On 29/01/24 10:07 pm, Greg KH wrote:
On Mon, Jan 29, 2024 at 09:57:40PM +0530, Harshit Mogalapalli wrote:
Hi Kovalev,
On 29/01/24 1:49 pm, kovalev@altlinux.org wrote:
29.01.2024 08:43, Harshit Mogalapalli wrote:
This patch is only for v5.10.y stable kernel. I have tested the patched kernel, after mounting it doesn't become unavailable.
Context: [1] https://lore.kernel.org/all/CAH2r5mv2ipr4KJfMDXwHgq9L+kGdnRd1C2svcM=PCoDjA7u...
Note to Greg: This is alternative way to fix by not taking commit eb3e28c1e89b ("smb3: Replace smb2pdu 1-element arrays with flex-arrays"). before applying this patch a patch in the queue needs to be removed: https://git.kernel.org/pub/scm/linux/kernel/git/stable/stable-queue.git/tree...
Maybe I don't understand something, but isn't there a goal when fixing bugs to keep the code of stable branches with upstream code as much as possible? Otherwise, the following fixes will not be compatible..
I agree, but at the same time we also should observe this: eb3e28c1e89b ("smb3: Replace smb2pdu 1-element arrays with flex-arrays") is not in 5.15.y so we probably shouldn't queue it up for 5.10.y.
Hi Greg,
It is queued up for 5.10, but not 5.15 for some reason?
Context: https://lore.kernel.org/all/472d92aa-1b49-43c9-a91f-80dfc8f25ad3@oracle.com/
I think the above commit eb3e28c1e89b ("smb3: Replace smb2pdu 1-element arrays with flex-arrays") got queued up from the above backport.
I did try applying on 5.15.y but I noticed many conflicts, so I asked which is the preferred way: 1. Resolving conflicts by backporting above commit(flex-arrays one), which touches more code. 2. Or go with one line change.
I thought one liner is is a simpler change although it is not a upstream commit. Steve French also agreed with approach 2(which Paul suggested here [1], so I made patch(approach 2) for 5.15.y and 5.10.y and tested both after patching, they work.
[1] https://lore.kernel.org/all/446860c571d0699ed664175262a9e84b@manguebit.com/
Thanks, Harshit
confused,
greg k-h