On Wed, Sep 03, 2025 at 09:35:58AM GMT, Pali Rohár wrote: > On Tuesday 02 September 2025 20:35:14 Rob Herring wrote: > > On Tue, Sep 2, 2025 at 3:32 PM Pali Rohár <pali@xxxxxxxxxx> wrote: > > > > > > These issues have been discussed more times for last 3 years. > > > Tested changes which are fixing real bugs and are improving the > > > pci-mvebu.c driver from people listed as M: in MAINTAINERS are being > > > rejected or silently ignored. And untested changes which are not going > > > to fix any bug are being accepted and causing new regressions. People > > > then reporting bug reports to M: people who cannot do anything with it. > > > The only advice which they can get is to not use mainline kernel. > > > This pci-mvebu.c driver in mainline kernel is broken and nobody wanted > > > to do anything with this situation for last 3 years, I was pointing for > > > it more times. Due to this I'm ignoring bug reports for pci-mvebu.c > > > driver for mainline kernel. > > > > This makes no sense. Mainline is broken, but yet there's a regression > > in mainline breaking the driver. The pci-mvebu.c maintainers have the > > ability to test patches, but refuse to test any submitted patches > > other than their own. I guess the only solution here is just remove > > the driver... > > > > Rob > > This is not the way how to get patches being tested. Ignoring other > people and taking untested patches because you did not wanted to > communicate with maintainers. I have no reason to test other patches if > my own are being ignored. This is fair. In past I spent lot of time on > addressing issues in this driver, preparing fixes for real bugs and > whole time was wasted. And if you expected that it would work in that, > then go ahead and do it, remove the driver. > Pali, as I communicated earlier, please send your fixes again and I'll try to get them merged if we do not have any blocking reasons and the patches make sense. I do not know what happened earlier with your patches and that's irrelevant for me. But if none of the other mainainters oppose your changes (IRQ in specific), there are no reasons to not merge them IMO. But if there are concerns, then it is your job to address them. That being said, if you do not want to send them again and the driver continues to being untested in mainline, we may need to mark it as Orphan/Obsolete and get it removed later. - Mani -- மணிவண்ணன் சதாசிவம்
