01:35 <haier> Hi, I when I compile my own linux kernel (instead of using the precompiled one) for linux-on-litex-vexriscv, I get a kernel panic at boot because it cannot read the scratch register from the soc controller. I placed a couple of printk's in litex_soc_ctrl.c and found out that the resource returned by platform_get_resource has correct name (

01:35 <haier> soc_controller@f0000000)and start address (0xf0000000), but an end address of res->end == 0x0, although the dts file specifies it should be 0xf000000c. And this is why I suspect the kernel panic (cant read scratch register) happens, because nothing of the IO space is actually mapped. Has anyone an idea of how to fix this? The build/sim/sim.dts reg

01:35 <haier> entry is correct: reg = <0xf0000000 0xc>;

11:21 <dkozel> _florent_: Wow. How are you connecting the SATA drive to the CLE?

11:22 <dkozel> That looks like one of the generic USB to PCIe bitcoin carriers

11:24 <dkozel> teknoman117: the LiteFury is the same PCB (as far as I know) to the Acorn CLE-215 which is also in the litex-boards repo

11:24 <dkozel> I'm running both the Aller and CLE-215+ with litepcie

11:56 <_florent_> teknoman117: hi, sorry i haven't seen your message here and answer on the issue, I'll provide an example for the Acorn/LiteFury with 128-bit data with PCIe in the next days

11:58 <_florent_> dkozel: yes with a modded PCIe riser, bitcoin mining is very useful to get cheap FPGA dev boards/hardware :)

17:50 <teknoman117> you're right, it does look like the same PCB. I didn't realize.