Re: [gentoo-user] Problems copmiling firefox 57.0 (linking phase)
On Thu, Nov 16, 2017 at 4:05 AM, wrote: > Hi, > > building firefox 57.0 failed on my system - it looks like > the last stage (linking) fails. > It doesn't fail at the last stage - that's just when the error is repeated after other parallel tasks in the pipeline are completed. The actual error you got starts around line 5520 and is: --- stderr thread '' panicked at 'Unable to find libclang: "the `libclang` shared library could not be opened: /usr/lib64/llvm/5/lib64/libclang.so.5.0"', src/libcore/result.rs:860 stack backtrace: Short of the file missing, no idea what could be the root cause.
Re: [gentoo-user] is multi-core really worth it?
On Wed, Nov 22, 2017 at 07:52:54AM +0100, Raffaele Belardi wrote: > Hi, > > rebuilding system and world with gcc-7.2.0 on a 6-core AMD CPU I have the > impression that > most of the ebuilds limit parallel builds to 1, 2 or 3 threads. I'm aware it > is only an > impression, I did not spend the night monitoring the process, but > nevertheless every time > I checked the load was very low. multi-core is definitely worth it. Any non-trivial package (wine,glibc,gcc,llvm,qt off the top of my head) will spend a long time building many files that can be built in parallel with multiple make jobs. That being said: if you do a world rebuild you will have lots of packages that spend ~40 seconds doing their autoconf run, only to build 2-3 sources files. On an 8-core machine at work, I get good results using parallel emerge jobs (emerge -jX). For your 6-core AMD CPU (assuming it actually has 12 threads) I'd start with 'emerge -j3' and MAKEOPTS='-j12 -l16'. That should get you a nice speedup, but may require a bit more ram.
Re: [gentoo-user] error while loading shared libraries: libstdc++.so.6:
On Tue, Jan 02, 2018 at 11:14:24AM -0700, the...@sys-concept.com wrote: > On 01/02/2018 11:01 AM, Mick wrote: > > > > This is a coincidence. Something *else* must have happened because > > untarring > > the brother archives did not interfere in any way with your @system > > packages. > > > > Did you move to the 17.1 amd64 profile as per portage news item '2017-12-26- > > experimental-amd64-17-1-profiles' and partly messed up the removal of the > > symlink to /lib64? > > No, I did not touch any of that staff > > There should be al lot of packages in: /usr/lib > but my dir is almost empty only one file left: > > -rwxr-xr-x 1 root root 53606 Sep 18 2008 libbrcomplpr2.so > > I think I have to bootstrap the system and start by unzipping untar the > stage-3 file. > How to untar it so to only replace missing files. > > -- > Joseph Had the same happen to me when i created a kernel `make tar-pkg` and then untarred it with `tar -C`. What happens is ./lib or ./usr/lib in the tarball are directories, but on a normal gentoo amd64 system (not 17.1) they are symlinks to lib64 and usr/lib64. The symlink gets removed, an empty directory created and filled with the tarball contents and that's what you have. Do # mv /usr/lib/* /usr/lib64/ # rm -rf /usr/lib/ # ln -s lib64 /usr/lib and your system should be back. Also, once you figure out what the problem is/was (empty /usr/lib), naturally you'll google for 'tar don't overwrite symlinks' or something of the sort. I'll save you the time: you want 'tar -h' when extracting.
Re: [gentoo-user] crossdev arm-unknown-linux-gnu failed
On Sun, May 27, 2018 at 05:06:43AM +0200, tu...@posteo.de wrote: > Hi, > > too feed a STM32F103C8T6 MCU (Core-M3) with some code to execute, > I want a compiler. For that I did a > > crossdev arm-unknown-linux-gnu > > . That one failed to build (gcc, binytils seem to be ok). > That triplet is not going to work for that hardware for two reasons: - bare-metal implies no kernel so 'linux' is wrong - gnu (== glibc) is not going to work on bare-metal What you're looking for is 'arm-none-eabi', that's what all the vendor supplied prebuilt toolchains use. That triplet will use 'newlib' as the libc, which is the correct choice for bare-metal. I wouldv'e said 'arm-unknown-linux-eabi' or the gentoo specific 'armv7m-softfloat-none-eabi' but for some reasons the binutils build chokes on that. However compared to arm-none-eabi the only thing these triplets do is change some of the configured defaults (arch/fpu/..., see /usr/portage/eclass/toolchain.eclass), so you should be fine with 'arm-none-eabi'.
