Re: [PATCH] lkdtm: do not depend on CONFIG_BLOCK

[Greg Kroah-Hartman]

On Fri, Nov 09, 2018 at 07:05:51AM +0000, Christophe Leroy wrote:
> Most parts of lkdtm don't require CONFIG_BLOCK.
> 
> This patch limits dependency to CONFIG_BLOCK in order to give embedded
> platforms which don't select CONFIG_BLOCK the opportunity to use LKDTM.
> 
> Fixes: fddd9cf82c9f ("make LKDTM depend on BLOCK")
> Signed-off-by: Christophe Leroy <christophe.leroy@xxxxxx>
> ---
>  drivers/misc/lkdtm/core.c | 7 ++++++-
>  lib/Kconfig.debug         | 1 -
>  2 files changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/misc/lkdtm/core.c b/drivers/misc/lkdtm/core.c
> index 2837dc77478e..bc76756b7eda 100644
> --- a/drivers/misc/lkdtm/core.c
> +++ b/drivers/misc/lkdtm/core.c
> @@ -40,9 +40,12 @@
>  #include <linux/interrupt.h>
>  #include <linux/hrtimer.h>
>  #include <linux/slab.h>
> -#include <scsi/scsi_cmnd.h>
>  #include <linux/debugfs.h>
>  
> +#ifdef CONFIG_BLOCK
> +#include <scsi/scsi_cmnd.h>
> +#endif

Why would this config option be needed to be checked just to be able to
include a .h file?

And shouldn't you be depending on SCSI instead?

> +
>  #ifdef CONFIG_IDE
>  #include <linux/ide.h>
>  #endif
> @@ -101,7 +104,9 @@ static struct crashpoint crashpoints[] = {
>  	CRASHPOINT("FS_DEVRW",		 "ll_rw_block"),
>  	CRASHPOINT("MEM_SWAPOUT",	 "shrink_inactive_list"),
>  	CRASHPOINT("TIMERADD",		 "hrtimer_start"),
> +# ifdef CONFIG_BLOCK
>  	CRASHPOINT("SCSI_DISPATCH_CMD",	 "scsi_dispatch_cmd"),
> +# endif

Again, scsi?

thanks,

greg k-h