half\-precision (full) in projects: linux-6.14.4

Project(s)

Full Search
Definition
Symbol
File Path
History
Type

Searched +full:half +full:- +full:precision (Results 1 – 25 of 99) sorted by relevance

12 3 4

/linux-6.14.4/tools/perf/pmu-events/arch/arm64/fujitsu/monaka/
D	fp_operation.json	5 … "BriefDescription": "This event counts architecturally executed floating-point move operations." 19 …"BriefDescription": "This event counts architecturally executed Advanced SIMD floating-point opera… 23 … "BriefDescription": "This event counts architecturally executed SVE floating-point operation." 27 …ion": "This event counts architecturally executed Advanced SIMD and SVE floating-point operations." 31 …"BriefDescription": "This event counts architecturally executed half-precision floating-point oper… 35 …ion": "This event counts architecturally executed Advanced SIMD half-precision floating-point oper… 39 …efDescription": "This event counts architecturally executed SVE half-precision floating-point oper… 43 …his event counts architecturally executed Advanced SIMD and SVE half-precision floating-point oper… 47 …"BriefDescription": "This event counts architecturally executed single-precision floating-point op… 51 …: "This event counts architecturally executed Advanced SIMD single-precision floating-point operat… [all …]
D	sve.json	4 …ing the Advanced SIMD scalar instructions and the instructions listed in Non-SIMD SVE instructions… 8 …cturally executed SVE instructions, including the instructions listed in Non-SIMD SVE instructions… 12 …cturally executed SVE instructions, including the instructions listed in Non-SIMD SVE instructions… 20 "BriefDescription": "This event counts all architecturally executed micro-operations." 28 …ations due to scalar, Advanced SIMD, and SVE instructions listed in Floating-point instructions se… 32 …on": "This event counts architecturally executed floating-point fused multiply-add and multiply-su… 36 …"BriefDescription": "This event counts architecturally executed floating-point reciprocal estimate… 40 …uted floating-point convert operations due to the scalar, Advanced SIMD, and SVE floating-point co… 60 …"BriefDescription": "This event counts architecturally executed SVE 64-bit integer divide operatio… 76 …"BriefDescription": "This event counts architecturally executed SVE integer 64-bit x 64-bit multip… [all …]
/linux-6.14.4/tools/perf/pmu-events/arch/arm64/arm/neoverse-n2-v2/
D	fp_operation.json	4 … "PublicDescription": "Counts speculatively executed half precision floating point operations." 8 … "PublicDescription": "Counts speculatively executed single precision floating point operations." 12 … "PublicDescription": "Counts speculatively executed double precision floating point operations." 16 …"PublicDescription": "Counts speculatively executed scalable single precision floating point opera… 20 …"PublicDescription": "Counts speculatively executed non-scalable single precision floating point o…
/linux-6.14.4/arch/x86/math-emu/
D	reg_round.S	1 /* SPDX-License-Identifier: GPL-2.0 / 3 /---------------------------------------------------------------------------+ 10 \| Australia. E-mail [email protected] \| 20 \| Return value is the tag of the answer, or-ed with FPU_Exception if \| 21 \| one was raised, or -1 on internal error. \| 26 +---------------------------------------------------------------------------/ 28 /---------------------------------------------------------------------------+ 47 \| must be non-zero. \| 48 \| If the significand extension is non-zero then the significand is \| 52 \| non-zero values: \| [all …]
D	README	1 +---------------------------------------------------------------------------+ 2 \| wm-FPU-emu an FPU emulator for 80386 and 80486SX microprocessors. \| 6 \| Australia. E-mail [email protected] \| 21 +---------------------------------------------------------------------------+ 25 wm-FPU-emu is an FPU emulator for Linux. It is derived from wm-emu387 27 msdos); wm-emu387 was in turn based upon emu387 which was written by 31 My target FPU for wm-FPU-emu is that described in the Intel486 40 wm-FPU-emu does not implement all of the behaviour of the 80486 FPU, 52 --Bill Metzenthen 56 ----------------------- Internals of wm-FPU-emu ----------------------- [all …]
