Re: [PATCH net v3 RESEND 0/3] Fix switch mode for offload and non-offload cases

From: Meghana Malladi

Date: Sat Nov 21 2026 - 10:08:30 EST


Hi Jakub,

On 11/20/26 06:14, Jakub Kicinski wrote:
On Thu, 12 Nov 2026 15:51:37 +0530 Meghana Malladi wrote:
v2: https://lore.kernel.org/all/20261027131732.3411403-1-m-malladi@xxxxxx/

What changed in v3? When was v3 posted (since this is a repost?)

I first posted v3 on November 10 and resent the full series on

November 12. Here's the original cover letter:



https://lore.kernel.org/all/20261110122700.1849634-1-m-malladi@xxxxxx/

I reposted because I lately realized patch 3/3 was missing (could me mail server glitch)

I included the v2-to-v3 changes in the individual patch notes. Here's the summary for the series:
* Patch 1/3 adds a TX timestamp cleanup fix so reserved ERR_PTR(-EBUSY)

slots are cleared without attempting to free them as skbs.

* Patch 2/3 changes the R30 race fix to stop and restart phylib around

the PRU firmware restart, so emac_adjust_link() cannot issue R30

commands while the firmware is stopped.

* Patch 3/3 records the UC/MC flooding flags before issuing the firmware

commands and restores the flooding state after firmware restarts on

bridge and HSR unlink. It also carries the NULL emac check from v2.

You are expected to reply to the sashiko review, even if just
to confirm that it's correct.

Yes, got it. Some series specifically ask for it and some don't. So it slipped my mind while posting. I will keep that in mind for my upcoming patches.

Thanks,
Meghana