Embedded in Academia

Category: Computer Science

  • Conference Hijacking

    The ACM Workshop on Programming Languages and Operating Systems (PLOS) is a small event where researchers interested in the intersection of these areas can share ideas and results. There have been five of these workshops and I’ve attended a few of them and presented a paper at one. It’s a fun event and a perfectly…

  • Why Not Mix Signed and Unsigned Values in C/C++?

    Most C/C++ programmers have been told to avoid mixing signed and unsigned values in expressions. However — at least in part because we usually follow this advice — many of us are not totally on top of the underlying issues. This program illustrates what can go wrong: #include <stdio.h> int main (void) {   long…

  • Giving a Talk at the GCC Summit

    Last month I proposed giving a paper presentation at the next GCC Developers’ Summit about our compiler bug finding work. Happily, it was accepted and in late October I’ll head up to Ottawa to tell them what we’ve been doing and what we plan to do.  Here’s the abstract: About 60 wrong-code and crash bugs…

  • Fun With Shift Optimizations

    It’s fun to see what a modern compiler can optimize and what it cannot. The other day, while working on a new piece about undefined behavior, I noticed some C compilers failing to optimize simple code based on shifts. Here are the functions: int shift1 (int x, int y) { if (x>=0) { return (x>>y)…

  • Static Analysis Fatigue

    My student Peng and I have been submitting lots of bug reports to maintainers of open source software packages. These bugs were found using Peng’s integer undefined behavior detector. We’ve found problems in OpenSSL, BIND, Perl, Python, PHP, GMP, GCC, and many others. As we reported these bugs, I noticed developers doing something funny: in…

  • Sensor Network Technology in Vinge’s A Deepness in the Sky

    An important function of science fiction is to help us understand sociological, technological, and other aspects of our future. A really good SF novel — like some of those produced by Asimov, Clarke, Heinlein, Le Guin, Niven, and Vinge — is so full of ideas and possibilities that the reader’s mind is expanded a little.…

  • The Future of Compiler Correctness

    Notes: This piece is mainly about compilers used for safety-critical or otherwise-critical systems. The intersection of formal methods and compilers for scripting languages is the empty set. Readers may be interested in a companion piece The Future of Compiler Optimization. A half-century of experience in developing compilers has not resulted in a body of widely-used…

  • The Future of Compiler Optimization

    Also see The Future of Compiler Correctness. Compiler optimizations are great: developers can write intuitive code in high-level languages, and still have them execute in a reasonably fast way. On the other hand, progress in optimization research is excruciatingly slow despite hundreds of papers being published on the topic every year. Proebsting’s Law speculates that…

  • Poll: Do You Want Compiler Bugs to Be Quiet or Loud?

    I’m preparing a longer piece on compiler correctness but thought this idea was worth putting into its own short post. The question is: Which would you choose: A compiler that crashes more often, or one that silently generates incorrect code more often? Of course, all real compilers have bugs, and all real compilers display both…

  • Knowing When to Quit

    In industry it’s often pretty easy to know when to stop working on a project: you might get moved off the project, it might get canceled, etc. In academia, it’s less clear: I can stop working on something after half an hour, or else I can work on basically the same idea until the end…