Re: [PATCH 6.1 000/481] 6.1.167-rc1 review

[Sun Yongjian]

在 2026/3/25 21:34, Theodore Tso 写道:
> I'm confused.  That commit 060913999d7a should be backported to
> 6.1.yy?  Since 6.1 lacks this patch, I dont understand your statement
> "the race condition doesn't exist there".  Or were you trying to say
> that the_fix_ for race condition is missing in 6.1?
>
> Or is there some other commit which fixes 060913999d7a that needs to
> be backported to 6.1?
Hi Ted,

Sorry for the confusion. To clarify: commit 060913999d7a is actually the 
regressing commit that introduced the race condition by reordering the 
migrate mapping and folio copy operations.

Because 6.1.y does not include this specific commit, the concurrency 
window it created does not exist in the 6.1 kernel. Therefore, 6.1.y is 
not affected by this particular issue and does not require this backport 
(ext4: fix e4b bitmap inconsistency reports).

     	       	   		     	     - Yongjian