[UrJTAG-dev] Added support for Kintex 7 xc7k325t-ffg900 chip

[Wojciech Z. <wz...@gm...>]

Hi,

DISCLAIMER: The procedure described below may be risky, as the source BSDL
file had to be modified. There is a risk, that removed sections contain
information important for safe control of the FPGA via JTAG. If you are
going to use the described procedure, you do it ON YOUR OWN RISK!

I had to use UrJTAG with our boards based on Kintex 7 xc7K325t-ffg900 chip.
Unfortunately this chip was not supported by UrJTAG.
I tried to generate the "jtag" file from the xc7k325t_ffg900.bsd provided
with Xilinx Vivado, but it resulted in error:

bsdl2jtag xc7k325t_ffg900.bsd xc7k325t_ffg900
error: -E- error: In Package STD_1149_6_2003, Line 755, Error in
User-Defined Package declarations.
error: -E- error: BSDL file 'xc7k325t_ffg900.bsd' contains errors in VHDL
stage, stopping
error: system error: Success Cannot open file STD_1149_6_2003 or
/usr/local/wzab/urjtag/share/urjtag/bsdl/STD_1149_6_2003

Basing on discussion:
https://sourceforge.net/p/urjtag/mailman/message/30062425/
I have removed the line 755:

use STD_1149_6_2003.all;

And then, retrying the conversion, I had to remove yet some sections:
1. The one starting from (near to line 1407):
> attribute PORT_GROUPING of XC7K325T_FFG900 : entity is
> "DIFFERENTIAL_VOLTAGE (" &
> "(MGTXRXP0_115, MGTXRXN0_115), " &
> [...]

2. The one starting from (near to line 3193):

> attribute AIO_COMPONENT_CONFORMANCE of XC7K325T_FFG900 : entity is
>       "STD_1149_6_2003";
>
> attribute AIO_EXTEST_Pulse_Execution of XC7K325T_FFG900 : entity is
>       "Wait_Duration TCK 15";
>
> attribute AIO_EXTEST_Train_Execution of XC7K325T_FFG900 : entity is
>       "train 30, maximum_time 120.0e-6";
>
> attribute AIO_Pin_Behavior of XC7K325T_FFG900 : entity is
> "MGTXRXP0_115 : LP_time=22.5e-9 HP_time=45.0e-9; " &
> "MGTXRXP0_116 : LP_time=22.5e-9 HP_time=45.0e-9; " &
> [...]

Afterwards the file converted cleanly, and UrJTAG was able to recognize the
chip.

Of course there is a risk, that removal of certain incompatible sections
may result in improper control of chip via JTAG, which may lead to
destruction of the chip.
I don't fully understand meaning of removed sections, so I do not provide
any warranty, that the described procedure is safe.
If you are going to use it, you do it on your own risk.

-- 
With best regards,
Wojtek


Wojciech M. Zabołotny
My GPG/PGP keys:
standard: 8192R/FE58A848 (0720 9430 85DB 7CCD F4C5  5F1E 5107 91FB FE58
A848)
confidential: 16384R/C76D2FB0 (C4E7 9597 CF22 7B5D 28BF  4656 FED7 A63F
C76D 2FB0)