Using advice from this forum, I got tap to click working in LXDE on my Powerbook G4 running Debian 8.7. I have since been experimenting with the MATE desktop, how do I get tap to click working in MATE? Thank you.

I'm attempting to build a 32-bit toolchain in a docker container with an aarch64 architecture. I have explicitly called --without-long-double-128 in my first pass of gcc but glibc is failing compile due to missing __uint128_t types. I've looked through libc configuration options and I don't see a way to explicitly avoid these types. Does anyone know a way around this? **First Error** 1.504 /opt/usr/include/asm/sigcontext.h:81:9: error: unknown type name '__uint128_t'; did you mean '__uint32_t'? 1.504 81 | __uint128_t vregs; 1.504 | ^~~~~~~~~~~ 1.504 | __uint32_t 1.514 In file included from ../sysdeps/unix/sysv/linux/powerpc/sys/user.h:24, 1.514 from ../sysdeps/unix/sysv/linux/sys/procfs.h:35, 1.514 from ../sysdeps/nptl/thread_db.h:28, 1.514 from ../nptl/descr.h:31: **GCC first pass config** RUN mkdir -p ${WORKROOT}/src/gcc/build WORKDIR ${WORKROOT}/src/gcc/build RUN ../configure --prefix=${WORKROOT}/tools \ --target=powerpc-linux-gnu \ --with-sysroot=${WORKROOT} \ --with-newlib \ --without-headers \ --enable-initfini-array \ --enable-default-pie \ --enable-default-ssp \ --disable-nls \ --disable-shared \ --disable-multilib \ --disable-decimal-float \ --disable-threads \ --disable-libatomic \ --disable-libgomp \ --disable-libquadmath \ --disable-libssp \ --disable-libvtv \ --disable-libstdcxx \ --disable-gcov \ --without-long-double-128 \ --enable-languages=c,c++ **glibc config** RUN ../configure \ --prefix=/usr \ --host=powerpc-linux-gnu \ --build=$(../scripts/config.guess) \ --enable-kernel=4.14 \ --with-headers=${WORKROOT}/usr/include \ libc_cv_slibdir=/usr/lib

I'm in the process of upgrading my desktop/server setup and am considering using POWER9 architecture. However, I'm a bit concerned about the level of support it receives from GNU/Linux distributions and applications. Can anyone provide insights on how well supported POWER9 is within the GNU/Linux ecosystem in 2024? Any advice or recommendations would be greatly appreciated!

I am working on a PowerPC board using kernel version 4.1.35 and encountering significant boot **delays**, specifically around the RTC clock (ds-1337) operations. During the boot process, a script (hwclock.sh) runs the hwclock --hctosys command, and I've added print statements to log its execution time in the kernel logs. Here’s a snippet from hwclock.sh: echo "running command hwclock --hctosys. - arij" | tee /dev/kmsg hwclock $tz --hctosys echo "ending hwclock --hctosys. - arij" | tee /dev/kmsg When the kernel configuration CONFIG_DYNAMIC_DEBUG is disabled, the hwclock command experiences significant delays, sometimes up to 50 seconds. However, when **CONFIG_DYNAMIC_DEBUG** is enabled, the delays are significantly reduced, with execution times ranging between 4 to 12 seconds. Here are some kernel log snippets from various boot cycles with **CONFIG_DYNAMIC_DEBUG** disabled: **Boot cycle 1:**

[    6.401707] running command hwclock --hctosys. - arij
[   39.206200] ending hwclock --hctosys. - arij

**Boot cycle 2:**

[    7.054404] running command hwclock --hctosys. - arij
[   14.218021] ending hwclock --hctosys. - arij

**Boot cycle 5:**

[   18.659482] running command hwclock --hctosys. - arij
[   50.005113] ending hwclock --hctosys. - arij

With CONFIG_DYNAMIC_DEBUG enabled, the delays decrease: Example with **CONFIG_DYNAMIC_DEBUG **enabled:

[   8.833980] running command hwclock --hctosys. - arij
[  14.223970] ending hwclock --hctosys. - arij

Additionally, when booting with the configuration CONFIG_PROFILING enabled, the hwclock command consistently takes no more than 2 seconds. **Questions** 1. What could be causing the drastic increase in boot time when CONFIG_DYNAMIC_DEBUG is disabled? 1. Why does enabling CONFIG_DYNAMIC_DEBUG reduce the time taken for the hwclock --hctosys command? 1. Could there be any underlying issues with the RTC driver or interactions between the hwclock command and kernel configurations? 1. How might CONFIG_PROFILING impact the timing of hwclock --hctosys? 1. Any insights into these behaviors would be greatly appreciated. **Update 1:** i have tried multiple kernel prints in the rtc drivers and it only showed that the mutex lock is waiting to be acquired. i cant understand whats causing the problem and how is it not appearing when dynamic debugging is enabled the prints i enabled are like this: pr_info("--------------------- %s: %s: Line %d - Mutex Locking interruptible - arij\n", __FILE__, __func__, __LINE__); int err = mutex_lock_interruptible(&rtc->ops_lock); pr_info("--------------------- %s: %s: Line %d - Mutex Locked interruptible - arij\n", __FILE__, __func__, __LINE__); and got prints: [ 17.450075] running command hwclock --hctosys. - arij [ 17.461239] rtc_core: --------------------- /home/pc/kernel-source/drivers/rtc/rtc-dev.c: rtc_dev_ioctl: Line 234 - Mutex locking- arij [ 23.737863] rtc_core: --------------------- /home/pc/kernel-source/drivers/rtc/rtc-dev.c: rtc_dev_ioctl: Line 236 - Mutex locked with cmd=1076129801 - arij [ 23.752437] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_read_time: Line 62 - Mutex Locking interruptible - arij [ 27.005861] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_read_time: Line 65 - Mutex Locked interruptible - arij [ 27.020196] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_read_time: Line 74 - Mutex Unlock interruptible - arij [ 27.034029] rtc_core: --------------------- /home/pc/kernel-source/drivers/rtc/rtc-dev.c: rtc_dev_ioctl: Line 234 - Mutex locking- arij [ 30.637859] rtc_core: --------------------- /home/pc/kernel-source/drivers/rtc/rtc-dev.c: rtc_dev_ioctl: Line 236 - Mutex locked with cmd=536899588 - arij [ 30.652312] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_update_irq_enable: Line 515 - Mutex Locking interruptible - arij [ 34.805855] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_update_irq_enable: Line 517 - Mutex Locked interruptible - arij [ 34.819887] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_update_irq_enable: Line 552 - Mutex unock in out: - arij [ 34.833291] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_update_irq_enable: Line 515 - Mutex Locking interruptible - arij [ 62.173853] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_update_irq_enable: Line 517 - Mutex Locked interruptible - arij [ 62.187884] --------------------- /home/pc/kernel-source/drivers/rtc/interface.c: rtc_update_irq_enable: Line 552 - Mutex unock in out: - arij [ 62.204506] ending hwclock --hctosys. - arij **Update 2:** I also changed the configuration to not get a timeless kernel. but i still observed the same delays CONFIG_HZ_PERIODIC=y CONFIG_NO_HZ_IDLE not set CONFIG_NO_HZ not set

Is there a tool similar to debootstrap for a recent Linux distribution that supports PPC so I could build a working system or have an automated process to cross-compile a base install?

