I don't have access to this hardware. I can't test the changes here.

This was superceded by @kib in r343617. The purpose of this revision was to provide a simple way to have module support for the various adapters supported by iflib and to cull iflib out when not needed.

I've been running this with no defects for quite some time.

Is the raw patch updated so I can test it?

Huh ... Settings->Keyboard->Typing Settings->Repeat Speed was reset to "1". I've bumped it locally back to 50 so I don't throw my keyboard across the room. :-)

Ugh, no. It does repeat but its on a 3 second or so output delay.

Does anyone else notice that holding a key in xfce-terminal and waiting for a repeat of characters doesn't seem to happen with this update?

This is what I have installed currently for the record.

sbruno@alice:~ % pkg info |grep xfce
libxfce4menu-4.14.1            Widgets library for the Xfce desktop environment
libxfce4util-4.14.0            Extension library for the Xfce desktop environment
xfce-4.14                      meta-port for the Xfce Desktop Environment
xfce-icons-elementary-0.13.1   Elementary icons, extended and maintained for Xfce
xfce4-appfinder-4.14.0         Application launcher and finder
xfce4-conf-4.14.1              D-Bus-based configuration storage system
xfce4-desktop-4.14.1           Xfce's desktop manager
xfce4-mixer-4.11.0_5           Volume control for the Xfce desktop
xfce4-notifyd-0.4.4_1          Visually-appealing notification daemon for Xfce
xfce4-panel-4.14.0             Xfce's panel
xfce4-screenshooter-plugin-1.9.5_1 Application and panel plugin to take screenshots
xfce4-session-4.14.0           Xfce's session manager
xfce4-settings-4.14.0          Xfce 4 settings application
xfce4-terminal-0.8.8_1         Terminal emulator for the X windowing system
xfce4-tumbler-0.2.7            Thumbnail service for Xfce desktop
xfce4-wm-4.14.0                Xfce's window manager

I've applied and compiled all this and it seems to do the right thing. A few icons on my tool bars lost their icons, but that was easily fixed.

Allrighty, thank you. This applied and I'm rebuilding now. I'll report back on behavior.

In D21278#462738, @madpilot wrote:

In D21278#462737, @sbruno wrote:

Hrm ... phabricator and patch really don't like to apply this to my local tree. It chokes on adding new stuff and I can't seem to work around it. Any suggestions?

Strange, I generated it using arc from command line.

You can grab a ports overlay here:

https://github.com/madpilot78/FreeBSD-xfce4.13

it's the same as the patch.

I also uploaded the result of svn diff here:

https://people.freebsd.org/~madpilot/xfce414.diff

In D21278#462738, @madpilot wrote:

In D21278#462737, @sbruno wrote:

Hrm ... phabricator and patch really don't like to apply this to my local tree. It chokes on adding new stuff and I can't seem to work around it. Any suggestions?

Strange, I generated it using arc from command line.

You can grab a ports overlay here:

https://github.com/madpilot78/FreeBSD-xfce4.13

it's the same as the patch.

I also uploaded the result of svn diff here:

https://people.freebsd.org/~madpilot/xfce414.diff

Hrm ... phabricator and patch really don't like to apply this to my local tree. It chokes on adding new stuff and I can't seem to work around it. Any suggestions?

I will add the following references to the commits:

While this review was in progress, 20190511 was released. Reset to that version.

Use GH_TAGNAME a bit more thoughout the build as suggested by swills.

Apparently arc doesn't use relative path by default. Delete unrelated tree changes.

Implement @mat suggestions for USE_GH

Yeah, just changing VREF_ENABLE_MASK was insufficient. I think I'll just commit this so people can use it for their newer desktops/laptops.

This seems like a good accounting change. Thanks.

Thanks for digging into this.

This looks good from the user side as the machine is humming along quite nicely. Thank you for digging into this!

Patched the build tree in the freebsd.org cluster and installed -current + this review on the pkg build server. You can track builds here when it starts spinning up in about an hour.

bump to see if phabricator shovels some emails.

trivial bump to generate email.

@shurd any objections from the peanut gallery?

Ports revision r489228 based on different patches has been committed.

r481915 | sbruno | 2018-10-12 11:51:37 -0600 (Fri, 12 Oct 2018) | 11 lines

In D18470#396630, @jeffrey.e.pieper_intel.com wrote:

In D18470#396629, @sbruno wrote:

In D18470#396626, @jeffrey.e.pieper_intel.com wrote:

We are seeing tx hangs with UDP using multiple clients when offloads are disabled. Cannot reproduce on 12.0-RELEASE. Reproducible on X520 and X550.

Huh? Does this review break things for you? Or are you bringing up a new problem?

Just reporting issues/regressions we found while testing this :)

In D18470#396626, @jeffrey.e.pieper_intel.com wrote:

We are seeing tx hangs with UDP using multiple clients when offloads are disabled. Cannot reproduce on 12.0-RELEASE. Reproducible on X520 and X550.

Doesn't this patch need to be conditional somehow (FreeBSD versions < 12 don't need this).

I guess this will wait for @jeffrey.e.pieper_intel.com to test or approve? This looks correct to my eye.

Agreed. Feel free to commit this at your leisure.

This panic is fairly reliable on our Power8 box. Is it related to this change or is it unrelated? I'm not convinced this panic should keep this review blocked as it resolves other things.

panic: Memory modified after free 0xc0000000850584a0(32) val=0 @ 0xc0000000850584a0

In D17806#393519, @leonardo.bianconi_eldorado.org.br wrote:

The reason of segmentation fault when treating the 0x1500 exception
was because it was being treated as a non-hypervisor trap, which
consider the registers srr0/srr1. As this trap is an hypervisor one,
the values were being placed in hsrr0/hsrr1.
Just changed the enter point for hypertrap, which saves the value of
hsrr0 and hsrr1 in srr0 and srr1.

In D10776#390549, @emaste wrote:

Is it acceptable if some binaries fail?

This has become a real issue - @sbruno found wc did not work on ERL, fixed in rS341140. This is somewhat independent of whether we have the #ifdefs or not though.

Oof. I got a different panic, that may or may not be useful.

Meaningless update to test that phab is still working

Running this on the Tyan Power8 hosts now. I'll report back in a day or two after we get some poudriere runs under our belt.

Is there any intention to rebase this to -current?

Nuke extra pkgname suffix