Problem with "Cleanup dma device in spl and move dma channel[0]" series and am64

[Santhosh Kumar K]

Hi, Tom,

On 30/12/24 23:12, Tom Rini wrote:
> Hey all,
> 
> As part of testing the SPI fixes, I tracked down a problem on my AM64X
> SR1.0 GP to the above series. With the series applied, I get:
> => sf probe
> jedec_spi_nor flash at 0: non-uniform erase sector maps are not supported yet.
> SF: Detected s28hs512t with page size 256 Bytes, erase size 256 KiB, total 64 MiB
> => sf read $loadaddr 0 4000
> device 0 offset 0x0, size 0x4000
> ti-udma dma-controller at 485c0000: k3_dmaring Ring probed rings:288,
> sci-dev-id:30
> ti-udma dma-controller at 485c0000: dma-ring-reset-quirk: disabled
> ............................... (never ends).
> 
> With the series reverted:
> => sf probe
> SF: Detected s28hs512t with page size 256 Bytes, erase size 256 KiB, total 64 MiB
> => sf read $loadaddr 0 4000
> device 0 offset 0x0, size 0x4000
> SF: 16384 bytes @ 0x0 Read: OK
> =>

A fix for this issue has already been provided (not merged yet).

Patch: 
https://lore.kernel.org/u-boot/20241218130045.2070314-1-p-mantena@ti.com/

With this fix:
=> sf probe
SF: Detected s28hs512t with page size 256 Bytes, erase size 256 KiB, 
total 64 MiB
=> sf read $loadaddr 0 4000
device 0 offset 0x0, size 0x4000
SF: 16384 bytes @ 0x0 Read: OK
=>

Regards,
Santhosh.

>