Hi Luca, Em Qui, 2007-05-03 às 15:37 +0200, Luca Olivetti escreveu: > En/na Mauro Carvalho Chehab ha escrit: > > Em Qui, 2007-05-03 às 13:55 +0200, Markus Rechberger escreveu: > >> Is there anything against it? > >> > >> Markus > >> > >> On 5/1/07, P. van Gaans <w3ird_n3rd@xxxxxxx> wrote: > >>> I managed to screw up my message for a second time so I'll just write a > >>> whole new message. Would it be possible to push the AF9005 drvier from > >>> /hg/~manu/af9005 to the main repo? I'm not sure Luca's line in his reply > >>> to me will get noticed so I thought this deserved a whole new message. > > > > For me, it is ok, providing that Luca or Manu send us a pull request. > > Is this a special mail or something, or can this message considered as a > pull request? ;-) This would be enough, providing that the patch is Ok. > > Most CodingStyle issues can easily by running a really small script, > > available at scripts/Lindent at the kernel tree. The script content is: > > > > #!/bin/sh > > indent -npro -kr -i8 -ts8 -sob -l80 -ss -ncs "$@" > > I ran Lindent before submitting the patch to Manu !?!? I don't think so. I saw some hexadecimal values on uppercase. Also some idents were wrong. Could you review it please, and then, send me the newest version c/c ML? Please, also mark the driver as EXPERIMENTAL, since this will be the first release. Hopefully, we can remove the tags at the next kernel. > > (of course, you need to have the indent package installed at the local > > machine) > > > > It is interesting also to mark newer drivers as EXPERIMENTAL for one or > > two kernel releases. > > > > If AF9005 is intended to be added on kernel 2.6.22, this should be done > > ASAP, since the merge window is closing. Otherwise, if the driver is not > > stable enough, the better is to wait to the end of the window, then add > > it to the tree. I prefer the last option. > > Well, I've been using my driver since december and I didn't suffer any > crash or problems. I had reports of funky usage number in lsmod, but I > think that's because of my use of symbol_request/symbol_put (though it > doesn't happen on my machine): I separated the remote decoding functions > in a standalone module, so I can substitute the implementation (in fact > I have a different one on my page that interfaces with lirc). > Other users have been using it with no problems (or at least with no > more problems than under windows, as P. van Gaans can testify ;-) > Anyway, do as you like, but if it's more visible someone more capable > than me can correct more bugs ;-) Even not being on Mainstream, it will be added at the project's main tree and at -mm series. This will help more people to see your driver, fixing issues and commenting it. > > Bye -- Cheers, Mauro _______________________________________________ linux-dvb mailing list linux-dvb@xxxxxxxxxxx http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
