Re: [PATCH for-next 09/12] RDMA/bnxt_re: Use GFP_KERNEL in non atomic context

On Mon, Sep 13, 2021 at 4:41 PM Leon Romanovsky [off-list ref] wrote:

On Sun, Sep 12, 2021 at 11:15:23AM -0700, Selvin Xavier wrote:

quoted

Use GFP_KERNEL instead of GFP_ATOMIC while allocating
control path structures which will be only called from
non atomic context

Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
---
 drivers/infiniband/hw/bnxt_re/qplib_rcfw.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c
index 947e8c5..3de8547 100644
--- a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c
+++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c

@@ -848,13 +848,13 @@ struct bnxt_qplib_rcfw_sbuf *bnxt_qplib_rcfw_alloc_sbuf(
 {
      struct bnxt_qplib_rcfw_sbuf *sbuf;

-     sbuf = kzalloc(sizeof(*sbuf), GFP_ATOMIC);
+     sbuf = kzalloc(sizeof(*sbuf), GFP_KERNEL);
      if (!sbuf)
              return NULL;

I think that you can do same change in bnxt_re_netdev_event() too.

Sure.. thanks for pointing it out.

quoted

      sbuf->size = size;
      sbuf->sb = dma_alloc_coherent(&rcfw->pdev->dev, sbuf->size,
-                                   &sbuf->dma_addr, GFP_ATOMIC);
+                                   &sbuf->dma_addr, GFP_KERNEL);
      if (!sbuf->sb)
              goto bail;

--
2.5.5