iMac G5 "windfarm"
Hi, it is my first time I run linux on an iMac G5 and have some doubts about its cooling. I noticed that while installing Debian, the fan spins at full speed and it really deserves the term "windfarm", it is as loud as a vacuum cleaner. Once Linux is fully installed, the system runs quiet when doing nothing (console prompt) and remains quiet when doing several things (also unexpected ones). However, e.g. opening GIMP, the fans whir up, just to slow down immediately. Then, running a filter is smooth (a bit unexpected). What I find strange is that fans run shortly full-speed, then stop after a second or sometimes two! I have never run MacOS on this system so I can't tell if this is normal and if the fan has any intermediate speeds, but this behavior seems strange. 1) if the CPU is under "slight load" I would expect the fans to stabilize to a low speed to keep it cool 2) if the CPU has high load indeed, then it probably takes more than just 1-2 seconds to clear it! in dmesg I see: [54879.528590] windfarm: Overtemp condition cleared ! [55034.523862] windfarm: Clamping CPU frequency to minimum ! [55034.570358] windfarm: Overtemp condition detected ! [55036.462349] windfarm: CPU frequency unclamped ! [55036.513668] windfarm: Overtemp condition cleared ! Here it looks that my CPU frequency is temporarily reduced (at the same time I hear the fan spinning) and if you see the intervals are very short (only one cycle copy-pasted here, dmesg is much longer) Riccardo
Re: iMac G5 "windfarm"
Hi Riccardo, On Wed, Dec 30, 2020 at 9:57 AM Riccardo Mottola wrote: > > Hi, > > it is my first time I run linux on an iMac G5 and have some doubts about > its cooling. There was a regression at some point, did you check: * https://lists.debian.org/debian-powerpc/2020/03/msg00035.html What's your kernel version ? > I noticed that while installing Debian, the fan spins at full speed and > it really deserves the term "windfarm", it is as loud as a vacuum cleaner. > > Once Linux is fully installed, the system runs quiet when doing nothing > (console prompt) and remains quiet when doing several things (also > unexpected ones). > > However, e.g. opening GIMP, the fans whir up, just to slow down > immediately. Then, running a filter is smooth (a bit unexpected). > > What I find strange is that fans run shortly full-speed, then stop after > a second or sometimes two! > > I have never run MacOS on this system so I can't tell if this is normal > and if the fan has any intermediate speeds, but this behavior seems strange. > > 1) if the CPU is under "slight load" I would expect the fans to > stabilize to a low speed to keep it cool > > 2) if the CPU has high load indeed, then it probably takes more than > just 1-2 seconds to clear it! > > > in dmesg I see: > > [54879.528590] windfarm: Overtemp condition cleared ! > [55034.523862] windfarm: Clamping CPU frequency to minimum ! > [55034.570358] windfarm: Overtemp condition detected ! > [55036.462349] windfarm: CPU frequency unclamped ! > [55036.513668] windfarm: Overtemp condition cleared ! > > Here it looks that my CPU frequency is temporarily reduced (at the same > time I hear the fan spinning) and if you see the intervals are very > short (only one cycle copy-pasted here, dmesg is much longer) > > > Riccardo >
G5 compilation speed vs G4
Hi, now that the iMac G5 is up and running, I did run the same build task as on my iBook G4 and PowerBook G4, where I already found some differences in speed, given a similar CPU and setup - We attribtuted them to the quite much faster disk in the iBook. Now comes the G5 (PPC970fx) which runs runs 64bit, has twice the RAM of the iBook, a SATA 7200rpm hard disk and 2.1GHz compared to 1.2GHz (although I know this is comparing two different CPUs, just noting that) Again? The iMac G5 is about as fast as the iBook G4! When using the computer, it feels definitely faster... but compiling ArcticFox takes about the same. Unexpected. Could be there a performance issue due to cooling, CPU scaling or such? When compiling, the iMac remains almost quiet spinning up the fans from time to time (see other post). My iBook instead has them whirring up full speed for hours! Riccardo
Re: iMac G5 "windfarm"
