Re: ixg(4) mutex problem (PR#49328)

[christos%zoulas.com@localhost (Christos Zoulas)]

On Feb 1, 10:00am, chuq%chuq.com@localhost (Chuck Silvers) wrote:
-- Subject: Re: ixg(4) mutex problem (PR#49328)

| I'm not familiar enough with the current state or planned future direction
| of our network stack to say which way makes more sense for us.

This driver will never work with a spin mutex and will require an
extensive re-write if we insist going down the spin mutex path.
It is better to change this now and see what breaks. Of course
this is orthogonal if we are allowed to call allocation routines
while holding a spin mutex.

The two problems are separate so let's change the driver to not
hold a spin mutex right now and see what happens.

christos