Kroah-Hartman: Remove Hyper-V Driver from Kernel?

Sep 07, 2009

Microsoft caused a sensation with its release of its Hyper-V virtualization driver -- now its inclusion in the Linux kernel could be in doubt.

In the gmane.linux.kernel newsgroup, lead driver developer Kroah-Hartman announced the state of drivers in the kernel staging tree, which ones will come and which will go. As he says, "If
no one steps up to maintain and work to get the code merged into the main portion of the kernel, the drivers will be removed," as drivers/staging/ is no "dumping ground for dead code." Next to some Android and WLAN drivers, Microsoft's Hyper-V driver is definitely among those that could be dropped.

Redmond's driver seems to have cost the kernel team quite a bit. Kroah-Hartman writes that it took over 200 patches to get the code "into semi-sane kernel coding style," apparently mostly through his personal effort ("someone owes me a bit [sic] bottle of rum for that work!"). The main problem, however, is the total unresponsiveness of the Microsoft team: "Unfortunately the Microsoft developers seem to have disappeared, and no one is answering my emails. If they do not show back up to claim this driver soon, it will be removed in the 2.6.33 release. So sad...."

The situation is especially sad for Kroah-Hartman in that it was mostly through his effort that Microsoft put the code under GPL. The Hyper-V driver will nevertheless still appear in the 2.6.32 kernel, as he writes in the newsgroup.

The linux-staging developer tree has been around since June 2008. It contains drivers in early development that are tested and cleaned up for the main kernel.

Related content

Comments

  • I'm clueless and even I understand

    I've never submitted a patch to the Kernel, but even so I know that every Open Source project including the Linux Kernel has coding standards that should be followed. Microsoft developers dumping code into the kernel and running away isn't appropriate and their code should not be included. If I submitted cruddy code to the Kernel would someone else make hundreds of edits to get it included? I think not.

    So, is Microsoft trying to play more games here? "See, we tried to participate in open source, but they would accept our contributions."
  • Remove Hyper-V driver from kernel?

    Classic case of the kernel team wanting to have their cake and eat it. If you wanted the Microsoft devs to be able to maintain the code, perhaps making over 200 changes wasn't a particularly wise way to go about it? You can't claim that proprietary drivers are bad and se that as reasoning for not providing a stable interface and then throw the toys out of the pram when third party code is written to third parties coding conventions.

    Ultimately it's those wanting to use Linux who suffer as a result of this mentality. Throwing the Hyper-V drivers out will send a very clear message to third parties that there is no point sharing driver code with the FOSS community, because its just going to get thrown back in your face. Meanwhile, Windows on Hyper-V looks a more and more promising server infrastructure, eating into the one market where Linux has traditionally been strong.
  • "gmane.linux.kernel newsgroup"

    The "gmane.linux.kernel newsgroup" is just a gateway between the regular Linux Kernel Mailing List (lkml) and INN. Good to see that the author is so intimately familiar with the Linux kernel development process.
comments powered by Disqus