Add codespell spell checker (makefile, CI)

.codespellignore

@@ -0,0 +1,4 @@
+ba
+fo
+opne
+optimyze

.codespellrc

@@ -0,0 +1,10 @@
+# https://github.com/codespell-project/codespell
+[codespell]
+builtin = clear,rare,informal
+check-filenames =
+check-hidden =
+ignore-words = .codespellignore
+interactive = 1
+skip = .git,AUTHORS.md,go.mod,go.sum,LICENSES,venv,zydis
+uri-ignore-words-list = *
+write =

.github/workflows/codespell.yml

@@ -0,0 +1,16 @@
+name: codespell
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+jobs:
+  codespell:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Install codespell
+        run: sudo apt-get install codespell
+      - name: Checkout Repo
+        uses: actions/checkout@v4
+      - name: Codespell
+        run: make codespell

.gitignore

@@ -6,3 +6,5 @@
 opentelemetry-ebpf-profiler
 ci-kernels
 
+# Installed by `make codespell`
+venv/

Makefile

@@ -1,5 +1,5 @@
 .PHONY: all all-common binary clean ebpf generate test test-deps protobuf docker-image agent legal \
-	integration-test-binaries lint linter-version
+	integration-test-binaries codespell lint linter-version
 
 SHELL := /usr/bin/env bash
 
@@ -114,3 +114,7 @@ legal:
 	@echo "Dependencies license summary (from deps.profiling-agent.csv):"
 	@echo "  Count License"
 	@tail -n '+2' deps.profiling-agent.csv | cut -d',' -f5 | sort | uniq -c | sort -k1rn
+
+
+codespell: 
+	@codespell

doc/gopclntab.md

@@ -398,7 +398,7 @@ struct runtime_inlinedCall {
 
 This section provides some examples on what structures need to be traversed to
 locate various function information. This omits details like relative offsets
-that depend on data from the header or moduledata to keep complexity managable.
+that depend on data from the header or moduledata to keep complexity manageable.
 
 ### Locating `runtime._func` for a PC
 

doc/symb-proto/README.md

@@ -49,7 +49,7 @@ frame symbol information are uploaded via `/api/symbols-ranges`. The "return
 pads" files that contain leaf frame symbol information are uploaded via
 `/api/symbols-returnpads`. Symbfiles may be split and uploaded in multiple
 chunks (in separate HTTP requests) for improved load balancing in the presence
-of muliple symbolizer services.
+of multiple symbolizer services.
 
 ### File metadata (request)
 

interpreter/dotnet/dotnet.go

@@ -50,7 +50,7 @@ package dotnet
 //     with its metadata maps. insert this into ebpf pid_page mappings.
 //  2. just use standard frame pointer unwinding in ebpf, and also locate the JIT function
 //     code header to get access to debug data and method descriptors
-//  3. in the host agent, the debug data and method descriptors are resolved and mappped to
+//  3. in the host agent, the debug data and method descriptors are resolved and mapped to
 //     PE (.dll) FileID, Method index, and the IL code offset (JIT code), or the PE FileID
 //     and Relative Virtual Address (RVA) (Ready to Run code)
 //  4. symbolizer can then map the above data to source code file and line
@@ -128,7 +128,7 @@ var (
 	// regex for the core language runtime
 	dotnetRegex = regexp.MustCompile(`/(\d+)\.(\d+).(\d+)/libcoreclr.so$`)
 
-	// The FileID used for Dotnet stub frames. Same FileID as in other interpeters.
+	// The FileID used for Dotnet stub frames. Same FileID as in other interpreters.
 	stubsFileID = libpf.NewFileID(0x578b, 0x1d)
 
 	// compiler check to make sure the needed interfaces are satisfied

interpreter/dotnet/method.go

@@ -95,7 +95,7 @@ func (m *dotnetMethod) mapPCOffsetToILOffset(pcOffset uint32, findCall bool) uin
 			lastCallILOffset = ilOffset
 		}
 
-		// NOTE: _DEBUG builds could have a 0xA nibble to indentify row change.
+		// NOTE: _DEBUG builds could have a 0xA nibble to identify row change.
 		log.Debugf(" %3d, native %3d -> IL %#03x, sourceFlags %#x",
 			i, nativeOffset, ilOffset, sourceFlags)
 	}
@@ -116,7 +116,7 @@ func (m *dotnetMethod) dumpBounds() {
 		nativeOffset += nr.Uint32()
 		ilOffset := uint32(int32(nr.Uint32()) + mappingTypeMaxValue)
 		sourceFlags := nr.Uint32()
-		// NOTE: _DEBUG builds could have a 0xA nibble to indentify row change.
+		// NOTE: _DEBUG builds could have a 0xA nibble to identify row change.
 