Hi Mathieu, On 12/30/20 10:22 AM, Mathieu Malaterre wrote: There was a regression at some point, did you check: *https://lists.debian.org/debian-powerpc/2020/03/msg00035.html What's your kernel version ? Right now, it apparently is an older version: Linux PPC970FX 4.16.0-1-powerpc64 #1 SMP Debian 4.16.5-1 (2018-04-29) ppc64 GNU/Linux which appears to be around the critical one cited there about issues. However, theere the issue was "windfarm not loaded". To me here it seems to be loaded, to be almost too aggressive. In lsmod I see. windfarm_cpufreq_clamp 3213 1 windfarm_smu_sat 7464 0 windfarm_smu_sensors 7223 6 windfarm_smu_controls 7878 3 windfarm_pm121 12261 0 windfarm_pid 2995 1 windfarm_pm121 I installed also a newer kernel: -rw-r--r-- 1 root root 169543 Apr 29 2018 config-4.16.0-1-powerpc64 lrwxrwxrwx 1 root root 28 Dec 29 18:21 initrd.img -> initrd.img-5.9.0-5-powerpc64 -rw-r--r-- 1 root root 24420098 Dec 29 18:22 initrd.img-4.16.0-1-powerpc64 -rw-r--r-- 1 root root 27346593 Dec 29 18:22 initrd.img-5.9.0-5-powerpc64 lrwxrwxrwx 1 root root 29 Dec 29 14:16 initrd.img.old -> initrd.img-4.16.0-1-powerpc64 drwx-- 2 root root 12288 Dec 29 14:09 lost+found -rw-r--r-- 1 root root 4126454 Apr 29 2018 System.map-4.16.0-1-powerpc64 lrwxrwxrwx 1 root root 25 Dec 29 18:21 vmlinux -> vmlinux-5.9.0-5-powerpc64 -rwxr-xr-x 1 root root 25371992 Apr 29 2018 vmlinux-4.16.0-1-powerpc64 lrwxrwxrwx 1 root root 26 Dec 29 14:16 vmlinux.old -> vmlinux-4.16.0-1-powerpc64 But yaboot doesn't find it, so I have to boot "old". Maybe this is part of the issue. TO me vimlinux seems as correct as vmlinux.old yaboot has: image=/vmlinux label=Linux read-only initrd=/initrd.img image=/vmlinux.old label=old read-only initrd=/initrd.img.old so it looks all fine to me, why should "old" work and Linux not? Riccardo
firmware - nonfree repository
Hi, I just noticed I have this issue seen in dmesg: [ 21.449826] radeon :04:00.0: firmware: failed to load radeon/R300_cp.bin (-2) [ 21.449835] radeon :04:00.0: Direct firmware load for radeon/R300_cp.bin failed with error -2 [ 21.450016] [drm:.r100_cp_init [radeon]] *ERROR* Failed to load firmware! [ 21.450096] radeon :04:00.0: failed initializing CP (-2). [ 21.450107] radeon :04:00.0: Disabling GPU acceleration I bet I simply miss the ATI firmware. Is this because I miss non-free software? or instead architecture-shared repositories? Right now, I have configured what Adrian suggested in a mail a couple of months ago: # binary default deb http://ftp.ports.debian.org/debian-ports/ unstable main deb http://incoming.ports.debian.org/buildd/ unstable main deb http://ftp.ports.debian.org/debian-ports/ unreleased main Riccardo
Re: firmware - nonfree repository
On 12/31/20 12:29 AM, Riccardo Mottola wrote: > I bet I simply miss the ATI firmware. Is this because I miss non-free > software? or instead architecture-shared repositories? > > Right now, I have configured what Adrian suggested in a mail a couple of > months ago: You need to add this line: deb [arch=all] http://ftp.debian.org/debian/ unstable contrib non-free Adrian -- .''`. John Paul Adrian Glaubitz : :' : Debian Developer - glaub...@debian.org `. `' Freie Universitaet Berlin - glaub...@physik.fu-berlin.de `-GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913
Re: iMac G5 "windfarm"
On Wed, Dec 30, 2020 at 2:55 PM Riccardo Mottola wrote: > But yaboot doesn't find it, so I have to boot "old". Maybe this is part > of the issue. > make sure you run ybin -v after making changes cam
Re: iMac G5 "windfarm"
You have vmlinux -> vmlinux-5.9.0-5-powerpc64 But there is no actual file named "vmlinux-5.9.0-5-powerpc64" in the directory. Maybe that's your problem? Rick On Wed, Dec 30, 2020, at 2:55 PM, Riccardo Mottola wrote: > Hi Mathieu, > > > On 12/30/20 10:22 AM, Mathieu Malaterre wrote: > > There was a regression at some point, did you check: > > > > *https://lists.debian.org/debian-powerpc/2020/03/msg00035.html > > > > What's your kernel version ? > > > Right now, it apparently is an older version: > > Linux PPC970FX 4.16.0-1-powerpc64 #1 SMP Debian 4.16.5-1 (2018-04-29) > ppc64 GNU/Linux > > which appears to be around the critical one cited there about issues. > However, theere the issue was "windfarm not loaded". To me here it seems > to be loaded, to be almost too aggressive. > > In lsmod I see. > > windfarm_cpufreq_clamp 3213 1 > windfarm_smu_sat 7464 0 > windfarm_smu_sensors 7223 6 > windfarm_smu_controls 7878 3 > windfarm_pm121 12261 0 > windfarm_pid 2995 1 windfarm_pm121 > > > I installed also a newer kernel: > > -rw-r--r-- 1 root root 169543 Apr 29 2018 config-4.16.0-1-powerpc64 > lrwxrwxrwx 1 root root 28 Dec 29 18:21 initrd.img -> > initrd.img-5.9.0-5-powerpc64 > -rw-r--r-- 1 root root 24420098 Dec 29 18:22 initrd.img-4.16.0-1-powerpc64 > -rw-r--r-- 1 root root 27346593 Dec 29 18:22 initrd.img-5.9.0-5-powerpc64 > lrwxrwxrwx 1 root root 29 Dec 29 14:16 initrd.img.old -> > initrd.img-4.16.0-1-powerpc64 > drwx-- 2 root root 12288 Dec 29 14:09 lost+found > -rw-r--r-- 1 root root 4126454 Apr 29 2018 System.map-4.16.0-1-powerpc64 > lrwxrwxrwx 1 root root 25 Dec 29 18:21 vmlinux -> > vmlinux-5.9.0-5-powerpc64 > -rwxr-xr-x 1 root root 25371992 Apr 29 2018 vmlinux-4.16.0-1-powerpc64 > lrwxrwxrwx 1 root root 26 Dec 29 14:16 vmlinux.old -> > vmlinux-4.16.0-1-powerpc64 > > > But yaboot doesn't find it, so I have to boot "old". Maybe this is part > of the issue. > > TO me vimlinux seems as correct as vmlinux.old > > yaboot has: > > image=/vmlinux > label=Linux > read-only > initrd=/initrd.img > > image=/vmlinux.old > label=old > read-only > initrd=/initrd.img.old > > > so it looks all fine to me, why should "old" work and Linux not? > > > Riccardo > >