Re: [gentoo-user] avahi/mdns-nss and ipv4/ipv6 while pinging
On Fri, Jun 29, 2018 at 12:13:39AM +0200, tu...@posteo.de wrote: > On 06/28 09:04, Mick wrote: > > On Thursday, 28 June 2018 17:55:12 BST tu...@posteo.de wrote: > > > Hi, > > > > > > I am "dicovering" (what a great word :) the IP address of my Raspberry > > > Pi now via ahavi/nss-mdns. > > > > > > When pinging my PC on my Raspberry Pi I get the IPv4-address (good!) - > > > when pinging my Raspberry I only get the IPv6-address(wierd-looking) . The > > > commands are identical. > > > > > > ifconfig on my Raspberry shows me, that the interface gets a ip4 and a > > > ipv6 address. > > > > > > What do I need where to tweak to get ipv4 address shown when pinging? > > > > > > Thanks a lot for any help in advance! > > > Cheers > > > Meino > > > > ping -4 -c 3 > > > > -- > > Regards, > > Mick > > > Hi Mick, > > thanks for your help! :) > > But...sorry I asked not exactly enough (no native speaker). > > I dont want to change the ping options...I am searching for > a global setting, which seems to different on both machines, > since the output of an unaltered ping call on each of them > produces different output... > > What config should I modify? > /etc/nsswitch.conf, the line that starts with 'host:'. Mine looks like this: hosts: files mdns_minimal [NOTFOUND=return] dns mdns you can specify mdns4_minimal/mdns4 instead of mdns_minimal/mdns to only get ipv4 results. With mdns either ipv4 or ipv6 will be returned, but apparently ipv6 is preferred.
Re: [gentoo-user] Re: Virtualbox-5.0.0 [wow!]
> On Wed, 15 Jul 2015 19:43:05 -0400 > Fernando Rodriguez wrote: > >> On Tuesday, July 14, 2015 6:53:43 PM walt wrote: >> > I'd like to know if anyone else is seeing spectacular speed >> > performance with vbox-5.0.0. >> > >> >> No noticeable performance improvement for me using hardware >> virtualization. >> Also here the paravirtualization additions are not having any noticeable effect on performance (checked with the Windows Experience Index on W8). What I am seeing though are various regressions: - fullscreen no longer goes fullscreen (with fluxbox doesn't cover the slit and hides behind toolbar) - weird sound/video problems with youtube (accelerated video, constant popping noises) - doesn't work at all since I updated to the 4.2.0-rc2 kernel I had great expectations but so far I'm disappointed.
Re: [gentoo-user] Re: Virtualbox-5.0.0 [wow!]
On Sat, 18 Jul 2015, Fernando Rodriguez wrote: > > Fullscreen works fine on kde and openbox. Works fine on fluxbox as will, but only with <=virtualbox-4.3.30 (I'm using the binary package btw). > I've had the sound problem with earlier versions and pulseaudio. IIRC I fixed > it by tuning PA fragment size. I'll check that but I have a feeling that the popping in the audio is connected to the accelerated video. But as I said, pulseaudio normally works fine, and everything works with the previous virtualbox version. > And it works fine with 4.2.0-rc2 for me. Did you remember to re-emerge app- > emulation/virtualbox-modules (after setting the /usr/src/linux symlink to the > new kernel source)? Yes and yes. However now I'm experiencing lockups whenever I start a VM in virtualbox 5 when running kernel 4.2.0-rc2, so I can't even check anything else. For the time being I'm back on 4.1.2; let me see again if I can make it work here. And Btw. I'm not talking about starting a fresh vm here, I have an existing one that I'm trying to migrate.
Re: [gentoo-user] [~amd64] Does the EGL useflag break mesa-progs-8.2.0?
