On Wed, 2010-06-02 at 10:07 +0200, Bjørn Mork wrote: > I'm attaching an alternate patchset consisting of a number of cherry > picked commits from (Linus') upstream. Most of these applied without > changes on top of 2.6.32.14 as most of them only touch new files under > drivers/media/dvb/ . This adds a lot of meta data, such as pointers to > upstream commits, which I guess is useful. But do note that the > resulting driver is still the same. > > This patchset replaces the first patch attached to this bug! Please > apply either one, but not both :-) > > A few of the commits had to be edited to avoid touching files not > related to the mantis driver, or to resolve conflicts in the upper level > Kconfig/Makefile. This has been noted in the commit message. > > I have deliberately ignored a few patches which touched the ir input > system, as this has been changed in 2.6.33 and newer. > > As before, the only existing file touched is tda10021.c (except for the > necessary Kconfig and Makefile changes). This is a simple bugfix to > allow tda10021 and tda10023 to be probed in that order. The upstream > commit 4af699c13a5fb7332255de5fa56442d065f547a3 is applied unmodified. > > I hope this set will make it easier for you to review and add this > driver to the Debian 2.6.32 kernel. Thanks. Thank you. While I don't know the DVB system, I will assume that this has been properly reviewed upstream and I can verify that it matches. Therefore I've applied these changes and enabled building these drivers as modules. [...] > 2) a MODULE_DEVICE_TABLE() entry has been added to the two PCI drivers > to allow them to autoload. This is a bugfix not yet in upstream, > taken from http://jusst.de/hg/mantis-v4l-dvb/raw-rev/3731f71ed6bf > I've chosen to include it because it's obviously correct and it does > improve the usability a lot. But I'm of course willing to let it go > if the deviation from upstream is a problem. The patch file name is > 0119-Mantis-hopper-use-MODULE_DEVICE_TABLE-use-the-macro-.patch > and it can safely be dropped if necessary That makes sense. [...] > I can provide a git repository with the patchset if necessary, but the > bandwidth is so low that I don't think it's useful. Let me know if you > still want to try. But you might just as well clone > git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.32.y.git > and run "git am" on the attached patches. We don't (yet) use git for maintenance of the linux-2.6 package, so it wouldn't help that much anyway. Ben. -- Ben Hutchings Once a job is fouled up, anything done to improve it makes it worse.
Attachment:
signature.asc
Description: This is a digitally signed message part