Right, I did not suggest this as a patch but just wanted to pinpoint the problematic issue which is that virt_to_page does not give you the correct pointer to the page.
-----Original Message----- From: Christoph Hellwig [mailto:h...@infradead.org] Sent: Monday, April 18, 2016 9:33 AM To: Eli Cohen <e...@mellanox.com> Cc: Sinan Kaya <ok...@codeaurora.org>; linux-r...@vger.kernel.org; ti...@codeaurora.org; c...@codeaurora.org; Yishai Hadas <yish...@mellanox.com>; netdev@vger.kernel.org; linux-ker...@vger.kernel.org Subject: Re: [PATCH V2] net: ethernet: mellanox: correct page conversion On Mon, Apr 18, 2016 at 09:54:47AM +0300, Eli Cohen wrote: > Sinan, > > if we get rid of the part this code: > > if (BITS_PER_LONG == 64) { > struct page **pages; > pages = kmalloc(sizeof *pages * buf->nbufs, gfp); > if (!pages) > goto err_free; > ... > ... > if (!buf->direct.buf) > goto err_free; > } > > Does that solve the arm issue? Not quite. You still have code in mlx4_en_map_buffer that performs this mapping later if it it wasn't mapped in mlx4_buf_alloc. You'll need to get rid of that by ensuring max_direct for all the cases currently using mlx4_en_map_buffer as well.
