How efficient is reading a file one byte at a time in C?

Question

After going through most of the book, "The C Programming Language," I think I have a decent grasp on programming in C. One common C idiom presented in that book is reading a file a single byte at a time, using functions like getchar() and fgetc(). So far, I've been using these functions to do all IO in my C programs.

My question is, is this an efficient way of reading a file? Does a call to get a single byte require a lot of overhead that can be minimized if I read multiple bytes into a buffer at a time, for instance by using the read() system call on Unix systems? Or do the operating system and C library handle a buffer behind the scenes to make it more efficient? Also, does this work the same way for writing to files a single byte at a time?

I would like to know how this generally works in C, but if it is implementation or OS specific, I would like to know how it works in GCC on common Unix-like systems (like macOS and linux).

Answer 1

Using getchar() etc is efficient because the standard I/O library uses buffering to read many bytes at once (saving them in a buffer) and doles them out one at a time when you call getchar().

Using read() to read a single byte at a time is much slower, typically, because it makes a full system call each time. It still isn't catastrophically slow, but it is nowhere near as fast as reading 512, or 4096, bytes into a buffer.

Those are broad, sweeping statements. There are many caveats that could be added, but they are a reasonable general outline of the performance of getchar(), etc.