On May 8, 2019 4:33 PM, "Richard Henderson" <richard.hender...@linaro.org> wrote: > > On 5/8/19 1:15 AM, Aleksandar Markovic wrote: > > > > On May 8, 2019 2:19 AM, "Richard Henderson" < richard.hender...@linaro.org > > <mailto:richard.hender...@linaro.org>> wrote: > >> > >> > >> > > > > This commit message doesnˊt explain the reason for the change, and why is this > > an improvement. The underlyng reason for distingishing between env_cpu and > > env_archcpu cases is not explained too. > > It's certainly explained in the preceeding patches that introduce those functions. >
A commit (code+message) should be as standalone as possible, and one should not be forced to resort to reverse-engineering and perusing mailing list or patchwork in order to reveal its true meaning in another commit message altogether. Thanks, Aleksandar > Are you suggesting that it is beneficial to copy-and-paste a common block > explanation into 21 commit messages for each of target/foo/? > > > r~