/linux-6.14.4/tools/perf/pmu-events/arch/arm64/
D	common-and-microarch.json	129 "PublicDescription": "Attributable Level 1 data cache write-back", 132 "BriefDescription": "Attributable Level 1 data cache write-back" 147 "PublicDescription": "Attributable Level 2 data cache write-back", 150 "BriefDescription": "Attributable Level 2 data cache write-back" 273 "PublicDescription": "Access to another socket in a multi-socket system", 276 "BriefDescription": "Access to another socket in a multi-socket system" 303 … "PublicDescription": "Attributable memory read access to another socket in a multi-socket system", 306 … "BriefDescription": "Attributable memory read access to another socket in a multi-socket system" 309 …"PublicDescription": "Level 1 data cache long-latency read miss. The counter counts each memory r… 312 "BriefDescription": "Level 1 data cache long-latency read miss" [all …]
/linux-6.14.4/tools/perf/pmu-events/arch/x86/graniterapids/
D	floating-point.json	77 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 81 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 86 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 90 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 95 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 99 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 104 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 108 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 113 …-bit packed single and 256-bit packed double precision FP instructions retired; some instructions … 117 …-bit packed single precision and 256-bit packed double precision floating-point instructions reti… [all …]
/linux-6.14.4/tools/perf/pmu-events/arch/x86/sapphirerapids/
D	floating-point.json	77 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 81 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 86 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 90 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 95 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 99 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 104 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 108 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 113 …-bit packed single and 256-bit packed double precision FP instructions retired; some instructions … 117 …-bit packed single precision and 256-bit packed double precision floating-point instructions reti… [all …]
/linux-6.14.4/tools/perf/pmu-events/arch/x86/emeraldrapids/
D	floating-point.json	77 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 81 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 86 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 90 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 95 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 99 …-bit packed double precision floating-point instructions retired; some instructions will count twi… 104 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 108 …-bit packed single precision floating-point instructions retired; some instructions will count twi… 113 …-bit packed single and 256-bit packed double precision FP instructions retired; some instructions … 117 …-bit packed single precision and 256-bit packed double precision floating-point instructions reti… [all …]
/linux-6.14.4/drivers/gpu/drm/i915/gt/
D	selftest_ring.c	1 // SPDX-License-Identifier: GPL-2.0 14 kref_init(&ring->ref); in mock_ring() 15 ring->size = sz; in mock_ring() 16 ring->wrap = BITS_PER_TYPE(ring->size) - ilog2(sz); in mock_ring() 17 ring->effective_size = sz; in mock_ring() 18 ring->vaddr = (void *)(ring + 1); in mock_ring() 19 atomic_set(&ring->pin_count, 1); in mock_ring() 39 result = -1; in check_ring_direction() 46 return -EINVAL; in check_ring_direction() 60 err \|= check_ring_direction(ring, prev, next, -1); in check_ring_step() [all …]
/linux-6.14.4/Documentation/devicetree/bindings/riscv/
D	extensions.yaml	`1 # SPDX-License-Identifier: (GPL-2.0 OR MIT) 3 --- 5 $schema: http://devicetree.org/meta-schemas/core.yaml# 7 title: RISC-V ISA extensions 10 - Paul Walmsley <[email protected]> 11 - Palmer Dabbelt <[email protected]> 12 - Conor Dooley <[email protected]> 15 RISC-V has a large number of extensions, some of which are "standard" 16 extensions, meaning they are ratified by RISC-V International, and others 36 Identifies the specific RISC-V instruction set architecture [all …]`
