Jump to content
funtoo forums

paddymac

Members
  • Content Count

    39
  • Joined

  • Last visited

  • Days Won

    4

paddymac last won the day on October 14

paddymac had the most liked content!

2 Followers

About paddymac

  • Rank
    Advanced Member

Recent Profile Visitors

301 profile views
  1. Funtoo is my favorite Linux distro. I've been using it enthusiastically since at least 2012. I've been tinkering a lot with FreeBSD for the past 3 years which has divided my attention, but I still use Funtoo as my primary desktop OS. Here's my concern. I'm using Funtoo 1.2. In the past few months, I've encountered a LOT of blockers when trying to emerge world. Sometimes the blockers are things that don't exist in the portage tree, sometimes they are things which need to be unmasked, and sometimes they are things which need to be unmasked by keyword. I went through a short period of time where I filed a lot of bugs about these blockers. But one bug I filed was dismissed as a "workaround" because I filed a bug about something which needed to be unmasked. My feeling on the matter is that, if I try to emerge a package that it unmasked by default, then it should emerge without any problem. It should NOT require me to unmask anything or unmask something by keyword. Only ebuilds which are themselves masked or keyworded should rely on other ebuilds which are masked or keyworded. That's my opinion, but if that is not the opinion of the Funtoo project, I ask that such opinions be made plain so that people like me who love the project and want to contribute do not become nuisances by filing bugs or complaining when that is not the goal of the project. But at the same time, if that is not the goal of the project, then why are such bug reports being closed as "workaround" instead of being properly fixed? I asked Daniel and the rest of the Funtoo devs to please make a clear statement about these sorts of things so that users like me who are trying to help can be a genuine help without being a nuisance.
  2. paddymac

    Can't switch user from Plasma 5 lock screen

    Well, a few minutes after I wrote this question, I found this other forum post: It reminded me that I forgot to use rc-update to disable consolekit from default and to enable elogind on boot. I did so, rebooted, and now it works. SDDM seems a little slow now, and I still don't know why it doesn't work with consolekit, but I think this issue is solved for me.
  3. For some reason, if I click "Switch User" from the Plasma 5 lock screen, if I click the button to start a new session, it simply brings up the password prompt for the currently logged in user. I can't switch to a new user. I found a thread on the Gentoo forums about this same issue, and it was suggested to disable the consolekit USE flag, enable the elogind USE flag, and emerge @world to recompile everything that had a changed USE flag setting. I did that, but I had no change in the situation. I've had a desktop with OpenSUSE Leap 42.2 installed, and user switching from the Plasma 5 lock screen worked fine on that system. Does anyone have any ideas on how to get this working? It wasn't a big deal when I was the only one using the computer, but now that I'm sharing it with someone I need this feature to work.
  4. I've been trying to get early boot Intel microcode updating working on my Funtoo install. I have verified with microcode-ctl that I can update my CPU's microcode manually from the command line. However, after following the instructions at https://wiki.gentoo.org/wiki/Intel_microcodeI have been unable to get it working at boot. Here's an entry from my /etc/boot.conf "Funtoo - Gentoo 4.8.6 + CK/MuQSS 0.135 + UKSM" { kernel vmlinuz-4.8.6-gentoo-ck6 initrd /lib/firmware/microcode.cpio initramfs-genkernel-x86-4.8.6-gentoo-ck6 params += real_root=auto rootfstype=auto params += mtrr_reg_cleanup=6 } and the corresponding section of /boot/grub/grub.cfg menuentry "Funtoo - Gentoo 4.8.6 + CK/MuQSS 0.135 + UKSM - vmlinuz-4.8.6-gentoo-ck6" { insmod part_msdos insmod btrfs set root=(hostdisk//dev/sda,msdos1) search --no-floppy --fs-uuid --set e050f0b2-8494-4e53-a19f-6c57f036cc41 linux /boot/vmlinuz-4.8.6-gentoo-ck6 rootfstype=auto mtrr_reg_cleanup=6 real_root=/dev/sda1 rootfstype=btrfs initrd /boot/initramfs-genkernel-x86-4.8.6-gentoo-ck6 set gfxpayload=keep } Notice that the "/lib/firmware/microcode.cpio" was removed from the initrd line when grub.cfg was processed by the boot-update utility. boot-update needs to have support for this added, or some other method for providing the microcode to the kernel at boot should be proposed.
  5. paddymac

