core: transition to c++17, deboostification

[kwvg]

No description provided.

[github-actions]

This pull request has conflicts, please rebase.

[PastaPastaPasta]

utACK for merging via merge commit

For 21064, can we do https://github.com/bitcoin/bitcoin/pull/21064/files#diff-dac26d1a1b4a405373f40090c92278a64cbda24c875d3394d5e7f81747367702L70-L71? git grep doesn't yield any results

Also, does it make sense to remove the cxx17 ci build now?

[UdjinM6]

1. agree with @PastaPastaPasta - we should drop linux64_cxx17 build (should we add --with-sanitizers=undefined to linux64 build?)
2. should apply 20471 to depends/packages/protobuf.mk

[kwvg]

@PastaPastaPasta, we could remove those two lines from the inter whitelist but they're not there to begin with so there's nothing to remove (which is kinda surprising but not too problematic)

@UdjinM6, bitcoin#20471 has been updated and linux64_cxx17 has been dropped

[UdjinM6]

utACK

[PastaPastaPasta]

Hmmmm, we shouldn't just do this b/c of --enable-werror --with-sanitizers=undefined which are not on linux64/32

Also the fact that this one doesn't do functional tests. We should probably keep this, but just rename it to be like linux64_nofun :D

[PastaPastaPasta]

(Otherwise everything does good)

[UdjinM6]

CI is broken:

• linux32: looks like smth that could be fixed by Fix Dash build for Arch Linux with GCC 11.1 #4398
• linux64: we probably hit some gitlab limit there ERROR: Uploading artifacts as "archive" to coordinator... too large archive

Let's maybe drop 63f0fc4 for now and implement it in a separate PR?

[PastaPastaPasta]

Let's maybe drop 63f0fc4 for now and implement it in a separate PR?

If we drop 63f0fc4, we should drop kwvg@73de8a6 as well

[kwvg]

Updated with rollbacks

[PastaPastaPasta]

utACK for merging via merge commit

[UdjinM6]

Not a huge fan of merging PRs with failed CI builds but these failures are unrelated, so

utACK