/linux-6.14.4/arch/m68k/fpsp040/
D	satanh.S	`9 \| Input: Double-extended number X in location pointed to 12 \| Output: The value arctanh(X) returned in floating-point register Fp0. 16 \| result is subsequently rounded to double precision. The 17 \| result is provably monotonic in double precision. 29 \| z := 2y/(1-y) 36 \| divide-by-zero by 63 \|--ATANH(X) = X FOR DENORMALIZED X 75 \|--THIS IS THE USUAL CASE, \|X\| < 1 76 \|--Y = \|X\|, Z = 2Y/(1-Y), ATANH(X) = SIGN(X) * (1/2) * LOG1P(Z). 80 fnegx %fp1 \| ...-Y [all …]`
D	ssin.S	9 \| Input: Double-extended number X in location pointed to 20 \| result is subsequently rounded to double precision. The 21 \| result is provably monotonic in double precision. 32 \| 2. If \|X\| >= 15Pi or \|X\| < 2(-40), go to 7. 40 \| 5. (k is odd) Set j := (k-1)/2, sgn := (-1)j. Return sgncos(r) 45 \| 6. (k is even) Set j := k/2, sgn := (-1)j. Return sgnsin(r) 52 \| 8. (\|X\|<2(-40)) If SIN is invoked, return X; otherwise return 1. 57 \| 1. If \|X\| >= 15Pi or \|X\| < 2(-40), go to 6. 64 \| 4. (k is odd) Set j1 := (k-1)/2, j2 := j1 (EOR) (k mod 2), i.e. 66 \| sgn1 := (-1)j1, sgn2 := (-1)j2. [all …]
D	stan.S	8 \| Input: Double-extended number X in location pointed to 11 \| Output: The value tan(X) returned in floating-point register Fp0. 15 \| result is subsequently rounded to double precision. The 16 \| result is provably monotonic in double precision. 24 \| 1. If \|X\| >= 15Pi or \|X\| < 2(-40), go to 6. 37 \| 4. (k is odd) Tan(X) = -cot(r). Since tan(r) is approximated by a 41 \| -Cot(r) = -V/U. Exit. 45 \| 7. (\|X\|<2(-40)) Tan(X) = X. Exit. 83 \|--NPI/2, -32 <= N <= 32, IN A LEADING TERM IN EXT. AND TRAILING 84 \|--TERM IN SGL. NOTE THAT PI IS 64-BIT LONG, THUS NPI/2 IS AT [all …]
/linux-6.14.4/drivers/net/ethernet/qualcomm/
D	Kconfig	`1 # SPDX-License-Identifier: GPL-2.0-only 59 supports IEEE 802.3-2002, half-duplex mode at 10/100 Mb/s, 60 full-duplex mode at 10/100/1000Mb/s, Wake On LAN (WOL) for 61 low power, Receive-Side Scaling (RSS), and IEEE 1588-2008 62 Precision Clock Synchronization Protocol.`
/linux-6.14.4/tools/perf/Documentation/
D	topdown.txt	2 --------------------- 11 perf stat --topdown implements this using available metrics that vary 14 % perf stat -a --topdown -I1000 84 int slots_fd = perf_event_open(&slots, 0, -1, -1, 0); 104 int metrics_fd = perf_event_open(&metrics, 0, -1, slots_fd, 0); 205 retiring_slots = GET_METRIC(metric_b, 0) * slots_b - retiring_slots_a 206 bad_spec_slots = GET_METRIC(metric_b, 1) * slots_b - bad_spec_slots_a 207 fe_bound_slots = GET_METRIC(metric_b, 2) * slots_b - fe_bound_slots_a 208 be_bound_slots = GET_METRIC(metric_b, 3) * slots_b - be_bound_slots_a 213 slots_delta = slots_b - slots_a [all …]
/linux-6.14.4/tools/testing/selftests/bpf/progs/
D	verifier_scalar_ids.c	1 // SPDX-License-Identifier: GPL-2.0 7 /* Check that precision marks propagate through scalar IDs. 10 * Check that precision mark for r0 causes precision marks for r{1,2} 153 /* Check that precision marks propagate through scalar IDs. 155 * precision information is propagated up the call stack. 161 __msg("frame2: last_idx 12 first_idx 11 subseq_idx -1 ") 233 * - bar frame r1 in precision_many_frames__bar() 234 * - foo frame r{1,6,7} in precision_many_frames__bar() 235 * - main frame r{1,6} in precision_many_frames__bar() 253 __msg("frame1: regs=r1 stack=-8,-16 before 8: (7b) (u64 )(r10 -16) = r1") [all …]
/linux-6.14.4/arch/powerpc/kernel/
D	vecemu.c	1 // SPDX-License-Identifier: GPL-2.0 36 * Computes an estimate of 2^x. The `s' argument is the 32-bit 37 * single-precision floating-point representation of x. 45 exp = ((s >> 23) & 0xff) - 127; in eexp2() 50 /* 2^-big = 0, 2^+big = +Inf */ in eexp2() 53 if (exp < -23) in eexp2() 61 pwr >>= -exp; in eexp2() 63 pwr = -pwr; in eexp2() 69 if (exp < -23) in eexp2() 85 exp = -exp; in eexp2() [all …]
/linux-6.14.4/arch/mips/dec/
