diff mbox series

rdma/siw: fix NOMMU build

Message ID 20190906151028.1064531-1-arnd@arndb.de
State New
Headers show
Series rdma/siw: fix NOMMU build | expand

Commit Message

Arnd Bergmann Sept. 6, 2019, 3:10 p.m. UTC
On kernels without CONFIG_MMU, we get a link error for the siw
driver:

drivers/infiniband/sw/siw/siw_mem.o: In function `siw_umem_get':
siw_mem.c:(.text+0x4c8): undefined reference to `can_do_mlock'

I don't know whether this driver is able to work at all without
an MMU, but it's easy to avoid the link problem by adding another
compile-time check.

Fixes: 2251334dcac9 ("rdma/siw: application buffer management")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

---
 drivers/infiniband/sw/siw/siw_mem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

-- 
2.20.0

Comments

Jason Gunthorpe Sept. 7, 2019, 7:34 a.m. UTC | #1
On Fri, Sep 06, 2019 at 05:10:10PM +0200, Arnd Bergmann wrote:
> On kernels without CONFIG_MMU, we get a link error for the siw

> driver:

> 

> drivers/infiniband/sw/siw/siw_mem.o: In function `siw_umem_get':

> siw_mem.c:(.text+0x4c8): undefined reference to `can_do_mlock'

> 

> I don't know whether this driver is able to work at all without

> an MMU, but it's easy to avoid the link problem by adding another

> compile-time check.

> 

> Fixes: 2251334dcac9 ("rdma/siw: application buffer management")

> Signed-off-by: Arnd Bergmann <arnd@arndb.de>

>  drivers/infiniband/sw/siw/siw_mem.c | 2 +-

>  1 file changed, 1 insertion(+), 1 deletion(-)

> 

> diff --git a/drivers/infiniband/sw/siw/siw_mem.c b/drivers/infiniband/sw/siw/siw_mem.c

> index e99983f07663..b3b614c07fdb 100644

> +++ b/drivers/infiniband/sw/siw/siw_mem.c

> @@ -374,7 +374,7 @@ struct siw_umem *siw_umem_get(u64 start, u64 len, bool writable)

>  	unsigned int foll_flags = FOLL_WRITE;

>  	int num_pages, num_chunks, i, rv = 0;

>  

> -	if (!can_do_mlock())

> +	if (!IS_ENABLED(CONFIG_MMU) || !can_do_mlock())

>  		return ERR_PTR(-EPERM);


I feel like !CONFIG_MMU should provide a dummy inline stub for can_do_mlock
instead?

Jason
Arnd Bergmann Sept. 7, 2019, 12:57 p.m. UTC | #2
On Sat, Sep 7, 2019 at 9:34 AM Jason Gunthorpe <jgg@mellanox.com> wrote:
> > @@ -374,7 +374,7 @@ struct siw_umem *siw_umem_get(u64 start, u64 len, bool writable)

> >       unsigned int foll_flags = FOLL_WRITE;

> >       int num_pages, num_chunks, i, rv = 0;

> >

> > -     if (!can_do_mlock())

> > +     if (!IS_ENABLED(CONFIG_MMU) || !can_do_mlock())

> >               return ERR_PTR(-EPERM);

>

> I feel like !CONFIG_MMU should provide a dummy inline stub for can_do_mlock

> instead?


Fair enough. I've added that to my test tree in place of my first patch
and will send it after some more build testing.

     Arnd
diff mbox series

Patch

diff --git a/drivers/infiniband/sw/siw/siw_mem.c b/drivers/infiniband/sw/siw/siw_mem.c
index e99983f07663..b3b614c07fdb 100644
--- a/drivers/infiniband/sw/siw/siw_mem.c
+++ b/drivers/infiniband/sw/siw/siw_mem.c
@@ -374,7 +374,7 @@  struct siw_umem *siw_umem_get(u64 start, u64 len, bool writable)
 	unsigned int foll_flags = FOLL_WRITE;
 	int num_pages, num_chunks, i, rv = 0;
 
-	if (!can_do_mlock())
+	if (!IS_ENABLED(CONFIG_MMU) || !can_do_mlock())
 		return ERR_PTR(-EPERM);
 
 	if (!len)