On Fri, Feb 28, 2020 at 10:06:45AM -0800, Joe Perches wrote:
On Fri, 2020-02-28 at 08:12 +0100, Greg Kroah-Hartman wrote:
On Thu, Feb 27, 2020 at 07:55:49PM -0800, Joe Perches wrote:
On Thu, 2020-02-27 at 14:35 +0100, Greg Kroah-Hartman wrote:
From: Nathan Chancellor natechancellor@gmail.com
I believe these sorts of whitespace only changes should not be applied to a stable branch unless it's useful for porting other actual defect fixes.
I want to get clang build warnings down to the same level that gcc build warnings are, so that they become useful in detecting new issues. That is why I add these types of patches to the stable trees.
New issues should be found in the current kernel.
Right, but we're talking about issues with backporting patches. We catch a lot of issues where a patch applied to an older version, but is missing stuff or is outright incorrect by having the compiler warn us about it.
We're not trying to debug upstream, but rather to catch broken backports that compile but are broken.
Backporting whitespace changes is value-free.
The value here is to make other errors obvious.