On Sun, Jun 03, 2018 at 02:26:23PM +0200, Greg Kroah-Hartman wrote: > There are pros and cons of dealing with tools in the kernel directory. > The pros are the fact that development happens fast, and new features > can be added to the kernel and the tools at the same times. The cons > are when dealing with backported kernel patches, it can be necessary to > backport parts of the tool changes as well. > > For 4.9.y so far, we have backported individual patches. That quickly > breaks down when there are minor differences between how backports were > handled, so grabbing 40+ patch long series can be difficult, not > impossible, but really frustrating to attempt. > > To help mitigate this mess, this patch series works to sync up the > objtool code with the version that is currently in 4.14.47. This > required a number of objtool old patches to be backported, a single big > "sync the world" patch, and a lot of include file updates to get > everything all working properly as well as some minor fixes for build > warnings. > > And at the end of this series, is a set of backported objtool patches > for gcc-8 that was the main reason this whole patch series was created. > They applied just fine, being the identical version that goes into the > 4.14.y stable tree, so it feels like the backport was successful. > > This has survivied my limited testing, and as the codebase is identical > to 4.14.47, I'm pretty comfortable dropping this big change in here in > 4.9.y. Hopefully all goes well... > > If anyone has any objections to this patch series, or finds anything I > messed up on, please let me know. Thanks Greg. Looks good to me. Acked-by: Josh Poimboeuf <jpoimboe@xxxxxxxxxx> -- Josh
