• Support for MMX, 3DNow!, SSE, and SSE2 instructions. Enable these with the -mmmx, -m3dnow, -msse, and -msse2 options.
• Intel C++ compatible MMX/3DNow!/SSE intrinsics are implemented.
• Targets were added for Intel Pentium® MMX, Pentium III, and Pentium 4 processors.
• For those targets that support it, -mfpmath=sse causes the compiler to generate SSE/SSE2 instructions for floating point math instead of x87 instructions. Usually, this will lead to quicker code, especially on the Pentium 4 processor. Note that only scalar floating point instructions are used and GCC does not exploit SIMD features yet.
• Prefetch support has been added to the Pentium III and Pentium 4 processors.
• Code generated for floating-point-to-integer-conversions has been improved, leading to better performance of many 3D applications.
• By default, gcc compiles code for the machine it is installed on. This is a configurable option, and different gcc configurations for various target machines can exist side-by-side.

Here is the basic syntax for invoking the C compiler:

$ gcc -o executable-name optimizations sourcefile.c 

A typical Hello World example:

$ gcc -o hello -Wall -march=i686 -ggdb -O2 hello.c

• -Wall means display the most useful warnings
• -march=i686 optimizes for Intel 686 CPU
• -ggdb adds debugging information that gdb, the GNU debugger, can read
• -O2 optimizes code size and increases execution speed. (Choices are O1-O3)

Intel CPU target options include the entire x86 family and 960. The O flag is interesting, O3 turns on automatic function inlining, which usually creates fatter code, as well as the fastest.

There are many compiler options, and here are some of the more commonly used ones:

-c = compile or assemble but do not link
-S = stop after compilation, do not assemble
-E = stop after preprocessing, do not compile
-fsyntax-only = check for syntax errors
-pedantic = issue all the warnings demanded by strict ISO C and ISO C++
-static = link against static libraries. The default is shared libraries.