On Tuesday 23 September 2014 17:18:43 Jingchang Lu wrote:
> I will reconsider this setting. BTW, why writel() can't be used
> for memory location since it's still mapped and registers space is
> also memory mapped? Thanks.
You cannot rely on an '__iomem' token to actually be a pointer, even
if tha
-edma: fixup reg offset and hw S/G support
in big-endian model
On Tue, Sep 23, 2014 at 05:15:19PM +0800, Jingchang Lu wrote:
> static u16 edma_readw(struct fsl_edma_engine *edma, void __iomem *addr)
> {
> - if (edma->big_endian)
> - return ioread16be(addr);
> -
reg offset and hw S/G support
in big-endian model
On Tuesday 23 September 2014 17:15:19 Jingchang Lu wrote:
> @@ -459,20 +480,27 @@ static void fill_tcd_params(struct fsl_edma_engine
> *edma,
> u16 csr = 0;
>
> /*
> -* eDMA hardware SGs require the TCD pa
On Tuesday 23 September 2014 17:15:19 Jingchang Lu wrote:
> @@ -459,20 +480,27 @@ static void fill_tcd_params(struct fsl_edma_engine
> *edma,
> u16 csr = 0;
>
> /*
> -* eDMA hardware SGs require the TCD parameters stored in memory
> -* the same endian as the eDMA
On Tue, Sep 23, 2014 at 05:15:19PM +0800, Jingchang Lu wrote:
> static u16 edma_readw(struct fsl_edma_engine *edma, void __iomem *addr)
> {
> - if (edma->big_endian)
> - return ioread16be(addr);
> - else
> + u32 dst;
This should be unsigned long, if it needs to exist.
>
The offset of all 8-/16-bit register in big-endian eDMA model are
swapped in a 32-bit size opposite those in the little-endian model.
The hardware Scatter/Gather requires the subsequent TCDs in memory
to be auto loaded should retain the same endian as the core independent
of the register endian mo
6 matches
Mail list logo
