08-24-2010 01:31 AM
in new to PR and need some Help.
I made an Design that looks similar to the one Shown in UG744. Everything works fine, even Running SW on the MicroBlaze and reconfiguration the PRR. But when I try to use the ICAP to reconfigurate the PRR the MicroBlaze gets destroyed and just an full Re-Programming gets it back to life.
I use the ICAP-Test Software from the UG744 and Stored my Bit Files in the CF Card. The .bit Files are found and it starts writing the data into the ICAP Buffer until "SectorNumber" 8 should be read. Then the System freezes and just the full Re-Programming helps.
Does Anyone know this Problem an may help me?
09-02-2010 04:54 AM
don't know which device you are using, however in case you are on a Virtex-5: have you constrained the nets going to and from the ICAP with a MAX_DELAY?
The ICAP in Virtex-5 is not considered as a synchronuous element even though it is clocked and hence the PERIOD-constraint does not cover its connected nets.
01-13-2011 10:03 AM
We have introduced a timing model for the ICAP in Virtex-6 (and beyond), so it can be covered by standard timing constraints (PERIOD, etc.). With Virtex-4 and Virtex-5, MAXDELAY constraints must be used. More information is given on page 38 of the Partial Reconfiguration User Guide:
01-24-2011 10:16 AM
I have exactly the same problem.
I fixed the following constraints (as recommended in UG702) :
NET "xps_hwicap_0/xps_hwicap_0/HWICAP_CTRL_I/icap_datain<*>" MAXDELAY = 2 ns;
NET "xps_hwicap_0/xps_hwicap_0/HWICAP_CTRL_I/icap_ce" MAXDELAY = 2 ns;
(I do not have "icap_statemachine_I1/Icap_datain<*>" nets)
I applied UG744 but with a different ngc file. The microblaze always hangs at the same place, no matter what is inside the PRR.
When the microblaze is hanged, if I perform a PR via JTAG, the microblaze continues and finishes the reconfiguration.
Do you have any idea about this problem?
Can it be a problem with the PLB bus?