Le 23/07/2023 à 22:34, Greg KH a écrit :
On Fri, Jul 21, 2023 at 07:47:41PM +0200, Christophe JAILLET wrote:
Since commit 235602146ec9 ("i2c-nomadik: turn the platform driver to an amba driver"), there is no more request_mem_region() call in this driver.
So remove the release_mem_region() call from the remove function which is likely a left over.
Fixes: 235602146ec9 ("i2c-nomadik: turn the platform driver to an amba driver") Cc: stable@vger.kernel.org # v3.6+ Acked-by: Linus Walleij linus.walleij@linaro.org Reviewed-by: Andi Shyti andi.shyti@kernel.org Signed-off-by: Christophe JAILLET christophe.jaillet@wanadoo.fr
The patch below that should fix a merge conflict related to commit 9c7174db4cdd1 ("i2c: nomadik: Use devm_clk_get_enabled()") has been HAND MODIFIED.
I don't understand, that commit is not in the stable trees. What do you mean by "hand modified"?
I mean that I took the file initially generated by git format-patch when I sent the patch against -next.
I updated the diff context from this file and left everything else as-is. I only added some (apparently unclear :) ) context below the ---.
Yes, precisely. My patch has been sent *after* 9c7174db4cdd1 in -next, and it depends on it.
The proposed modified patch for backport tries to deal with this missing part in older branches.
If 9c7174db4cdd1 was also backported, I think that my patch would apply cleanly.
I hope it is fine, but is provided as-is. Especially line numbers should be wrong, but 'patch' should be able to deal with it. (sorry if it does not apply)
I guess that it should also apply to all previous branches.
I've left the commit description as it was. Not sure what to do with A-b and R-b tags.
Why isn't this needed in Linus's tree?
I never said that.
It is one of the first time, if not the first one, I update a patch to ease backport. I only sent something for 6.4.y, but it must be valid for other branches as well. (i.e. # v3.6+)
The way I generated this backport proposal is likely not the correct approach. I choose to hand modify the initial git format-patch output. I think that I should have switched to the expected branch and apply a clean process from there.
Personalty, I don't really care if backported or not. I don't use these older kernel. I don't use this hardware (AFAIK :)). I proposed an updated patch to help, should it be helpful for you or any one else.
If I did it wrong, sorry for the noise.
No hard feeling, but I won't update this backport proposal, should something need to be fixed.
I consider I've already done my part of the job.
CJ
confused,
greg k-h