void main(void)

Calls, Returns and In-Between.

February 2007

Alan Anderson continues with part two in his series on optimizing C code for DSP. This time, he analyzes some of the trade-offs involved in trying to get the compiler to convert C code into as few instructions as possible. There are ways (such as using intrinsics) to increase the efficiency of your C, but you lose some of the benefits of using C in the first place.

Posted by Kevin Carlson at 12:42 PM  Permalink |

Digital signal processing is one domain where optimization is king. That's why DSP code has been written in assembly for ages. But in the real world, hand-tuned assembly can be a liability if staff and time resources are limited. So there's a growing need to keep DSP code written in C. This allows you to have more time to experiment with algorithms. Alan Anderson begins a five-part series on DSP programming in C with "Programming and Optimizing C Code: Part 1."

Posted by Kevin Carlson at 12:57 PM  Permalink |

By now, most C/C++ programmers are familiar with the dangers of allowing untrusted user input into a format string. But just because languages like Ruby, Perl, Python and Java are less susceptible to buffer overflow doesn't mean that format strings in these languages are safe. You should know the risks in order to minimize the danger. Hal Burch and Robert Seacord lay it all out in "Programming Language Format String Vulnerabilities."

Posted by Kevin Carlson at 01:06 PM  Permalink |

Sometimes for good reasons, and sometimes for bad, exception handling isn't always present in code you need to make use of. If you're dealing with a C API, or code that runs in a managed environment, exceptions aren't necessarily an option. Yet you know exceptions are your error-handling mechanism of choice. What's the best way to approach this problem? As Gigi Sayfan explains in "Practical C++ Error Handling in Hybrid Environments," a big part of the solution is picking your battles intelligently.

Posted by Kevin Carlson at 01:08 PM  Permalink |