Cory Fields
Cory Fields
Post-merge utACK fa46cc22bc696e6845915ae91d6b68e36bf4c242. Nice cleanup :)
> Seems to suggest that it is a good thing for these symbols to be exported so the library and its callers use the same symbols, particularly type_info symbols We...
Adding one more bit to complicate and maybe explain why I described the boundary as a firewall... We build `libbitcoinconsensus.so` with guix using `-static-libstdc++` (side-note, we should perhaps move this...
Any reason not to do `Rotl` in sha3 and `ROTL` in siphash as well?
Concept ACK. I think another way of looking at this is that `script/bitcoinconsensus.cpp` is moving out into its own convenience lib, it just happens to be a single file. We...
@hebasto #24994 counted on linker side-effects to fix things IIRC. As do your other suggestions as far as I can see. If there's an issue with a compiler/stdlib symbol visibility,...
From what I can tell, we shouldn't need this after clang-14: https://github.com/bitcoin/bitcoin/blob/master/build-aux/m4/bitcoin_runtime_lib.m4 Is that the case? Noticed while reviewing https://github.com/hebasto/bitcoin/pull/43/commits/910fd7a64802380fe912446cd3c938d56e470f0b#diff-b937c829d4a0cf9a653f5361656ea8196b6ecef40fd033aa41a6da945e987bcdR146 @maflcko Can we just skip porting that to CMake, or...
Does ` __attribute__ ((visibility ("default")))` override this? Otherwise I assume we'd need to make sure this flag doesn't make it to shared libs.
> Does ` __attribute__ ((visibility ("default")))` override this? Otherwise I assume we'd need to make sure this flag doesn't make it to shared libs. Oh whoops, already asked and answered....
Ping @TheCharlatan @hebasto I went around in circles several times on this. I think it accomplishes what we need, but please sanity check me :)