ext * +fp_fneg(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "fneg\n"); + + fp_monadic_check(dest, src); + + dest->sign = !dest->sign; + + return dest; +} + +/* Now, the slightly harder ones */ + +/* fp_fadd: Implements the kernel of the FADD, FSADD, FDADD, FSUB, + FDSUB, and FCMP instructions. */ + +struct fp_ext * +fp_fadd(struct fp_ext *dest, struct fp_ext *src) +{ + int diff; + + dprint(PINSTR, "fadd\n"); + + fp_dyadic_check(dest, src); + + if (IS_INF(dest)) { + /* infinity - infinity == NaN */ + if (IS_INF(src) && (src->sign != dest->sign)) + fp_set_nan(dest); + return dest; + } + if (IS_INF(src)) { + fp_copy_ext(dest, src); + return dest; + } + + if (IS_ZERO(dest)) { + if (IS_ZERO(src)) { + if (src->sign != dest->sign) { + if (FPDATA->rnd == FPCR_ROUND_RM) + dest->sign = 1; + else + dest->sign = 0; + } + } else + fp_copy_ext(dest, src); + return dest; + } + + dest->lowmant = src->lowmant = 0; + + if ((diff = dest->exp - src->exp) > 0) + fp_denormalize(src, diff); + else if ((diff = -diff) > 0) + fp_denormalize(dest, diff); + + if (dest->sign == src->sign) { + if (fp_addmant(dest, src)) + if (!fp_addcarry(dest)) + return dest; + } else { + if (dest->mant.m64 < src->mant.m64) { + fp_submant(dest, src, dest); + dest->sign = !dest->sign; + } else + fp_submant(dest, dest, src); + } + + return dest; +} + +/* fp_fsub: Implements the kernel of the FSUB, FSSUB, and FDSUB + instructions. + + Remember that the arguments are in assembler-syntax order! */ + +struct fp_ext * +fp_fsub(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "fsub "); + + src->sign = !src->sign; + return fp_fadd(dest, src); +} + + +struct fp_ext * +fp_fcmp(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "fcmp "); + + FPDATA->temp[1] = *dest; + src->sign = !src->sign; + return fp_fadd(&FPDATA->temp[1], src); +} + +struct fp_ext * +fp_ftst(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "ftst\n"); + + (void)dest; + + return src; +} + +struct fp_ext * +fp_fmul(struct fp_ext *dest, struct fp_ext *src) +{ + union fp_mant128 temp; + int exp; + + dprint(PINSTR, "fmul\n"); + + fp_dyadic_check(dest, src); + + /* calculate the correct sign now, as it's necessary for infinities */ + dest->sign = src->sign ^ dest->sign; + + /* Handle infinities */ + if (IS_INF(dest)) { + if (IS_ZERO(src)) + fp_set_nan(dest); + return dest; + } + if (IS_INF(src)) { + if (IS_ZERO(dest)) + fp_set_nan(dest); + else + fp_copy_ext(dest, src); + return dest; + } + + /* Of course, as we all know, zero * anything = zero. You may + not have known that it might be a positive or negative + zero... */ + if (IS_ZERO(dest) || IS_ZERO(src)) { + dest->exp = 0; + dest->mant.m64 = 0; + dest->lowmant = 0; + + return dest; + } + + exp = dest->exp + src->exp - 0x3ffe; + + /* shift up the mantissa for denormalized numbers, + so that the highest bit is set, this makes the + shift of the result below easier */ + if ((long)dest->mant.m32[0] >= 0) + exp -= fp_overnormalize(dest); + if ((long)src->mant.m32[0] >= 0) + exp -= fp_overnormalize(src); + + /* now, do a 64-bit multiply with expansion */ + fp_multiplymant(&temp, dest, src); + + /* normalize it back to 64 bits and stuff it back into the + destination struct */ + if ((long)temp.m32[0] > 0) { + exp--; + fp_putmant128(dest, &temp, 1); + } else + fp_putmant128(dest, &temp, 0); + + if (exp >= 0x7fff) { + fp_set_ovrflw(dest); + return dest; + } + dest->exp = exp; + if (exp < 0) { + fp_set_sr(FPSR_EXC_UNFL); + fp_denormalize(dest, -exp); + } + + return dest; +} + +/* fp_fdiv: Implements the "kernel" of the FDIV, FSDIV, FDDIV and + FSGLDIV instructions. + + Note that the order of the operands is counter-intuitive: instead + of src / dest, the result is actually dest / src. */ + +struct fp_ext * +fp_fdiv(struct fp_ext *dest, struct fp_ext *src) +{ + union fp_mant128 temp; + int exp; + + dprint(PINSTR, "fdiv\n"); + + fp_dyadic_check(dest, src); + + /* calculate the correct sign now, as it's necessary for infinities */ + dest->sign = src->sign ^ dest->sign; + + /* Handle infinities */ + if (IS_INF(dest)) { + /* infinity / infinity = NaN (quiet, as always) */ + if (IS_INF(src)) + fp_set_nan(dest); + /* infinity / anything else = infinity (with appropriate sign) */ + return dest; + } + if (IS_INF(src)) { + /* anything / infinity = zero (with appropriate sign) */ + dest->exp = 0; + dest->mant.m64 = 0; + dest->lowmant = 0; + + return dest; + } + + /* zeroes */ + if (IS_ZERO(dest)) { + /* zero / zero = NaN */ + if (IS_ZERO(src)) + fp_set_nan(dest); + /* zero / anything else = zero */ + return dest; + } + if (IS_ZERO(src)) { + /* anything / zero = infinity (with appropriate sign) */ + fp_set_sr(FPSR_EXC_DZ); + dest->exp = 0x7fff; + dest->mant.m64 = 0; + + return dest; + } + + exp = dest->exp - src->exp + 0x3fff; + + /* shift up the mantissa for denormalized numbers, + so that the highest bit is set, this makes lots + of things below easier */ + if ((long)dest->mant.m32[0] >= 0) + exp -= fp_overnormalize(dest); + if ((long)src->mant.m32[0] >= 0) + exp -= fp_overnormalize(src); + + /* now, do the 64-bit divide */ + fp_dividemant(&temp, dest, src); + + /* normalize it back to 64 bits and stuff it back into the + destination struct */ + if (!temp.m32[0]) { + exp--; + fp_putmant128(dest, &temp, 32); + } else + fp_putmant128(dest, &temp, 31); + + if (exp >= 0x7fff) { + fp_set_ovrflw(dest); + return dest; + } + dest->exp = exp; + if (exp < 0) { + fp_set_sr(FPSR_EXC_UNFL); + fp_denormalize(dest, -exp); + } + + return dest; +} + +struct fp_ext * +fp_fsglmul(struct fp_ext *dest, struct fp_ext *src) +{ + int exp; + + dprint(PINSTR, "fsglmul\n"); + + fp_dyadic_check(dest, src); + + /* calculate the correct sign now, as it's necessary for infinities */ + dest->sign = src->sign ^ dest->sign; + + /* Handle infinities */ + if (IS_INF(dest)) { + if (IS_ZERO(src)) + fp_set_nan(dest); + return dest; + } + if (IS_INF(src)) { + if (IS_ZERO(dest)) + fp_set_nan(dest); + else + fp_copy_ext(dest, src); + return dest; + } + + /* Of course, as we all know, zero * anything = zero. You may + not have known that it might be a positive or negative + zero... */ + if (IS_ZERO(dest) || IS_ZERO(src)) { + dest->exp = 0; + dest->mant.m64 = 0; + dest->lowmant = 0; + + return dest; + } + + exp = dest->exp + src->exp - 0x3ffe; + + /* do a 32-bit multiply */ + fp_mul64(dest->mant.m32[0], dest->mant.m32[1], + dest->mant.m32[0] & 0xffffff00, + src->mant.m32[0] & 0xffffff00); + + if (exp >= 0x7fff) { + fp_set_ovrflw(dest); + return dest; + } + dest->exp = exp; + if (exp < 0) { + fp_set_sr(FPSR_EXC_UNFL); + fp_denormalize(dest, -exp); + } + + return dest; +} + +struct fp_ext * +fp_fsgldiv(struct fp_ext *dest, struct fp_ext *src) +{ + int exp; + unsigned long quot, rem; + + dprint(PINSTR, "fsgldiv\n"); + + fp_dyadic_check(dest, src); + + /* calculate the correct sign now, as it's necessary for infinities */ + dest->sign = src->sign ^ dest->sign; + + /* Handle infinities */ + if (IS_INF(dest)) { + /* infinity / infinity = NaN (quiet, as always) */ + if (IS_INF(src)) + fp_set_nan(dest); + /* infinity / anything else = infinity (with approprate sign) */ + return dest; + } + if (IS_INF(src)) { + /* anything / infinity = zero (with appropriate sign) */ + dest->exp = 0; + dest->mant.m64 = 0; + dest->lowmant = 0; + + return dest; + } + + /* zeroes */ + if (IS_ZERO(dest)) { + /* zero / zero = NaN */ + if (IS_ZERO(src)) + fp_set_nan(dest); + /* zero / anything else = zero */ + return dest; + } + if (IS_ZERO(src)) { + /* anything / zero = infinity (with appropriate sign) */ + fp_set_sr(FPSR_EXC_DZ); + dest->exp = 0x7fff; + dest->mant.m64 = 0; + + return dest; + } + + exp = dest->exp - src->exp + 0x3fff; + + dest->mant.m32[0] &= 0xffffff00; + src->mant.m32[0] &= 0xffffff00; + + /* do the 32-bit divide */ + if (dest->mant.m32[0] >= src->mant.m32[0]) { + fp_sub64(dest->mant, src->mant); + fp_div64(quot, rem, dest->mant.m32[0], 0, src->mant.m32[0]); + dest->mant.m32[0] = 0x80000000 | (quot >> 1); + dest->mant.m32[1] = (quot & 1) | rem; /* only for rounding */ + } else { + fp_div64(quot, rem, dest->mant.m32[0], 0, src->mant.m32[0]); + dest->mant.m32[0] = quot; + dest->mant.m32[1] = rem; /* only for rounding */ + exp--; + } + + if (exp >= 0x7fff) { + fp_set_ovrflw(dest); + return dest; + } + dest->exp = exp; + if (exp < 0) { + fp_set_sr(FPSR_EXC_UNFL); + fp_denormalize(dest, -exp); + } + + return dest; +} + +/* fp_roundint: Internal rounding function for use by several of these + emulated instructions. + + This one rounds off the fractional part using the rounding mode + specified. */ + +static void fp_roundint(struct fp_ext *dest, int mode) +{ + union fp_mant64 oldmant; + unsigned long mask; + + if (!fp_normalize_ext(dest)) + return; + + /* infinities and zeroes */ + if (IS_INF(dest) || IS_ZERO(dest)) + return; + + /* first truncate the lower bits */ + oldmant = dest->mant; + switch (dest->exp) { + case 0 ... 0x3ffe: + dest->mant.m64 = 0; + break; + case 0x3fff ... 0x401e: + dest->mant.m32[0] &= 0xffffffffU << (0x401e - dest->exp); + dest->mant.m32[1] = 0; + if (oldmant.m64 == dest->mant.m64) + return; + break; + case 0x401f ... 0x403e: + dest->mant.m32[1] &= 0xffffffffU << (0x403e - dest->exp); + if (oldmant.m32[1] == dest->mant.m32[1]) + return; + break; + default: + return; + } + fp_set_sr(FPSR_EXC_INEX2); + + /* We might want to normalize upwards here... however, since + we know that this is only called on the output of fp_fdiv, + or with the input to fp_fint or fp_fintrz, and the inputs + to all these functions are either normal or denormalized + (no subnormals allowed!), there's really no need. + + In the case of fp_fdiv, observe that 0x80000000 / 0xffff = + 0xffff8000, and the same holds for 128-bit / 64-bit. (i.e. the + smallest possible normal dividend and the largest possible normal + divisor will still produce a normal quotient, therefore, (normal + << 64) / normal is normal in all cases) */ + + switch (mode) { + case FPCR_ROUND_RN: + switch (dest->exp) { + case 0 ... 0x3ffd: + return; + case 0x3ffe: + /* As noted above, the input is always normal, so the + guard bit (bit 63) is always set. therefore, the + only case in which we will NOT round to 1.0 is when + the input is exactly 0.5. */ + if (oldmant.m64 == (1ULL << 63)) + return; + break; + case 0x3fff ... 0x401d: + mask = 1 << (0x401d - dest->exp); + if (!(oldmant.m32[0] & mask)) + return; + if (oldmant.m32[0] & (mask << 1)) + break; + if (!(oldmant.m32[0] << (dest->exp - 0x3ffd)) && + !oldmant.m32[1]) + return; + break; + case 0x401e: + if (oldmant.m32[1] & 0x80000000) + return; + if (oldmant.m32[0] & 1) + break; + if (!