/freebsd/lib/libpmc/pmu-events/arch/arm64/fujitsu/a64fx/ |
H A D | other.json | 3 "PublicDescription": "This event counts the occurrence count of the micro-operation split.", 4 "EventCode": "0x139", 6 "BriefDescription": "This event counts the occurrence count of the micro-operation split." 9 …"PublicDescription": "This event counts every cycle that no operation was committed because the ol… 10 "EventCode": "0x180", 12 …"BriefDescription": "This event counts every cycle that no operation was committed because the old… 15 …"PublicDescription": "This event counts every cycle that no instruction was committed because the … 16 "EventCode": "0x181", 18 …"BriefDescription": "This event counts every cycle that no instruction was committed because the o… 21 …"PublicDescription": "This event counts every cycle that no instruction was committed because the … [all …]
|
/freebsd/contrib/llvm-project/llvm/include/llvm/ADT/ |
H A D | GenericCycleImpl.h | 1 //===- GenericCycleImpl.h -------------------------------------*- C++ -*---===// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 18 /// - llvm/lib/IR/CycleInfo.cpp 19 /// - llvm/lib/CodeGen/MachineCycleAnalysis.cpp 21 //===----------------------------------------------------------------------===// 30 #define DEBUG_TYPE "generic-cycle-impl" 39 if (Depth > C->Depth) in contains() 41 while (Depth < C->Depth) in contains() [all …]
|
H A D | GenericCycleInfo.h | 1 //===- GenericCycleInfo.h - Info for Cycles in any IR ------*- C++ -*------===// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 10 /// \brief Find all cycles in a control-flow graph, including irreducible loops. 15 /// - A cycle is a generalization of a loop which can represent 17 /// - Cycles identified in a program are implementation defined, 19 /// - Cycles are well-nested, and form a forest with a parent-child 21 /// - In any choice of DFS, every natural loop L is represented by a 22 /// unique cycle C which is a superset of L. [all …]
|
H A D | GenericUniformityImpl.h | 1 //===- GenericUniformityImpl.h -----------------------*- C++ -*------------===// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 17 // - UniformityAnalysis.cpp 23 //===----------------------------------------------------------------------===// 30 /// divergence (whose discovery must be implemented by a CFG- or even 31 /// target-specific derived class), divergence of values is propagated from 32 /// definition to uses in a straight-forward way. The main complexity lies in 42 //===----------------------------------------------------------------------===// [all …]
|
/freebsd/lib/libpmc/pmu-events/arch/arm64/arm/cortex-a55/ |
H A D | pipeline.json | 9 … operation issued due to the frontend, cache miss.This event counts every cycle the DPU IQ is empt… 10 "EventCode": "0xE1", 12 … operation issued due to the frontend, cache miss.This event counts every cycle the DPU IQ is empt… 15 …No operation issued due to the frontend, TLB miss.This event counts every cycle the DPU IQ is empt… 16 "EventCode": "0xE2", 18 …No operation issued due to the frontend, TLB miss.This event counts every cycle the DPU IQ is empt… 21 …ion issued due to the frontend, pre-decode error.This event counts every cycle the DPU IQ is empty… 22 "EventCode": "0xE3", 24 …ion issued due to the frontend, pre-decode error.This event counts every cycle the DPU IQ is empty… 27 …"No operation issued due to the backend interlock.This event counts every cycle that issue is stal… [all …]
|
/freebsd/lib/libpmc/pmu-events/arch/arm64/arm/cortex-a510/ |
