rust: support build under rosetta2

[lukaso]

Description

Support building rust under rosetta2 (for x86_64). Otherwise it fails.

See bug report: https://trac.macports.org/ticket/68185

Type(s)

• bugfix
• enhancement
• security fix

Tested on

macOS 13.6 22G120 arm64
Xcode 15.0.1 15A507

Verification

Have you

• followed our Commit Message Guidelines?
• squashed and minimized your commits?
• checked that there aren't other open pull requests for the same change?
• referenced existing tickets on Trac with full URL?
• checked your Portfile with port lint --nitpick?
• tried existing tests with sudo port test?
• tried a full install with sudo port -vst install?
• tested basic functionality of all binary files?
• checked that the Portfile's most important variants haven't been broken?

[macportsbot]

Notifying maintainers:
@MarcusCalhoun-Lopez for port rust.

[kencu]

I am suspicious this error is probably coming from our old friend, the misconfiguration of jemalloc running under rosetta due to LG_VADDR not being detected correctly when running under the rosetta emulator.

The block that fixes this in the rust portfile:

macports-ports/lang/rust/Portfile

Line 156 in 746b2f5

if { ${os.arch} eq "arm" } {

will not be activated when you build it with the terminal set to an x86_64 arch, as IIRC, {${os.arch} eq "arm" } does not return true when you run macports in that fashion on an M1.

So another way around this might be to force those env vars that fix this to activate...

build.env-append JEMALLOC_SYS_WITH_LG_VADDR=48
destroot.env-append JEMALLOC_SYS_WITH_LG_VADDR=48

[lukaso]

I'll try this.

[lukaso]

I've tried this and it's not worked.

[kencu]

I've tried this and it's not worked.

That's odd -- because we 100% know that fixes the build of rust when running under rosetta2 and building muniversal with jemalloc enabled...I wonder if the env vars applied...

When @MarcusCalhoun-Lopez does come around again, I can't help but feel fairly sure this is going to be reverted, if it is committed.

But he's not here, and I'm not going to set up another installation of macports running under an x86_64 terminal just to work on this...

Are we sure this PR doesn't break the universal build of rust we have now? We don't want that build to change, for example by building rust two different ways (jemalloc on and jemalloc off).

[lukaso]

I'll be honest, once I changed the Portfile, it stopped building under any circumstances, so I'm not confident anymore. Now I'm trying to figure out how to get any debugging info out of Portfiles. using ui_debug doesn't work. What is the best way to see what is happening when a Portfile is executed? Is there a way to print debugging info to a screen or log?

[lukaso]

I'll be honest, once I changed the Portfile, it stopped building under any circumstances, so I'm not confident anymore. Now I'm trying to figure out how to get any debugging info out of Portfiles. using ui_debug doesn't work. What is the best way to see what is happening when a Portfile is executed? Is there a way to print debugging info to a screen or log?

To answer my own question: use ui_error and the message you want to print. This seems to work in getting a printout without a super excessive debug log.

[kencu]

ui_debug output is visible when you build with the “-d” flag added (debug)

[pmetzger]

How do we either finish this or close it?

[kencu]

This needs to be re-examined from a fresh start.

[pmetzger]

@kencu So does that mean we close it and someone starts over? (What is the next concrete action that should be taken?)

[kencu]

sure, I guess.