Messages in this thread | | | Date | Tue, 25 Jul 2000 12:03:33 -0400 (EDT) | From | Stephen Frost <> | Subject | Re: Direct access to hardware |
| |
On Tue, 25 Jul 2000, James Sutherland wrote:
> On Tue, 25 Jul 2000, Stephen Frost wrote: > > > On Tue, 25 Jul 2000, James Sutherland wrote: > > > > > The proper interface will allow the very small handful of utilities which > > > need to bypass these checks (to update firmware or whatever) to do so, > > > while preventing other access. > > > > The proper interface... You mean like the nice clean interface > > we currently have that will just be cluttered up by these attempts to > > filter it? > > There is NOTHING "nice [and] clean" about the current abomination of "do > everything in userland, however stupid". In the long term, it won't be > "cluttered up", it will be removed and replaced by a sane interface which > doesn't require the keys to the universe just to change the drive > powersave settings.
We *don't* do *everything* in userland. We just allow access to the raw side of the world from userland so that a program doesn't have to do worse things to try and get the same access. The way it currently is there is the possibility that a vendor wouldn't have to muck around in the kernel to be able to upgrade the flash. This keeps the vendor specific *junk* out of the kernel and in userland. The kernel provides a nice clean interface to devices which conform to the spec. Note that such raw access is, from what I can tell, part of the spec, just the specific data sent using it isn't specified in the spec and has been used by vendors to provide vendor-specific hooks, which reminds me of 'SCSI generic'...
Stephen
- 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/
| |