 		log.Debugf(" %3d, native %3d -> IL %#03x, sourceFlags %#x",
 			i, nativeOffset, ilOffset, sourceFlags)

interpreter/dotnet/pe.go

@@ -165,7 +165,7 @@ type StreamHeader struct {
 	Size   uint32
 }
 
-// table* variables are ECMA-335 II.22 defined Metdata table numbers
+// table* variables are ECMA-335 II.22 defined Metadata table numbers
 const (
 	tableModule                 = 0x00
 	tableTypeRef                = 0x01

interpreter/hotspot/stubs.go

@@ -87,7 +87,7 @@ func findStubBounds(vmd *hotspotVMData, bias libpf.Address,
 	for i := 0; i < len(stubs); i++ {
 		cur := &stubs[i]
 
-		// Some stubs re-use the code from another stub. Skip elements until
+		// Some stubs reuse the code from another stub. Skip elements until
 		// we detected the next stub that doesn't occupy the same address.
 		for i < len(stubs) {
 			if i != len(stubs)-1 {

interpreter/php/decode_amd64.c

@@ -69,7 +69,7 @@ int retrieveJITBufferPtr(const uint8_t * const code, const size_t codesize,
 // which allows us to recover accurate PC data for JIT code
 int retrieveExecuteExJumpLabelAddress(const uint8_t * const code, const size_t codesize,
                                       const uint64_t rip_base, uint64_t * const out) {
-  // The raison d'etre for this function is described in the php8 unwinding doc,
+  // The purpose of this function is described in the php8 unwinding doc,
   // in particular in the "disassembling execute_ex" section.
   ZydisDecoder decoder;
   ZydisDecoderInit(&decoder, ZYDIS_MACHINE_MODE_LONG_64, ZYDIS_STACK_WIDTH_64);

libpf/basehash/hash128_test.go

@@ -136,23 +136,23 @@ func TestHash128Format(t *testing.T) {
 		0xCC, 0xDD, 0xEE, 0xFF})
 
 	tests := map[string]struct {
-		formater string
-		expected string
+		formatter string
+		expected  string
 	}{
-		"v":  {formater: "%v", expected: "{4822678189205111 9843086184167632639}"},
-		"d":  {formater: "%d", expected: "{4822678189205111 9843086184167632639}"},
-		"x":  {formater: "%x", expected: "112233445566778899aabbccddeeff"},
-		"X":  {formater: "%X", expected: "112233445566778899AABBCCDDEEFF"},
-		"#v": {formater: "%#v", expected: "0x112233445566778899aabbccddeeff"},
-		"#x": {formater: "%#x", expected: "0x112233445566778899aabbccddeeff"},
-		"#X": {formater: "%#X", expected: "0x112233445566778899AABBCCDDEEFF"},
+		"v":  {formatter: "%v", expected: "{4822678189205111 9843086184167632639}"},
+		"d":  {formatter: "%d", expected: "{4822678189205111 9843086184167632639}"},
+		"x":  {formatter: "%x", expected: "112233445566778899aabbccddeeff"},
+		"X":  {formatter: "%X", expected: "112233445566778899AABBCCDDEEFF"},
+		"#v": {formatter: "%#v", expected: "0x112233445566778899aabbccddeeff"},
+		"#x": {formatter: "%#x", expected: "0x112233445566778899aabbccddeeff"},
+		"#X": {formatter: "%#X", expected: "0x112233445566778899AABBCCDDEEFF"},
 	}
 
 	for name, test := range tests {
 		name := name
 		test := test
 		t.Run(name, func(t *testing.T) {
-			output := fmt.Sprintf(test.formater, h)
+			output := fmt.Sprintf(test.formatter, h)
 			assert.Equal(t, test.expected, output)
 		})
 	}

libpf/pfelf/pfelf.go

@@ -25,7 +25,7 @@ import (
 
 // SafeOpenELF opens the given ELF file in a safely way in that
 // it recovers from panics inside elf.Open().
-// Under cirumstances we see fatal errors from inside the runtime, which
+// Under circumstances we see fatal errors from inside the runtime, which
 // are not recoverable, e.g. "fatal error: runtime: out of memory".
 func SafeOpenELF(name string) (elfFile *elf.File, err error) {
 	defer func() {

