svn commit: r357051 - head/sys/dev/bge

[Gleb Smirnoff]

On Thu, Jan 23, 2020 at 04:08:32PM -0700, Ian Lepore wrote:
I> On Thu, 2020-01-23 at 15:05 -0800, Gleb Smirnoff wrote:
I> > On Thu, Jan 23, 2020 at 02:17:33PM -0500, Ryan Stone wrote:
I> > R> What is a driver's responsibility now for entering/leaving the net epoch now?
I> > 
I> > For drivers that are 'special', entering the net epoch is necessary. Special
I> > usually means running if_input outside network interrupt context.
I> > 
I> > However, there is plan to generalize entering/exiting epoch for taskqueues
I> > and callouts.
I> > 
I> 
I> That sounds every bit as horrible and out-of-place as the recent hack
I> (and it does feel very much like a horrible hack) that put network-
I> specific code into the guts of interrupt dispatching.

It isn't really a network code. You just include <sys/epoch.h> which
declares the epoch KPI and a few globally recognized epochs. For now
the only one is the network epoch. 

If you want to have for a example a disk epoch also supported
by interrupt code, that would add one extra declaration to epoch.h,
and again nothing really disk related.

-- 
Gleb Smirnoff