Does bloaty have any support for reporting waste via duplication?
I'm thinking of cases when bloaty takes in something like an .a or a list of .o, and then reports symbols (for example) which have the most waste, defined as (N - 1) * size for a symbol that occurs N times. I.e., you only need one copy of a given symbol, the rest are waste (this usually happens for things defined in header and instantiated in many TUs).
Open to ideas on post-processing bloaty's output to get this info too.
Does bloaty have any support for reporting waste via duplication?
I'm thinking of cases when bloaty takes in something like an .a or a list of .o, and then reports symbols (for example) which have the most waste, defined as
(N - 1) * sizefor a symbol that occurs N times. I.e., you only need one copy of a given symbol, the rest are waste (this usually happens for things defined in header and instantiated in many TUs).Open to ideas on post-processing bloaty's output to get this info too.