On 04-02-24, 07:59, Andrea della Porta wrote: > This patchset aims to update the dma engine for BCM* chipset with respect > to current advancements in downstream vendor tree. In particular: > > * Added support for BCM2712 DMA. > * Extended DMA addressing to 40 bit. Since BCM2711 also supports 40 bit addressing, > it will also benefit from the update. > * Handled the devicetree node from vendor dts (e.g. "dma40"). > > The only difference between the application of this patch and the relative code > in vendor tree is the dropping of channel reservation for BCM2708 DMA legacy > driver, that seems to have not made its way to upstream anyway, and it's > probably used only from deprecated subsystems. > > Compile tested and runtime tested on RPi4B only. > > Dom Cobley (4): > bcm2835-dma: Support dma flags for multi-beat burst > bcm2835-dma: Need to keep PROT bits set in CS on 40bit controller > dmaengine: bcm2835: Rename to_bcm2711_cbaddr to to_40bit_cbaddr > bcm2835-dma: Fixes for dma_abort > > Maxime Ripard (2): > dmaengine: bcm2835: Use to_bcm2711_cbaddr where relevant > dmaengine: bcm2835: Support DMA-Lite channels > > Phil Elwell (6): > bcm2835-dma: Add support for per-channel flags > bcm2835-dma: Add proper 40-bit DMA support > bcm2835-dma: Add NO_WAIT_RESP, DMA_WIDE_SOURCE and DMA_WIDE_DEST flag > bcm2835-dma: Advertise the full DMA range > bcm2835-dma: Derive slave DMA addresses correctly > dmaengine: bcm2835: Add BCM2712 support > > drivers/dma/bcm2835-dma.c | 701 ++++++++++++++++++++++++++++++++------ > 1 file changed, 588 insertions(+), 113 deletions(-) Everything is modifying one file and still you have 2 different tags for patches, why? Consistency is a good thing, right... -- ~Vinod
