cxgbe(4): Integrate with PNP PCIID-based autoload

[cem]

Summary

PR: 236886

[imp]

jhb specifically asked me not to do this driver. Please be sure to get sign off from him.

[ae]

Note, that automatic loading for cxgbe can do unexpected firmware update when user does first boot.

[jhb]

It's really @np who should sign off on this. I can blacklist the driver on my dev boxes since since there's now an rc.conf variable to exempt drivers from devmatch which would be sufficient for my needs.

[cem]

In D19754#424087, @jhb wrote:

It's really @np who should sign off on this. I can blacklist the driver on my dev boxes since since there's now an rc.conf variable to exempt drivers from devmatch which would be sufficient for my needs.

Yep, I agree, and that's what I've been holding off for.

If people perceive the auto-upgrading of firmware as deleterious, maybe that's an indication of a feature that should be disabled? It could always be made a tunable to give users the option to opt in? It's not my driver -- I think np@ gets to make that call.

[ae]

We have single machine, that after automatic firmware upgrade fails to attache the driver:

t5nex0: <Chelsio T580-LP-SO-CR> mem 0xfb300000-0xfb37ffff,0xfa000000-0xfaffffff,0xfb984000-0xfb985fff irq 64 at device 0.4 numa-domain 1 on pci9
t5nex0: firmware on card (1.16.63.0) is older than the version bundled with this driver, installing firmware 1.19.1.0 on card.
t5nex0: fw init failed: 5.
t5nex0: error during attach, adapter is now in recovery mode.
device_attach: est5 attach returned 6

t5nex0: <Chelsio T580-LP-SO-CR> mem 0xfb300000-0xfb37ffff,0xfa000000-0xfaffffff,0xfb984000-0xfb985fff irq 64 at device 0.4 numa-domain 1 on pci9
t5nex0: firmware on card (1.19.1.0) is older than the version bundled with this driver, installing firmware 1.23.0.0 on card.
t5nex0: fw init failed: 5.
t5nex0: error during attach, adapter is now in recovery mode.

t5nex0: <Chelsio T580-LP-SO-CR> mem 0xfb300000-0xfb37ffff,0xfa000000-0xfaffffff,0xfb984000-0xfb985fff irq 64 at device 0.4 numa-domain 1 on pci9
t5nex0: failed to connect to the firmware: 16, 963208048.
t5nex0: error during attach, adapter is now in recovery mode.

So, it can be dangerous...

[cem]

In D19754#429906, @ae wrote:

We have single machine, that after automatic firmware upgrade fails to attache the driver:
...
So, it can be dangerous...

Ok, that seems like maybe a misfeature in cxgbe :-(. I won't commit this with that unresolved (or without np@'s approval).

[zlei]

In D19754#429998, @cem wrote:

In D19754#429906, @ae wrote:

We have single machine, that after automatic firmware upgrade fails to attache the driver:
...
So, it can be dangerous...

Ok, that seems like maybe a misfeature in cxgbe :-(. I won't commit this with that unresolved (or without np@'s approval).

There's an tunable hw.cxgbe.fw_install which is default to 1 .

hw.cxgbe.fw_install
             0 prohibits the driver from installing a firmware on the card.  1
             allows the driver to install a new firmware if internal driver
             heuristics indicate that the new firmware is preferable to the
             one already on the card.  2 instructs the driver to always
             install the new firmware on the card as long as it is compatible
             with the driver and is a different version than the one already
             on the card.  The default is 1.

Is it better to set it default to 0, and then enable PNP PCIID-based autoload ?