Saturday, 2021-06-05

*** tpb <[email protected]> has joined #symbiflow00:00
tpbtpb has joined on libera00:01
*** gsmecher <[email protected]> has quit IRC (Ping timeout: 258 seconds)00:11
*** FFY00_ <FFY00_!~FFY00@archlinux/trusteduser/ffy00> has quit IRC (Ping timeout: 245 seconds)01:27
*** FFY00_ <FFY00_!~FFY00@archlinux/trusteduser/ffy00> has joined #symbiflow02:02
tpbTMM_ has quit libera (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)03:41
tpbTMM_ has joined on libera03:41
*** proteusguy <[email protected]> has quit IRC (Ping timeout: 252 seconds)04:49
*** proteusguy <[email protected]> has joined #symbiflow05:01
tpbcr1901 has quit libera (Ping timeout: 272 seconds)05:57
*** ayazar2 <[email protected]> has quit IRC (Quit: The Lounge - https://thelounge.chat)07:49
*** ayazar2 <[email protected]> has joined #symbiflow07:51
*** ayazar2 <[email protected]> has quit IRC (Quit: The Lounge - https://thelounge.chat)07:58
*** ayazar2 <[email protected]> has joined #symbiflow07:59
sf-slack<ajinkyaraghuwanshi> I am trying to run the test model in the SymbiFlow-arch-defs, I created a test model ,ran this code and got this output: " make -f CMakeLists.txt CMakeLists.txt:1: *** missing separator.  Stop.". Now I am not able to understand what was the supposed output as I am getting the same output for all the test files in it.09:22
sf-slack<ajinkyaraghuwanshi> I am getting this same output for all the make files in SymbiFlow-arch-defs09:28
sf-slack<ajinkyaraghuwanshi> I am also not able to build the pre-existing models after making the env, in the getting started manual of the SymbiFlow-arch-defs09:30
sf-slack<ajinkyaraghuwanshi> Can anybody help?09:31
sf-slack<fdarling> @ajinkyaraghuwanshi maybe I am misinterpreting what you've written, but you should't run `make` directly on a CMakeLists.txt file, but rather you should run `cmake` on it. You typically create a `build` subdirectory, and run `cmake ..` which pulls in the `CMakeLists.txt` file from the parent directory.12:40
sf-slack<ajinkyaraghuwanshi> yes that I misspelled but the problem that none of my make files are working12:47
sf-slack<ajinkyaraghuwanshi> basically after the make env , none of the build and other make files are working, I have also  tried to clean and repeated the process which took hours but I am still getting some errors12:52
sf-slack<ajinkyaraghuwanshi>12:53
sf-slack<fdarling> @ajinkyaraghuwanshi walk me through what you're trying to do, I just cloned https://github.com/SymbiFlow/symbiflow-arch-defs.git myself so I can test whatever it may be12:59
tpb<https://x0.no/4uwwk> (at github.com)13:00
sf-slack<ajinkyaraghuwanshi> I am trying to make a test program in the system to be ran by yosys13:02
sf-slack<fdarling> is your goal to set up Yosys for a particular family of FPGA, or is it to test something specifically within that git repository?13:03
sf-slack<ajinkyaraghuwanshi> I want to test the in the repo itself later it will be based for artix7 board13:05
sf-slack<fdarling> all right, so you are using that same repository that I linked you to, right?13:07
sf-slack<fdarling> I am confused because in that log_inst.txt you posted, you run `make -f Makefile` (equivalent of just `make`) first, and then you do `make all_demos` before running `make env`, even though the README.md file says to run `make env` first13:09
sf-slack<ajinkyaraghuwanshi> Yes13:11
sf-slack<ajinkyaraghuwanshi> I have followed the same steps before also13:12
sf-slack<fdarling> for me `make env` runs a *lot* of stuff, downloading git repositories, installing python pip dependencies, etc.13:12
sf-slack<ajinkyaraghuwanshi> here I was trying to again run them so it looks as if I am trying to run all_demos before env13:13
sf-slack<ajinkyaraghuwanshi> yes it is same for me too13:13
sf-slack<ajinkyaraghuwanshi> These are the final output error after the build that I find:"2021-06-05 16:55:26.704730 Track aliveness committed Dumping database to '/home/ajinkya/GSOC2021/symbiflow-arch-defs/build/xc/xc7/archs/zynq7_z020/devices/xc7z020-virt/channels.db'  87% (796900 of 914185) |#############   | Elapsed Time: 0:03:29 ETA:   0:00:05Traceback (most recent call last):   File13:16
