This first thing that needs to be done is compile the source. When compiling, the compiler needs to know the processor type so the -mmcu option is specified. The -Os option will tell the compiler to optimize the code for efficient space usage (at the possible expense of code execution speed). The -g is used to embed debug info. The debug info is useful for disassemblies and doesn't end up in the .hex files, so I usually specify it. Finally, the -c tells the compiler to compile and stop -- don't link. This demo is small enough that we could compile and link in one step. However, real-world projects will have several modules and will typically need to break up the building of the project into several compiles and one link.

$ avr-gcc -g -Os -mmcu=atmega8 -c demo.c

The compilation will create a demo.o file. Next we link it into a binary called demo.elf.

$ avr-gcc -g -mmcu=atmega8 -o demo.elf demo.o

It is important to specify the MCU type when linking. The compiler uses the -mmcu option to choose start-up files and run-time libraries that get linked together. If this option isn't specified, the compiler defaults to the 8515 processor environment, which is most certainly what you didn't want.