    Illegal instruction for sandybridge subarch

    I have to agree with Oleg that, if you're going to run a source-based distro, you should be familiar with your hardware. Are you getting the errors during compilation or during runtime? Generally if you're compiling software with cpu flags your processor doesn't recognize, the error will appear when you attempt to run the program -- not during compilation. I've seen the error plenty of times since I use some old hardware that doesn't support sse2 which is often required nowadays.
  6. I got a secondhand Dell Dimension 2400 that apparently has a buggy BIOS. I can't complain too much because I got it for free. It only has 256 MB RAM, but it'll support up to 2 GB RAM, and PC2700 is dirt cheap these days. My plan was to upgrade the RAM and use it as a transparent web filtering proxy with a Squid/e2guardian setup. The only functional problem I've encountered is that sometimes it seems a little slow. I've got Funtoo installed, and I thought perhaps the low 256 MB RAM was the problem (it may be), but even when RAM usage reported by top is only around 20% by a process and disk usage is normal, it takes longer for this Pentium 4 at 2.4 GHz to do the same thing as an AMD Sempron at 2 GHz (which has 768 MB RAM). I'm wondering if the slowness is due to the timer bug reported by dmesg. In any case, I'll post below the dmesg output that contains error messages. Does anyone know if this is something I can work around? I'm currently running the Debian sources 4.2.6-3. I upgraded the kernel thinking a newer version would fix the issues reported under kernel 3.19.3, but they remained. Maybe someone knows some fixes? [ 0.000000] ACPI: Early table checksum verification disabled [ 0.000000] ACPI: RSDP 0x00000000000FEB80 000014 (v00 DELL ) [ 0.000000] ACPI: RSDT 0x00000000000FD22A 000034 (v01 DELL 2400 00000007 ASL 00000061) [ 0.000000] ACPI: FACP 0x00000000000FD25E 000074 (v01 DELL 2400 00000007 ASL 00000061) [ 0.000000] ACPI: DSDT 0x00000000FFFCC0F1 002404 (v01 DELL dt_ex 00001000 MSFT 0100000D) [ 0.000000] ACPI: FACS 0x000000000FE74000 000040 [ 0.000000] ACPI: SSDT 0x00000000FFFCE632 0000BA (v01 DELL st_ex 00001000 MSFT 0100000D) [ 0.000000] ACPI: APIC 0x00000000000FD2D2 00006C (v01 DELL 2400 00000007 ASL 00000061) [ 0.000000] ACPI: BOOT 0x00000000000FD33E 000028 (v01 DELL 2400 00000007 ASL 00000061) [ 0.000000] ACPI: Local APIC address 0xfee00000 [ 0.105736] Simple Boot Flag value 0x87 read from CMOS RAM was invalid [ 0.105792] Simple Boot Flag at 0x7a set to 0x1 [ 0.128197] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S2_] (20150619/hwxface-580) [ 0.128355] ACPI: (supports S0 S1 S3 S4 S5) [ 0.128407] ACPI: Using IOAPIC for interrupt routing [ 0.128585] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug [ 0.173283] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff]) [ 0.173362] acpi PNP0A03:00: _OSC: OS supports [ASPM ClockPM Segments MSI] [ 0.173428] acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM [ 0.175058] acpi PNP0A03:00: fail to add MMCONFIG information, can't access extended PCI configuration space under this bridge. [ 0.175517] PCI host bridge to bus 0000:00 [ 0.175582] pci_bus 0000:00: root bus resource [bus 00-ff] [ 0.175636] pci_bus 0000:00: root bus resource [io 0x0000-0x0cf7 window] [ 0.175691] pci_bus 0000:00: root bus resource [io 0x0d00-0xffff window] [ 0.175745] pci_bus 0000:00: root bus resource [mem 0x000a0000-0x000bffff window] [ 0.175807] pci_bus 0000:00: root bus resource [mem 0x80860000-0xfebfffff window] [ 0.175891] pci 0000:00:00.0: [8086:2560] type 00 class 0x060000 [ 0.175917] pci 0000:00:00.0: reg 0x10: [mem 0xf8000000-0xfbffffff pref] [ 0.176266] pci 0000:00:02.0: [8086:2562] type 00 class 0x030000 [ 0.176304] pci 0000:00:02.0: reg 0x10: [mem 0xf0000000-0xf7ffffff pref] [ 0.176316] pci 0000:00:02.0: reg 0x14: [mem 0xfeb80000-0xfebfffff] [ 0.176651] pci 0000:00:1d.0: [8086:24c2] type 00 class 0x0c0300 [ 0.176716] pci 0000:00:1d.0: reg 0x20: [io 0xff80-0xff9f] [ 0.176907] pci 0000:00:1d.0: System wakeup disabled by ACPI [ 0.177087] pci 0000:00:1d.1: [8086:24c4] type 00 class 0x0c0300 [ 0.177150] pci 0000:00:1d.1: reg 0x20: [io 0xff60-0xff7f] [ 0.177366] pci 0000:00:1d.1: System wakeup disabled by ACPI [ 0.177541] pci 0000:00:1d.2: [8086:24c7] type 00 