So, I'm having an old PowerPC based IBM server, RS/6000 F50- that I'd like to get running simply for my own fun and hobby. There are several different OSes that can be installed on it and I was going to try out the Debian 8.10.0. The server boots from the CD-ROM, but after the installer loads and asks first few basic questions, it stops with an error: No common CD-ROM drive was detected. After that if offers me to load a driver from the floppy, select cdrom device manually or I can run installer shell. I tried to find out more in the shell. I see that there is a device /dev/cdrom that links to /dev/sr0. I tried to read some data from /dev/sr0 using a dd and I see data coming from it. blkid detects CD FS as iso9660. I tried to mount device /dev/sr0 to /cdrom with no luck. Command: mount -t iso9660 /dev/sr0 /cdrom/ fails with "Invalid argument". Is it possible that installer does not support the file system that's on its own CD? What could be the possible solutions to mount the CD-ROM drive? I also tried netinst image, but it stops the same way. I tried to install AIX 4.3.3, it installs but with some other unrelated issues. I have some older Debian images that I'll try out too. The server is in minimal configuration - no additional cards, one HDD, one CD-ROM, no video, console only.

I have an old IBM 9111-285 with AIX 7.1 Consider that Aix 7.2 don't boot on oldest ppc cpu's (this is a power5!) and Aix 7.1 is near EOL as I know I want to install Linux on it. Actually I have those "solutions" Linux Debian 11: WORKS ONLY ON PPC64LE, this is BE Adelie Linux: DON'T BOOT Linux Debian 12 (is a port for ppc64): boot but..kernel panic. Centos 6 ppc64: is EOL Centos 7 ppc64: kernel panic Gentoo ppc64: Boot and hang after 1 minute with no error, probably a panic. The distribution Opensuse Tumbleweed has dvd iso images for ppc64 and ppc64le I boot the cd and..kernel panic because cannot mount root image, I think is probably because my workstation has IDE DVD and a lot of Linux distro now don't support old ide with default kernel So I try the network boot, I have setup a dhcp/tftp/bootp server, qemu-system-ppc64 can boot (go to openfirmware prompt and after type boot net I get the grub menu and installation start). With my Ibm workstation I go to openfirmware prompt and... 0 > boot net BOOTP: chosen-network-type = ethernet,auto,rj45,auto BOOTP: server IP = 0.0.0.0 BOOTP: requested filename = BOOTP: client IP = 0.0.0.0 BOOTP: client HW addr = ********* BOOTP: gateway IP = 0.0.0.0 BOOTP: device /pci@800000020000004/pci@2,4/ethernet@1 BOOTP: loc-code U787F.001.DPM5MVH-P1-T5 BOOTP: wait 60 seconds for Spanning Tree ... BOOTP: BOOTP request fail: 0 BOOTP: BOOTP request fail: 1 BOOTP: BOOTP request fail: 2 BOOTP ERROR: BOOTP request failed, QUIT !BA010004 ! I try to specify all parameters and... 0 > boot net:192.168.0.88,core.elf,192.168.0.10,192.168.0.1 BOOTP: chosen-network-type = ethernet,auto,rj45,auto BOOTP: server IP = 192.168.0.88 BOOTP: requested filename = core.elf BOOTP: client IP = 192.168.0.10 BOOTP: client HW addr = **************** BOOTP: gateway IP = 192.168.0.1 BOOTP: device /pci@800000020000004/pci@2,4/ethernet@1 BOOTP: loc-code U787F.001.DPM5MVH-P1-T5 BOOTP: wait 60 seconds for Spanning Tree ... icmp 5 : redirect icmp 3 : port unreachable ERROR: icmp 3 BOOTP: BOOTP request fail: 0 icmp 3 : port unreachable ERROR: icmp 3 BOOTP: BOOTP request fail: 1 icmp 3 : port unreachable ERROR: icmp 3 BOOTP: BOOTP request fail: 2 BOOTP ERROR: BOOTP request failed, QUIT !BA010004 ! Any solution? EDIT1: Ubuntu 16.04 load fine with this option (but don't go over the lang selection) boot: install-powerpc64 console=hvc0 console=tty0 console=tty1 console=ttyS0,9600,8,n,1 console=ttyS1,9600,8,n,1

My AIX machine is currently running on 6.1 OS. I would like to know whether it is on PPC OR PPC64, Is there any command other than arch which gives the information about PPC ?

