Entry #4: Internals of GCC

-
"Internals of GCC" is a 53 minutes long podcast produced in 2007 by Software Engineering Radio with Morgan Deters as guest. GCC is a set of compilers for various languages including ADA, C, C++, Fortran, Objective C and Java at one point. 

GCC provides the whole infrastructure to build software in those languages mentioned before, going from Source Code to Assembly. GCC works in three main sections, each one works in a different aspect of the compiler:
  • Front-End: In this section, the compiler acknowledges the language we are using to generate a tree structure that shows the program description and that will pass to the next section.
  • Middle-End: In this section, the optimization of the tree passed from the Front-End section occurs. This section makes the code more efficient and makes a transformation to a more generic code, as it takes this generic form and converts it into low-level structures called Register Transfer Language (RTL). During this process, some optimizations and workarounds are made, like Object-Oriented Languages must be converted into something that C can understand. Finally, the RTL representation is passed to the Back-End and it converts it into a target architecture.
  • Back-End: In this section, after the tree is optimized, it will be sent to the architecture it needs and it will become target code.
Morgan Deters mentions that GCC is a portable compiler, as it runs on most platforms available today, and it can produce outputs for many types of processors. GCC is not only a native compiler, but it can also cross-compile any program, producing executable files for a different system from the one used by GCC itself. This feature allows the software to be compiled for embedded systems which are not capable of running a compiler. 

At the end of the podcast, Morgan Deters was asked about his favorite language and he said that he doesn’t have one. He analyzes the problem and tries to look for the best approach as every language has advantages and disadvantages, but the compiler is the one that knows how to use most of the hardware.

Podcast: Software Engineering Radio. (2007). Episode 61: Internals of GCC. Recovered from http://www.se-radio.net/2007/07/episode-61-internals-of-gcc/

Comentarios

Publicar un comentario

Entradas populares de este blog

Entry #7: Building Server-Side Web Language Processors

-
“Building Server-Side Web Language Processors” is an article written by Ariel Ortiz, who is my professor during the Compilers Design course. The article talks about the advantages of building a language processor that actually runs on the web, instead of a processor that just runs on a command-line shell. Creating a web language processor can be a great challenge, as it requires to have knowledge in: Compilers design, Web server creation, Front-end design, and Security Management. The strategy used by the author consists of programming a web server from scratch and should be able to serve static resources and host a language processor that generates dynamic content based on the input and send the result to the client.  The implementation uses JavaServer Pages, which allows the processor to use custom tags, that are evaluated and produce an output. One of the advantages of this implementation is that the language processor has access to all the facilities provided by the hostin
Leer más

Entry #2: Making Compiler Design Relevant for Students

-
“Making Compiler Design Relevant for Students who will (Most Likely) Never Design a Compiler“ is an article written by Saumya Debray in which he invites the students to understand why the Compiler Design course is so important and explains that the professors should teach in a way that their students develop skills instead of the idea of just teaching knowledge. The article shows that a compiler can be seen as a translator or as a process that can transform data expressed in one language into the same data expressed in another language. This comparison illustrates the similarity of the behavior, but also considering all the technical translation problems that can be found in common components of a language, such as ambiguity, context, and metaphors. Based on this idea of a translator, we can better understand the phases that need to be executed to obtain a result these are: Lexical Analysis and Parsing, Semantic Analysis, Code Generation, and Code Parsing. The Lexical Anal
Leer más