Re: [casper] Problem reading DRAM in ROACH2
Ok, so I updated the libraries and I'm having the same problem, however I notice that now I have two new registers: 'ddr3_ctrl', 'ddr3_mem'. I suppose those are the registers that I'm missing changed names. There was a name changing in the library so that it make incompatible with the katcp_wrapper script? thanks, Franco On 02/12/16 12:42, Franco wrote: No, I'm using the tutorial version. That might be the problem. However I'm reluctant to update since this might cause incompatibilities with my previous models. I suppose that if I what to use the DRAM I don't have other choice. Do you have any suggestions? Many thanks, Franco Jason Manley wrote Are you running the latest version of the toolflow? Earlier versions did not include the CPU interface to the DRAM. Jason On 02 Dec 2016, at 15:30, Franco> wrote: > The register is not present in the bitstream. I thought that the register was inside the dram yellow block though. I have to add it manually or something? > > Here is the listdev just in case: > > ['acc_len', > 'adc5g_controller', > 'cnt_rst', > 'dout0_0', > 'dout0_1', > 'dout0_2', > 'dout0_3', > 'dout1_0', > 'dout1_1', > 'dout1_2', > 'dout1_3', > 'dout2_0', > 'dout2_1', > 'dout2_2', > 'dout2_3', > 'filter_gain', > 'filter_on', > 'snapshot0_bram', > 'snapshot0_ctrl', > 'snapshot0_status', > 'snapshot0_trig_offset', > 'snapshot1_bram', > 'snapshot1_ctrl', > 'snapshot1_status', > 'snapshot1_trig_offset', > 'sync_gen_sync', > 'sync_gen_sync_period_sel', > 'sync_gen_sync_period_var', > 'sys_board_id', > 'sys_clkcounter', > 'sys_rev', > 'sys_rev_rcs', > 'sys_scratchpad', > 'write_done'] > > thanks, > > Franco > > > On 02/12/16 10:02, Marc Welz wrote: >> Then there is no bitstream programmed or the particular bitstream >> doesn't contain that register. Type >> >> ?listdev >> >> on that connection to find out >> >> regards >> >> marc >> >> >> On Fri, Dec 2, 2016 at 12:52 PM, Franco > wrote: >>> It says: #log error 1018802834854 raw >>> register\_dram_controller\_not\_defined >>> >>> >>> Franco >>> >>> >>> >>> On 02/12/16 04:32, Marc Welz wrote: telnet to the roach on port 7147 while repeating the operation - what do the #log messages say ? On Thu, Dec 1, 2016 at 3:43 PM, Franco > wrote: > Hi! > > I tried to test the dram of ROACH2 by porting an example model I had for > ROACH1. > > However when I tried to read the dram I got the following error: > > RuntimeError: Request write failed. >Request: ?write dram_controller 0 \0\0\0\0 >Reply: !write fail. > > Someone knows what's the problem? > > > thanks, > > Franco Curotto > > > >
Re: [casper] Problem reading DRAM in ROACH2
No, I'm using the tutorial version. That might be the problem. However I'm reluctant to update since this might cause incompatibilities with my previous models. I suppose that if I what to use the DRAM I don't have other choice. Do you have any suggestions? Many thanks, Franco Jason Manley wrote >Are you running the latest version of the toolflow? Earlier versions did not >include the CPU interface to the DRAM. > >Jason > > >On 02 Dec 2016, at 15:30, Francowrote: > >> The register is not present in the bitstream. I thought that the register >> was inside the dram yellow block though. I have to add it manually or >> something? >> >> Here is the listdev just in case: >> >> ['acc_len', >> 'adc5g_controller', >> 'cnt_rst', >> 'dout0_0', >> 'dout0_1', >> 'dout0_2', >> 