From: Jason Wever <[EMAIL PROTECTED]> Date: Wed, 17 Aug 2005 22:15:59 -0600
> 1) The max frequency for the CPUs appears to be off by 3 decimal places. > Originally, I had checked with cpufreq-info from the cpufrequtils > package[1], and then later confirmed by checking the > /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq file by hand. > The values I was getting were 502000000 (502.00 GHz) on a 500 MHz Blade > 100 and 750000000 (750.00 GHz) on either CPU in a dual 750 MHz Blade > 1000. I'm guessing this is a problem with how the speeds are being > reported to sysfs? That's definitely a bug. I'll try to figure out how I'm off by a factor of 1000 :-) > 2) The minimum CPU frequencies possible being reported for the Blade 100 > and Blade 1000 are 62750000 and 23437500 respectively. Are these > accurate (even with the possibility mentioned above being taken into > account)? They seem a bit low to me (based on what I've seen from the > x86 world). It looks right. The UltraSPARC-III has three clock divisor values you can use, 1, 2, and 32. I know that seems weird, but that is what the chip supports. UltraSPARC-IIe supports divisors 1, 2, 4, 6, and 8, so the IIe number looks correct as well. > 3) On x86, amd64 and PPC, there exists a file called > /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_cur_freq which shows the > current running frequency of the CPU. For both the IIe and III cpufreq > drivers, this is missing. Is there any reason for this (like no good > way to poll the CPUs) or is it just a possible change in cpufreq stuff > that wasn't synced up with sparc? Strange, I'll try to figure out this bug as well. Thanks for the report. - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
