On April 13, 2014 8:04:57 PM EDT, Greg Donald <[email protected]> wrote: >On Sun, Apr 13, 2014 at 4:46 PM, Greg Freemyer ><[email protected]> wrote: >> In general use checkpatch.pl on code you are submitting or around >code >> you are already patching. > >But I see patches that do more than one thing get turned down every >day. > >> Sending in standalone coding style patches 9 times in 10 will result >> in a rejected patch. > >Why would 9/10 coding style patches be rejected simply for being >standalone? I thought standalone patches were preferred.
Most patches I see are sent in as a series. A coding style patch might be 1 of 2 and then 2 of 2 might be an actual code fix. Most of the patch series I see tend to have 5 or more patches in the series. So each patch should accomplish a single goal. Code clean-up would be a single goal and thus a good discrete patch. But a series that accomplished nothing beyond code cleanup is rejected by most of the subsystem maintainers. They figure there are too many patches in progress floating around that will have to be rebased for the value of the cleanup to be worthwhile. Greg -- Sent from my Android phone with K-9 Mail. Please excuse my brevity. _______________________________________________ Kernelnewbies mailing list [email protected] http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
