[Bug 246630] stable/11 regression: base.txz reproducibility depends on number of cpu cores

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=246630

--- Comment #20 from Dimitry Andric <dim at FreeBSD.org> ---
(In reply to Fabian Keil from comment #19)
> Which stable/11 revision are you currently testing on?

On head, base r361410 (as of 2020-05-23) in fact.

> I tried the work around with a system based on r360986 and still ended up with unreproducible rescue binaries like before.

Yes, unfortunately upstream replaced the CLANG_SPAWN_CC1 environment variable
right after llvm 10 was branched, in
https://github.com/llvm/llvm-project/commit/c4a134a5107c1926262ee887057036ca53aa7265:

Author: Nico Weber <thakis at chromium.org>
Date:   Wed Jan 15 10:45:02 2020 -0500

    Replace CLANG_SPAWN_CC1 env var with a driver mode flag

    Flags are clang's default UI is flags.

    We can have an env var in addition to that, but in D69825 nobody has yet
    mentioned why this needs an env var, so omit it for now.  If someone
    needs to set the flag via env var, the existing CCC_OVERRIDE_OPTIONS
    mechanism works for it (set CCC_OVERRIDE_OPTIONS=+-fno-integrated-cc1
    for example).

    Also mention the cc1-in-process change in the release notes.

    Also spruce up the test a bit so it actually tests something :)

    Differential Revision: https://reviews.llvm.org/D72769

    (cherry picked from commit 8e5018e990b701391e6c33ba85b012343df67272)

It's a bit awkward to add -fno-integrated-cc1 to each CC invocation, but it be
done either with the mechanism shown in the commit mesage above, or by setting
CFLAGS="-O2 -fno-integrated-cc1" in make.conf.

Alternatively, I could flip the default in
lib/clang/include/clang/Config/config.h.

In any case, I haven't yet gotten to the bottom of why this causes changed
behavior.

-- 
You are receiving this mail because:
You are the assignee for the bug.