On Wed, Jul 27, 2011 at 09:02:55AM -0700, Greg KH wrote: > On Wed, Jul 27, 2011 at 12:28:09PM +0100, Tony Vroon wrote: > > On 07/19/11 16:46, Greg KH wrote: > > > No one has ever sent me a patch :( > > > > There is mention of it here: > > http://permalink.gmane.org/gmane.linux.network/192229 > > > > Would you be able to pick it up for staging as-is? Perhaps Dave's review > > can be put in a TODO file? > > If someone sent me a patch for drivers/staging/ as-is, yes, I could take > it. But it requires someone to be responsible for cleaning it up and > working to resolve the issues and move it to the "main" portion of the > kernel tree. Without that, I would just drop it from the > drivers/staging/ tree in 2 releases. Sorry about disappearing; this fell off my radar. Here's my status: 1) The driver is in the chromium-os tree, still in use, and still under somewhat heavy development; the current sources are at <http://git.chromium.org/gitweb/?p=chromiumos/third_party/kernel.git;a=tree;f=drivers/net/usb/gobi;h=a84aacad41decaa4d4a7c634350cbad3278c53ad;hb=HEAD> 2) Qualcomm released their userspace under the 3-clause BSD license, which means that the Gobi 3000 is now useable by anyone using only free code (although the firmware images on the device are nonfree, the Gobi 3000 comes preloaded with firmware). Dan Williams (dcbw@xxxxxxxxxx, the networkmanager/modemmanager maintainer, CCed) owns libqmi <http://cgit.freedesktop.org/libqmi>, which is to be a BSD-licensed rewrite of Qualcomm's userspace. I am currently waiting for approval from Google Legal to contribute code and documentation to this. 3) As a consequence of (2), we (chromium-os) are currently open-sourcing all of our Gobi-related userspace work. This is not done yet but ought to be done within a couple of weeks. 4) As a further consequence of (2), we can redesign the userspace to use a single file descriptor and a single thread instead of the current atrocity, which will allow a lot of simplification of the kernel driver. I've been holding off on submitting the driver upstream again until this was done, but I can resubmit it now if you like. I am happy to become the maintainer of the driver and accept patches against it. Tom Tuttle (ttuttle@xxxxxxxxxx, CCed) is taking ownership of much of the cellular infrastructure from me and would probably also be glad to help. In particular, having a fully open userspace ought to make it much easier to get the driver into the tree :). > greg k-h -- elly
Attachment:
signature.asc
Description: Digital signature
