On Fri, 2009-12-18 at 12:54 -0800, Yinghai Lu wrote: > this patche set is trying to update pci bridge BAR when that BAR is big enough. > > default it is disabled. > > could use pci=try=2 to enable it. Thanks for posting this as a nice new series. It was getting hard to figure out what went where. I think you mean "when the BAR is *not* big enough." And strictly speaking, I think you're concerned with the bridge *window*, which isn't actually a bridge BAR. I don't like the new "pci=try=2" parameter. Linux should be smart enough to do the right thing without requiring the user to do something special. Using a parameter to enable the new code makes it feel like "here's some new functionality, but I'm not sure it's safe enough for everybody to use, so try this parameter if you need it." The result is that the new code will be rarely used and poorly tested, yet it remains a burden to all future PCI maintainers, who will have to understand it and try not to break it. Can you please put the "-v2"-type comments about your development history in the [0/n] email? They are useful while reviewing the different versions, but they don't need to be in the commit logs. Also, some of the patches in this series have "-v2" in subject, others have "-v3", others have nothing. It's easier to follow development of the series if the version applies to the series as a whole, not to individual patches. The version doesn't need to be on the patch subject line, because that will go into the commit log, where it won't be relevant. Tools like stgit put it where it will be automatically discarded, e.g., "[PATCH -v2 x/n]" Bjorn -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
