Damien objected off list. It is his driver so his vote counts. I'll send you his comments off list. If he wanted them out in public he would have sent it here.
On Tue, Jan 26, 2010 at 04:56:55AM +0000, Roland Dreier wrote: > Marco Peereboom <slash <at> peereboom.us> writes: > > > Objections were made. Apparently this patch only works for AP and does > > funky stuff to the hardware. So back to the drawing board on this one. > > Who made these objections? I didn't see anything on the mailing list. > (And it would be nice if I could be kept cc'ed on this discussion, as the > original author of the patch) > > I honestly can't be bothered to test in STA mode at this point but it's > hard for me to see how my patch breaks that -- I don't particularly know > much about how this hardware works but it seems to me that all the things > about the low-level TX interrupt handling that I'm changing work exactly > the same in AP or STA mode. According to Damien this part is broken. > And "does funky stuff to the hardware"?? I don't even know what that > means, and I can't see how a patch that does nothing but simplify the > code and close races in interrupt handling could be described that way. > > The current rt2661 code is seriously buggy, and has been buggy ever since > the driver was committed. Multiple people find rt2661 essentially unusable > because the interface gets stuck with OACTIVE set. I think I understand > the root cause of that, and described it when I sent my original patch; if > you guys want to leave rt2661 broken because "objections were made" in > some private way that I can't even see, that's fine but I'm afraid it will > be someone other than me going back to the drawing board. > > - Roland
