execute work-ver/Variane_testharness $RISCV/riscv64-unknown-elf/bin/pk hello.elf

[Shuangxie0708]

Is there an existing CVA6 bug for this?

• I have searched the existing bug issues

Bug Description

When I running User-Space Applications..I execute work-ver/Variane_testharness $RISCV/riscv64-unknown-elf/bin/pk hello.elf There is below error that failed.
But I run work-ver/Variane_testharness rv64um-v-divuw, it is successful.
Could you help me solve this problem that why I failed when running user-space applications. Thanks!

root@samsung-Precision-5820-Tower-X-Series:/home/samsung/riscv/cva6# work-ver/Variane_testharness $RISCV/riscv64-unknown-elf/bin/pk hello.elf
This emulator compiled with JTAG Remote Bitbang client. To enable, use +jtag_rbb_enable=1.
Listening on port 38919
*** [rvf_tracer] WARNING: No valid address of 'tohost' (tohost == 0x00000000000000), termination possible only by timeout or Ctrl-C!

bbl loader

/home/samsung/riscv/install_riscv/riscv64-unknown-elf/bin/pk *** FAILED *** (tohost = 2147483647) after 2000012 cycles

CPU time used: 256327.08 ms
Wall clock time passed: 256427.19 ms
root@samsung-Precision-5820-Tower-X-Series:/home/samsung/riscv/cva6#

[JeanRochCoulon]

@Shuangxie0708 uses pk, @eyssartk as you are the closest person I know to the issue, I assigned it to you. Do not hesitate to forward the "hot potato" if you think somebody else could help. The objective is to find the right person to give feedback !

[retina007]

I try to booting Linux-64 CV64A6 using Verilator or Questa simulator. is this the command to use?（work-ver/Variane_testharness $RISCV/riscv64-unknown-elf/bin/pk hello.elf）

[Shuangxie0708]

yes, you need use the command (work-ver/Variane_testharness $RISCV/riscv64-unknown-elf/bin/pk hello.elf)

[rishikakushwah16]

Hey @Shuangxie0708 , I am having the same issue, is there any solution for the problem?

work-ver/Variane_testharness $RISCV/riscv64-unknown-elf/bin/pk hello.elf
This emulator compiled with JTAG Remote Bitbang client. To enable, use +jtag_rbb_enable=1.
Listening on port 40083
*** [rvf_tracer] WARNING: No valid address of 'tohost' (tohost == 0x00000000000000), termination possible only by timeout or Ctrl-C!

bbl loader
/home/rkushwah/riscv-tools/riscv64-unknown-elf/bin/pk *** FAILED *** (tohost = 2147483647) after 2000012 cycles
CPU time used: 471939.98 ms
Wall clock time passed: 471956.68 ms.

If I remove the print statement from the hello.c file, it simulates successfully. I am not sure why it is not working with the print statement. Also it simulates with spike, so I am assuming that the proxy kernel environment is set correctly.

[yulong-lan]

I am wondering where this "FAILED" message comes from, which I have also got.

Because when you pass another command -v hello.vcd in between to dump a vcd file, as below:
work-ver/Variane_testharness -v hello.vcd $RISCV/riscv64-unknown-elf/bin/pk hello.elf

It does generate a non-empty hello.vcd file, and open it you will see all the signals, which indicates that the programme has been actually running.

Beware though! Notwithstanding a very tiny "hello" programme, the size of vcd file ends up more than 15 GB, I had to terminate the simulation halfway since it was completely running up my disk space.

[allensted]

try to use puts() instread of printf()， that works for me.

i guess the env or the core or maybe the pk can not support printf() well.

int main() {
//printf("Hello CVA6!\n");
puts("Hello CVA6!\n");
return 0;
}

wpf@ubuntu:~/Desktop/cva6$ riscv64-unknown-elf-gcc hello.c -o hello.elf
wpf@ubuntu:~/Desktop/cva6$ work-ver/Variane_testharness $RISCV/riscv64-unknown-elf/bin/pk hello.elf
This emulator compiled with JTAG Remote Bitbang client. To enable, use +jtag_rbb_enable=1.
Listening on port 40007
*** [rvf_tracer] WARNING: No valid address of 'tohost' (tohost == 0x00000000000000), termination possible only by timeout or Ctrl-C!

bbl loader
Hello CVA6!\n
/opt/riscv/riscv64-unknown-elf/bin/pk *** SUCCESS *** (tohost = 0) after 1469420 cycles
CPU time used: 44124.90 ms
Wall clock time passed: 44530.56 ms
wpf@ubuntu:~/Desktop/cva6$ 

[JeanRochCoulon]

Related to #1539

[tjt7a]

I run into the same problem; I think the issue is that the simulator won't simulate beyond 2000012 cycles, even after I set the max_cycles parameter in the makefile:
max_cycles ?= 100000000

[JeanRochCoulon]

Indeed, there is a RTL timeout implemented in rvfi_tracer.sv in SIM_FINISH variable.

[valentinThomazic]

@0ena, did you manage to solve your issue ? If you did, could you close it ?
This may help you use the proxy kernel otherwise : #1358 (comment)
Please note that the support for the pk has been dropped and simulation should be ran using cva6.py

[JeanRochCoulon]

As the discussion seems to be over, I will close the issue.