maccess/maccess_arm64.go

@@ -114,7 +114,7 @@ func CopyFromUserNoFaultIsPatched(codeblob []byte, _, _ uint64) (bool, error) {
 					//   0 otherwise
 					// - C
 					//   1 if CMP results in carry condition, like unsigned overflow
-					//   0 otherweise
+					//   0 otherwise
 					expectedInstructionTracker ^= stepB
 					continue
 				}

periodiccaller/periodiccaller_test.go

@@ -55,7 +55,7 @@ func isSelfOrRuntime(t *testing.T, stack *[32]uintptr, self string) bool {
 			if funcName == self {
 				return true
 			}
-			// Go runtime specifc filters
+			// Go runtime specific filters
 			if !strings.HasPrefix(funcName, "runtime.") &&
 				!strings.HasPrefix(funcName, "runtime/") &&
 				!strings.HasPrefix(funcName, "testing.") &&

reporter/config.go

@@ -16,7 +16,7 @@ type Config struct {
 	// Name defines the name of the agent.
 	Name string
 
-	// Version defines the vesion of the agent.
+	// Version defines the version of the agent.
 	Version string
 
 	// CollAgentAddr defines the destination of the backend connection.

support/ebpf/native_stack_trace.ebpf.c

@@ -630,7 +630,7 @@ static inline ErrorCode copy_state_regs(UnwindState *state,
   // https://github.com/torvalds/linux/blob/2ef5971ff3/arch/x86/entry/entry_64.S#L847
   state->return_address = interrupted_kernelmode && regs->orig_ax != -1;
 #elif defined(__aarch64__)
-  // For backwards compatability aarch64 can run 32-bit code.
+  // For backwards compatibility aarch64 can run 32-bit code.
   // Check if the process is running in this 32-bit compat mod.
   if (regs->pstate & PSR_MODE32_BIT) {
     return ERR_NATIVE_AARCH64_32BIT_COMPAT_MODE;

support/ebpf/php_tracer.ebpf.c

@@ -87,7 +87,7 @@ int process_php_frame(PerCPURecord *record, PHPProcInfo *phpinfo, bool is_jitted
     }
 
     // Get execute_data->This.type_info. This reads into the `type_info` argument
-    // so we can re-use it in walk_php_stack
+    // so we can reuse it in walk_php_stack
     if (bpf_probe_read_user(type_info, sizeof(u32),
                             execute_data + phpinfo->zend_execute_data_this_type_info)) {
       DEBUG_PRINT("Failed to read execute_data->This.type_info (0x%lx)",

support/ebpf/tracemgmt.h

@@ -127,7 +127,7 @@ bool pid_event_ratelimit(u32 pid, int ratelimit_action) {
 
   // Update the map entry. Technically this is not SMP safe, but doing
   // an atomic update would require EBPF atomics. At worst we send an
-  // extra sync event and the likelyhood for this race is very low, so
+  // extra sync event and the likelihood for this race is very low, so
   // we can live with this.
   int err = bpf_map_update_elem(&reported_pids, &pid, &token, BPF_ANY);
   if (err != 0) {
@@ -314,7 +314,7 @@ static bool is_kernel_address(u64 addr) {
 
 // resolve_unwind_mapping decodes the current PC's mapping and prepares unwinding information.
 // The state text_section_id and text_section_offset are updated accordingly. The unwinding program
-// index that should be used is writen to the given `unwinder` pointer.
+// index that should be used is written to the given `unwinder` pointer.
 static ErrorCode resolve_unwind_mapping(PerCPURecord *record, int* unwinder) {
   UnwindState *state = &record->state;
   pid_t pid = record->trace.pid;

support/ebpf/types.h

@@ -614,7 +614,7 @@ typedef struct RubyUnwindState {
 typedef struct DotnetUnwindScratchSpace {
   // Buffer to read nibble map to locate code start. One map entry allows seeking backwards
   // 32*8 = 256 bytes of code. This defines the maximum size for a JITted function we
-  // can reconize: 256 bytes/element * 128 elements = 32kB function size.
+  // can recognize: 256 bytes/element * 128 elements = 32kB function size.
   u32 map[128];
   // Extra space to read to map fixed amount of bytes, but to dynamic offset.
   u32 extra[128];

tools/coredump/testsources/node/async.js

@@ -19,7 +19,7 @@ const notifyWaiter = (id) => {
     setTimeout(() => {
       console.log(`Order for customer at table #${id} processed....`);
       resolve({ customerId: id, customerOrder: "Pizza" });
-      // reject(new Error("Error occured with waiter"));
+      // reject(new Error("Error occurred with waiter"));
     }, 1);
   });
 }