In D35107#853078, @rouslan_rshell.net wrote:

In D35107#798673, @jrm wrote:

In D35107#798672, @rouslan_rshell.net wrote:

Yes, I have capability to test it on my GCP account. Send me the binary, I’m running FreeBSD 13.0 from GCP marketplace.

Thanks. Here is a binary. https://ftfl.ca/misc/google_cloud_ops_agent_engine

Hi,
Sorry for loooooong delay. I've installed 13.1-RELEASE in GCP, please provide your ssh public key and I'll grant access to you.
Best regards,
Rouslan

In D35107#798672, @rouslan_rshell.net wrote:

In D35107#798671, @jrm wrote:

I contacted Google to see about getting credits for testing, but they said that they will require a credit card. Rouslan, would it be easier for you to test if I send you a binary? If so, what version/arch of FreeBSD are you running?

Yes, I have capability to test it on my GCP account. Send me the binary, I’m running FreeBSD 13.0 from GCP

Are you saying to remove the quotes and add a comma between en and es. I'm checking because there is an example just after Example 9 (which for some reason isn't labelled like the other examples) in the same chapter that says make DOC_LANG="en fr" pdf-books. Also, make DOC_LANG="en es" HUGO_ARGS="--verbose --debug" works.

Don't touch _index.po. It's update is scripted.

CC @me_igalic.co who is doing some cloud-init work. Maybe there is some overlap or potential for collaboration.

Remove directory from URL

I'll commit just after @salvadore commits D37266. I'll also bump PORTREVSION.

Ensure GCC version 11+ is used when building with NATIVECOMP

Remove whitespace change

The change here doesn't solve the problem with Emacs not finding the jit headers. After dc67e7d3009de2b78d96136c8ad09cf186e553df include/libgccjit.h and include/libgccjit++.h) are no longer installed.

To be clear, I'm just referring to the style changes mostly to conform to Chapter 15. Order of Variables in Port Makefiles and a few other style issues. It's kind of like style(9) for ports. No problem if you prefer to leave things as is, but could you clarify?

Also, if you are good with these changes, I can do the same for devel/valgrind.

Show the revision relative to the update discussed in D37197.

Please let me know if you would like the update to devel/valdrind-devel committed.

Both the patched devel/valgrind-devel and HEAD (aed1e501) of upstream build successfully in a CURRENT jail built from 0f97a0d425cf (2022-09-15). I also successfully built the patched devel/valgrind-devel in 12.3 and 13.1 i386/amd64 jails.

Sure, whatever you think makes most sense.

Full build tests and run-time tests look good here. Thanks. I'll commit.

In D36645#832058, @jfree wrote:

During my testing of this port, I found that poudriere-devel has a testport -b [branch] flag that
prefetches most dependencies as packages through the pkg repos.

This was especially useful for this port because Rust takes ~1-2 hours to build on my
machine. Perhaps you already know about this flag, but I thought I would bring it up
because it made my life a lot easier.

Looks great other than these little nit-picks. An entry in audio/Makefile is also required.

The problems with japanese/lookup on CURRENT are not related to this update and are probably not with the port itself, but with Mk/Scripts/qa.sh.

With the default flavor everything built successfully except japanese/lookup. There seems to be a core dump during the QA stage. I'll take a closer look, but it might take me a day.

Testing on FreeBSD 14.0-CURRENT #0 main-n258024-0f97a0d425cf: http://pkg.ftfl.ca/build.html?mastername=14amd64-default&build=2022-09-19_13h53m54s

The only other minor changes I plan to make before committing:

Jake, are you satisfied this is ready to be committed? If so, I'll give it another look and commit if all looks good on my end.

In D36213#823515, @jfree wrote:

In D36213#823470, @jrm wrote:

Consider running portfmt -D Makefile.

Joe,

Did portfmt generate these suggestions or are they from you?

USES=gnome
USE_GNOME= cairo glib20

USES=xorg
USE_XORG=pixman

Jake, I'm moving our email discussion here so others can weigh in if they have anything helpful to add. It looks like the build issues on 12.3 and 13.0 are related to changes in 160b4b922b6021848b6b48afc894d16b879b7af2 (and subsequent commits?) not available in those releases. If you are going to wait until September 1 to have this committed you don't have to worry about 13.0, however, 12.3 is still supported until three months after 12.4 is released. You may be able to patch your code by looking at /usr/include/sched.h. If it's not straightforward to work around the issue, you can add BROKEN_FreeBSD_12=sched.h definitions missing in 12.3. You might be able to come up with a better message.

Consider running portfmt -D Makefile.

In D36213#823036, @jfree wrote:

• Added patch to POSIX-ify igt bash scripts