Greetings,Just an update: I've been running a (slightly) customize 6.3.0.y kernel from the RPi linux repo and while it's still a wee bit sluggish (compared to F-37), chronyd keeps perfect time and the CPU core's enter into "WAIT" a *whole* lot less than before. Planning a source refresh and another deep dive into the .config. I've upgraded my x86_64 laptop and the VirtualBox VM to 38 and they're blazingly fast. Not sure what the hiccup is with this aarch64 install I'm struggling with, but have made substantial progress...
That is all. :D On 4/26/2023 7:39 PM, Jeremy Linton wrote:
Hi, On 4/25/23 16:40, Randy DuCharme wrote:Am I the only one having difficulties with this? It's worse than half as fast as f37 was, NTP/Chrony refuses to work..... I could go on. Afraid I'm going to have to ditch this. Kernel's been building for 7 hours now. Clocking, over-clocking, governors, schedulers make no difference. Btrace, strace points to nothing. CPU load through the roof though yet I don't see anything especially noteworthy.Thoughts?F38 is behaving similarly to f37 on my rpi's so...You might try installing the kernel tools, and flipping on the debuginfos and running something like `perf top --vmlinux=/usr/lib/debug/lib/modules/`uname -r`/vmlinux` if you see its spending a lot of time in the kernel.What process is taking all the cpu time? You might turn on the mem/cpu/io pressure stall fields in htop and see what is being reported as the major bottleneck.. but i'm guessing your on a SD or a USB disk?Its a big weakness of that platform that it doesn't have a reasonable disk interface (aka sata/nvme/etc). So one of the failure modes I've seen with bad SD/USB disks is that they get incredibly slow (think KB/sec) before they die. The stress of downloading a couple GB, delta RPM'ing it, installing, etc, could have pushed your disk closer to failure. I've killed a fair number of disks before I switched to USB->sata enclosures with proper SATA SSDs*, and one of the first indications is the whole machine starts just acting slow. 15 second login times from the shell, seconds to start things like htop, etc.I guess it could also be thermal throttling if your building kernels as you report below. I've got a fairly beefy heatsink on mine in order to keep the temps below the throttling limits.As far as time does something like `systemctrl restart systemd-timesyncd` fix your time issues?* (I've recently started using samsung T7's which are blisteringly fast on proper USB3 controllers, which sadly doesn't include the rpi4 though)_______________________________________________ arm mailing list -- arm@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to arm-leave@xxxxxxxxxxxxxxxxxxxxxxxFedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelinesList Archives: https://lists.fedoraproject.org/archives/list/arm@xxxxxxxxxxxxxxxxxxxxxxx Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
-- Randall DuCharme (Radio AD5GB) Powered entirely by Open Source software.
Attachment:
OpenPGP_0x4BC19D71C1381AA3.asc
Description: OpenPGP public key
Attachment:
OpenPGP_signature
Description: OpenPGP digital signature
_______________________________________________ arm mailing list -- arm@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to arm-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/arm@xxxxxxxxxxxxxxxxxxxxxxx Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
