Re: [PATCH -next] init/Kconfig: fix unmet direct dependencies

From: Ren Zhijie
Date: Wed Sep 28 2022 - 03:53:55 EST



在 2022/9/28 15:20, Lukas Bulwahn 写道:
On Wed, Sep 28, 2022 at 9:01 AM Sebastian Andrzej Siewior
<bigeasy@xxxxxxxxxxxxx> wrote:
On 2022-09-28 06:49:34 [+0000], Ren Zhijie wrote:
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -1273,6 +1273,7 @@ endif # NAMESPACES

config CHECKPOINT_RESTORE
bool "Checkpoint/restore support"
+ select PROC_FS
Couldn't this become a depends?

It could also be a depends (to resolve the warning).

It is just the question whether:

When PROC_FS is not set, should the CHECKPOINT_RESTORE still be
visible as a config option to add (and then automatically add
PROC_FS)? Then select is right here.

or:

When PROC_FS is not set, should the CHECKPOINT_RESTORE not be visible
as a config option to add? Instead the user first needs to add
PROC_FS, then CHECKPOINT_RESTORE becomes visible as an option to add,
and then the user can add it. Then depends would be right.

For me, both seem reasonable. So, I assume Ren considered select the
better choice.

But maybe Ren can confirm.

My consider is that if CHECKPOINT_RESTORE depends on PROC_FS , when PROC_FS is not set the user have no chance to set it on.

Thanks,

Ren


A kernel build configuration without PROC_FS is quite special
anyway... and then being interested in CHECKPOINT_ RESTORE for such a
system is really really special. I wonder if that user then really
knows what he or she is configuring at that point.


Lukas
.
.