zhaisitong

@saaramahmoudi tight floor plan testcases in nightly_test5 "vpr_tight_floor_plan"

> @zhaisitong: can you take a look at this one? We can discuss Thursday. How could I run to the failed place? What is the exact command or what regressions...

@MohamedElgammal I didn't find the architecture file in the "Archive.zip". I tried the stratixiv_arch.xml in the vtr_flow, but it doesn't work. Could you upload the architecture file again?

@MohamedElgammal Thanks for your architecture file. Which branch and commit id could I use to reproduce this problem? Could I use the newest codes on the master branch?

@MohamedElgammal When did you merge your API for sfkhalid's codes? Could the codes work after you merged your API? Could you specify me the commit id of your merge?

I traced back and found this problem comes from @jmah76 committed on Jun 8, the commit id is d22193d4c8030a24daa73884b60d9bdb8d0aec0b. I think @MohamedElgammal had pushed all the relevant code and worked...

@vaughnbetz Yes, it is weird. But I run 5 times, the results are the same. When I checked out to each commits, I found the codes are very different. I...

I think this might be because @jmah76 forgot to update codes before pushing them. After I recover all relevant codes, the three commands mentioned above works. Above is incorrect, this...

The problem might come from a14647570ac98df63b1871508a73496b64fbf31f. In function vpr/src/pack/cluster_util.cpp: get_highest_gain_molecule, before finding unpacked molecules based on attraction group of the current cluster, missing the condition "cur_pb->pb_stats->num_feasible_blocks == 0". After adding...

This issue has been solved with PR #2160 .