sf-slack"/home/ajinkya/GSOC2021/symbiflow-arch-defs/xc/common/utils/prjxray_create_edges.py", line 103, in <module>     main()   File "/home/ajinkya/GSOC2021/symbiflow-arch-defs/xc/common/utils/prjxray_create_edges.py", line 72, in main     ccio_sites = create_edges(args)   File "/home/ajinkya/GSOC2021/symbiflow-arch-defs/xc/common/utils/prjxray_edge_library.py", line 2312, in create_edges     mark_track_liveness(conn, input_only_nodes,13:16
sf-slackoutput_only_nodes)   File "/home/ajinkya/GSOC2021/symbiflow-arch-defs/xc/common/utils/prjxray_db_cache.py", line 61, in __exit__     self.file_connection, pages=100, progress=self._progress sqlite3.OperationalError: database or disk is full 100% (914185 of 914185) |################| Elapsed Time: 0:03:31 Time:  0:03:3113:16
sf-slackxc/xc7/archs/zynq7_z020/devices/xc7z020-virt/CMakeFiles/file_xc_xc7_archs_zynq7_z020_devices_xc7z020-virt_channels.db.dir/build.make:84: recipe for target 'xc/xc7/archs/zynq7_z020/devices/xc7z020-virt/channels.db' failed make[2]: *** [xc/xc7/archs/zynq7_z020/devices/xc7z020-virt/channels.db] Error 1 make[2]: *** Deleting file 'xc/xc7/archs/zynq7_z020/devices/xc7z020-virt/channels.db' CMakeFiles/Makefile2:92950: recipe for target13:16
sf-slack'xc/xc7/archs/zynq7_z020/devices/xc7z020-virt/CMakeFiles/file_xc_xc7_archs_zynq7_z020_devices_xc7z020-virt_channels.db.dir/all' failed make[1]: *** [xc/xc7/archs/zynq7_z020/devices/xc7z020-virt/CMakeFiles/file_xc_xc7_archs_zynq7_z020_devices_xc7z020-virt_channels.db.dir/all] Error 2 Makefile:159: recipe for target 'all' failed make: *** [all] Error 2:"13:16
sf-slack<fdarling> `make env`, I am trying the following: ```cd build make all_demos``` It seemed to hang at first, but it was actually just taking a really long time to get started.13:16
sf-slack<ajinkyaraghuwanshi> These are final lines from the previous log file that I sent.13:17
sf-slack<fdarling> is your hard drive full? you might want to check with `df -h`13:17
sf-slack<ajinkyaraghuwanshi> yes I completed it13:17
sf-slack<ajinkyaraghuwanshi> no13:17
sf-slack<ajinkyaraghuwanshi> I checked it is still 3.3GB free for that drive13:18
sf-slack<fdarling> "sqlite3.OperationalError: database or disk is full" that is concerning13:18
sf-slack<ajinkyaraghuwanshi> yes I thought too but when I checked it showed 3.3GB space avaiable13:18
sf-slack<fdarling> would you please show me the output of `df -h` ?13:19
sf-slack<ajinkyaraghuwanshi> ok it will take 2 min13:19
sf-slack<ajinkyaraghuwanshi>13:23
sf-slack<ajinkyaraghuwanshi>13:23
sf-slack<fdarling> you say that there was free space, but if you notice it says "make[2]: *** Deleting file 'xc/xc7/archs/zynq7_z020/devices/xc7z020-virt/channels.db'" after the error message, so it may have cleaned up after itself and hidden the fact that it was using 100% disk space13:24
sf-slack<fdarling> I assume you're operating with a VirtualBox VM. If so, I suggest shutting it down and growing the virtual disk size to be larger, then use a "live CD" with `gparted` to resize/move the partitions to take up the newly available space13:30
sf-slack<fdarling> so far the symbiflow-arch-defs directory on my system is taking up 6.1GB, but it seems a long way off from being finished13:31
sf-slack<fdarling> I'll let it finish and report how much space it took up13:31
sf-slack<ajinkyaraghuwanshi> Actually I am using ubuntu, in my dual boot system13:34
sf-slack<ajinkyaraghuwanshi> so increasing the space is a problem13:34
sf-slack<fdarling> I am running Xubuntu 20.04 for the record13:36
sf-slack<ajinkyaraghuwanshi> I am working on the project-XRAY and I wanted to know that the git prjxray and the folder in the symbiflow-arch-defs, are they same?13:37
sf-slack<ajinkyaraghuwanshi> So, did you complete the installation?14:28
