Merge pull request #6 from ashvardanian/refactor

Latency Hiding, CPU Ports, and Huge Pages 🐧

Author: ashvardanian

.clang-format

@@ -1,3 +1,69 @@
+Language: Cpp
 BasedOnStyle: LLVM
 IndentWidth: 4
-ColumnLimit: 120
+TabWidth: 4
+NamespaceIndentation: None
+ColumnLimit: 120
+ReflowComments: true
+UseTab: Never
+
+AlignConsecutiveAssignments: false
+AlignConsecutiveDeclarations: false
+AlignEscapedNewlines: true
+AlignOperands: true
+AlignTrailingComments: true
+
+AllowAllArgumentsOnNextLine: false
+AllowAllConstructorInitializersOnNextLine: true
+AllowAllParametersOfDeclarationOnNextLine: true
+AllowShortBlocksOnASingleLine: Always
+AllowShortIfStatementsOnASingleLine: Always
+AllowShortCaseLabelsOnASingleLine: true
+AllowShortFunctionsOnASingleLine: true
+AllowShortLambdasOnASingleLine: true
+AllowShortLoopsOnASingleLine: true
+AlwaysBreakTemplateDeclarations: Yes
+AlwaysBreakAfterReturnType: None
+PenaltyReturnTypeOnItsOwnLine: 200
+
+BreakBeforeBraces: Custom
+BraceWrapping:
+  AfterCaseLabel: false
+  AfterClass: false
+  AfterControlStatement: false
+  AfterEnum: false
+  AfterExternBlock: false
+  AfterFunction: false
+  AfterStruct: false
+  AfterNamespace: false
+  AfterUnion: false
+  BeforeCatch: true
+  BeforeElse: true
+  SplitEmptyFunction: false
+  SplitEmptyRecord: false
+  SplitEmptyNamespace: false
+  IndentBraces: false
+
+SortIncludes: true
+SortUsingDeclarations: true
+
+SpaceAfterCStyleCast: false
+SpaceAfterLogicalNot: false
+SpaceAfterTemplateKeyword: true
+SpaceBeforeAssignmentOperators: true
+SpaceBeforeCpp11BracedList: true
+SpaceBeforeCtorInitializerColon: true
+SpaceBeforeInheritanceColon: true
+SpaceBeforeParens: ControlStatements
+SpaceBeforeRangeBasedForLoopColon: true
+SpaceInEmptyParentheses: false
+SpacesBeforeTrailingComments: 1
+SpacesInAngles: false
+SpacesInCStyleCastParentheses: false
+SpacesInContainerLiterals: false
+SpacesInParentheses: false
+SpacesInSquareBrackets: false
+
+BinPackArguments: true
+BinPackParameters: true
+PenaltyBreakBeforeFirstCallParameter: 1

.vscode/launch.json

@@ -10,8 +10,33 @@
             "preLaunchTask": "Build Debug",
             "program": "${workspaceFolder}/build_debug/reduce_bench",
             "args": [
-                "--benchmark_filter=cuda"
-            ]
+                "--benchmark_filter=sse"
+            ],
+            "setupCommands": [
+                { // Display content in STL containers pretty
+                    "description": "Enable pretty-printing for gdb",
+                    "text": "-enable-pretty-printing",
+                    "ignoreFailures": true
+                }
+            ],
+            "environment": [
+                {
+                    "name": "ASAN_OPTIONS",
+                    "value": "detect_leaks=0:atexit=1:strict_init_order=1:strict_string_checks=1"
+                }
+            ],
+            "stopAtEntry": false,
+            "linux": {
+                "MIMode": "gdb"
+            },
+            "osx": {
+                "MIMode": "lldb"
+            },
+            "windows": {
+                "program": "${workspaceFolder}\\build_debug\\reduce_bench.exe",
+                "MIMode": "gdb",
+                "miDebuggerPath": "C:\\MinGw\\bin\\gdb.exe"
+            }
         },
         {
             "name": "CUDA C++: Launch",

.vscode/settings.json

@@ -2,17 +2,25 @@
     "cSpell.words": [
         "ashvardanian",
         "blas",
+        "blasint",
+        "cblas",
         "CCCL",
         "constexpr",
         "cublas",
         "CUDA",
+        "GTEST",
+        "HUGEPAGE",
+        "HUGETLB",
         "Kahan",
         "METALLIB",
+        "NUMA",
+        "opencl",
         "openmp",
         "shfl",
         "SPIR",
         "STREQUAL",
         "threadgroup",
+        "unseq",
         "Vardanian",
         "wmma"
     ],
@@ -21,8 +29,13 @@
         "*.ipp": "cpp",
         "*.metal": "cpp",
         "*.tcc": "cpp",
+        "__bit_reference": "cpp",
+        "__config": "cpp",
         "__debug": "cpp",
+        "__hash_table": "cpp",
         "__nullptr": "cpp",
+        "__split_buffer": "cpp",
+        "__tree": "cpp",
         "algorithm": "cpp",
         "any": "cpp",
         "array": "cpp",
@@ -57,13 +70,16 @@
         "deque": "cpp",
         "exception": "cpp",
         "expected": "cpp",
+        "filesystem": "cpp",
+        "format": "cpp",
         "forward_list": "cpp",
         "fstream": "cpp",
         "functional": "cpp",
         "future": "cpp",
         "hash_map": "cpp",
         "hash_set": "cpp",
         "initializer_list": "cpp",
+        "inplace_vector": "cpp",
         "iomanip": "cpp",
         "iosfwd": "cpp",
         "iostream": "cpp",
@@ -100,6 +116,7 @@
         "string_view": "cpp",
         "strstream": "cpp",
         "system_error": "cpp",
+        "text_encoding": "cpp",
         "thread": "cpp",
         "tuple": "cpp",
         "type_traits": "cpp",

.vscode/tasks.json

@@ -6,12 +6,12 @@
         {
             "label": "Build Debug",
             "type": "shell",
-            "command": "cmake -B ./build_debug && cmake --build ./build_debug --config Debug"
+            "command": "cmake -B ./build_debug -DCMAKE_BUILD_TYPE=Debug && cmake --build ./build_debug --config Debug"
         },
         {
             "label": "Build & Run Release",
             "type": "shell",
-            "command": "cmake -B ./build_release && cmake --build ./build_release --config Release && ./build_release/reduce_bench"
+            "command": "cmake -B ./build_release -DCMAKE_BUILD_TYPE=Release && cmake --build ./build_release --config Release && ./build_release/reduce_bench"
         }
     ]
 }