search

Progressive-Learning-Platform / PLPTool6

Prototype for PLPTool frontend interface overhaul
2 stars 4 forks source link

[Core] In PLP6 , need a "nop" at the end of the code compared to PLP5 which is not compulsory. #88

Open abhilash576 opened 7 years ago

abhilash576 commented 7 years ago

In PLP6, if you write the code and dont put "nop" at the end then the code runs continuously and updates any register values used in the program for infinite times which is not expeccted. The same code when being run in PLP5 works fine.

To Generate the bug using the below code:

.org 0x10000000 li $t0, 10 li $t1, 15 addu $t0, $t0, $t1

After simulation, check the value of $t0 in watcher window , it will be continuously incrementing.

tobielf commented 7 years ago

I think this bug will happen under "run" simulation. However under the "step" simulation, there is no problem. @abhilash576 Please confirm this.

abhilash576 commented 7 years ago

The problem is observed only with "Run Simulation" and with "Step Simulation" its works fine.

harsha-kadekar commented 6 years ago

Looking into the issue. As of now current delivery of the fix will be on October 27th.

harsha-kadekar commented 6 years ago

The issue is reproducible even with "Step Simulation". Run simulation is internally calling Step simulation function. Initial analysis looks like problem with not clearing of the buffers after reaching the end of instructions. Further analysis is going on.