On 12/27/20 9:24 AM, Philippe Mathieu-Daudé wrote: > On 12/27/20 5:40 PM, BALATON Zoltan via wrote: >> On Sun, 27 Dec 2020, Philippe Mathieu-Daudé wrote: >>> On 12/25/20 12:23 AM, BALATON Zoltan wrote: >>>> From: Guenter Roeck <li...@roeck-us.net> >>>> >>>> Fuloong2e needs to use legacy mode for IDE support to work with Linux. >>>> Add property to via-ide driver to make the mode configurable, and set >>>> legacy mode for Fuloong2e. >>>> >>> >>> Fixes: 4ea98d317eb ("ide/via: Implement and use native PCI IDE mode")? >> >> Not really. That patch did what it said (only emulating (half) native >> mode instead of only emulating legacy mode) so it wasn't broken per se >> but it turned out that approach wasn't good enough for all use cases so >> this now takes a different turn (emulating either legacy or half-native >> mode based on option property). Therefore. I don't think Fixes: applies >> in this case. It fixes an issue with a guest but replaces previous patch >> with different approach. (Even though it reuses most of it.) > > Well, if Linux guest got broken by this commit, why not name it a "fix"? > Anyway I don't mind how it is called. I find important to refer to the > commit hash to help navigating between commits while reviewing history. > > What about: > > ''' > The legacy mode for IDE support has been removed in commit 4ea98d317eb > ("ide/via: Implement and use native PCI IDE mode"). When using a Linux > guest, the Fuloong2e machine requires the legacy mode. > Add property to via-ide driver to make the mode configurable, and set > legacy mode for Fuloong2e. > ''' > > Guenter, is that OK with you? (I can update when applying this series > via the MIPS tree). >
Sure, I don't really care about the commit message as long as the problem is fixed. Guenter