1. 15 Feb, 2018 1 commit
  2. 08 Jan, 2018 1 commit
  3. 28 Dec, 2017 1 commit
  4. 27 Dec, 2017 1 commit
  5. 27 Oct, 2017 6 commits
    • Cedric Roux's avatar
      T: minor: close socket · 4d1705e3
      Cedric Roux authored
      Sometimes when the user does ctrl+c to quit, it may not quit
      if nothing comes on the socket. Closing the socket in the
      signal handler fixes this.
      4d1705e3
    • Cedric Roux's avatar
      T: cleanup · faae3c8c
      Cedric Roux authored
      Do not link the X stuff for tracers not needing it.
      faae3c8c
    • Cedric Roux's avatar
      T tracer: add a tracer to dump to VCD file · 7e7fd3c4
      Cedric Roux authored
      Later on, use this file with gtkwave.
      To be used to check realtime behaviour of the system.
      7e7fd3c4
    • Cedric Roux's avatar
      T: minor update of .gitignore · 2beac6fe
      Cedric Roux authored
      2beac6fe
    • Cedric Roux's avatar
      T: add extract_output_subframe · 8dd3a2bd
      Cedric Roux authored
      This tracer extracts output subframes generated by the eNodeB,
      so they can be analyzed with some external tools.
      8dd3a2bd
    • Cedric Roux's avatar
      T: add a tracer: extract · 6322010e
      Cedric Roux authored
      This tracer extracts the content of a buffer field of an event
      that was previously saved using the tracer 'record'.
      
      For example, to extract the channel estimation done in
      frame 924 and subframe 2 as saved in the file input_record.raw
      and to store it in the file output.raw, do:
      
      ./extract -d ../T_messages.txt -o output.raw input_record.raw ENB_PHY_UL_CHANNEL_ESTIMATE chest_t -f frame 924 -f subframe 2
      6322010e
  6. 23 Oct, 2017 2 commits
  7. 18 Oct, 2017 1 commit
    • Cedric Roux's avatar
      T tracer: improve a bit enb tracer · 2516216d
      Cedric Roux authored
      - UE_id 0 is for rnti 65535 (sib)
      - UE_id 1 is for rnti 65534 (this is a hack)
      - UE_id 2 is for rnti 2     (RA RNTI as of today's code/config)
      - we start enb by displaying UE_id 3
      - there was a bug: we need to lock when resetting UE ids (reset_ue_ids)
      2516216d
  8. 17 Oct, 2017 1 commit
  9. 12 Oct, 2017 1 commit
  10. 11 Oct, 2017 1 commit
  11. 09 Oct, 2017 1 commit
  12. 07 Oct, 2017 1 commit
  13. 06 Oct, 2017 1 commit
  14. 04 Oct, 2017 1 commit
  15. 02 Oct, 2017 1 commit
  16. 01 Oct, 2017 1 commit
  17. 29 Sep, 2017 2 commits
  18. 27 Sep, 2017 1 commit
    • Cedric Roux's avatar
      update T tracer · e6936f1b
      Cedric Roux authored
      not finished, UE may not work, many things bad/to fix properly
      e6936f1b
  19. 14 Sep, 2017 1 commit
  20. 29 Aug, 2017 2 commits
  21. 20 Aug, 2017 1 commit
  22. 16 Jun, 2017 1 commit
  23. 19 May, 2017 1 commit
  24. 15 May, 2017 1 commit
  25. 17 Apr, 2017 1 commit
  26. 23 Mar, 2017 1 commit
    • Cedric Roux's avatar
      fix issue 227 - UE IP settings disrupts realtime · cff91499
      Cedric Roux authored
      see https://gitlab.eurecom.fr/oai/openairinterface5g/issues/227
      
      When the UE connects to the eNodeB and receives its IP address from the
      network, it calls system() to set it in the linux kernel world. This call
      is not done in a realtime thread, but in the NAS, which uses its own thread,
      independent of the realtime processing.
      
      In some situations this totally disrupts realtime processing.
      
      It is difficult to know precisely why that happens, but it seems that calling
      fork(), as system() does, in a multi-threaded program is not a good idea. (So
      say several people on the internet.) It is not clear why the softmodem is
      impacted, but it seems that fork() is really what triggers the disruption.
      Several tests lead to that conclusion.
      
      To fix the problem, we create a child background process very early in main()
      (before anything else basically). Then instead of calling system(), the main
      process sends the string to the background process. The background process
      gets...
      cff91499
  27. 20 Feb, 2017 1 commit
  28. 17 Feb, 2017 1 commit
    • Cedric Roux's avatar
      T: let it compile with C++ · f9b5dafe
      Cedric Roux authored
      We now have 4 versions of T_HEADER:
       - bad quality C++ version with time
       - good quality C version with time
       - bad quality C++ version without time
       - good quality C version without time
      f9b5dafe
  29. 09 Feb, 2017 1 commit
  30. 02 Feb, 2017 1 commit
  31. 30 Jan, 2017 2 commits