We have a setup where two Xilinx FPGA devices are individually connected to a computer. We use Vivado Tcl Console to program the FPGAs in our automation environment. I am able to successfully program the two devices one after the other. However, I am trying to reduce the programming time by launching two vivado instances in parallel and program each device simultaneously. While doing so, I get errors like below:
Targets(s) ", jsn-JTAG-SMT2-XXXXXXXXjsn-JTAG-HS3-XXXXXXXX" may be locked by another hw_server.
Error Message: ERROR: [Common 17-39] 'get_hw_targets' failed due to earlier errors.
Here is my python script that launches hw_server and cs_server on different ports and programs FPGAs parallely.
def program(hw_server_cmd, cs_server_cmd, cmd):
cs_server_process = subprocess.Popen(cs_server_cmd, stdin=subprocess.PIPE, stdout=subprocess.PIPE, encoding='utf8')
cs_server_process.wait()
print("cs_server started")
hw_server_process = subprocess.Popen(hw_server_cmd, stdin=subprocess.PIPE, stdout=subprocess.PIPE, encoding='utf8')
hw_server_process.wait()
print("hw_server started")
p = subprocess.Popen(cmd, stdin=subprocess.PIPE, stdout=subprocess.PIPE, encoding='utf8')
p.wait()
stdout, stderr = p.communicate()
p.wait()
hw_server_process.kill()
cs_server_process.kill()
print('Command [{}] return value: {}'.format(cmd, p.returncode))
print('stdout for cmd [{}]: {}'.format(cmd, stdout))
print('stderr for cmd [{}]: {}'.format(cmd, stderr))
if __name__ == '__main__':
cs_server_cmd_1 = r"C:\Xilinx\Vivado\2020.1\bin\cs_server.bat -d -stcp::3041"
hw_server_cmd_1 = r"C:\Xilinx\Vivado\2020.1\bin\hw_server.bat -d -stcp::3121"
cmd_1 = r'C:\Xilinx\Vivado\2020.1\bin\vivado.bat -mode tcl -source ProgFPGA_1.tcl'
cs_server_cmd_2 = r"C:\Xilinx\Vivado\2020.1\bin\cs_server.bat -d -stcp::3042"
hw_server_cmd_2 = r"C:\Xilinx\Vivado\2020.1\bin\hw_server.bat -d -stcp::3122"
cmd_2 = r'C:\Xilinx\Vivado\2020.1\bin\vivado.bat -mode tcl -source ProgFPGA_2.tcl'
fpga_prog_threads = {}
fpga_prog_threads[0] = threading.Thread(name=f'Programming FPGA on Platform 1', target=program, args=[hw_server_cmd_1, cs_server_cmd_1, cmd_1])
fpga_prog_threads[1] = threading.Thread(name=f'Programming FPGA on Platform 2', target=program, args=[hw_server_cmd_2, cs_server_cmd_2, cmd_2])
for t in fpga_prog_threads.values():
t.start()
#t.join()
join_all_threads(list(fpga_prog_threads.values()), 3600)
Here is one of the tcl script I used. The other file also contains same set of commands but different port numbers and Serial number.
open_hw_manager
connect_hw_server -url 127.0.0.1:3121 -cs_url 127.0.0.1:3041 -allow_non_jtag
current_hw_server 127.0.0.1:3121
current_hw_target [get_hw_targets */xilinx_tcf/Digilent/XXXXXXXXXX]
open_hw_target
current_hw_device [lindex [get_hw_devices] 0]
refresh_hw_device -update_hw_probes false [lindex [get_hw_devices] 0]
set_property PROBES.FILE {} [lindex [get_hw_devices] 0]
set_property FULL_PROBES.FILE {} [lindex [get_hw_devices] 0]
set_property PROGRAM.FILE {C:/temp/bitfile2.bit} [lindex [get_hw_devices] 0]
program_hw_devices [lindex [get_hw_devices] 0]
refresh_hw_device [lindex [get_hw_devices] 0]
exit
Could someone point out what I am doing wrong here that is stopping me from programming the devices simultaneously?
P.S: Initially we used Xilinx ISE. I tried to program the devices simultaneously using impact.exe. In that case one FPGA or the other always returned error message like "Pin did not go HIGH".