Get the latest tech news
Fast Shadow Stacks for Go
Learn how shadow stacks could deliver up to 8x faster stack trace capturing than frame pointer unwinding.
However, for legacy reasons going back to the days when we had only 32bit CPUs with limited registers, compilers other than Go generally omit frame pointers for optimized builds (-O1 or greater). Since DWARF unwind tables are expressed in terms of opcodes for a Turing-complete VM (I wish I was making this up), it's pretty common to preprocess this data for better lookup performance. This is probably still many years away, but once this future arrives, we should be able to get rid of DWARF unwinding and perhaps also frame pointers and enjoy blazing fast stack trace capturing across the board.
Or read this on Hacker News