Re: [PATCH v1] net/e1000: do not update link status in secondary process

[Bruce Richardson <bruce.richardson@intel.com>]

On Sun, Jul 14, 2024 at 04:26:26PM +0800, Jun Wang wrote:
>    >> The code to update link status is not safe in secondary process.
>    >> If called from secondary it will crash, example from dumpcap:
>    >>     eth_em_link_update
>    >>
>    >> Signed-off-by: Jun Wang <junwang01@cestc.cn>
>    >
>    > Wouldn't it be better to fix the code in e1000_check_link to work in
>    > secondary process. There are network virtual appliances that use
>    > secondary process for all processing.
> 
>    Yes, the e1000 virtual network card currently does not work properly
> 
>    in the secondary process. After skipping eth_em_link_update, I tested
> 
>    the e1000 card and it was able to capture packets normally. For the
> 
>    secondary process, I think eth_em_link_update is not necessary.
>    __________________________________________________________________
> 

Hi Jun,

can you provide some instructions as to how I can go about reproducing the
issue? I used a VM with emulated e1000 NICs on it and was able to run
testpmd and dumpcap side by side. Similarly, I was able to run two
instances of the symmetric_mp app side by size without seeing any crashes.
I'd just like to verify the issue and confirm this fixes a problem before
merging.

Thanks,
/Bruce