Facing issue while loading bare metal elf file through gdb debugger with openocd + gdb over virtual Jtag setup on actual fpga emulator on RISCV soc.

Question

Facing issue while loading bare metal elf file through gdb debugger with openocd + gdb over virtual Jtag setup on actual fpga emulator on RISCV soc.

Closed this issue 9 months ago · 1 comments

tushar1277 commented 9 months ago

Background Work

Yes, I searched the mailing list
Yes, I searched prior issues
Yes, I searched the documentation

Chipyard Version and Hash

Release: [1.11.0]
Hash: [ac58f38]

OS Setup

Fpga emulator : Federa linux OS with kernel 4.18.0 and arch of type x86_64.
rtl generated with chipyard source release 1.11.0 on Ubuntu 22.04.
Generated rtl is being used to design custom riscv based soc which is running on actual fpga emulator.
I have also attached .dts file after rtl generation.
For debugging we have virtual jtag + openocd +gdb setup to debug the soc on fpga emulator

Other Setup

NA

Current Behavior

From the logs we can see the connection between virtual jtag + openocd and gdb is done successfully. but when the moment I try to load the .elf file using gdb using load command then I am seeing lot of timeout issues on openocd setup. attaching the generated rtl dts file in chipyard repo.

chipyard.harness.TestHarness.RocketConfig.txt

please suggest what could be the possible root cause of the issue here ?

GDB logs
(gdb) tar ext:3333
Remote debugging using :3333
0x0000000000010040 in ?? ()
(gdb) monitor targets
TargetName Type Endian TapName State

0* riscv.cpu riscv little riscv.cpu halted

(gdb) load
Loading section .text.init, size 0x204 lma 0x80000000
Loading section .tohost, size 0x48 lma 0x80001000
Loading section .text, size 0x3c4c lma 0x80001048
Ignoring packet error, continuing...
Loading section .text.startup, size 0x14 lma 0x80004c94
Load failed
(gdb)

openocd logs

Info : Examined RISC-V core; found 1 harts
Info : hart 0: XLEN=64, misa=0x800000000094112d
Info : starting gdb server for riscv.cpu on 3333
Info : Listening on port 3333 for gdb connections
Info : accepting 'gdb' connection on tcp/3333
TargetName Type Endian TapName State

0* riscv.cpu riscv little riscv.cpu halted

Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001008). Increase the timeout with riscv set_command_timeout_sec.
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Warn : keep_alive() was not invoked in the 1000 ms timelimit. GDB alive packet not sent! (8222 ms). Workaround: increase "set remotetimeout" in GDB
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Error: Failed to read priv register.
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Warn : negative acknowledgment, but no packet pending
Warn : negative acknowledgment, but no packet pending
Warn : negative acknowledgment, but no packet pending
Warn : keep_alive() was not invoked in the 1000 ms timelimit. GDB alive packet n
Warn : keep_alive() was not invoked in the 1000 ms timelimit. GDB alive packet not sent! (8222 ms). Workaround: increase "set remotetimeout" in GDB
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Error: Failed to read priv register.
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Warn : negative acknowledgment, but no packet pending
Warn : negative acknowledgment, but no packet pending
Warn : negative acknowledgment, but no packet pending
Warn : keep_alive() was not invoked in the 1000 ms timelimit. GDB alive packet not sent! (5999 ms). Workaround: increase "set remotetimeout" in GDB
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Error: Failed to read priv register.
Error: Abstract command ended in error 'busy' (abstractcs=0x10001108)
Error: Timed out after 2s waiting for busy to go low (abstractcs=0x10001108). Increase the timeout with riscv set_command_timeout_sec.
Warn : negative acknowledgment, but no packet pending
Warn : keep_alive() was not invoked in the 1000 ms timelimit. GDB alive packet not sent! (7006 ms). Workaround: increase "set remotetimeout" in GDB
Warn : keep_alive() was not invoked in the 1000 ms timelimit. GDB alive packet n

Expected Behavior

elf file loading at memory (0x80000000) through gdb should be done successfully without any error/warning from openocd.

Other Information

No response

Answer 1 · 2024-06-19T06:12:07.000Z

The issue was inside our custom generated rtl hence closing this bug