Switch to C89 comments in prep for making the whole codebase C89 compatible. #105

pull gmaxwell wants to merge 1 commits into bitcoin-core:master from gmaxwell:c89_comments changing 38 files +803 −717
  1. gmaxwell commented at 3:34 PM on November 15, 2014: contributor

    This should be whitespace/comment only changes and should produce the same object code.

  2. Switch to C89 comments in prep for making the whole codebase C89 compatible.
    This should be whitespace/comment only changes and should produce the same
    object code.
    71712b27e5
  3. gmaxwell commented at 3:34 PM on November 15, 2014: contributor

    Lets see if this gets me strangled. Its the most stupid of the required changes for C89.

  4. sipa commented at 8:16 PM on November 15, 2014: contributor

    utACK

  5. sipa commented at 8:20 PM on November 15, 2014: contributor

    Verified that the binary is identical.

  6. sipa commented at 9:38 PM on November 16, 2014: contributor

    No clue how I ended up finding that the resulting binary is identical; I can't reproduce that anymore...

  7. gmaxwell commented at 9:42 PM on November 16, 2014: contributor

    Cool. Yea, I didn't check that yet I assumed you would. Interesting, okay, I'll go find my mistake then. (or at least see if the binaries are identical for me)

  8. sipa commented at 9:49 PM on November 16, 2014: contributor

    I think I'm just not doing enough to make the result deterministic. I'm configuring/compiling with CFLAGS="-O0 -g0 -frandom-seed=bla".

  9. gmaxwell commented at 11:02 PM on November 16, 2014: contributor

    The non-determinism is, at a minimum, from macros using _ _ LINE _ _. (figured this out just by diffing the objdump -d output). I suggest you compare versions that have _ _ LINE _ _ replaced with 0? You shouldn't need -O0, and if you objdump you shouldn't need -g0.

    I confirm that it's determinstic and unchanged for me once the line number is removed from the macro. Perhaps we should have that controlled by a define for easier testing?

  10. sipa commented at 12:25 PM on November 17, 2014: contributor

    Tried rebasing this on top of #107, and compiling with -DDETERMINISTIC does result in identical binaries.

  11. sipa merged this on Nov 17, 2014
  12. sipa closed this on Nov 17, 2014

  13. sipa referenced this in commit c48cdc4cb3 on Nov 17, 2014

github-metadata-mirror

This is a metadata mirror of the GitHub repository bitcoin-core/secp256k1. This site is not affiliated with GitHub. Content is generated from a GitHub metadata backup.
generated: 2026-04-14 11:15 UTC

This site is hosted by @0xB10C
More mirrored repositories can be found on mirror.b10c.me