Hi
On 11/11/2024 14:12, Alexander Shishkin wrote:
Uwe Kleine-König u.kleine-koenig@baylibre.com writes:
After commit 0edb555a65d1 ("platform: Make platform_driver::remove() return void") .remove() is (again) the right callback to implement for platform drivers.
Convert all platform drivers below drivers/hwtracing to use .remove(), with the eventual goal to drop struct platform_driver::remove_new(). As .remove() and .remove_new() have the same prototypes, conversion is done by just changing the structure member name in the driver initializer.
Also adapt some whitespace to make indention consistent.
Signed-off-by: Uwe Kleine-König u.kleine-koenig@baylibre.com
Acked-by: Alexander Shishkin alexander.shishkin@linux.intel.com
Hello,
I did a single patch for all of drivers/hwtracing. While I usually prefer to do one logical change per patch, this seems to be overengineering here as the individual changes are really trivial and shouldn't be much in the way for stable backports. But I'll happily split the patch if you prefer it split. Maybe split for coresight vs. intel_th? Also if you object the indentation stuff, I can rework that.
I'm fine with it as is.
This is based on today's next, if conflicts arise when you apply it at some later time and don't want to resolve them, feel free to just drop the changes to the conflicting files. I'll notice and followup at a later time then. Or ask me for a fixed resend. (Having said that, I recommend b4 am -3 + git am -3 which should resolve most conflicts just fine.)
Does anybody want to pick this up or should I? I'm fine either way, but if there are any conflicts they won't be from my end of things, so it might make sense to take it via the coresight path.
I am happy to take them via coresight tree and queue them for v6.14
Suzuki
Thanks,
Alex