'dout0_3', >> 'dout1_0', >> 'dout1_1', >> 'dout1_2', >> 'dout1_3', >> 'dout2_0', >> 'dout2_1', >> 'dout2_2', >> 'dout2_3', >> 'filter_gain', >> 'filter_on', >> 'snapshot0_bram', >> 'snapshot0_ctrl', >> 'snapshot0_status', >> 'snapshot0_trig_offset', >> 'snapshot1_bram', >> 'snapshot1_ctrl', >> 'snapshot1_status', >> 'snapshot1_trig_offset', >> 'sync_gen_sync', >> 'sync_gen_sync_period_sel', >> 'sync_gen_sync_period_var', >> 'sys_board_id', >> 'sys_clkcounter', >> 'sys_rev', >> 'sys_rev_rcs', >> 'sys_scratchpad', >> 'write_done'] >> >> thanks, >> >> Franco >> >> >> On 02/12/16 10:02, Marc Welz wrote: >>> Then there is no bitstream programmed or the particular bitstream >>> doesn't contain that register. Type >>> >>> ?listdev >>> >>> on that connection to find out >>> >>> regards >>> >>> marc >>> >>> >>> On Fri, Dec 2, 2016 at 12:52 PM, Franco wrote: It says: #log error 1018802834854 raw register\_dram_controller\_not\_defined Franco On 02/12/16 04:32, Marc Welz wrote: > telnet to the roach on port 7147 while repeating the operation - what > do the #log messages say ? > > On Thu, Dec 1, 2016 at 3:43 PM, Franco wrote: >> Hi! >> >> I tried to test the dram of ROACH2 by porting an example model I had for >> ROACH1. >> >> However when I tried to read the dram I got the following error: >> >> RuntimeError: Request write failed. >>Request: ?write dram_controller 0 \0\0\0\0 >>Reply: !write fail. >> >> Someone knows what's the problem? >> >> >> thanks, >> >> Franco Curotto >> >> >> >> >
Re: [casper] Problem reading DRAM in ROACH2
Are you running the latest version of the toolflow? Earlier versions did not include the CPU interface to the DRAM. Jason On 02 Dec 2016, at 15:30, Francowrote: > The register is not present in the bitstream. I thought that the register was > inside the dram yellow block though. I have to add it manually or something? > > Here is the listdev just in case: > > ['acc_len', > 'adc5g_controller', > 'cnt_rst', > 'dout0_0', > 'dout0_1', > 'dout0_2', > 'dout0_3', > 'dout1_0', > 'dout1_1', > 'dout1_2', > 'dout1_3', > 'dout2_0', > 'dout2_1', > 'dout2_2', > 'dout2_3', > 'filter_gain', > 'filter_on', > 'snapshot0_bram', > 'snapshot0_ctrl', > 'snapshot0_status', > 'snapshot0_trig_offset', > 'snapshot1_bram', > 'snapshot1_ctrl', > 'snapshot1_status', > 'snapshot1_trig_offset', > 'sync_gen_sync', > 'sync_gen_sync_period_sel', > 'sync_gen_sync_period_var', > 'sys_board_id', > 'sys_clkcounter', > 'sys_rev', > 'sys_rev_rcs', > 'sys_scratchpad', > 'write_done'] > > thanks, > > Franco > > > On 02/12/16 10:02, Marc Welz wrote: >> Then there is no bitstream programmed or the particular bitstream >> doesn't contain that register. Type >> >> ?listdev >> >> on that connection to find out >> >> regards >> >> marc >> >> >> On Fri, Dec 2, 2016 at 12:52 PM, Franco wrote: >>> It says: #log error 1018802834854 raw >>> register\_dram_controller\_not\_defined >>> >>> >>> Franco >>> >>> >>> >>> On 02/12/16 04:32, Marc Welz wrote: telnet to the roach on port 7147 while repeating the operation - what do the #log messages say ? On Thu, Dec 1, 2016 at 3:43 PM, Franco wrote: > Hi! > > I tried to test the dram of ROACH2 by porting an example model I had for > ROACH1. > > However when I tried to read the dram I got the following error: > > RuntimeError: Request write failed. >Request: ?write dram_controller 0 \0\0\0\0 >Reply: !write fail. > > Someone knows what's the problem? > > > thanks, > > Franco Curotto > > > >
