Hello ports@, I've got computers of various vintage running OpenBSD and I'd like to get a better idea what kind of workload could the machine handle compared to others.
I came across package called ubench, but it behaves weird on these AMD CPUs: - cpu0: AMD GX-217GA SOC with Radeon(tm) HD Graphics, 1647.11 MHz, 16-00-01 - cpu0: AMD Athlon(tm) II Neo N36L Dual-Core Processor, 1298.10 MHz, 10-06-03 Description at (http://www.phystech.com/download/ubench.html) and even in pkg_info ubench says: "Ubench will spawn about 2 concurrent processes for each CPU available on the system. This ensures all available raw CPU horsepower is used." This is not the case for AMD CPUs, as there's just one process jumping around those two dual-core systems. (could be seen in sysstat or htop) Intel CPUs work as expected, I can see multiple processes in top/htop and all cores are busy and I get a "score" after run. The run ends with either ubench score of 0 or exiting on signal 28 $ ubench -c Unix Benchmark Utility v.0.3 Copyright (C) July, 1999 PhysTech, Inc. Author: Sergei Viznyuk <s...@phystech.com> http://www.phystech.com/download/ubench.html OpenBSD 6.9 GENERIC.MP#0 amd64 Ubench CPU: 0 $ ubench -c Unix Benchmark Utility v.0.3 Copyright (C) July, 1999 PhysTech, Inc. Author: Sergei Viznyuk <s...@phystech.com> http://www.phystech.com/download/ubench.html OpenBSD 6.9 GENERIC.MP#0 amd64 **** cpubench: exiting on signal 28 # - seems to be triggered by resizing terminal window running or not running as root makes no difference. Any idea what's going on here? The code is from 2000, should the port be decomissioned? What do people generally use to benchmark hardware these days? Thank you, Jan
