1142352 - configure should test for the GCC version of the host compiler when cross compiling

Status: RESOLVED FIXED

(Firefox Build System :: General, defect)

Description

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

As discussed in https://groups.google.com/d/msg/mozilla.dev.platform/ZeUgSSGHBPk/53tLVVjE90UJ we don't appear to have any tests for the compiler version of the host compiler.

Since I didn't see any tests for clang version in the target-compiler section, I didn't add any tests for clang version in the host-compiler section either.  However, I did add (with a little bit of copying):
 * the test that the host C and C++ compilers are the same
 * the test that the host compiler meets the minimum gcc version (although the test is now in two places)

Try run at:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=f4e1c85f1419

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Attachment: Add a configure test for the gcc version of the host compiler when cross compiling

The duplication of the code higher up is a little bit annoying, but I
don't see an easy way to avoid that.  It's also still quite far from
duplicating everything.

I tested locally with a Fennec build that if I bump the requirement from
4.6 to 4.9, I get the expected build error.

Comment 2

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

I suspect we should also check that the C and C++ compiler versions are the same (for both host and target).

Comment 3

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

(In reply to David Baron [:dbaron] (UTC-8) from comment #2)
> I suspect we should also check that the C and C++ compiler versions are the
> same (for both host and target).

I made that bug 1142420.

Comment 4

[Mike Hommey [:glandium]]

Comment on attachment 8576486 [details] [diff] [review]
Add a configure test for the gcc version of the host compiler when cross compiling

Review of attachment 8576486 [details] [diff] [review]:
-----------------------------------------------------------------

It's possible to factor to avoid duplication, but I won't demand you learn m4 just for that.

Comment 5

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/e3a4467dc9df

Comment 6

[Carsten Book [:Tomcat]]

sorry had to back this out in https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&revision=e76c33c4d452 because this cset caused https://treeherder.mozilla.org/logviewer.html#?job_id=7575053&repo=mozilla-inbound

Comment 7

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

I'm not sure how you could tell from the log that it was gcc 4.4 -- but perhaps it's worth landing now with gcc 4.4 as the minimum, and bumping it when we can?

Comment 8

[Mike Hommey [:glandium]]

It's 4.4 because that's what cc/c++ are in our build environment, and that's what was being used. Hopefully bug 1142908 can just be fixed quickly.

Comment 9

[Bob Owen (:bobowen)]

I've just landed the patches to change our minimum supported target compiler to gcc-4.7.
This is really just recognising the reality that Gecko 38+ doesn't compile with 4.6 anyway.

Comment 10

[Mike Hommey [:glandium]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/a0f267629f11

Comment 11

[Mike Hommey [:glandium]]

(In reply to Mike Hommey [:glandium] from comment #10)
> https://hg.mozilla.org/integration/mozilla-inbound/rev/a0f267629f11

I made it check for 4.7 instead of 4.6 as the previous landing.

Comment 12

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/a0f267629f11