Looking at the code of te_priority and the flow chart of the spec. It's infered that pkts are generated on the basis of tc (this commit) and what's nc (next commit) or lc (last commit). I have faced issue on pkts generation. The fix that worked for me is on line 578 in the image below. There are mainly three stages of commits. 1st stage is nc, second stage is tc, third stage is lc. Mainly tc (this commit : 2nd stage) is the focal point according to the flow chart of the spec. Here on line 578 before the valid_i was being driven into the te_priority which actually is the nc_valid, i have now connected valid1_q which is aligned with tc.
Another fix for the same second stage valid connection was for te_branch_map inputs. On line 498 and 499 below i have connected second stage signals which are aligned with tc.
The above fixes worked for me and seems aligned with spec as well. I was facing problem on generation of format 3 subformat 0 pkts on first qualified instr which got resolved with this fix.
Looking at the code of te_priority and the flow chart of the spec. It's infered that pkts are generated on the basis of tc (this commit) and what's nc (next commit) or lc (last commit). I have faced issue on pkts generation. The fix that worked for me is on line 578 in the image below. There are mainly three stages of commits. 1st stage is nc, second stage is tc, third stage is lc. Mainly tc (this commit : 2nd stage) is the focal point according to the flow chart of the spec. Here on line 578 before the valid_i was being driven into the te_priority which actually is the nc_valid, i have now connected valid1_q which is aligned with tc.
Another fix for the same second stage valid connection was for te_branch_map inputs. On line 498 and 499 below i have connected second stage signals which are aligned with tc.
The above fixes worked for me and seems aligned with spec as well. I was facing problem on generation of format 3 subformat 0 pkts on first qualified instr which got resolved with this fix.