Re: 3ware driver
On Sat, 12 Jun 2004 07:45, Gunnar Wolf <firstname.lastname@example.org> wrote:
> As a sidenote, I would like to ask you: Why are you approaching a
> specific (or many separate) Linux distributions, instead of submitting
> the module for the inclusion in the Linux kernel? I assume you are
> aware that, in order to include it in most distributions (and
> certainly in Debian), the module should be thoroughly tested in order
> to ensure it does not destabilize the system and performs reliably
For a long time Debian has had an aggressive policy of limiting the number of
kernel patches applied to the kernel.org source. Patches for random pieces
of hardware generally didn't get into the Debian kernel source tree unless
they were in the kernel.org kernel, or the hardware was very popular.
The Debian kernel maintenance process is changing at the moment, but I still
don't expect them to start accepting patches for new hardware. When you've
sold a million RAID cards there may be some pressure to get it in, but for a
new device there is no interest.
Adopting any kernel patches requires forward-porting the patches to newer
kernels (so a kernel upgrade does not reduce functionality), and there is
risk of things breaking which takes precious debugging time away from other
The situation in Red Hat for Fedora kernels is quite similar, there is a plan
to avoid patching the kernel.org source without a really good reason.
Generally I recommend not even bothering to talk to distributions about
accepting such kernel patches. Send the source with a GPL license statement
to the Linux kernel mailing list, if the kernel gurus accept the code as
worthy of inclusion then all the distributions will accept it into their
kernel source tree quickly. Distributions that aren't using the latest
kernel will usually be happy to back-port a driver from a newer kernel (it's
no pain for them as they know that when they move to a newer kernel they
won't have to worry about the driver).
Getting drivers on install disks is another issue (there are space problems),
but really it's not worth even trying that until kernel.org accepts the code.
http://www.coker.com.au/selinux/ My NSA Security Enhanced Linux packages
http://www.coker.com.au/bonnie++/ Bonnie++ hard drive benchmark
http://www.coker.com.au/postal/ Postal SMTP/POP benchmark
http://www.coker.com.au/~russell/ My home page