H A D | pipeline.json | 21 …operation issued due to the frontend, cache miss. This event counts every cycle that the Data Proc… 22 "EventCode": "0xE1", 24 …operation issued due to the frontend, cache miss. This event counts every cycle that the Data Proc… 27 …o operation issued due to the frontend, TLB miss. This event counts every cycle that the DPU instr… 28 "EventCode": "0xE2", 30 …o operation issued due to the frontend, TLB miss. This event counts every cycle that the DPU instr… 33 "PublicDescription": "No operation issued due to the frontend, pre-decode error", 34 "EventCode": "0xE3", 36 "BriefDescription": "No operation issued due to the frontend, pre-decode error" 39 …No operation issued due to the backend interlock. This event counts every cycle where the issue of… [all …]
|
/freebsd/lib/libpmc/pmu-events/arch/x86/amdzen3/ |
H A D | floating-point.json | 4 "EventCode": "0x00", 6 … Each increment represents a one- cycle dispatch event. This event is a speculative event. Since t… 7 "UMask": "0x0f" 11 "EventCode": "0x00", 13 … Each increment represents a one-cycle dispatch event. This event is a speculative event. Since th… 14 "UMask": "0x08" 18 "EventCode": "0x00", 20 … Each increment represents a one- cycle dispatch event. This event is a speculative event. Since t… 21 "UMask": "0x04" 25 "EventCode": "0x00", [all …]
|
/freebsd/lib/libpmc/pmu-events/arch/x86/amdzen1/ |
H A D | floating-point.json | 4 "EventCode": "0x00", 5 "BriefDescription": "Total number multi-pipe uOps assigned to all pipes.", 6 …-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the F… 7 "UMask": "0xf0" 11 "EventCode": "0x00", 12 "BriefDescription": "Total number multi-pipe uOps assigned to pipe 3.", 13 …-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the F… 14 "UMask": "0x80" 18 "EventCode": "0x00", 19 "BriefDescription": "Total number multi-pipe uOps assigned to pipe 2.", [all …]
|
/freebsd/usr.bin/gprof/ |
H A D | arcs.c | 1 /*- 2 * SPDX-License-Identifier: BSD-3-Clause 56 count , parentp -> name , childp -> name ); in addarc() 60 if ( arcp != 0 ) { in addarc() 67 arcp -> arc_count , count ); in addarc() 70 arcp -> arc_count += count; in addarc() 76 arcp -> arc_parentp = parentp; in addarc() 77 arcp -> arc_childp = childp; in addarc() 78 arcp -> arc_count = count; in addarc() 82 arcp -> arc_childlist = parentp -> children; in addarc() [all …]
|
H A D | gprof.h | 1 /*- 2 * SPDX-License-Identifier: BSD-3-Clause 43 #define OFFSET_OF_CODE 0 52 #define FALSE 0 93 struct arcstruct *arc_parentlist; /* parents-of-this-child list */ 94 struct arcstruct *arc_childlist; /* children-of-this-parent list */ 95 struct arcstruct *arc_next; /* list of arcs on cycle */ 104 #define DEADARC 0x01 /* time should not propagate across the arc */ 105 #define ONLIST 0x02 /* arc is on list of arcs in cycles */ 127 int toporder; /* graph call chain top-sort order */ [all …]
|
/freebsd/contrib/bmake/unit-tests/ |
H A D | dotwait.mk | 5 TESTS= simple recursive shared cycle 11 # Ignore "--- target ---" lines printed by parallel make. 14 @${.MAKE} -f ${THISMAKEFILE} -j4 $t 2>&1 | grep -v "^--- " 18 # Within each test, the names of the sub-targets follow these 34 PARALLEL_TARG= ${.TARGET:C/\.[a-z]/.*/g:Q} 49 # shared: both shared.1.99 and shared.2.99 depend on shared.0. 50 # shared.0 must be made first, even though it is a child of 53 shared.1.99: shared.0 _ECHOUSE 54 shared.2.99: shared.2.1 shared.0 _ECHOUSE 56 # cycle: the cyclic dependency must not cause infinite recursion [all …]
|
/freebsd/contrib/llvm-project/llvm/lib/CodeGen/ |
H A D | ScoreboardHazardRecognizer.cpp | 1 //===- ScoreboardHazardRecognizer.cpp - Scheduler Support -----------------===// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 10 // encapsultes hazard-avoidance heuristics for scheduling, based on the 13 //===----------------------------------------------------------------------===// 18 #include "llvm/Config/llvm-config.h" 36 // the scoreboard. We always make the scoreboard at least 1 cycle deep to in ScoreboardHazardRecognizer() 39 if (ItinData && !ItinData->isEmpty()) { in ScoreboardHazardRecognizer() 40 for (unsigned idx = 0; ; ++idx) { in ScoreboardHazardRecognizer() [all …]
