On Tue, 2021-02-23 at 15:54 +0100, Timo Rothenpieler wrote: > On 23.02.2021 15:51, Trond Myklebust wrote: > > On Tue, 2021-02-23 at 15:19 +0100, Timo Rothenpieler wrote: > > > This follows what was done in > > > 8c2fabc6542d9d0f8b16bd1045c2eda59bdcde13. > > > With the default being m, it's impossible to build the module > > > into > > > the > > > kernel. > > > > > > Signed-off-by: Timo Rothenpieler <timo@xxxxxxxxxxxxxxxx> > > > --- > > > fs/nfs/Kconfig | 2 +- > > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > > > diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig > > > index e2a488d403a6..14a72224b657 100644 > > > --- a/fs/nfs/Kconfig > > > +++ b/fs/nfs/Kconfig > > > @@ -127,7 +127,7 @@ config PNFS_BLOCK > > > config PNFS_FLEXFILE_LAYOUT > > > tristate > > > depends on NFS_V4_1 && NFS_V3 > > > - default m > > > + default NFS_V4 > > > > > > > Let's just make that > > > > default y > > > > ...and let the dependencies work out whether or not this needs to > > be a > > 'y', 'm' or 'N'. Tying it to NFS_V4 just makes the Kconfig harder > > to > > read, with the result being the same anyway. > > Sure. Should I change the other two PNFS_ configs to default to y as > well while at it? > Oh hang on... Brain fart... I thought the dependency on NFS_V4_1 would suffice to make the correct choice, but it looks like that is a 'bool' and so it won't propagate the 'm' choice from NFS_V4. That is probably why Christoph did the NFS_V4 default. Never mind, then. Let's keep your first patch. -- Trond Myklebust Linux NFS client maintainer, Hammerspace trond.myklebust@xxxxxxxxxxxxxxx