I've been doing some research on ppc64le Linux and am unclear on how the POWER CPU is set to little endian. As I understand it the machine and CPU state will run as big endian by default in firmware. But what is responsible for switching into the intended endian when launching the kernel? Without getting into specifics launching a kernel on POWER/PPC would be summarised as follows: - Firmware executes on machine reset and resets machine state. - Firmware loads bootloader and passes over control. - Boot loader loads kernel and launches it. I've searched online and found no answers. I've looked into boot loader code such as grub, qemu-vof and slof, though there is a lot to sift through, and it was unclear. I've looked into Linux startup code and can see it reads the MSR_LE state at times but I cannot see it setting it on startup. So, what is responsible for setting endian? The boot loader or the kernel? Or even something else? My answer would be the boot loader since it is responsible for loading in the ELF image and launching it in the intended endian. But I so far haven't found any ASM sources where this is clear when launching into an entry point. I'd appreciate finding out what is responsible for setting endian and any source listings detailing how it is done.

The [IBM AIX Documentation](https://www.ibm.com/docs/en/aix/7.3?topic=set-svc-supervisor-call-instruction) for the svc (Supervisor Call) instruction has a note that states > To ensure correct operation, an svc instruction must be preceded by an unconditional branch or a CR instruction. Why is this required?

My desire is conceptually simple, I have a file (really a PCIe resource file from /sys/bus/pci/device/.... but that isn't too relevant) on the host that I want to make available somewhere in guest memory, so that changes from either side get reflected to each other. Since my goal was to actually map a limited segment of PCIe address space in the host, I couldn't productively map the entire guest RAM. The base command that I am trying to add is listed below. The goal is to get memory id "bar0.ram" mapped *somewhere* in guest memory. qemu-system-ppc -M ppce500 -cpu e500 -m 64M -d guest_errors,unimp -bios $PWD/test.elf -s -object memory-backend-file,size=1m,id=bar0.ram,mem-path=/sys/bus/pci/devices/0000\:04\:00.0/resource0,share=on -monitor telnet:127.0.0.1:4999,server,nowait -nographic Perhaps this would be easier on ARM or x86, but PPC doesn't offer persistent memory, nvram, multiple memory slots backed by different files, or similar tricks (that I could figure out how to get working). It does offer ivshmem, but I was unable to figure out how to get that to be transparently mapped into the guest address space. Vaguely useful/related resources: - https://unix.stackexchange.com/questions/616596/mapping-guest-ram-to-file-in-qemu - https://superuser.com/questions/1795238/qemu-system-sparc-shared-memory-between-host-and-guest - https://blog.reds.ch/?p=1379

I am using Debian 8.6 LXDE on a Powerbook G4 15" 1.67GHz and would like to enable tap to click on the touchpad. It is already double scrolling but tap to click would help to save the ageing mouse button. Two fingered tap for left click would be the icing on the cake, is this possible?

I have an old Apple PowerBook on which I dual boot Void Linux and Mac OSX Leopard. I had some hardware problems around July 2021 and it took until now for me to get it fixed. When I got it working again, it booted straight into Void Linux and I (unwisely) decided to run the sudo xbps-install -Su command to get it back up to date. I got the message that xbps needed to update first so I ran the appropriate command and then sudo xbps-install -Suagain. The update began running as normal but then I started getting error messages about things breaking like NTP daemon and WPA supplicant. Now it won't boot into a graphical environment (it was booting into Mate) but hangs at loading b43 firmware version 666.2 after about 35 seconds. If I use ctrl-alt-f1 at this point I can get to a command line interface and log in but then there are repeated error messages openntpd:segfault (11) at 4 nip a7b63f74 lr 74f4dc code 1 in libc-2.32.so[a7a2f0001c2000]... These keep repeating with different numbers but only if the ethernet cable is plugged in. I would be grateful for any advice on salvaging the install, I had it working really nicely and the hardware problem was unrelated to the Void Install. Or is it better just to start again? Sadly, the Void PPC port has just been deprecated this month so this would be a temporary solution while I decide what to install instead...