|
H A D | MachineCycleAnalysis.cpp | 1 //===- MachineCycleAnalysis.cpp - Compute CycleInfo for Machine IR --------===// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 22 char MachineCycleInfoWrapperPass::ID = 0; 29 INITIALIZE_PASS_BEGIN(MachineCycleInfoWrapperPass, "machine-cycles", 30 "Machine Cycle Info Analysis", true, true) 31 INITIALIZE_PASS_END(MachineCycleInfoWrapperPass, "machine-cycles", 32 "Machine Cycle Info Analysis", true, true) 48 OS << "MachineCycleInfo for function: " << F->getName() << "\n"; in print() [all …]
|
/freebsd/sys/contrib/device-tree/Bindings/regulator/ |
H A D | pwm-regulator.txt | 7 predefined voltage <=> duty-cycle values must be 10 Intermediary duty-cycle values which would normally 13 the user if the assumptions made in continuous-voltage 18 regulator-{min,max}-microvolt properties to calculate 19 appropriate duty-cycle values. This allows for a much 21 voltage-table mode above. This solution does make an 22 assumption that a %50 duty-cycle value will cause the 27 -------------------- 28 - compatible: Should be "pwm-regulator" 30 - pwms: PWM specification (See: ../pwm/pwm.txt) [all …]
|
H A D | pwm-regulator.yaml | 1 # SPDX-License-Identifier: GPL-2.0 3 --- 4 $id: http://devicetree.org/schemas/regulator/pwm-regulator.yaml# 5 $schema: http://devicetree.org/meta-schemas/core.yaml# 10 - Brian Norris <briannorris@chromium.org> 11 - Lee Jones <lee@kernel.org> 12 - Alexandre Courbot <acourbot@nvidia.com> 19 duty-cycle values must be provided via DT. Limitations are that the 21 Intermediary duty-cycle values which would normally allow finer grained 23 is given to the user if the assumptions made in continuous-voltage mode do [all …]
|
/freebsd/share/doc/psd/18.gprof/ |
H A D | postp.me | 60 .ce 0 88 In these cases, we discover strongly-connected 92 We use a variation of Tarjan's strongly-connected 94 that discovers strongly-connected components as it is assigning 99 For example, a self-recursive routine 100 (a trivial cycle in the call graph) 109 Time is not propagated from one member of a cycle to another, 112 In addition, children of one member of a cycle 113 must be considered children of all members of the cycle. 114 Similarly, parents of one member of the cycle must inherit [all …]
|
/freebsd/contrib/llvm-project/llvm/include/llvm/CodeGen/ |
H A D | ModuloSchedule.h | 1 //===- ModuloSchedule.h - Software pipeline schedule expansion ------------===// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 18 // A schedule is, for every instruction in a block, a Cycle and a Stage. Note 19 // that we only support single-block loops, so "block" and "loop" can be used 22 // The Cycle of an instruction defines a partial order of the instructions in 23 // the remapped loop. Instructions within a cycle must not consume the output 24 // of any instruction in the same cycle. Cycle information is assumed to have 26 // lock-step (for example in a VLIW ISA). [all …]
|
H A D | MachinePipeliner.h | 1 //===- MachinePipeliner.h - Machine Software Pipeliner Pass -------------===// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 18 // "Swing Modulo Scheduling: A Lifetime-Sensitive Approach", by J. Llosa, 22 // "Lifetime-Sensitive Modulo Scheduling in a Production Environment", by J. 28 // Urbana-Champaign, 2005. 39 //===----------------------------------------------------------------------===// 78 unsigned II_setByPragma = 0; 122 unsigned MII = 0; [all …]
|