class 0x0c0300 [ 0.177606] pci 0000:00:1d.2: reg 0x20: [io 0xff40-0xff5f] [ 0.177801] pci 0000:00:1d.2: System wakeup disabled by ACPI [ 0.177997] pci 0000:00:1d.7: [8086:24cd] type 00 class 0x0c0320 [ 0.178041] pci 0000:00:1d.7: reg 0x10: [mem 0xffa80800-0xffa80bff] [ 0.178150] pci 0000:00:1d.7: PME# supported from D0 D3hot D3cold [ 0.178394] pci 0000:00:1e.0: [8086:244e] type 01 class 0x060400 [ 0.178608] pci 0000:00:1e.0: System wakeup disabled by ACPI [ 0.178798] pci 0000:00:1f.0: [8086:24c0] type 00 class 0x060100 [ 0.178812] * The chipset may have PM-Timer Bug. Due to workarounds for a bug, * this clock source is slow. If you are sure your timer does not have * this bug, please use "acpi_pm_good" to disable the workaround [ 0.179020] pci 0000:00:1f.0: can't claim BAR 13 [io 0x0800-0x087f]: address conflict with ACPI CPU throttle [io 0x0810-0x0815] [ 0.201975] pci 0000:00:1d.7: can't claim BAR 0 [mem 0xffa80800-0xffa80bff]: no compatible bridge window [ 0.242275] pnp 00:04: [dma 0 disabled] [ 0.242887] pnp 00:04: Plug and Play ACPI device, IDs PNP0401 (active) [ 0.244947] system 00:05: [io 0x0800-0x085f] could not be reserved [ 0.245012] system 00:05: [io 0x0c00-0x0c7f] has been reserved [ 0.245067] system 00:05: [io 0x0860-0x08ff] could not be reserved [ 0.245131] system 00:05: Plug and Play ACPI device, IDs PNP0c01 (active) [ 0.245968] pnp: PnP ACPI: found 6 devices [ 0.246036] PnPBIOS: Disabled by ACPI PNP [ 0.289731] clocksource: acpi_pm: mask: 0xffffff max_cycles: 0xffffff, max_idle_ns: 2085701024 ns [ 0.289845] clocksource: Switched to clocksource acpi_pm [ 0.289925] pci 0000:01:09.0: can't claim BAR 6 [mem 0xfea00000-0xfea03fff pref]: address conflict with 0000:01:04.0 [mem 0xfea00000-0xfea1ffff pref] [ 0.290025] pci 0000:00:1f.0: BAR 13: [io size 0x0080] has bogus alignment [ 12.710629] [drm:intel_parse_bios [i915]] *ERROR* General definiton block child device size is too small. NOTE: The above drm error is the only message that appears in red text in dmesg output. [ 22.965671] smsc47m1: Found SMSC LPC47M10x/LPC47M112/LPC47M13x [ 22.965700] ACPI Warning: SystemIO range 0x0000000000000C04-0x0000000000000C04 conflicts with OpRegion 0x0000000000000C04-0x0000000000000C08 (\WSTX) (20150619/utaddress-254) [ 22.965714] ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver [ 25.816244] b44 ssb0:0 eth1: Link is up at 100 Mbps, full duplex [ 25.816256] b44 ssb0:0 eth1: Flow control is off for TX and off for RX [ 25.816430] IPv6: ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready [ 701.073343] smsc47m1: Found SMSC LPC47M10x/LPC47M112/LPC47M13x [ 701.073376] ACPI Warning: SystemIO range 0x0000000000000C04-0x0000000000000C04 conflicts with OpRegion 0x0000000000000C04-0x0000000000000C08 (\WSTX) (20150619/utaddress-254) [ 701.073390] ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver Also, even though smsc47m1 is found, the kernel module is not automatically loaded. And if I attempt to load it manually with modprobe, I get the error message "modprobe: ERROR: could not insert 'smsc47m1': Device or resource busy"
  7. I've been working with a lot of software and text files from the old DOS days of BBSing before the rise of the internet. The primary form of graphics in that era was ANSI - a glorious 16 colors and 256 characters. The first 0-127 characters, or low ASCII, are not a problem even now, but apparently characters 128-255, or high ASCII, were not actually standardized. Linux terminals do seem to have partial ANSI emulation down. Colors seem to display alright, for example, but certain features apparently don't work correctly so that sometimes lines of text don't end when they should so that what should appear on multiple lines continues unabated on a single line. Or sometimes text appears misaligned either a little or quite badly. When accessing online BBS systems via telnet, putty, gcomm, or other such programs, the graphics don't appear correctly. There are a handful of specialized terminals that handle everything internally and get it right. So at least there are a few viable options here. But when I open such files in an editor such as nano, vi, or try to display them with cat, more, or other programs, they do not display correctly. Following are a few screenshots to try to explain the issue visually.
  8. paddymac

