Closed mmueller-kaffeeschluerfercom closed 3 years ago
Hi @mmueller-kaffeeschluerfercom,
this is probably an issue in LiteX that has been fixed with https://github.com/enjoy-digital/litex/pull/962. Updating LiteX should fix it.
Ah, yes, building rgmii_test.py now works (btw, I had to comment out the line, ./litex_setup.py update
did not update the file litex/soc/cores/clock/xilinx_s6.py
)
Thanks a lot!
The problem with sdram_test.py persists, though.
The size of the BIOS probably increased a little bit. You can just remove the integrated_rom_size
configuration here: https://github.com/q3k/chubby75/blob/master/rv901t/sdram_test.py#L60 since LiteX now automatically reduces it or keep it and increase it.
Yes, thank you! One step further :-)
Unfortunately, now, I get another error:
Phase 9.8 Global Placement
ERROR:Place:543 - This design does not fit into the number of slices available
...
Phase 11.9 Local Placement Optimization
ERROR:Place:543 - This design does not fit into the number of slices available
...
ERROR:Place:120 - There were not enough sites to place all selected components.
...
ERROR:Pack:1654 - The timing-driven placement phase encountered an error.
I tried export XIL_PAR_ENABLE_LEGALIZER=1
but with same errors.
It's possible the resource usage increased a little bit and that the design is now too big.
With:
SoCSDRAM.__init__(self, platform,
cpu_type = "vexriscv",
cpu_variant = "minimal",
clk_freq = sys_clk_freq,
integrated_rom_size = 0x8000)
it's building fine.
Your first issue is also fixed with https://github.com/enjoy-digital/litex/commit/7bd06d178f33c2bdce28f648e553dcd1d87d7878.
excellent! thank you very much!
@mmueller-kaffeeschluerfercom, rgmii_test.py work? Did you try ping?
I just ran a quick test. The ping doesn't work. When I start pinging, I get a broadcast ARP request from the FPGA which is answered from my computer. But then, nothing more. Unfortunately, right now I don't have time to look at this further...
@mmueller-kaffeeschluerfercom thank you for your reply!
Has anyone an actual build with working ping?
I've been able to test successfully the LiteX-Boards design: https://github.com/litex-hub/litex-boards/blob/master/litex_boards/targets/linsn_rv901t.py with./linsn_rv901t.py --cpu-type=None --integrated-main-ram-size=0x100 --with-etherbone --build
and ping it at 192.168.1.50
.
The issue with rgmii_test.py
here is probably that the tx_delay
of the PHY now need to be specified since the default value has changed:
https://github.com/litex-hub/litex-boards/blob/master/litex_boards/targets/linsn_rv901t.py#L78
If you still have troubles, the hardware will be on my desk for the next couple of days, so feel free to ask.
Yes, this was it: just add tx_delay = 0e-9
to the PHY and ping works! Thank you very much!! I'll create a pull request.
Just installed ISE 14.7 and LiteX as described here. I can successfully build the blink example in ISE and configure the FPGA. But when trying to build sdram_test.py, I get following error:
litex/riscv64-unknown-elf-gcc-8.3.0-2019.08.0-x86_64-linux-ubuntu14/bin/../lib/gcc/riscv64-unknown-elf/8.3.0/../../../../riscv64-unknown-elf/bin/ld: bios.elf section `.rodata' will not fit in region `rom'
litex/riscv64-unknown-elf-gcc-8.3.0-2019.08.0-x86_64-linux-ubuntu14/bin/../lib/gcc/riscv64-unknown-elf/8.3.0/../../../../riscv64-unknown-elf/bin/ld: region `rom' overflowed by 840 bytes
collect2: error: ld returned 1 exit status
and when trying to build rgmii_test.py
Elaborating module <PLL_ADV(BANDWIDTH="OPTIMIZED",CLKFBOUT_MULT=6'b100101,CLKFBOUT_PHASE=0.0,CLKIN1_PERIOD=40.0,CLKIN2_PERIOD=0.0,CLKOUT0_DIVIDE=3'b111,CLKOUT0_DUTY_CYCLE=0.5,CLKOUT0_PHASE=0.0,CLK_FEEDBACK="CLKFBOUT",COMPENSATION="INTERNAL",DIVCLK_DIVIDE=1'b1,REF_JITTER=0.01,SIM_DEVICE="SPARTAN6")>. ERROR:HDLCompiler:267 - "/home/user/fpga/chubby75/rv901t/build/gateware/platform.v" Line 5398: Cannot find port PWRDWN on this module
Has anyone an idea of what I am doing wrong?