H A D | ScheduleHazardRecognizer.h | 1 //=- llvm/CodeGen/ScheduleHazardRecognizer.h - Scheduling Support -*- C++ -*-=// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 10 // hazard-avoidance heuristics for scheduling. 12 //===----------------------------------------------------------------------===// 22 /// HazardRecognizer - This determines whether or not an instruction can be 23 /// issued this cycle, and whether or not a noop needs to be inserted to handle 27 /// MaxLookAhead - Indicate the number of cycles in the scoreboard 29 /// MaxLookAhead=0 identifies a fake recognizer, allowing the client to [all …]
|
H A D | ScoreboardHazardRecognizer.h | 1 //=- llvm/CodeGen/ScoreboardHazardRecognizer.h - Schedule Support -*- C++ -*-=// 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 7 //===----------------------------------------------------------------------===// 10 // encapsulates hazard-avoidance heuristics for scheduling, based on the 13 //===----------------------------------------------------------------------===// 30 // Scoreboard to track function unit usage. Scoreboard[0] is a 31 // mask of the FUs in use in the cycle currently being 32 // schedule. Scoreboard[1] is a mask for the next cycle. The 33 // Scoreboard is used as a circular buffer with the current cycle [all …]
|
/freebsd/lib/libpmc/pmu-events/arch/x86/ivybridge/ |
H A D | floating-point.json | 4 "Counter": "0,1,2,3", 5 "CounterHTOff": "0,1,2,3", 7 "EventCode": "0xCA", 11 "UMask": "0x1e" 15 "Counter": "0,1,2,3", 16 "CounterHTOff": "0,1,2,3,4,5,6,7", 17 "EventCode": "0xCA", 21 "UMask": "0x10" 25 "Counter": "0,1,2,3", 26 "CounterHTOff": "0,1,2,3,4,5,6,7", [all …]
|
/freebsd/lib/libpmc/pmu-events/arch/x86/ivytown/ |
H A D | floating-point.json | 4 "Counter": "0,1,2,3", 5 "CounterHTOff": "0,1,2,3", 7 "EventCode": "0xCA", 11 "UMask": "0x1e" 15 "Counter": "0,1,2,3", 16 "CounterHTOff": "0,1,2,3,4,5,6,7", 17 "EventCode": "0xCA", 21 "UMask": "0x10" 25 "Counter": "0,1,2,3", 26 "CounterHTOff": "0,1,2,3,4,5,6,7", [all …]
|
/freebsd/sys/contrib/device-tree/Bindings/input/ |
H A D | pwm-vibrator.yaml | 1 # SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) 3 --- 4 $id: http://devicetree.org/schemas/input/pwm-vibrator.yaml# 5 $schema: http://devicetree.org/meta-schemas/core.yaml# 10 - Sebastian Reichel <sre@kernel.org> 14 strength increases based on the duty cycle of the enable PWM channel 15 (100% duty cycle meaning strongest vibration, 0% meaning no vibration). 18 driven at fixed duty cycle. If available this is can be used to increase 23 const: pwm-vibrator 25 pwm-names: [all …]
|
H A D | pwm-vibrator.txt | 4 strength increases based on the duty cycle of the enable PWM channel 5 (100% duty cycle meaning strongest vibration, 0% meaning no vibration). 8 driven at fixed duty cycle. If available this is can be used to increase 12 - compatible: should contain "pwm-vibrator" 13 - pwm-names: Should contain "enable" and optionally "direction" 14 - pwms: Should contain a PWM handle for each entry in pwm-names 17 - vcc-supply: Phandle for the regulator supplying power 18 - direction-duty-cycle-ns: Duty cycle of the direction PWM channel in 26 pinctrl-single,pins = < 27 OMAP4_IOPAD(0x1ce, PIN_OUTPUT | MUX_MODE1) /* dmtimer8_pwm_evt (gpio_27) */ [all …]
|
/freebsd/lib/libpmc/pmu-events/arch/x86/amdzen2/ |
H A D | floating-point.json | 4 "EventCode": "0x00", 6 … Each increment represents a one- cycle dispatch event. This event is a speculative event. Since t… 7 "UMask": "0x0f" 11 "EventCode": "0x00", 13 … Each increment represents a one-cycle dispatch event. This event is a speculative event. Since th… 14 "UMask": "0x08" 18 "EventCode": "0x00", 20 … Each increment represents a one- cycle dispatch event. This event is a speculative event. Since t… 21 "UMask": "0x04" 25 "EventCode": "0x00", [all …]
|