[srslte-users] Debug build and LV_HAVE_SSE and LV_HAVE_AVX

[Kewin Rausch]

Hey there,

You are not the only one with this problem, since some time ago I
encountered the same problem (no connection when srsENB was built in Debug
mode... well, the Ue was not even able to detect the signal and synchronize
with it). What I did to overcome this problem was to build the software in
release mode, but I added "-g" flag in the Release mode profile (somewhere
in the top-most CMakeFiles).

That way the eNB work fine AND with debugging symbols, which allowed to
perform backtrace and identify a couple of exceptions.

Cheers,
Kewin R.

On Wed, Feb 7, 2018 at 8:39 PM, J Giovatto <jgiovatto at adjacentlink.com>
wrote:

> Hi,
>
> I was running a Debug build to bring in debug symbols for a crash I was
> seeing.
>
> To make a long story short I was not able to make a connection between
> the ue and enb when using n_prb 6 or 15, while 25 did work.
>
> The only thing I've noticed so far is this in ...
>
> "lib/src/phy/fec/rm_turbo.c"
>
> #ifdef DEBUG_MODE
> #warning FIXME: Disabling SSE/AVX turbo rate matching
> #undef LV_HAVE_SSE
> #undef LV_HAVE_AVX
> #endif
>
> One caveat here is that I'm an running with a faux_rf (no hardware)
> meaning the calls to the rf_device for send/recv are actually a call to
> a unix socket send/recv, so
>
> the samples in should be exactly the same as samples in almost all cases
> with the exception of a few timing hiccups at a 1msec TTI, so I kind of
> expected that
>
> if n_prb 25 worked, 6 or 15 would be supported with even less processing
> overhead.
>
> I have NOT tried the same on real hardware so cant say for sure.
>
> Anyway here is a portion of the ue log...
>
> 4:11:13.241316 [MAC ] [I] [10235] SCHED: Searching for DL grant for
> SI-RNTI window_st=0, window_len=20
> 14:11:13.242349 [MAC ] [I] [10236] SCHED: Searching for DL grant for
> SI-RNTI window_st=0, window_len=20
> 14:11:13.243398 [MAC ] [I] [10237] SCHED: Searching for DL grant for
> SI-RNTI window_st=0, window_len=20
> 14:11:13.246231 [MAC ] [I] [00000] BSR:   QUEUE status: 0 (0), 0 (0), 0
> (0), 0 (0), 0 (0), 0 (0),
> 14:11:13.255872 [MAC ] [I] [00009] SCHED: Searching for DL grant for
> SI-RNTI window_st=160, window_len=20
> 14:11:13.405340 [RRC ] [I] RRC Cell Selecting: timeout expired. Starting
> Cell Search...
> 14:11:13.405373 [PHY0] [W] [00161] SYNC:  Can't start cell search
> procedure while camping on cell
> 14:11:13.415759 [MAC ] [I] [00169] SCHED: Searching for DL grant for
> SI-RNTI window_st=320, window_len=20
> 14:11:13.575418 [MAC ] [I] [00329] SCHED: Searching for DL grant for
> SI-RNTI window_st=480, window_len=20
> 14:11:13.735650 [MAC ] [I] [00489] SCHED: Searching for DL grant for
> SI-RNTI window_st=640, window_len=20
> 14:11:13.747209 [MAC ] [I] [00501] BSR:   QUEUE status: 0 (0), 0 (0), 0
> (0), 0 (0), 0 (0), 0 (0),
> 14:11:13.895623 [MAC ] [I] [00649] SCHED: Searching for DL grant for
> SI-RNTI window_st=800, window_len=20
>
> I tried to remove the above check but got a build error
>
> /lib/src/phy/fec/rm_turbo.c:373:19: error: selector must be an integer
> constant in the range 0..7
>            int16_t x  = (int16_t)  _mm_extract_epi16(xVal,   j);
>                    ^
> so there is more of a difference in a Debug vs Release build that I
> figured might be worth mentioning.
>
> In Release builds I am able to connect using n_pbr 6,15,25 and 50.
>
> Regards
>
> Joe
>
>
> _______________________________________________
> srslte-users mailing list
> srslte-users at lists.softwareradiosystems.com
> http://www.softwareradiosystems.com/mailman/listinfo/srslte-users
>

-- 
--
Le informazioni contenute nella presente comunicazione sono di natura privata 
e come tali sono da considerarsi riservate ed indirizzate esclusivamente ai 
destinatari indicati e per le finalità strettamente legate al relativo 
contenuto. Se avete ricevuto questo messaggio per errore, vi preghiamo di 
eliminarlo e di inviare una comunicazione all’indirizzo e-mail del mittente.
--
The information transmitted is intended only for the person or entity to 
which it is addressed and may contain confidential and/or privileged 
material. If you received this in error, please contact the sender and 
delete the material.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20180207/90887170/attachment.htm>