Here's a practical, minimal guide for using GDB in Ubuntu terminal — from launching to setting breakpoints and navigating your code. 🧰 Step 1: Compile with Debug Symbols Use -g and disable optimizations ( -O0 preferred for debugging): bash Copy Edit g++ -g -O0 -o myprog myprog.cpp 🚀 Step 2: Launch GDB bash Copy Edit gdb ./myprog You'll see a (gdb) prompt. 🎯 Step 3: Set Breakpoints You can set breakpoints in three main ways : ✅ a. At a function gdb Copy Edit (gdb) break main ✅ b. At a specific file and line gdb Copy Edit (gdb) break myprog.cpp:42 ✅ c. At a method in a class gdb Copy Edit (gdb) break MyClass::doSomething If the symbol is mangled (C++), GDB usually demangles automatically, especially if compiled with -g . ▶️ Step 4: Run the Program gdb Copy Edit (gdb) run 🧭 Step 5: Navigate Command Meaning next or n Step over current line step or s Step into function call finish Run until current function returns continue or c Continue until...
Question:
suppose
string fn1(){ char * cstr=... string str(cstr); free(cstr); return str; } shared_ptr<char*,&free> fn2(){ char *cstr=... shared_ptr<char*,&free> ptr=... return ptr; } unique_ptr<char*,&free> fn3(){ char *cstr=... unique_ptr<char*,&free> ptr=... return ptr; } Compare above 3 Fns wrt memory require, performance upon calling in main(){...}Answer:
Let's compare the three functions `fn1()`, `fn2()`, and `fn3()` in terms of memory requirements and performance.
1. `fn1()`:
```cpp
string fn1() {
char* cstr = ... // Assuming cstr is allocated dynamically, e.g., using malloc
string str(cstr);
free(cstr); // Release the memory after creating the string
return str;
}
```
In this function, memory is dynamically allocated for `cstr`, and then it is used to create a `string` object. After creating the `string`, the memory is explicitly released using `free()`. This approach has an additional overhead of manual memory management (allocating and deallocating memory using `malloc` and `free`). It is prone to errors like forgetting to free memory or using the `cstr` pointer after it has been freed.
2. `fn2()`:
```cpp
shared_ptr<char*, &free> fn2() {
char* cstr = ... // Assuming cstr is allocated dynamically, e.g., using malloc
shared_ptr<char*, &free> ptr(cstr, &free); // Using a shared_ptr to manage the memory
return ptr;
}
```
In this function, a `shared_ptr` is used to manage the memory. The `shared_ptr` is initialized with a custom deleter function `&free`, which will be called automatically when the last `shared_ptr` referring to the same memory is destructed. This approach automatically takes care of memory management and reduces the risk of memory leaks. It also provides shared ownership of the memory, which can be useful in certain scenarios.
3. `fn3()`:
```cpp
unique_ptr<char*, &free> fn3() {
char* cstr = ... // Assuming cstr is allocated dynamically, e.g., using malloc
unique_ptr<char*, &free> ptr(cstr, &free); // Using a unique_ptr to manage the memory
return ptr;
}
```
In this function, a `unique_ptr` is used to manage the memory. The `unique_ptr` guarantees that there will be only one owner of the memory, and the memory will be automatically released when the `unique_ptr` goes out of scope. This approach provides better memory safety compared to raw pointers or `shared_ptr` because it ensures exclusive ownership of the memory.
Comparison:
- Memory Requirement: The memory requirement for `fn1()` is the same as `fn2()` and `fn3()` because they all use the same dynamically allocated memory for `cstr`.
- Performance: In terms of performance, `fn2()` and `fn3()` are better choices than `fn1()`. The reason is that `fn2()` and `fn3()` use smart pointers (`shared_ptr` and `unique_ptr`), which perform automatic memory management without the need for manual calls to `free()`. This eliminates the overhead of manual memory management and reduces the risk of memory leaks and other memory-related errors.
Conclusion:
In general, using `shared_ptr` or `unique_ptr` to manage dynamically allocated memory is preferred over manual memory management with raw pointers and explicit calls to `free()` or `delete`. Among `fn2()` and `fn3()`, the choice depends on your specific use case:
- If you need shared ownership of the memory (i.e., the memory is shared among multiple `shared_ptr` instances), use `fn2()`.
- If you only need exclusive ownership of the memory (i.e., the memory is managed by a single `unique_ptr` instance), use `fn3()`.
For most cases, `fn3()` with `unique_ptr` is recommended as it provides better memory safety and performance compared to manual memory management.
Comments
Post a Comment
If you have any doubts, please let me know.