Re: IB: INFINIBAND should depend on HAS_DMA
From: Jason Gunthorpe
Date: Fri Dec 01 2017 - 19:01:37 EST
On Sun, Nov 19, 2017 at 07:58:30PM +0100, Geert Uytterhoeven wrote:
> If NO_DMA=y:
>
> ERROR: "bad_dma_ops" [net/sunrpc/xprtrdma/rpcrdma.ko] undefined!
> ERROR: "bad_dma_ops" [net/smc/smc.ko] undefined!
> ERROR: "bad_dma_ops" [net/rds/rds_rdma.ko] undefined!
> ERROR: "bad_dma_ops" [net/9p/9pnet_rdma.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/nvme/target/nvmet-rdma.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/nvme/host/nvme-rdma.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/infiniband/ulp/srpt/ib_srpt.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/infiniband/ulp/srp/ib_srp.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/infiniband/ulp/isert/ib_isert.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/infiniband/ulp/ipoib/ib_ipoib.ko] undefined!
> ERROR: "bad_dma_ops" [drivers/infiniband/core/ib_core.ko] undefined!
>
> Before, this was handled implicitly by the dependency on PCI.
> Add an explicit dependency on HAS_DMA to fix this.
>
> Fixes: 931bc0d91639f8fb ("IB: Move PCI dependency from root KConfig to HW's KConfigs")
> Signed-off-by: Geert Uytterhoeven <geert@xxxxxxxxxxxxxx>
> Reviewed-by: Leon Romanovsky <leonro@xxxxxxxxxxxx>
> drivers/infiniband/Kconfig | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
Applied to for-rc, thanks
Jason