Re: [casper] Problem reading DRAM in ROACH2
The register is not present in the bitstream. I thought that the register was inside the dram yellow block though. I have to add it manually or something? Here is the listdev just in case: ['acc_len', 'adc5g_controller', 'cnt_rst', 'dout0_0', 'dout0_1', 'dout0_2', 'dout0_3', 'dout1_0', 'dout1_1', 'dout1_2', 'dout1_3', 'dout2_0', 'dout2_1', 'dout2_2', 'dout2_3', 'filter_gain', 'filter_on', 'snapshot0_bram', 'snapshot0_ctrl', 'snapshot0_status', 'snapshot0_trig_offset', 'snapshot1_bram', 'snapshot1_ctrl', 'snapshot1_status', 'snapshot1_trig_offset', 'sync_gen_sync', 'sync_gen_sync_period_sel', 'sync_gen_sync_period_var', 'sys_board_id', 'sys_clkcounter', 'sys_rev', 'sys_rev_rcs', 'sys_scratchpad', 'write_done'] thanks, Franco On 02/12/16 10:02, Marc Welz wrote: Then there is no bitstream programmed or the particular bitstream doesn't contain that register. Type ?listdev on that connection to find out regards marc On Fri, Dec 2, 2016 at 12:52 PM, Francowrote: It says: #log error 1018802834854 raw register\_dram_controller\_not\_defined Franco On 02/12/16 04:32, Marc Welz wrote: telnet to the roach on port 7147 while repeating the operation - what do the #log messages say ? On Thu, Dec 1, 2016 at 3:43 PM, Franco wrote: Hi! I tried to test the dram of ROACH2 by porting an example model I had for ROACH1. However when I tried to read the dram I got the following error: RuntimeError: Request write failed. Request: ?write dram_controller 0 \0\0\0\0 Reply: !write fail. Someone knows what's the problem? thanks, Franco Curotto
Re: [casper] Problem reading DRAM in ROACH2
Then there is no bitstream programmed or the particular bitstream doesn't contain that register. Type ?listdev on that connection to find out regards marc On Fri, Dec 2, 2016 at 12:52 PM, Francowrote: > It says: #log error 1018802834854 raw > register\_dram_controller\_not\_defined > > > Franco > > > > On 02/12/16 04:32, Marc Welz wrote: >> >> telnet to the roach on port 7147 while repeating the operation - what >> do the #log messages say ? >> >> On Thu, Dec 1, 2016 at 3:43 PM, Franco wrote: >>> >>> Hi! >>> >>> I tried to test the dram of ROACH2 by porting an example model I had for >>> ROACH1. >>> >>> However when I tried to read the dram I got the following error: >>> >>> RuntimeError: Request write failed. >>>Request: ?write dram_controller 0 \0\0\0\0 >>>Reply: !write fail. >>> >>> Someone knows what's the problem? >>> >>> >>> thanks, >>> >>> Franco Curotto >>> >>> >
Re: [casper] Problem reading DRAM in ROACH2
telnet to the roach on port 7147 while repeating the operation - what do the #log messages say ? On Thu, Dec 1, 2016 at 3:43 PM, Francowrote: > Hi! > > I tried to test the dram of ROACH2 by porting an example model I had for > ROACH1. > > However when I tried to read the dram I got the following error: > > RuntimeError: Request write failed. > Request: ?write dram_controller 0 \0\0\0\0 > Reply: !write fail. > > Someone knows what's the problem? > > > thanks, > > Franco Curotto > >
[casper] Problem reading DRAM in ROACH2
Hi! I tried to test the dram of ROACH2 by porting an example model I had for ROACH1. However when I tried to read the dram I got the following error: RuntimeError: Request write failed. Request: ?write dram_controller 0 \0\0\0\0 Reply: !write fail. Someone knows what's the problem? thanks, Franco Curotto