• iflib: Free remaining interrupt resources on failure in iflib_device_register()
• iflib: Free RX queue TCP LRO memory on unload

• ixl/ixlv(4): Fix panic when receiving jumbo frame (>4k mtu) traffic

Several outstanding issues:

• ixl(4): Fix compile error when PCI_IOV isn't defined

Changes from D16428:

• iflib: Prevent kernel panic on detach in _task_fn_admin by checking if in detach
• iflib: Add STATE_LOCK() around flag setting in iflib_request_reset()

New changes for D16429:

• ixl(4): Add tunable to enable VF loopback; enable by default
• ixl/ixlv(4): Fix dev.ixl/v.#.vsi.rxq#.bytes sysctl not incrementing

• iflib: Prevent kernel panic on detach in _task_fn_admin by checking if in detach
• iflib: Add STATE_LOCK() around flag setting in iflib_request_reset()

• ixl(4): Lie to the VF about configuring promiscuous mode if VF isn't allowed to configure it

@gallatin, I address your specific comments, but I would need to look through ixlvc.c some more to find more code that can be shared. There is almost certainly some in there that can be shared between the PF/VF drivers.

• ixl(4): Combine PF-only/VF-only ixl_disable_rings() in PF driver into one function.
• Remove a space

I need to add the state lock to iflib_request_reset

In general, I looked around at ixlv.c, and it looks like there are many functions for the main pf which are identical, or nearly identical for the VF. Can we please work on collapsing them down to common code?

• ixl(4): Fix module Makefile when PCI_IOV is enabled

• ixlv(4): Use same TX DMA seg sizes as the ones in ixl(4)
• ixl/ixlv(4): Preserve code removal done in r336313

• ixlv(4): Remove old/dead code

• ixlv(4): Re-add driver to GENERIC and build files
• ixl/ixlv(4): Fix module Makefiles

Sure.. Those checks for TCP in the TSO case should probably turn into asserts eventually..

@gallatin I'm assuming you no longer have any objections to this patch?

• iflib: Move stats increment in IPv6 case to match IPv4 case

• iflib: Remove comment and restore/fix setting CSUM_IP6_TCP flag during TSO6

In D15558#328462, @gallatin wrote:

In D15558#328453, @sbruno wrote:

Was this the patch that we discussed on our call? Does this fix your TX checksum problems on IXL?

Yes and Yes.

Unfortunately, I've now proceeded to the "entire box hangs" portion of the game. Even serial console break-to-debugger is ignored. I'm going to try NMIs next..

Looks good, but not forcing CSUM_TCP_IPV6 for TSO6 may need a closer look.

• iflib: Fixup changes made to record TCP checksum info in iflib for ixl

In D15558#328462, @gallatin wrote:

In D15558#328453, @sbruno wrote:

Was this the patch that we discussed on our call? Does this fix your TX checksum problems on IXL?

Yes and Yes.

Unfortunately, I've now proceeded to the "entire box hangs" portion of the game. Even serial console break-to-debugger is ignored. I'm going to try NMIs next..

In D15558#328453, @sbruno wrote:

Was this the patch that we discussed on our call? Does this fix your TX checksum problems on IXL?

Was this the patch that we discussed on our call? Does this fix your TX checksum problems on IXL?

In D7802#284240, @sbruno wrote:

@erj Do you want to abandon this review in light of your work on newer releases?

@erj Do you want to abandon this review in light of your work on newer releases?

Yep, looks good. Thanks!

LGTM

Looks fine to me.

This seems ugly to me, but that's the only way it works. ${CHMOD} -R doesn't do the right thing on 10.

I was able to reproduce the build issue as well using poudriere 3.2.
The inconsistency in behavior is caused by this commit, which is in 11, but not in 10: https://svnweb.freebsd.org/base?view=revision&revision=283997
After the distfile is unpacked - extract-fixup-modes target from bsd.port.mk tries to fix the permissions on the directories, but on 10 it does that only for the top level directory, so it can't look into the inner one.

looking at the packages from Intel Download Center, I think this is a tool build issue. It looks like file permissions changed between 5.05 and 6.01. In 5.05 the directory structure for nvmupdate is 755, but in 6.01 it is 644, which for some reason 10 doesn't like.

Builds. Untested at this time.

Odd, I get a wierd chmod error on FreeBSD 10 when building this port. On 11/12 it builds fine:

Updated to a newer, silently updated version. Ready for review.

Updated to 1.8.1, unbroken. Ready for review.

bump. What's the plan here?

bump. What's up here?

Needs additional testing.

Needs additional testing.

To document my understanding.. this is safe because interrupts are disabled while handling rx, including passing work to the 'que_task' if processing limits are present.

Better to MFC to 1.7.12 now...