Embedding net_device into structures prohibits the usage of flexible
arrays in the net_device structure. For more details, see the discussion
at [1].
Un-embed the net_device from struct hfi1_netdev_rx by converting it
into a pointer. Then use the leverage alloc_netdev() to allocate the
net_device object at hfi1_alloc_rx().
[1] https://lore.kernel.org/all/[email protected]/
Acked-by: Dennis Dalessandro <[email protected]>
Signed-off-by: Breno Leitao <[email protected]>
---
Changelog
v5:
* Basically replaced the old alloc_netdev() by the new helper
alloc_netdev_dummy().
v4:
* Fix the changelog format
v3:
* Re-worded the comment, by removing the first paragraph.
v2:
* Free struct hfi1_netdev_rx allocation if alloc_netdev() fails
* Pass zero as the private size for alloc_netdev().
* Remove wrong reference for iwl in the comments
---
drivers/infiniband/hw/hfi1/netdev.h | 2 +-
drivers/infiniband/hw/hfi1/netdev_rx.c | 9 +++++++--
2 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/infiniband/hw/hfi1/netdev.h b/drivers/infiniband/hw/hfi1/netdev.h
index 8aa074670a9c..07c8f77c9181 100644
--- a/drivers/infiniband/hw/hfi1/netdev.h
+++ b/drivers/infiniband/hw/hfi1/netdev.h
@@ -49,7 +49,7 @@ struct hfi1_netdev_rxq {
* When 0 receive queues will be freed.
*/
struct hfi1_netdev_rx {
- struct net_device rx_napi;
+ struct net_device *rx_napi;
struct hfi1_devdata *dd;
struct hfi1_netdev_rxq *rxq;
int num_rx_q;
diff --git a/drivers/infiniband/hw/hfi1/netdev_rx.c b/drivers/infiniband/hw/hfi1/netdev_rx.c
index 720d4c85c9c9..8608044203bb 100644
--- a/drivers/infiniband/hw/hfi1/netdev_rx.c
+++ b/drivers/infiniband/hw/hfi1/netdev_rx.c
@@ -188,7 +188,7 @@ static int hfi1_netdev_rxq_init(struct hfi1_netdev_rx *rx)
int i;
int rc;
struct hfi1_devdata *dd = rx->dd;
- struct net_device *dev = &rx->rx_napi;
+ struct net_device *dev = rx->rx_napi;
rx->num_rx_q = dd->num_netdev_contexts;
rx->rxq = kcalloc_node(rx->num_rx_q, sizeof(*rx->rxq),
@@ -360,7 +360,11 @@ int hfi1_alloc_rx(struct hfi1_devdata *dd)
if (!rx)
return -ENOMEM;
rx->dd = dd;
- init_dummy_netdev(&rx->rx_napi);
+ rx->rx_napi = alloc_netdev_dummy(0);
+ if (!rx->rx_napi) {
+ kfree(rx);
+ return -ENOMEM;
+ }
xa_init(&rx->dev_tbl);
atomic_set(&rx->enabled, 0);
@@ -374,6 +378,7 @@ void hfi1_free_rx(struct hfi1_devdata *dd)
{
if (dd->netdev_rx) {
dd_dev_info(dd, "hfi1 rx freed\n");
+ free_netdev(dd->netdev_rx->rx_napi);
kfree(dd->netdev_rx);
dd->netdev_rx = NULL;
}
--
2.43.0
On Fri, Apr 26, 2024 at 01:56:05AM -0700, Breno Leitao wrote:
> Embedding net_device into structures prohibits the usage of flexible
> arrays in the net_device structure. For more details, see the discussion
> at [1].
>
> Un-embed the net_device from struct hfi1_netdev_rx by converting it
> into a pointer. Then use the leverage alloc_netdev() to allocate the
> net_device object at hfi1_alloc_rx().
>
> [1] https://lore.kernel.org/all/[email protected]/
>
> Acked-by: Dennis Dalessandro <[email protected]>
> Signed-off-by: Breno Leitao <[email protected]>
> ---
> Changelog
>
> v5:
> * Basically replaced the old alloc_netdev() by the new helper
> alloc_netdev_dummy().
> v4:
> * Fix the changelog format
> v3:
> * Re-worded the comment, by removing the first paragraph.
> v2:
> * Free struct hfi1_netdev_rx allocation if alloc_netdev() fails
> * Pass zero as the private size for alloc_netdev().
> * Remove wrong reference for iwl in the comments
> ---
>
> drivers/infiniband/hw/hfi1/netdev.h | 2 +-
> drivers/infiniband/hw/hfi1/netdev_rx.c | 9 +++++++--
> 2 files changed, 8 insertions(+), 3 deletions(-)
Dennis,
Do you plan to send anything to rdma-next which can potentially create
conflicts with netdev in this cycle?