    xf86-input-evdev patch

    You might not be doing anything wrong. It could be an issue with the patch itself. I noticed a discussion regarding Arch Linux's implementation, and the same problem is being reported there, too: https://aur.archlinux.org/packages/xf86-input-evdev-debounce/ Hmm... On second thought, you say that it doesn't work right even if the use flag is disabled, right? Sometimes something as simple as having eclasses inherited in the wrong order can cause issues. Maybe move eutils to the beginning of the line. I'm really not sure just from looking at the ebuilds.
  9. paddymac

    GCC 5

    Eventually, gcc will be updated. Funtoo just tends to be conservative with the toolchain since Funtoo's primary focus is on servers. Still, I agree that I'd like to see newer toolchain packages available for those who want them, such as developers. In any case, you can write/modify ebuilds yourself or even use Gentoo's ebuilds. You can also use the toolchain overlay if you wish. There could be potential issues, but I've personally not run into any. In any case, I've written a Funtoo-style, self-contained gcc 5.3.0 ebuild for my own use which I can share. Iit's based on Funtoo's gcc-4.9.3-r1, so it doesn't apply the plethora of Gentoo patches and, as a result, probably won't work properly on a hardened system. I did try using 4.9.3-r2 as a base (for compatibility with hardened), but I ran into a mysterious bug resulting in failure to apply all of the Gentoo patchset.
  10. I'm sure it can be done, but I don't know how. Google "Gentoo FreeBSD" and you'll come up with some somewhat outdated info. Since it can be done on Gentoo, it should also be doable on Funtoo. But you need to speak to someone more knowledgeable than me to get any help with it.
  11. paddymac

    kde4 how to stay with it...

    When you emerge konsole, just specify the correct slot for KDE4. e.g. "emerge -va konsole:4". That *should* ensure you're just pulling in KDE4/QT4 stuff. Because if you don't specify a slot, it'll automatically prefer the higher version number which is of course going to be what's in slot 5.
  12. paddymac

    Automatic system updates?

    I was thinking earlier today that it would be nice if there were some way to have a Funtoo system automatically update periodically. It wouldn't have to be anything complex -- maybe just a shell script periodically run by cron. In fact, it would be easy to set up a cron job to periodically sync the Portage tree and run 'emerge -1uDNv --with-bdeps=y @world' or a similar command. But what if there were some kind of error that required human interaction? Then it would need to detect that error and notify the user -- perhaps by sending a system message to the root user or something. I was just wondering if something like that was already available or if any users had set up anything of the sort on their own systems.
  13. paddymac

    Can't start X as a regular user

    If you're trying to stat plain X, you must install twm, xterm, and xclock. If you're not trying to start plain X, edit /etc/init.d/xdm and put your preferred display mananger in there. The error about radeonsi_dri.so could be that you don't have that file. Did you enable the radeon use flag when compiling Mesa? Do you have VIDEO_CARDS="radeon" in /etc/portage/make.conf ?
  14. paddymac

    Clang and Portage

    The Gentoo tracker bug is just for Gentoo FreeBSD. That's probably because Clang is the default compiler for FreeBSD. Keep in mind that the primary reason FreeBSD switched to Clang as the default compiler was because Clang is BSD-licensed instead of GPL-licensed. It wasn't because of technical reasons. GCC is still a more reliable compiler. That said, I'm assuming that the current attitude toward package compatibility with Clang is that it's primarily an upstream issue. In other words, if a package fails to compile with Clang, bugs should be filed either with the developers of Clang if it is a compiler bug or with the developers of specific software if it is an issue with the code. Gentoo would probably accept patches which allow packages to compile with Clang if they were submitted, but those patches should also be submitted upstream -- assuming the package is still being developed.
  15. The vulnerability only comes into play when someone has physical access to the computer, but pressing the backspace key 28 times causes a buffer underflow which allows someone to bypass the Grub bootloader password if enabled. http://www.csoonline.com/article/3016100/security/vulnerability-in-popular-bootloader-puts-locked-down-linux-computers-at-risk.html
×