[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Emacs-diffs] master d826037 3/3: Remove the need for temacs.in

From: Daniel Colascione
Subject: Re: [Emacs-diffs] master d826037 3/3: Remove the need for temacs.in
Date: Thu, 11 Apr 2019 15:15:18 -0700
User-agent: SquirrelMail/1.4.23 [SVN]

>> Please don't do this. Computing a fingerprint over temacs.in factors
>> link
>> layout information into the fingerprint hash. Your approach doesn't.
>> It's
>> possible to link Emacs in different ways from the same object files and
>> produce different binaries. I don't think a little build speedup is
>> worth
>> the safety loss.
> BTW, regarding this fingerprint: following this discussion, I see that
> we actually don't know for sure what is preserved between two different
> runs of `ld` and more to the point, there is no guarantee.
> So, how'bout we don't compute a fingerprint at all.
> Instead we generate a UUID: this way we can have our cake (no need for
> a separate temacs.in) and eat it too (even very subtle changes `ld` does
> between two different runs will lead to different "fingerprints").

I thought about just making it random --- but I like reproducible and
deterministic builds. We could use the build GUID that some linkers can
compute from the build inputs (a bit like how we compute the fingerprint),
but there's no portable way to get it at runtime, and I really want to get
out of the ELF- and PE-parsing business.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]