Dynamic modules

2015-05-04 Thread Kamil Rytarowski
I have got a few questions. 1. luapmf and luasystm src/sys/modules/luapmf/luapmf.c src/sys/modules/luasystm/luasystm.c These modules are empty for the !_MODULE build. Why? Can we make them available for the builtin mode. 2. luactl(8) $ sudo luactl help usage: luactl [-cq] luactl [-cq] c

Re: Dynamic modules

2015-05-04 Thread Paul Goyette
On Mon, 4 May 2015, Kamil Rytarowski wrote: 3. Is it possible to automatically create a device file in /dev from a module? Not really, at least not from a loaded kernel module. On the other hand, modules are also applicable to "rump" environments, and you _can_ dynamically create the /de

Re: Dynamic modules

2015-05-04 Thread Kamil Rytarowski
> Sent: Monday, May 04, 2015 at 11:39 AM > From: "Paul Goyette" > To: "Kamil Rytarowski" > Cc: tech-kern@netbsd.org > Subject: Re: Dynamic modules > > On Mon, 4 May 2015, Kamil Rytarowski wrote: > > > > 3. Is it possible to automatically create a device file in /dev from a > >module? > >

re: Dynamic modules

2015-05-04 Thread matthew green
> > 3. Is it possible to automatically create a device file in /dev from a > >module? > > Not really, at least not from a loaded kernel module. sure you can. ipfilter was doing this back with LKMs in the 90s :-) > BTW, devsw_attach() doesn't "automatically generate the major". It does >

re: Dynamic modules

2015-05-04 Thread matthew green
> > BTW, devsw_attach() doesn't "automatically generate the major". It does > > a lookup in the majors table. If there is no entry in the table, you > > get an error. > > see eg this line in devsw_attach(): > > newptr = kmem_zalloc(new * DEVSWCONV_SIZE, KM_NOSLEEP); > > for

re: Dynamic modules

2015-05-04 Thread matthew green
matthew green writes: > > > > BTW, devsw_attach() doesn't "automatically generate the major". It does > > > a lookup in the majors table. If there is no entry in the table, you > > > get an error. > > > > see eg this line in devsw_attach(): > > > > newptr = kmem_zalloc(new *

Splitting the compat module?

2015-05-04 Thread Paul Goyette
Folks, The next modularization task I plan to attack is the SYSV_* options. But first, I need to separate the SYSV* compat syscalls out of the "standard" compat module, and into a separate module. Otherwise, anything that requires any compat functionality will bring in all of a new SYSV* module