How do I get wifi to work on my powerbook g4 with xfce debian? I have tried sudo apt-get install bcm43xx-fwcutter but it comes up with E: Unable to locate package bcm43xx-fwcutter. What do I do?

I have small device that starts up a tiny VM of PowerPC linux that I need to SSH into from a script on a tiny version of Yocto linux. The PowerPC VM SSH passwords stay the same, but it resets _most_ of its core files, including regenerating its ssh keys, each restart. The Yocto linux installation doesn't have the 'expect' command, nor can it install 'sshpass'. It has an extremely limited version of Python. I'd like to be able to complete the password requirement from a basic 'ssh' prompt using only a shell script. Is this possible?

I am trying to cross-compile Linux-PAM from a Ubuntu host machine to powerpc architecture (running linux kernel 2.6.34). First, I run ./configure command, giving the --host path to the cross compiler and --prefix path to a local directory called "target" as options to the configure command; after that, I run make and make install. The make install command ends apparently without any error. At the end of the install process few files and directories are created in my "target" folder, but the /etc/pam.d is missing: I would have expected to find that folder, containing the PAM modules. A pam.conf file is missing too. Am I doing something wrong? Or maybe the lack of both /etc/pam.d and pam.conf is an expected behavior? these are the commands I tried to cross-compile: export PATH=/path/to/crosscompiler/$PATH ./configure --prefix=/home/target --host=powerpc-unknown-linux-gnu make make install Thanks a lot.

I have a qemu in which I have debian-10.0-powerpc-NETINST-1 plus sid as package source. I need powerpc to as a reference big endian architecture (maybe there's something easier?) My problem is that trying to install packages end me up with NO_PUBKEY errors, which I can't fight: apt install debian-archive-keyring did not do anything about it and I can't install gpg since I can't install **anything** due to broken apt-get with obsolete database which gives 404 for most of files. Please tell me any options to ignore NO_PUBKEY for good. Since it's a throwaway qemu I don't care in the slightest, just let me ignore the stuff and install packages.

I am working with an embedded device which communicates with my PowerPC CPU via PCIe. Due to the SDK constraints i have to use linux kernel version 4.1.8. This version doesn't have functions related to cache operations in $KERNEL_SOURCE/arch/powerpc/include/asm/, such as: - "invalidate_dcache_range()" - "flush_dcache_range()" - "clean_dcache_range()" (in this directory "cacheflush.h" header just contains declaration of mentioned functions.) Also, my embedded device's SDK needs to call these functions to prepare DMA access. Note that Kernel versions higher than 4.5 provide declaration of these functions. Now i have some questions: 1. I can use and copy functions which are implemented in newer kernels, in my old kernel and rebuild it, but modifying the kernel source doesn't make sense, right? 2. Can you suggest some workaround to resolve it? 3. How can i test that cache invalidated or flushed correctly? is there any way to read cache blocks in userspace? Thanks,

Debian dropped support for my PowerPC CPU. For this reason my security updates are now no longer granted and I am clearly not happy about that. Is there any distro out there that supports PowerPC and provides an automatic compilation of updates on my PC? In my hopes I can get a distro that in the moment I run something like apt upgrade downloads the packages from source, compiles locally and install. Is there such a thing?

I am trying to install CentOS 7.3 bare-metal on a ppc64le machine. After completing the installation process using vncviewer, the last screen says the installation is completed and to reboot the system. However when the system reboots, there are no boot entries for the disk that I installed CentOS on. EDIT: This is also happened when installing RHEL 7.2, as they both use Anaconda.