For reference, I typically install the dependencies for building UHD as described here: https://kb.ettus.com/Building_and_Installing_the_USRP_Open-Source_Toolchain_(UHD_and_GNU_Radio)_on_Linux#Update_and_Install_dependencies
Can you check which version of make you're using? I'm using GNU Make 4.2.1 on Ubuntu 20.04. Perhaps you could post the console output somewhere and the build.log file? I'm still not clear on the initial problem. The dev_config.json file isn't required. Wade On Fri, Aug 25, 2023 at 9:50 AM Seal, Ryan L. [US-US] < ryan.s...@dynetics.com> wrote: > * CAUTION: *This email originated from outside of Leidos. Be cautious > when clicking or opening content. > > Hey Wade, > > Here is the sequence of events: > > - git clone from github > - git checkout UHD-4.4 > - cd to x400 folder > - source setupenv.sh (vivado 2021.1 with the AR patch applied) > - make cleanall > - make X410_X4_200 SYNTH=1 GUI=1 (<-- added those options so I could > open the block design and take a peek at the clocks and RFDC config) > > I am building on Linux Mint 20.3 Una, which is essentially Ubuntu 20.04 > from a package perspective. I cut/pasted the dependency list from one of > the setup docs, all of the packages were pulled in correctly, and I can > talk to the X410 and run some of the example files. > > I didn't make any modifications to the repo, I want to get a default > image successfully built before going any further. I've gotten through > synthesis now, but haven't tried implementation and bitstream generation > yet. I'm working from the GUI, so let me go back and check the > RFNOC_EDGE_TBL_FILE definition and see exactly what's going on there. > > Also, the "dev_config.json" file was missing from the x400 folder, not > sure if that's another generated file, or just missing from the branch. I > copied that from the x300 folder. > > Thanks, > > Ryan > > ------------------------------ > *From:* Wade Fife <wade.f...@ettus.com> > *Sent:* Thursday, August 24, 2023 11:12 PM > *To:* Seal, Ryan L. [US-US] <ryan.s...@dynetics.com> > *Cc:* usrp-users@lists.ettus.com <usrp-users@lists.ettus.com> > *Subject:* EXTERNAL: Re: [USRP-users] X410 FPGA build failure with UHD-4.4 > > > *CAUTION: *This email originated from outside of Leidos. Be cautious when > clicking or opening content. > > Hi Ryan, > > What was the problem/error message that suggested that folder wasn't being > pulled in properly? The "syntax error" implies that the > `RFNOC_EDGE_TBLE_FILE definition isn't being set correctly, but the > makefile should pass that in based on the target you're building. > > Did you make modifications to the repo or run rfnoc_image_builder? If so, > maybe try building the unmodified branch first to make sure everything is > working. For example: > > source setupenv.sh > make cleanall > make X410_X4_200 > > Also, what OS are you using to build on? > > Wade > > On Thu, Aug 24, 2023 at 11:16 AM <ryan.s...@dynetics.com> wrote: > > I attempted to build an FPGA image on the UHD-4.4 branch with Vivado > v2021.1_AR76780 (64-bit) using: > > *make X410_X4_200 SYNTH=1 GUI=1* > > When the IP cores are generated, the following folder is not pulled in > properly: > > *uhd/fpga/usrp3/top/x400/build-ip/xczu28drffvg1517-1e/xge_pcs_pma_ex/imports > * > > As a temporary work-around, I was able to generate the example files from > the xci file in > > *uhd/fpga/usrp3/top/x400/ip/xge_pcs_pma* > > and then copied the imports into my build directory and kicked off > synthesis once again via GUI. > > There was also a syntax error in x410_200_rfnoc_image_core.v. The forward > ticks on line 182 were causing an issue: > > *localparam EDGE_TBL_FILE = `"`RFNOC_EDGE_TBL_FILE`";* > > Can someone provide some guidance? I’m not sure if I made a mistake when I > pulled something in, or I’m one of the few that attempted to build the > image. > > Going forward, I need to make clock modifications and wanted to get a look > at the clock structure in vivado to find the best path forward. To > summarize, I need to change the 245.76 clocks to 204.8. I would also like > to use the RFDC directly if possible (static configuration is fine). That > would give me what I need without the DUC/DDC RFNoC blocks. If this isn’t > feasible then I’ll need some rate-changing filters and additional clocks to > adjust downstream. > > - Ryan > _______________________________________________ > USRP-users mailing list -- usrp-users@lists.ettus.com > To unsubscribe send an email to usrp-users-le...@lists.ettus.com > >
_______________________________________________ USRP-users mailing list -- usrp-users@lists.ettus.com To unsubscribe send an email to usrp-users-le...@lists.ettus.com
