OpenCores

checking errors

Back to bugtracker overview.

Information:
Type :: BUG
Status :: CLOSED
Assigned to :: Joris, van Rantwijk

Description:
On the spwlink.vhd, line 111: when r.state = S_ErrorWait , you test the fallowing errors:

ESCAPE ERROR, PARITY ERROR AND CHARACTER SEQUENCE ERROR,

but by the standard (ECSS-E-50-12A - 8.5.3.9.3b and 8.5.3.9.4b and 8.5.3.11b) it's says that this errors must be checked only after the first NULL had been received. And it's not the case. The same error repeat for r.state = S_Ready .

Thank you,

Rafael Corsi Ferrão

student of Mauá Institute of technology.

Comments:

Schlemm, Philipp May 8, 2018
Checking for character sequence errors before the first NULL is received is valid as far as I know.

Character sequence errors are N-Chars and L-Chars received before first NULL is received OR once a NULL is received ? FCT received before a NULL is sent OR N-Chars or time-codes received before both a NULL and FCT is received. All this is only during initialization. (see page 70 of ECSS-E-ST-50-12C)
van Rantwijk, Joris Sep 9, 2010
Indeed these errors should not be tested before the first NULL. I agree that it seems as if I am testing for these errors too early.

However I believe that the code in spwrecv.vhd is such that these conditions are never generated before the receiver has seen the first NULL. Note that the code that recognizes tokens is inactive until "null_seen" has been set. Do you agree?

I think this issue should be better clarified in the code. I will at least add some code comments to explain the subtle dependency between spwlink and spwrecv. Another option may be to add an explicit test on recvo.got_null, but I don't like to use unnecessary FPGA resources.

Post a comment:
Login to post comments!

Back to bugtracker overview.

© copyright 1999-2019 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.