Home
last modified time | relevance | path

Searched refs:loads (Results 1 – 25 of 396) sorted by relevance

12345678910>>...16

/openbsd-src/gnu/llvm/llvm/docs/
H A DAtomics.rst95 Note that speculative loads are allowed; a load which is part of a race returns
101 For cases where simple loads and stores are not sufficient, LLVM provides
106 non-atomic loads and stores, but provide additional guarantees in situations
144 location, loads from that location return undef.
152 The rule is essentially that all memory accessed with basic loads and stores
156 variable. Note that NotAtomic volatile loads and stores are not properly
162 Introducing loads to shared variables along a codepath where they would not
193 platforms for loads of a native width, but can be expensive or unavailable for
194 wider loads, like a 64-bit store on ARM. (A frontend for Java or other "safe"
200 single load into multiple loads, transforms a store into multiple stores,
[all …]
H A DSpeculativeLoadHardening.md54 branches and/or loads inside especially risky software (most notably various OS
63 The specific technique we propose is to cause loads to be checked using
66 predicate guarding potentially invalid loads:
153 * To achieve reasonable performance, many loads may not be checked, such as
160 * [Hardened loads](#hardening-the-address-of-the-load) may still load data from
167 block misspeculated loads was developed as part of a HACS 2018 discussion
231 use that validation to check that loads cannot leak information along
267 *implicitly* loads the return address off the stack. However, the
302 Mitigating this requires hardening loads from these locations, or mitigating
329 checking the loads, and checking control transfers between procedures.
[all …]
H A DBigEndianNEON.rst11 Generating code for big endian ARM processors is for the most part straightforward. NEON loads and …
13 The aim of this document is to explain the problem with NEON loads and stores, and the solution tha…
71 It may seem that ``LD1`` should suffice to peform vector loads on a big endian machine. However the…
84 … throughout this section we only mention loads. Stores have exactly the same problems as their ass…
159 …ructions so that they are never allowed to be selected to generate vector loads and stores. The ex…
/openbsd-src/gnu/llvm/llvm/lib/Analysis/models/
H A Dlog_reader.py81 header = json.loads(f.readline())
87 event = json.loads(event_str)
98 score_header = json.loads(f.readline())
/openbsd-src/gnu/llvm/llvm/lib/Target/WebAssembly/
H A DWebAssemblyInstrAtomics.td118 // Atomic loads
129 // Select loads
133 // Extending loads. Note that there are only zero-extending atomic loads, no
134 // sign-extending loads.
141 // Fragments for extending loads. These are different from regular loads because
146 // Unlike regular loads, extension to i64 is handled differently than i32.
149 // Extension to i32 is elided by SelectionDAG as our atomic loads are
161 // We don't have single sext atomic load instructions. So for sext loads, we
162 // match bare subword loads (for 32-bit results) and anyext loads (for 64-bit
170 // Select zero-extending loads
[all …]
/openbsd-src/distrib/notes/i386/
H A Dcontents26 cdbr First-stage CD bootstrap (boot sector), which loads
33 then loads the bsd.rd installation ramdisk kernel.
/openbsd-src/gnu/llvm/llvm/lib/Target/Hexagon/MCTargetDesc/
H A DHexagonShuffler.cpp374 if (Summary.loads == 1 && Summary.loads == Summary.memory && in restrictStoreLoadOrder()
390 else if (Summary.loads >= 1 && isMemReorderDisabled()) { // }:mem_noshuf in restrictStoreLoadOrder()
414 (Summary.loads == 0) || in restrictStoreLoadOrder()
497 ++Summary.loads; in GetPacketSummary()
523 ++Summary.loads; in GetPacketSummary()
535 ++Summary.loads; in GetPacketSummary()
/openbsd-src/distrib/notes/amd64/
H A Dcontents26 cdbr First-stage CD bootstrap (boot sector), which loads
33 then loads the bsd.rd installation ramdisk kernel.
/openbsd-src/gnu/llvm/llvm/lib/Target/AArch64/
H A DAArch64InstrAtomics.td22 // Atomic loads
53 // 8-bit loads
55 // 16-bit loads
57 // 32-bit loads
59 // 64-bit loads
63 // 8-bit loads
79 // 16-bit loads
95 // 32-bit loads
111 // 64-bit loads
127 // FP 32-bit loads
[all …]
/openbsd-src/gnu/llvm/lldb/bindings/python/
H A Dlldb-python12 info = json.loads(info_json)
/openbsd-src/gnu/llvm/llvm/utils/
H A Dmerge-stats.py22 data = json.loads(text)
/openbsd-src/gnu/usr.bin/perl/cpan/Math-BigInt/t/
H A Duse_lib1.t
H A Duse_lib6.t
H A Duse_lib5.t
H A Duse_lib2.t
H A Duse_lib4.t
/openbsd-src/gnu/llvm/lldb/source/Plugins/DynamicLoader/Darwin-Kernel/
H A DDynamicLoaderDarwinKernelProperties.td7 Desc<"Automatically loads kext images when attaching to a kernel.">;
/openbsd-src/gnu/llvm/llvm/docs/Frontend/
H A DPerformanceTips.rst67 Avoid loads and stores of large aggregate type
70 LLVM currently does not optimize well loads and stores of large :ref:`aggregate
103 ARM ISAs) then the hardware does not handle unaligned loads and stores, and
106 sequence of loads, shifts and masks (or load-right + load-left on MIPS) for
120 Note that if your loads and stores are atomic, the backend will be unable to
122 As a result, alignment is mandatory for atomic loads and stores.
153 be limited for loads not in the header. (This is due to the fact that LLVM
/openbsd-src/gnu/usr.bin/gcc/gcc/testsuite/gcc.c-torture/execute/
H A D20020720-1.x6 # insn performs abs, 1 insn loads 0.0, 1 insn sets CCmode flags based
/openbsd-src/gnu/llvm/llvm/lib/Target/ARM/
H A DARMScheduleM55.td29 // loads) actually take 1 cycle (amazingly).
41 // But a sequence of VLDR;VLDRB;VADD;VSTR because the loads cannot overlap,
314 // Normal loads
316 // Pre/post inc loads
318 // Gather loads
322 // Interleaving loads
324 // Interleaving loads with wb
/openbsd-src/gnu/llvm/llvm/lib/Target/Mips/
H A DRelocation.txt81 "lw" loads a 32 bit value from memory. On MIPS64, "lw" loads a 32 bit
/openbsd-src/gnu/llvm/llvm/include/llvm/IR/
H A DIntrinsicsBPF.td13 // Specialized loads from packet
/openbsd-src/gnu/llvm/llvm/include/llvm/Target/GlobalISel/
H A DSelectionDAGCompat.td32 // SelectionDAG has one node for all loads and uses predicates to
166 // G_LOAD handles both atomic and non-atomic loads where as SelectionDAG had
167 // separate nodes for them. This GINodeEquiv maps the non-atomic loads to
/openbsd-src/gnu/llvm/clang/docs/HLSL/
H A DResourceTypes.rst23 specifies the expected return type of resource loads, and the expected parameter
/openbsd-src/gnu/llvm/llvm/utils/llvm-locstats/
H A Dllvm-locstats.py12 from json import loads
210 json_parsed = loads(cmd_stdout)

12345678910>>...16