On Thu, Nov 14, 2024 at 11:25:59AM +0800, jiang.kun2@xxxxxxxxxx wrote:
> From: tuqiang <tu.qiang35@xxxxxxxxxx>
>
> The MR/QP restrack also needs to be released when delete it, otherwise it
> cause memory leak as the task struct won't be released.
>
> Refer to commit dac153f2802d ("RDMA/restrack: Release MR restrack when delete").
>
> Fixes: 13ef5539def7 ("RDMA/restrack: Count references to the verbs objects")
> Signed-off-by: tuqiang <tu.qiang35@xxxxxxxxxx>
> Signed-off-by: Jiang Kun <jiang.kun2@xxxxxxxxxx>
> Cc: xu xin <xu.xin16@xxxxxxxxxx>
> Cc: Doug Ledford <dledford@xxxxxxxxxx>
> Cc: Jason Gunthorpe <jgg@xxxxxxxx>
> Cc: Leon Romanovsky <leon@xxxxxxxxxx>
> ---
> drivers/infiniband/core/restrack.c | 2 --
> 1 file changed, 2 deletions(-)
This is not proper way to submit patches to stable@.
Please read and Documentation/process/stable-kernel-rules.rst
Thanks
>
> diff --git a/drivers/infiniband/core/restrack.c b/drivers/infiniband/core/restrack.c
> index bbbbec5b1593..d5a69c4a1891 100644
> --- a/drivers/infiniband/core/restrack.c
> +++ b/drivers/infiniband/core/restrack.c
> @@ -326,8 +326,6 @@ void rdma_restrack_del(struct rdma_restrack_entry *res)
> rt = &dev->res[res->type];
>
> old = xa_erase(&rt->xa, res->id);
> - if (res->type == RDMA_RESTRACK_MR || res->type == RDMA_RESTRACK_QP)
> - return;
> WARN_ON(old != res);
> res->valid = false;
>
> --
> 2.18.4
