Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-03 Thread Or Gerlitz
On Tue, May 3, 2016, David Miller wrote: > From: Or Gerlitz >> The patch changes the driver to do single allocation for potentially >> very large HW WQE descriptor buffers such as those used by the RDMA >> (mlx5_ib) driver. > I know exactly what this

Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-03 Thread David Miller
From: Haggai Abramovsky Date: Wed, 27 Apr 2016 10:07:00 +0300 > @@ -419,7 +419,8 @@ static int set_rq_size(struct mlx4_ib_dev *dev, struct > ib_qp_cap *cap, > } > > static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, > -

Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-03 Thread David Miller
From: Or Gerlitz Date: Tue, 3 May 2016 10:45:43 +0300 > The patch changes the driver to do single allocation for potentially > very large HW WQE descriptor buffers such as those used by the RDMA > (mlx5_ib) driver. I know exactly what this patch does and how, but thanks

Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-03 Thread Or Gerlitz
On Wed, Apr 27, 2016 at 10:07 AM, Haggai Abramovsky wrote: > The dma_alloc_coherent() function returns a virtual address which can > be used for coherent access to the underlying memory. On some > architectures, like arm64, undefined behavior results if this memory is > also

Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-01 Thread David Miller
From: Sinan Kaya Date: Sun, 1 May 2016 16:12:13 -0400 > I just wanted to make sure that this patch is moving in the right > direction. You never need to wonder about this. If the patch is in an active state in the patchwork queue, which this one is, there is absolutely

Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-01 Thread Sinan Kaya
On 5/1/2016 4:07 PM, David Miller wrote: > From: Sinan Kaya > Date: Sun, 1 May 2016 11:43:54 -0400 > >> Can we get this queued for 4.7? Mellanox with arm64 has been broken over 1.5 >> years. > > Can you not quote an entire huge patch just to say something like > this? >

Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-01 Thread David Miller
From: Sinan Kaya Date: Sun, 1 May 2016 11:43:54 -0400 > Can we get this queued for 4.7? Mellanox with arm64 has been broken over 1.5 > years. Can you not quote an entire huge patch just to say something like this? I'm waiting for the people involved all to give feedback.

Re: [PATCH v1 net] net/mlx4: Avoid wrong virtual mappings

2016-05-01 Thread Sinan Kaya
On 4/27/2016 3:07 AM, Haggai Abramovsky wrote: > The dma_alloc_coherent() function returns a virtual address which can > be used for coherent access to the underlying memory. On some > architectures, like arm64, undefined behavior results if this memory is > also accessed via virtual mappings