Why does Linux nfs not set EXCHGID4_FLAG_SUPP_FENCE_OPS in NFSv4.2 mode?

Why does Linux nfs not set EXCHGID4_FLAG_SUPP_FENCE_OPS in NFSv4.2 mode?

[Takeshi Nishimura]

Why does Linux nfs not set EXCHGID4_FLAG_SUPP_FENCE_OPS in NFSv4.2 mode?

From https://datatracker.ietf.org/doc/html/rfc7862#section-14.1.4
   A client SHOULD request the EXCHGID4_FLAG_SUPP_FENCE_OPS capability
   when it sends an EXCHANGE_ID operation.  The server SHOULD set this
   capability in the EXCHANGE_ID reply whether the client requests it or
   not.  It is the server's return that determines whether this
   capability is in effect.  When it is in effect, the following will
   occur:

Why is Linux not doing this?
-- 
Internationalization&localization dev / 大阪大学
Takeshi Nishimura <takeshi.nishimura.linux@gmail.com>

Re: Why does Linux nfs not set EXCHGID4_FLAG_SUPP_FENCE_OPS in NFSv4.2 mode?

[Takeshi Nishimura]

?

On Tue, Jan 7, 2025 at 4:41 PM Takeshi Nishimura
<takeshi.nishimura.linux@gmail.com> wrote:
>
> Why does Linux nfs not set EXCHGID4_FLAG_SUPP_FENCE_OPS in NFSv4.2 mode?
>
> From https://datatracker.ietf.org/doc/html/rfc7862#section-14.1.4
>    A client SHOULD request the EXCHGID4_FLAG_SUPP_FENCE_OPS capability
>    when it sends an EXCHANGE_ID operation.  The server SHOULD set this
>    capability in the EXCHANGE_ID reply whether the client requests it or
>    not.  It is the server's return that determines whether this
>    capability is in effect.  When it is in effect, the following will
>    occur:
>
> Why is Linux not doing this?
> --
> Internationalization&localization dev / 大阪大学
> Takeshi Nishimura <takeshi.nishimura.linux@gmail.com>



-- 
Internationalization&localization dev / 大阪大学
Takeshi Nishimura <takeshi.nishimura.linux@gmail.com>

Re: Why does Linux nfs not set EXCHGID4_FLAG_SUPP_FENCE_OPS in NFSv4.2 mode?

[Trond Myklebust]

On Thu, 2026-04-23 at 19:07 +0200, Takeshi Nishimura wrote:
> ?
> 
> On Tue, Jan 7, 2025 at 4:41 PM Takeshi Nishimura
> <takeshi.nishimura.linux@gmail.com> wrote:
> > 
> > Why does Linux nfs not set EXCHGID4_FLAG_SUPP_FENCE_OPS in NFSv4.2
> > mode?
> > 
> > From https://datatracker.ietf.org/doc/html/rfc7862#section-14.1.4
> >    A client SHOULD request the EXCHGID4_FLAG_SUPP_FENCE_OPS
> > capability
> >    when it sends an EXCHANGE_ID operation.  The server SHOULD set
> > this
> >    capability in the EXCHANGE_ID reply whether the client requests
> > it or
> >    not.  It is the server's return that determines whether this
> >    capability is in effect.  When it is in effect, the following
> > will
> >    occur:
> > 
> > Why is Linux not doing this?

   1. Minor versions are not allowed to introduce a feature and make it
      mandatory
   2. Why bother, since the server is going to set the flag whether or
      not the client requested it?


-- 
Trond Myklebust Linux NFS client maintainer, Hammerspace
trondmy@kernel.org, trond.myklebust@hammerspace.com