Hello, On Mon, Feb 29, 2016 at 02:44:16PM +0100, Geert Uytterhoeven wrote: > sata_rcar ee300000.sata: DMA-API: cpu touching an active dma mapped > cacheline [cln=0x000000000090ff40] > ------------[ cut here ]------------ > WARNING: at lib/dma-debug.c:604 > Modules linked in: > > CPU: 2 PID: 329 Comm: udevd Not tainted ... > [<ffffff8004222b00>] debug_dma_assert_idle+0x180/0x1d0 > [<ffffff800412b0fc>] wp_page_copy.isra.109+0xfc/0x57c > [<ffffff800412f118>] handle_mm_fault+0x125c/0x1300 > [<ffffff8004092894>] do_page_fault+0x184/0x328 > [<ffffff8004082250>] do_mem_abort+0x40/0x9c So, somehow udev has mmapped a page which is dma mapped and trying to falut that in? > Mapped at: > [<ffffff8004220a08>] dma_entry_alloc+0x98/0xac > [<ffffff8004222390>] debug_dma_alloc_coherent+0x48/0xe0 > [<ffffff80042926d8>] dmam_alloc_coherent+0xa8/0xf8 > [<ffffff80042c77dc>] ata_bmdma_port_start+0x40/0x5c > [<ffffff80042b6974>] ata_host_start+0xf8/0x19c And that region was originally allocated by ata_bmdma_port_start()? The only thing which gets allocated there is the PRD table which is the scatter-gather list for the DMA engine and has no reason to be ever accessed from userland. Any chance you can find out what udev is doing at that point? Thanks. -- tejun -- To unsubscribe from this list: send the line "unsubscribe dmaengine" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html