Jason has been asking for patches to pass clang-format-patch so I've
been cleaning up the code near functional changes since it doesn't
like extra spaces such as for vertical alignment.
If I could figure out how ib_umem_works there is a chance that it
would fail if it couldn't map all the user space virtual memory into
kernel virtual addresses. But so far I have failed. It's fairly
complex.
Bob
On Mon, Jul 5, 2021 at 3:35 AM Zhu Yanjun <zyjzyj2000@xxxxxxxxx> wrote:
>
> On Mon, Jul 5, 2021 at 4:16 PM Haakon Bugge <haakon.bugge@xxxxxxxxxx> wrote:
> >
> >
> >
> > > On 5 Jul 2021, at 05:42, Zhu Yanjun <zyjzyj2000@xxxxxxxxx> wrote:
> > >
> > > On Mon, Jul 5, 2021 at 6:37 AM Bob Pearson <rpearsonhpe@xxxxxxxxx> wrote:
> > >>
> > >> In rxe_mr_init_user() in rxe_mr.c at the third error the driver fails to
> > >> free the memory at mr->map. This patch adds code to do that.
> > >> This error only occurs if page_address() fails to return a non zero address
> > >> which should never happen for 64 bit architectures.
> > >
> > > If this will never happen for 64 bit architectures, is it possible to
> > > exclude 64 bit architecture with some MACROs or others?
> > >
> > > Thanks,
> > >
> > > Zhu Yanjun
> > >
> > >>
> > >> Fixes: 8700e3e7c485 ("Soft RoCE driver")
> > >> Reported by: Haakon Bugge <haakon.bugge@xxxxxxxxxx>
> > >> Signed-off-by: Bob Pearson <rpearsonhpe@xxxxxxxxx>
> > >> ---
> > >> drivers/infiniband/sw/rxe/rxe_mr.c | 41 +++++++++++++++++-------------
> > >> 1 file changed, 24 insertions(+), 17 deletions(-)
> > >>
> > >> diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c
> > >> index 6aabcb4de235..f49baff9ca3d 100644
> > >> --- a/drivers/infiniband/sw/rxe/rxe_mr.c
> > >> +++ b/drivers/infiniband/sw/rxe/rxe_mr.c
> > >> @@ -106,20 +106,21 @@ void rxe_mr_init_dma(struct rxe_pd *pd, int access, struct rxe_mr *mr)
> > >> int rxe_mr_init_user(struct rxe_pd *pd, u64 start, u64 length, u64 iova,
> > >> int access, struct rxe_mr *mr)
> > >> {
> > >> - struct rxe_map **map;
> > >> - struct rxe_phys_buf *buf = NULL;
> > >> - struct ib_umem *umem;
> > >> - struct sg_page_iter sg_iter;
> > >> - int num_buf;
> > >> - void *vaddr;
> > >> + struct rxe_map **map;
> > >> + struct rxe_phys_buf *buf = NULL;
> > >> + struct ib_umem *umem;
> > >> + struct sg_page_iter sg_iter;
> > >> + int num_buf;
> > >> + void *vaddr;
> >
> > This white-space stripping must be another issue, not related to the memleak?
> >
> > >> int err;
> > >> + int i;
> > >>
> > >> umem = ib_umem_get(pd->ibpd.device, start, length, access);
> > >> if (IS_ERR(umem)) {
> > >> - pr_warn("err %d from rxe_umem_get\n",
> > >> - (int)PTR_ERR(umem));
> > >> + pr_warn("%s: Unable to pin memory region err = %d\n",
> > >> + __func__, (int)PTR_ERR(umem));
> > >> err = PTR_ERR(umem);
> > >> - goto err1;
> > >> + goto err_out;
> > >> }
> > >>
> > >> mr->umem = umem;
> > >> @@ -129,15 +130,15 @@ int rxe_mr_init_user(struct rxe_pd *pd, u64 start, u64 length, u64 iova,
> > >>
> > >> err = rxe_mr_alloc(mr, num_buf);
> > >> if (err) {
> > >> - pr_warn("err %d from rxe_mr_alloc\n", err);
> > >> - ib_umem_release(umem);
> > >> - goto err1;
> > >> + pr_warn("%s: Unable to allocate memory for map\n",
> > >> + __func__);
> > >> + goto err_release_umem;
> > >> }
> > >>
> > >> mr->page_shift = PAGE_SHIFT;
> > >> mr->page_mask = PAGE_SIZE - 1;
> > >>
> > >> - num_buf = 0;
> > >> + num_buf = 0;
> >
> > White-space change.
>
> Yeah. It seems that some white-space changes in this commit.
>
> Zhu Yanjun
>
> >
> > Otherwise:
> >
> > Reviewed-by: Håkon Bugge <haakon.bugge@xxxxxxxxxx>
> >
> >
> > Thxs, Håkon
> >
> >
> >
> > >> map = mr->map;
> > >> if (length > 0) {
> > >> buf = map[0]->buf;
> > >> @@ -151,10 +152,10 @@ int rxe_mr_init_user(struct rxe_pd *pd, u64 start, u64 length, u64 iova,
> > >>
> > >> vaddr = page_address(sg_page_iter_page(&sg_iter));
> > >> if (!vaddr) {
> > >> - pr_warn("null vaddr\n");
> > >> - ib_umem_release(umem);
> > >> + pr_warn("%s: Unable to get virtual address\n",
> > >> + __func__);
> > >> err = -ENOMEM;
> > >> - goto err1;
> > >> + goto err_cleanup_map;
> > >> }
> > >>
> > >> buf->addr = (uintptr_t)vaddr;
> > >> @@ -177,7 +178,13 @@ int rxe_mr_init_user(struct rxe_pd *pd, u64 start, u64 length, u64 iova,
> > >>
> > >> return 0;
> > >>
> > >> -err1:
> > >> +err_cleanup_map:
> > >> + for (i = 0; i < mr->num_map; i++)
> > >> + kfree(mr->map[i]);
> > >> + kfree(mr->map);
> > >> +err_release_umem:
> > >> + ib_umem_release(umem);
> > >> +err_out:
> > >> return err;
> > >> }
> > >>
> > >> --
> > >> 2.30.2
> > >>
> >
