THIS IS AN AUTOMATED MESSAGE, DO NOT REPLY. The following task has been changed. The new details are below. For full information about what has changed, visit the URL and click the History tab.
FS#246 - Odd pthread_mutex_test behavior User who did this: - Ali Saidi (saidi) Attached to Project - M5 Bugs Summary - Odd pthread_mutex_test behavior Task Type - Bug Category - Testing Status - New Assigned To - Operating System - All Severity - Very Low Priority - Low Reported Version - 1.1 Due in Version - Due Date - Undecided Percent Complete - 0% Details - When testing pthread_mutex_test in FS mode, there are some odd behaviors to it. Specifically if this script is used with the TimingSimpleCPU, and caches (build/ALPHA_FS/m5.opt configs/example/fs.py -t --caches -bMutexTest): #!/bin/sh cd /benchmarks/tests /sbin/m5 resetstats ./pthread_mutex_test 4 20000 /sbin/m5 exit The simulation completes in the expected timeframe. However if this script is used, same command line: #!/bin/sh cd /benchmarks/tests /sbin/m5 resetstats ./pthread_mutex_test 4 20000 /sbin/m5 exit The simulation takes 3-4 times longer in terms of cycles, and commits many more cycles. Note the extra spaces after the "20000", compared to the first script. However, if in the second script 20000 iterations is changed to 19999 iterations, and the spaces are still included, the simulation completes in the expected timeframe. I'm still trying to track down the cause of these differences. It may just be something odd with the scheduler going on. Trying similar numbers and spaces in the SE case does not seem to impact the cycle time. More information can be found at the following URL: http://www.m5sim.org/flyspray/task/246 You are receiving this message because you have requested it from the Flyspray bugtracking system. You can be removed from future notifications by visiting the URL shown above. _______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