If not, it will be safe to apply this patch directly to net-next.
Thanks
On 4/30/24 8:50 AM, Leon Romanovsky wrote:
> On Fri, Apr 26, 2024 at 01:56:05AM -0700, Breno Leitao wrote:
>> Embedding net_device into structures prohibits the usage of flexible
>> arrays in the net_device structure. For more details, see the discussion
>> at [1].
>>
>> Un-embed the net_device from struct hfi1_netdev_rx by converting it
>> into a pointer. Then use the leverage alloc_netdev() to allocate the
>> net_device object at hfi1_alloc_rx().
>>
>> [1] https://lore.kernel.org/all/[email protected]/
>>
>> Acked-by: Dennis Dalessandro <[email protected]>
>> Signed-off-by: Breno Leitao <[email protected]>
>> ---
>> Changelog
>>
>> v5:
>> * Basically replaced the old alloc_netdev() by the new helper
>> alloc_netdev_dummy().
>> v4:
>> * Fix the changelog format
>> v3:
>> * Re-worded the comment, by removing the first paragraph.
>> v2:
>> * Free struct hfi1_netdev_rx allocation if alloc_netdev() fails
>> * Pass zero as the private size for alloc_netdev().
>> * Remove wrong reference for iwl in the comments
>> ---
>>
>> drivers/infiniband/hw/hfi1/netdev.h | 2 +-
>> drivers/infiniband/hw/hfi1/netdev_rx.c | 9 +++++++--
>> 2 files changed, 8 insertions(+), 3 deletions(-)
>
> Dennis,
>
> Do you plan to send anything to rdma-next which can potentially create
> conflicts with netdev in this cycle?
>
> If not, it will be safe to apply this patch directly to net-next.
>
> Thanks
Nothing right now. Should be safe to sent to net-next.
FYI, since I talked about it publicly at the OFA Workshop recently. We will be
starting the upstream of support for our new HW, soon.
-Denny
On Tue, Apr 30, 2024 at 10:03:49AM -0400, Dennis Dalessandro wrote:
> On 4/30/24 8:50 AM, Leon Romanovsky wrote:
> > On Fri, Apr 26, 2024 at 01:56:05AM -0700, Breno Leitao wrote:
> >> Embedding net_device into structures prohibits the usage of flexible
> >> arrays in the net_device structure. For more details, see the discussion
> >> at [1].
> >>
> >> Un-embed the net_device from struct hfi1_netdev_rx by converting it
> >> into a pointer. Then use the leverage alloc_netdev() to allocate the
> >> net_device object at hfi1_alloc_rx().
> >>
> >> [1] https://lore.kernel.org/all/[email protected]/
> >>
> >> Acked-by: Dennis Dalessandro <[email protected]>
> >> Signed-off-by: Breno Leitao <[email protected]>
> >> ---
> >> Changelog
> >>
> >> v5:
> >> * Basically replaced the old alloc_netdev() by the new helper
> >> alloc_netdev_dummy().
> >> v4:
> >> * Fix the changelog format
> >> v3:
> >> * Re-worded the comment, by removing the first paragraph.
> >> v2:
> >> * Free struct hfi1_netdev_rx allocation if alloc_netdev() fails
> >> * Pass zero as the private size for alloc_netdev().
> >> * Remove wrong reference for iwl in the comments
> >> ---
> >>
> >> drivers/infiniband/hw/hfi1/netdev.h | 2 +-
> >> drivers/infiniband/hw/hfi1/netdev_rx.c | 9 +++++++--
> >> 2 files changed, 8 insertions(+), 3 deletions(-)
> >
> > Dennis,
> >
> > Do you plan to send anything to rdma-next which can potentially create
> > conflicts with netdev in this cycle?
> >
> > If not, it will be safe to apply this patch directly to net-next.
> >
> > Thanks
>
> Nothing right now. Should be safe to sent to net-next.
Jakub, can you please take this patch?
>
> FYI, since I talked about it publicly at the OFA Workshop recently. We will be
> starting the upstream of support for our new HW, soon.
Great, thanks
>
> -Denny
On Tue, 30 Apr 2024 17:10:39 +0300 Leon Romanovsky wrote:
> > Nothing right now. Should be safe to sent to net-next.
>
> Jakub, can you please take this patch?
We'll need a repost, it wasn't CCed to netdev.
On Tue, Apr 30, 2024 at 07:55:34AM -0700, Jakub Kicinski wrote:
> On Tue, 30 Apr 2024 17:10:39 +0300 Leon Romanovsky wrote:
> > > Nothing right now. Should be safe to sent to net-next.
> >
> > Jakub, can you please take this patch?
>
> We'll need a repost, it wasn't CCed to netdev.
I can repost it and CC netdev.