D	time.c	1 // SPDX-License-Identifier: GPL-2.0 6 * This file contains the time handling details for PC-style clocks as 15 #include <asm/cpu-features.h> 55 year += real_year - 72 + 2000; in read_persistent_clock64() 57 ts->tv_sec = mktime64(year, mon, day, hour, min, sec); in read_persistent_clock64() 58 ts->tv_nsec = 0; in read_persistent_clock64() 96 if (((abs(real_minutes - cmos_minutes) + 15) / 30) & 1) in update_persistent_clock64() 97 real_minutes += 30; /* correct for half hour time zone */ in update_persistent_clock64() 100 if (abs(real_minutes - cmos_minutes) < 30) { in update_persistent_clock64() 111 retval = -1; in update_persistent_clock64() [all …]
/linux-6.14.4/drivers/net/dsa/mv88e6xxx/
D	chip.h	1 /* SPDX-License-Identifier: GPL-2.0-or-later / 3 Marvell 88E6xxx Ethernet switch single-chip definition 30 /* PVT limits for 4-bit port and 5-bit switch / 112 enum mv88e6xxx_edsa_support - Ethertype DSA tag support level 152 * ports 2-4 are not routet to pins. 155 /* Multi-chip Addressing Mode. 157 * when it is non-zero, and use indirect access to internal registers. 160 /* Dual-chip Addressing Mode 161 * Some chips respond to only half of the 32 SMI addresses, 364 /* Handles automatic disabling and re-enabling of the PHY [all …]
/linux-6.14.4/arch/m68k/
D	Kconfig.cpu	1 # SPDX-License-Identifier: GPL-2.0 13 applications, and are all System-On-Chip (SOC) devices, as opposed 66 System-On-Chip devices (eg 68328, 68302, etc). It does not contain 138 Select the type of ColdFire System-on-Chip (SoC) that you want 340 At some point in the future, this will cause floating-point math 342 floating-point math coprocessor. Thrill-seekers and chronically 343 sleep-deprived psychotic hacker types can say Y now, everyone else 347 bool "Math emulation extra precision" 361 This option prevents any floating-point instructions from being 364 kernel will only be usable on machines without a floating-point [all …]
/linux-6.14.4/arch/arm/vfp/
D	vfpmodule.c	1 // SPDX-License-Identifier: GPL-2.0-only 37 * Dual-use variable. 38 * Used in startup: set to non-zero if VFP checks fail 67 * choice here as bottom half processing is always in thread context on RT 68 * kernels so it implicitly prevents bottom half processing as well. 88 * Must be called from non-preemptible context. 93 if (thread->vfpstate.hard.cpu != cpu) in vfp_state_in_hw() 96 return vfp_current_hw_state[cpu] == &thread->vfpstate; in vfp_state_in_hw() 102 * clear vfp_current_hw_state. Must be called from non-preemptible context. 111 thread->vfpstate.hard.cpu = NR_CPUS; in vfp_force_reload() [all …]
/linux-6.14.4/kernel/cgroup/
D	cpuset-v1.c	1 // SPDX-License-Identifier: GPL-2.0-or-later 3 #include "cpuset-internal.h" 14 * Frequency meter - How fast is some event occurring? 18 * fmeter_init() - initialize a frequency meter. 19 * fmeter_markevent() - called each time the event happens. 20 * fmeter_getrate() - returns the recent rate of such events. 21 * fmeter_update() - internal routine used to update fmeter. 28 * The filter is single-pole low-pass recursive (IIR). The time unit 29 * is 1 second. Arithmetic is done using 32-bit integers scaled to 30 * simulate 3 decimal digits of precision (multiplied by 1000). [all …]
/linux-6.14.4/arch/parisc/math-emu/
D	dfsub.c	1 // SPDX-License-Identifier: GPL-2.0-or-later 3 * Linux/PA-RISC Project (http://www.parisc-linux.org/) 5 * Floating-point emulation code 6 * Copyright (C) 2001 Hewlett-Packard (Paul Bame) <[email protected]> 15 * Double_subtract: subtract two double precision values. 33 * Double_subtract: subtract two double precision values. 266 diff_exponent = result_exponent - right_exponent; in dbl_fsub() 313 /* Must have been "x-x" or "x+(-x)". */ in dbl_fsub() 319 result_exponent--; in dbl_fsub() 355 if((result_exponent -= 8) <= 0 && !underflowtrap) in dbl_fsub() [all …]
D	sfadd.c	1 // SPDX-License-Identifier: GPL-2.0-or-later 3 * Linux/PA-RISC Project (http://www.parisc-linux.org/) 5 * Floating-point emulation code 6 * Copyright (C) 2001 Hewlett-Packard (Paul Bame) <[email protected]> 15 * Single_add: add two single precision values. 33 * Single_add: add two single precision values. 259 diff_exponent = result_exponent - right_exponent; in sgl_fadd() 305 /* Must have been "x-x" or "x+(-x)". */ in sgl_fadd() 311 result_exponent--; in sgl_fadd() 347 if((result_exponent -= 8) <= 0 && !underflowtrap) in sgl_fadd() [all …]

12 3 4