This work is a little tricky -- the code I have to change was written to make a simple non-recursive DNS cache -- and progressing slowly. However, in the process of writing this code, I figured out a way to make the Deadwood binary noticeably smaller, by making a frequently called inline sequence its own function.
This code size optimization is useful for embedded applications of Deadwood; one of my goals is to keep the Deadwood Windows binary under 65,536 bytes in size. This optimization reduces the size of the Deadwood binary from 65,536 bytes to 61,952 bytes in size, and allows the tiny version of the binary to have filter_rfc1918 functionality again.
With this change, there is now 3,584 bytes to play with, which should be enough space to add all of the code I will add to handle CNAME chains better and for some other speed optimizations.
To compile a smaller binary, be sure TINY_BINARY is defined; this removes some of the parsing rules used for minimizing the amount of support emails the list gets.
I have, in addition to the source code, the tiny Windows binary. It can be downloaded here:djbdns users who still believe the myth that djbdns is magically completely secure and never needs patching.
To post a comment about an entry, send me an email and I may or may not post your comment (with or without editing)