On 03/01/2017 05:07 PM, Rush, Jason A. wrote:
Dinh Nguyen wrote:
Hi Jason,
On 03/01/2017 10:38 AM, Rush, Jason A. wrote:
Add the 'cdns,trigger-address' property to the cadence QSPI device node for Altera SoC devices.
Signed-off-by: Jason A. Rush jason.rush@gd-ms.com
Changed in v2:
- renamed trigger-base to trigger-address
arch/arm/dts/socfpga.dtsi | 1 + 1 file changed, 1 insertion(+)
diff --git a/arch/arm/dts/socfpga.dtsi b/arch/arm/dts/socfpga.dtsi index 8588221e57..ac26d05722 100644 --- a/arch/arm/dts/socfpga.dtsi +++ b/arch/arm/dts/socfpga.dtsi @@ -646,6 +646,7 @@ num-cs = <4>; fifo-depth = <128>; sram-size = <128>;
trigger-address = <0x00000000>;According to the documentation I see, the indaddrtrig on SoCFPGA is at offset of 0x1C.
Yes, the indaddrtrig register is at offset 0x1C, but the 'trigger-address' property is the value written to that address.
In older versions of U-Boot (where the CQSPI worked) the value was Written into the indaddrtrig register with the following code:
((u32)plat->ahbbase & CQSPI_INDIRECTTRIGGER_ADDR_MASK)
ahbbase is 0xffa00000 and the mask is 0xFFFFF, which resulted in writing a 0x0 into the indaddrtrig register.
Later changes to the CQSPI code removed that mask, so the indaddrtrig register was programmed with 0xffa00000 which resulted in the CPU resetting whenever a read/write was performed.
You can see this same change has been incorporated into the 4.10 kernel: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/ar... https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/drivers...
Ah ok. Thanks for the explanation, and sorry for the noise.
Dinh