On Wed, May 30 2018, Arnd Bergmann wrote:
> Something in recent linux-next kernels caused linux/highmem.h to
> no longer be included implicitly from o2iblnd_cb.c, causing a build
> failure:
>
> drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c: In function 'kiblnd_kvaddr_to_page':
> drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c:549:15: error: 'PKMAP_BASE' undeclared (first use in this function); did you mean 'RTM_BASE'?
> if (vaddr >= PKMAP_BASE &&
> ^~~~~~~~~~
> RTM_BASE
> drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c:549:15: note: each undeclared identifier is reported only once for each function it appears in
> drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c:550:28: error: 'LAST_PKMAP' undeclared (first use in this function); did you mean 'AT_HWCAP'?
> vaddr < (PKMAP_BASE + LAST_PKMAP * PAGE_SIZE)) {
> ^~~~~~~~~~
> AT_HWCAP
>
> This adds back an explicit include for the header.
>
> Signed-off-by: Arnd Bergmann <arnd@xxxxxxxx>
Thanks for finding and fixing this. I did recently reduce the number
of include files that were automatically included everywhere. Clearly
some config combinations didn't get tested properly.
Reviewed-by: NeilBrown <neilb@xxxxxxxx>
Thanks,
NeilBrown
> ---
> drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c
> index 47eb8b4c28db..65b7a62943ad 100644
> --- a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c
> +++ b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c
> @@ -35,6 +35,7 @@
> * Author: Eric Barton <eric@xxxxxxxxxxxxxxxxxx>
> */
>
> +#include <linux/highmem.h>
> #include "o2iblnd.h"
>
> #define MAX_CONN_RACES_BEFORE_ABORT 20
> --
> 2.9.0
Attachment:
signature.asc
Description: PGP signature
_______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