sf-slack<ajinkyaraghuwanshi> So was the installation complete?14:33
*** lopsided98 <lopsided98!~quassel@2601:18a:0:85e0:ba27:ebff:fe5e:6b6e> has quit IRC (Quit: Disconnected)14:53
tpblopsided98 has quit libera (Quit: Disconnected)14:53
tpblopsided98 has joined on libera14:55
*** lopsided98 <lopsided98!~quassel@2601:18a:0:85e0:ae35:7964:bde7:2765> has joined #symbiflow14:56
*** lopsided98 <lopsided98!~quassel@2601:18a:0:85e0:ae35:7964:bde7:2765> has left #symbiflow14:57
tpblopsided98 has quit libera (Quit: Disconnected)15:02
tpblopsided98 has joined on libera15:03
tpbRaito_Bezarius has quit libera (Ping timeout: 268 seconds)15:16
tpbRaito_Bezarius has joined on libera15:28
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has joined #symbiflow15:31
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has quit IRC (Ping timeout: 258 seconds)15:36
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has joined #symbiflow15:48
tpbcr1901 has joined on libera16:26
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has quit IRC (Ping timeout: 258 seconds)17:21
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has joined #symbiflow19:46
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has quit IRC (Ping timeout: 258 seconds)19:59
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has joined #symbiflow20:11
tpbTMM_ has quit libera (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)20:14
tpbTMM_ has joined on libera20:15
sf-slack<nelson> I saw above some comments yesterday or the day before regarding initialization of flip flops to either a 0 or 1 on power up with the tool flow, and  I wanted to weigh in.  We are seeing similar issues --- in the designs my student is running through the symbiflow-examples flow they have confirmed that if a Verilog signal is declared with an initializer as in:  “reg[3:0] cnt = 0;” and that signal is the register for20:20
sf-slacka counter, then the counter will power up to 0000.  But, if they just do: “reg[3:0] cnt;” the counter will power up to 1111.  I don’t know if any standards document describes the required behavior.  But, it is in contrast to other FPGA design tools I have used over many years  with Xilinx FPGAs.  I am not stating that I rely on this but just that it is different from anything else I have ever seen.  I would be curious20:20
sf-slackregarding others’ thoughts on why this was chosen as the behavior for the tool flow (or if it wasn’t chosen and this is perhaps a bug).20:20
tpb<gatecat> I think this is probably happening  because a synchronous set is being inferred for the flipflop. Older Xilinx FPGAs couldn't have the SR and init values set differently; so an async set means a default init of 1 if not overriden in Yosys, this reduces the code path differences between the different Xilinx families20:26
tpb<gatecat> Initialisation in synthesis isn't really defined at all (because it doesn't apply to ASICs); but it's certainly the case that in simulation an uninitialised register will initialise to x's20:27
tpb<gatecat> there's a whole host of behaviour in practice across different tooling, including in one case ignoring the initial value and always initialising flops to 020:27
*** kraiskil <kraiskil!kraiskil@gateway/vpn/mullvad/kraiskil> has quit IRC (Ping timeout: 260 seconds)20:46
sf-slack<fdarling> mine ultimately failed here:21:05
sf-slack<fdarling> ```$ du -hcs symbiflow-arch-defs/ 23Gsymbiflow-arch-defs/ 23Gtotal```21:05
sf-slack<fdarling> so it did grow to quite a large size first21:05
sf-slack<fdarling> ```$ find -type f -name channels.db | xargs du -hcs 4.3G./build/xc/xc7/archs/artix7_100t/devices/xc7a100t-virt/channels.db 1.5G./build/xc/xc7/archs/artix7_100t/channels/xc7a100tfgg676-1/channels.db```21:07
sf-slack<fdarling> those channels.db files can get pretty big, I seriously think it may have been the problem21:08
*** maartenBE <[email protected]> has joined #symbiflow22:10
tpbadjtm has joined on libera22:15

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!