Hello. On 04-01-2011 14:47, Mian Yousaf Kaukab wrote:
Gadget MUSB driver handles dma mappings in musb_gadget_queue(). Where as it is possible for dma code to reject the usb request later at ->channel_program() called from txstate()/rxstate()
For example ->channel_program in tusb6010_omap.c:
static int tusb_omap_dma_program(struct dma_channel *channel, u16 packet_sz, u8 rndis_mode, dma_addr_t dma_addr, u32 len) { ... if (unlikely(dma_addr & 0x1) || (len < 32) || (len > packet_sz)) return false; ... if (dma_addr & 0x2) return false; ... }
In this case, usb request will be handled in PIO mode which renders dma mapping operations unnecessary.
This patch adds an api to allow dma code to indicate incompatibility with usb request. Gadget musb driver call this api, if available, before dma mappings to avoid any unnecessary mapping operations.
Signed-off-by: Mian Yousaf Kaukab<mian.yousaf.kaukab@xxxxxxxxxxxxxx> --- This patch is based on Greg's usb-next branch
Changelog from previous version: - API addition and use is combined into one patch - Arguments for is_compatable() are now similar to channel_program()
Althought the arguments for is_compatible() are now valid for both host and gadget modes, there is currently no use of this API in host mode. As Sergei mentioned, host mode buffers are mapped by the higher layers. Suggestions are welcome on how to avoid unnecessary dma mappings for the host side.
drivers/usb/musb/musb_dma.h | 3 +++ drivers/usb/musb/musb_gadget.c | 14 ++++++++++++++ 2 files changed, 17 insertions(+), 0 deletions(-)
diff --git a/drivers/usb/musb/musb_dma.h b/drivers/usb/musb/musb_dma.h index 916065b..3a97c4e 100644 --- a/drivers/usb/musb/musb_dma.h +++ b/drivers/usb/musb/musb_dma.h @@ -169,6 +169,9 @@ struct dma_controller { dma_addr_t dma_addr, u32 length); int (*channel_abort)(struct dma_channel *); + int (*is_compatible)(struct dma_channel *channel, + u16 maxpacket, + void *buf, u32 length);
Are you going to actually implement this method? WBR, Sergei -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html