(oldmant.m32[1] << 1)) + return; + break; + case 0x401f ... 0x403d: + mask = 1 << (0x403d - dest->exp); + if (!(oldmant.m32[1] & mask)) + return; + if (oldmant.m32[1] & (mask << 1)) + break; + if (!(oldmant.m32[1] << (dest->exp - 0x401d))) + return; + break; + default: + return; + } + break; + case FPCR_ROUND_RZ: + return; + default: + if (dest->sign ^ (mode - FPCR_ROUND_RM)) + break; + return; + } + + switch (dest->exp) { + case 0 ... 0x3ffe: + dest->exp = 0x3fff; + dest->mant.m64 = 1ULL << 63; + break; + case 0x3fff ... 0x401e: + mask = 1 << (0x401e - dest->exp); + if (dest->mant.m32[0] += mask) + break; + dest->mant.m32[0] = 0x80000000; + dest->exp++; + break; + case 0x401f ... 0x403e: + mask = 1 << (0x403e - dest->exp); + if (dest->mant.m32[1] += mask) + break; + if (dest->mant.m32[0] += 1) + break; + dest->mant.m32[0] = 0x80000000; + dest->exp++; + break; + } +} + +/* modrem_kernel: Implementation of the FREM and FMOD instructions + (which are exactly the same, except for the rounding used on the + intermediate value) */ + +static struct fp_ext * +modrem_kernel(struct fp_ext *dest, struct fp_ext *src, int mode) +{ + struct fp_ext tmp; + + fp_dyadic_check(dest, src); + + /* Infinities and zeros */ + if (IS_INF(dest) || IS_ZERO(src)) { + fp_set_nan(dest); + return dest; + } + if (IS_ZERO(dest) || IS_INF(src)) + return dest; + + /* FIXME: there is almost certainly a smarter way to do this */ + fp_copy_ext(&tmp, dest); + fp_fdiv(&tmp, src); /* NOTE: src might be modified */ + fp_roundint(&tmp, mode); + fp_fmul(&tmp, src); + fp_fsub(dest, &tmp); + + /* set the quotient byte */ + fp_set_quotient((dest->mant.m64 & 0x7f) | (dest->sign << 7)); + return dest; +} + +/* fp_fmod: Implements the kernel of the FMOD instruction. + + Again, the argument order is backwards. The result, as defined in + the Motorola manuals, is: + + fmod(src,dest) = (dest - (src * floor(dest / src))) */ + +struct fp_ext * +fp_fmod(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "fmod\n"); + return modrem_kernel(dest, src, FPCR_ROUND_RZ); +} + +/* fp_frem: Implements the kernel of the FREM instruction. + + frem(src,dest) = (dest - (src * round(dest / src))) + */ + +struct fp_ext * +fp_frem(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "frem\n"); + return modrem_kernel(dest, src, FPCR_ROUND_RN); +} + +struct fp_ext * +fp_fint(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "fint\n"); + + fp_copy_ext(dest, src); + + fp_roundint(dest, FPDATA->rnd); + + return dest; +} + +struct fp_ext * +fp_fintrz(struct fp_ext *dest, struct fp_ext *src) +{ + dprint(PINSTR, "fintrz\n"); + + fp_copy_ext(dest, src); + + fp_roundint(dest, FPCR_ROUND_RZ); + + return dest; +} + +struct fp_ext * +fp_fscale(struct fp_ext *dest, struct fp_ext *src) +{ + int scale, oldround; + + dprint(PINSTR, "fscale\n"); + + fp_dyadic_check(dest, src); + + /* Infinities */ + if (IS_INF(src)) { + fp_set_nan(dest); + return dest; + } + if (IS_INF(dest)) + return dest; + + /* zeroes */ + if (IS_ZERO(src) || IS_ZERO(dest)) + return dest; + + /* Source exponent out of range */ + if (src->exp >= 0x400c) { + fp_set_ovrflw(dest); + return dest; + } + + /* src must be rounded with round to zero. */ + oldround = FPDATA->rnd; + FPDATA->rnd = FPCR_ROUND_RZ; + scale = fp_conv_ext2long(src); + FPDATA->rnd = oldround; + + /* new exponent */ + scale += dest->exp; + + if (scale >= 0x7fff) { + fp_set_ovrflw(dest); + } else if (scale <= 0) { + fp_set_sr(FPSR_EXC_UNFL); + fp_denormalize(dest, -scale); + } else + dest->exp = scale; + + return dest; +} + -- cgit v1.2.3