Change theme
Help
Press space for more information.
Show links for this issue (Shortcut: i, l)
Copy issue ID
Previous Issue (Shortcut: k)
Next Issue (Shortcut: j)
Sign in to use full features.
Vote: I am impacted
Notification menu
Refresh (Shortcut: Shift+r)
Go home (Shortcut: u)
Use Markdown for this comment
Set severity, which reflects how much the issue affects the use of the product
Change issue status back to 'Assigned'
Pending code changes (auto-populated)
[ID: 84651]
Story points rate the relative effort of work in a Fibonacci-like format: 0, 0.5, 1, 2, 3, 5, 8, 13, 20, 40, 100. Each team will estimate work on a slightly different scale, which means the values in this field are likely only meaningful to the team that owns the Buganizer component in which the issue resides.
See Atlassian's Agile Coach for more information on how to use story points for estimation: https://www.atlassian.com/agile/project-management/estimation [ID: 746686]
Set the version(s) of the product affected by this issue (comma-separated list)
Set the version(s) of the product in which the issue should be fixed (comma-separated list)
Set the version(s) of the product in which the issue fix was verified (comma-separated list)
Set if this issue occurs in production
[ID: 85206]
Set Reporter
Set Type
Set priority, which reflects how soon the issue should be fixed
Set Status
Set Assignee
Set Verifier
Remove item
View or edit staffing
View issue level access limits(Press Alt + Right arrow for more information)
Description
05-22 22:30:07.284 6029 6029 D Benchmark: cpu4 CoreDir(online=true, availableFreqs=[300000, 345600, 422400, 499200, 576000, 652800, 729600, 806400, 902400, 979200, 1056000, 1132800, 1190400, 1267200, 1344000, 1420800, 1497600, 1574400, 1651200, 1728000, 1804800, 1881600, 1958400, 2035200, 2112000, 2208000, 2265600, 2323200, 2342400, 2361600, 2457600], currentMinFreq=300000)
Looking at logcat, libperfmgr seems to be trying to change clocks, but failing a lot of the time:
05-22 22:30:11.954 806 2521 W libperfmgr: Failed to write to node: /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq with value: 1248000, fd: 10
05-22 22:30:11.954 806 2521 W libperfmgr: Failed to write to node: /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq with value: 1900800, fd: 11
05-22 22:30:11.954 806 2521 W libperfmgr: Failed to write to node: /sys/devices/system/cpu/cpu4/cpufreq/scaling_min_freq with value: 2457600, fd: 13
05-22 22:30:12.454 806 2521 W libperfmgr: Failed to write to node: /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq with value: 1248000, fd: 13
05-22 22:30:12.454 806 2521 W libperfmgr: Failed to write to node: /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq with value: 1900800, fd: 14
05-22 22:30:12.454 806 2521 W libperfmgr: Failed to write to node: /sys/devices/system/cpu/cpu4/cpufreq/scaling_min_freq with value: 2457600, fd: 10
Talked with wvw@, who mentioned they disable based on cpu governor, specifically cpu 0:
On walleye, we disable CPU 0 without bothering to set the governor. Fix is simply to set the governor even though we'll be disabling the CPU.
This issue will likely affect any devices with big cores numerically after little cores in the core list.