Fwd: [PATCHv2 1/1] RDMA/rxe: Disable ipv6 features when ipv6.disable set in cmdline

Fwd: [PATCHv2 1/1] RDMA/rxe: Disable ipv6 features when ipv6.disable set in cmdline

[Zhu Yanjun]

From: Zhu Yanjun <zyjzyj2000@gmail.com>

When ipv6.disable=1 is set in cmdline, ipv6 is actually disabled
in the stack. As such, the operations of ipv6 in RXE will fail.
So ipv6 features in RXE should also be disabled in RXE.

Fixes: 8700e3e7c4857 ("Soft RoCE driver")
Reported-by: Yi Zhang <yi.zhang@redhat.com>
Signed-off-by: Zhu Yanjun <zyjzyj2000@gmail.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Tested-by: Yi Zhang <yi.zhang@redhat.com>
---
V1->V2: Modify the pr_info messages
---
 drivers/infiniband/sw/rxe/rxe_net.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/infiniband/sw/rxe/rxe_net.c
b/drivers/infiniband/sw/rxe/rxe_net.c
index 0701bd1ffd1a..6ef092cb575e 100644
--- a/drivers/infiniband/sw/rxe/rxe_net.c
+++ b/drivers/infiniband/sw/rxe/rxe_net.c
@@ -72,6 +72,11 @@ static struct dst_entry *rxe_find_route6(struct
net_device *ndev,
        struct dst_entry *ndst;
        struct flowi6 fl6 = { { 0 } };

+       if (!ipv6_mod_enabled()) {
+               pr_info("IPv6 is disabled");
+               return NULL;
+       }
+
        memset(&fl6, 0, sizeof(fl6));
        fl6.flowi6_oif = ndev->ifindex;
        memcpy(&fl6.saddr, saddr, sizeof(*saddr));
@@ -608,6 +613,10 @@ static int rxe_net_ipv4_init(void)
 static int rxe_net_ipv6_init(void)
 {
 #if IS_ENABLED(CONFIG_IPV6)
+       if (!ipv6_mod_enabled()) {
+               pr_info("IPv6 is disabled");
+               return 0;
+       }

        recv_sockets.sk6 = rxe_setup_udp_tunnel(&init_net,
                                                htons(ROCE_V2_UDP_DPORT), true);
--
2.25.1

Re: Fwd: [PATCHv2 1/1] RDMA/rxe: Disable ipv6 features when ipv6.disable set in cmdline

[Jason Gunthorpe]

On Sun, Mar 07, 2021 at 12:59:07AM +0800, Zhu Yanjun wrote:
> From: Zhu Yanjun <zyjzyj2000@gmail.com>
> 
> When ipv6.disable=1 is set in cmdline, ipv6 is actually disabled
> in the stack. As such, the operations of ipv6 in RXE will fail.
> So ipv6 features in RXE should also be disabled in RXE.
> 
> Fixes: 8700e3e7c4857 ("Soft RoCE driver")
> Reported-by: Yi Zhang <yi.zhang@redhat.com>
> Signed-off-by: Zhu Yanjun <zyjzyj2000@gmail.com>
> Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
> Tested-by: Yi Zhang <yi.zhang@redhat.com>
> V1->V2: Modify the pr_info messages
>  drivers/infiniband/sw/rxe/rxe_net.c | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/drivers/infiniband/sw/rxe/rxe_net.c
> b/drivers/infiniband/sw/rxe/rxe_net.c
> index 0701bd1ffd1a..6ef092cb575e 100644
> +++ b/drivers/infiniband/sw/rxe/rxe_net.c
> @@ -72,6 +72,11 @@ static struct dst_entry *rxe_find_route6(struct
> net_device *ndev,
>         struct dst_entry *ndst;
>         struct flowi6 fl6 = { { 0 } };
> 
> +       if (!ipv6_mod_enabled()) {
> +               pr_info("IPv6 is disabled");
> +               return NULL;
> +       }

Why doesn't a similar touch of ipv6_stub in
drivers/infiniband/core/addr.c need a similar protection?

Also do not print.

Jason

Re: Fwd: [PATCHv2 1/1] RDMA/rxe: Disable ipv6 features when ipv6.disable set in cmdline

[Zhu Yanjun]

On Wed, Mar 24, 2021 at 3:53 AM Jason Gunthorpe <jgg@nvidia.com> wrote:
>
> On Sun, Mar 07, 2021 at 12:59:07AM +0800, Zhu Yanjun wrote:
> > From: Zhu Yanjun <zyjzyj2000@gmail.com>
> >
> > When ipv6.disable=1 is set in cmdline, ipv6 is actually disabled
> > in the stack. As such, the operations of ipv6 in RXE will fail.
> > So ipv6 features in RXE should also be disabled in RXE.
> >
> > Fixes: 8700e3e7c4857 ("Soft RoCE driver")
> > Reported-by: Yi Zhang <yi.zhang@redhat.com>
> > Signed-off-by: Zhu Yanjun <zyjzyj2000@gmail.com>
> > Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
> > Tested-by: Yi Zhang <yi.zhang@redhat.com>
> > V1->V2: Modify the pr_info messages
> >  drivers/infiniband/sw/rxe/rxe_net.c | 9 +++++++++
> >  1 file changed, 9 insertions(+)
> >
> > diff --git a/drivers/infiniband/sw/rxe/rxe_net.c
> > b/drivers/infiniband/sw/rxe/rxe_net.c
> > index 0701bd1ffd1a..6ef092cb575e 100644
> > +++ b/drivers/infiniband/sw/rxe/rxe_net.c
> > @@ -72,6 +72,11 @@ static struct dst_entry *rxe_find_route6(struct
> > net_device *ndev,
> >         struct dst_entry *ndst;
> >         struct flowi6 fl6 = { { 0 } };
> >
> > +       if (!ipv6_mod_enabled()) {
> > +               pr_info("IPv6 is disabled");
> > +               return NULL;
> > +       }
>
> Why doesn't a similar touch of ipv6_stub in
> drivers/infiniband/core/addr.c need a similar protection?

Got it. This patch is for RXE.

I will make a similar commit of ipv6_stub in
drivers/infiniband/core/addr.c soon.

>
> Also do not print.

I will remove print in the next patch.

Zhu Yanjun

>
> Jason