Messages in this thread | | | From | Keith Owens <> | Subject | Re: [PATCH] final support for MODULE_PARAM as kernel commandline | Date | Tue, 14 Sep 1999 22:01:31 -0700 |
| |
On Tue, 14 Sep 1999 13:35:26 +0100 (BST), Alan Cox <alan@lxorguk.ukuu.org.uk> wrote: >Richard Guenther wrote >> I dont think there is any use in different parameters for module/in kernel >> use. I basically think one should enforce a new policy with >> - just renaming MODULE_PARAM to SETUP_PARAM (this would of course >> confuse people using the other MODULE_ macros...) >> - force anyone to use a suitable prefix for the parameters >> rather than inventing a new "prefix macro" > >That breaks too many tools and admin setups. Its out of the question. >Allowing the short form (ie insmod 3c501 io=0x340 as well as >insmod 3c501 3c501.io=0x340) solves a lot
I recommend that modules built into kernel use modname.parmname. Use the same syntax when loaded as a separate module but change insmod so
insmod modname parm1=val1 parm2=val2
is tried "as is" first then retried as
insmod modname modname.parm1=val1 modname.parm2=val2
if the parameter names do not exist in the module. It would require 2.3.18-ac users to upgrade to a new modutils but that is nothing new. Parameter files would be unchanged and both code and parm files would be backwards compatible.
Shall I upgrade insmod to automatically insert a modname prefix on parameters if the raw parameter name does not exist?
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |