It builds fine at my side with poudriere: build of sysutils/uefi-edk2-qemu | uefi-edk2-qemu-x86_64-g20190307_1 ended at Wed Sep 11 09:26:10 CST 2019

In D21580#470256, @bcran wrote:

In D21580#470252, @araujo wrote:

You have my approve to commit it! Thank you!

Committed. But I just realized, should I also bump PORTREVISION?

In D21580#470251, @bcran wrote:

In D21580#470248, @araujo wrote:

USE_GCC will set build-time and run-time dependency, if we need only build-time dependency the patch looks good to me.

Thanks! Since I'm not a ports committer, could you commit the change please?

USE_GCC will set build-time and run-time dependency, if we need only build-time dependency the patch looks good to me.

In D21053#459032, @markj wrote:

In D21053#458959, @jhb wrote:

I worry this might be a bit premature. New features for bhyve are often developed without capsicum because capsicum is a pain to work with. The debug server didn't use it at first, save/restore currently disables it as does the 9p patch. Removing WITHOUT_CAPSICUM makes coordinating that existing work harder. It also adds a pretty good barrier to entry when trying to add new things that add new ioctls, etc.

It's not much more work to add a new right in a driver that already works under capsicum than it is to add -DWITHOUT_CAPSICUM to the build flags.

In D19400#458961, @rob.fx907_gmail.com wrote:

@araujo, sounds like you are taking over this review?

In D19400#458949, @jhb wrote:

I wonder if it makes more sense to do this in vmrun.sh? OTOH, it's probably fine to do this. We should also probably do it in bhyverun.c since bhyveload isn't used in the UEFI case. Another option would be to do this in the vm_create() wrapper in libvmmapi in which case that would fix bhyveload, grub2-bhyve, and bhyve all in one place.

Committed already at r349335

I'm commandeer this revision because this patch were committed already by @scottl: https://svnweb.freebsd.org/base?view=revision&sortby=date&revision=349335

@jhb Perhaps do the same with bhyverun.c?

Ok, I'm gonna do it tomorrow!!!

@jhb Any objection if I proceed to commit this patch? I just hit the same situation "forgot to load vmm" and this patch does fix it.

ping!!!

In D21053#456794, @oshogbo wrote:

The problem here is that you will not be able to build this code on other platforms like Solaris.
I'm not sure if this is applicaple for bhyve, if not then the change looks good to me.

Remove CASPER from Makefile as we don't use any service.

Lgtm!

Lgtm

LGTM! Thank you!

double accept this revision!

Any news about this patch?

I'm fine with that too! We can have it as it is and later evolve it.

• Also remove SHEBANGFIX not needed.

• Update ONLY_FOR_ARCHS that was replaced wrong.

Address @bcran suggestions!

LGTM!

I did only smoke tests!! Do you mind do additional tests with different guests?

In D19869#426696, @bcran wrote:

• Rename the port and files to include a -x86_64 suffix

It looks good! Thanks again @chuck

In D19042#422183, @jhixson wrote:

Sooooooooo. Is anybody going to help me fix this or approve this? ;-) I am still waiting.

LGTM!

LGTM!

If anybody else is interested on this patch, ping me and I can share my latest changes. However I'm not interested on it anymore.

I have added @mav as a reviewer, I know he has worked with the hda drivers before.

I still feel locked out of usr.sbin/bhyve, mostly of the commits made there in the past year were made by myself or reviewed by myself. Would be fair have my name explicit in that area, I don't want bypass reviews, it is more about the effort I have put in that area and appreciation of it, otherwise doesn't make sense for me to keep contributing with bhyve.

Please, reopen the review, when I'm back from holidays I will help you.

It is basically a cosmetic change, so I have committed it already: r343634.