> On Wed, Jul 15, 2015 at 01:53:50PM -0700, walt wrote: > > Just try emerging mesa-progs-8.2.0 with the EGL useflag set. > > Works fine here, compiles cleanly. This really requires atleast an 'emerge --info' to be able to tell more, but build logs may also be necessary. On Thu, 16 Jul 2015, wraeth wrote: > Attempting to compile =x11-apps/mesa-progs-8.2.0[egl,-gles1,-gles2] I got: > > Makefile:489: recipe for target 'eglut.lo' failed > make: *** [eglut.lo] Error 1 > Makefile:489: recipe for target 'eglut_screen.lo' failed > make: *** [eglut_screen.lo] Error 1 I'd look above these lines for the cause.
Re: [gentoo-user] gru2-mkconfig tries to read the extended partition ??
On Thu, 23 Jul 2015, Walter Dnes wrote: > Here's my question to you... why should we junk software that works, > i.e. it does what we need? I want a real reason. The fact that it's > old is not a justification. Neither is "teh shiney". This thread seems to prove that it doesn't "just work" and causes some people problems that simply don't exist with other partitioning schemes. GPT (if it's supported by your system) is much simpler to set up and has none of the primary-extended partition crap. gottl...@nyu.edu wrote: > It finds linux (all kernels) and the stub windows partition. But then I > get messages that both ext4-fs and FAT-fs have trouble with /dev/sda4, What exactly is the error message? Did you run fsck? It sounds like the system boots properly anyway, right? Do you have the partition mounted when your run `grub2-mkconfig` and is it present in fstab?
Re: [gentoo-user] Re: [~amd64] Does the EGL useflag break mesa-progs-8.2.0?
On Thu, 23 Jul 2015, walt wrote: > Interesting. Maybe you could add some helpful ideas to this existing > bug report: https://bugs.gentoo.org/show_bug.cgi?id=555186 > Added what I found upstream, should be enough to get it to work. > Alpine? I used Alpine a few years ago, but I thought UW had stopped > development on it for lack of users of it. If ya can't use it on an > iPhone these days the students won't use it... > When I finally made the move to commandline email not long ago, I had to choose between alpine and mutt. Since I was somewhat familiar with alpine and couldn't get used to mutt at all, I stuck with alpine. Lack of development doesn't bother me since it's pretty much complete (for my needs). Also someone took up the development and maintenence some time ago: http://repo.or.cz/alpine.git (the project website is currently offline, but worked some days ago)
Re: [gentoo-user] Wifi: Hardware out of luck ?
On Sat, 25 Jul 2015, meino.cra...@gmx.de wrote: > > ...under which name/config option? > > If I grep through 4.1.3 sources I dont find anything, > which is named 'mtu760*' or which contains that... > You have the answer here: > Mick [15-07-25 15:12]: > > Which apparently was merged into mainline kernel from 4.2 onwards. > you can try it by emerging git-sources-4.2_rc3, you should then find it under: -> Device Drivers -> Network device support -> Wireless LAN -> Mediatek Wireless LAN support -> MediaTek MT7601U (USB) support
Re: [gentoo-user] SOLVED xen hvm USE flag disabled (xen: How to enable a non-existant USE flag? (xen doesn't work))
On Wed, 15 Jul 2015, hw wrote: > Hm, I'm not sure what to make of this bug. I switched profile to one that > doesn't say 'no-multilib', and the flag became available. However, I could > not compile xen-tools with hvm because stub-32.h was missing. FWIW, it is not enough to change the profile to activate it, you have to also adjust your existing packages by re-emerging under the new profile. So what you should've tried is: emerge --update --ask --deep --newuse @world I don't know how well switching from no-multilib to multilib is supported, but this would've required atleast a recompilation of glibc (which provides stub-32.h), and gcc, probably not much more than that (but I could be wrong). Also, a multilib profile doesn't automatically mean your system starts building two binaries, for that you have the ABI_X86 flags which you can control on a per-package basis. > So I ended up re-installing the server to a different profile and finally got > it to work. Since rebuilding gcc and glibc takes a while, it's possible that this was actually the best (quickest) way out.
Re: [gentoo-user] New Firefox-38.1.0 headers, or is Google getting smarter?
On Sat, Aug 1, 2015 at 9:27 AM, Mick wrote: > I tried to connect using IMAP4 while overseas. So this tells me that Google > are also logging the IP addresses I am connecting from and check my geographic > location for security purposes. If you log into gmail, scroll to the bottom and on the right you will find something along the lines of Last account activity: 0 minutes ago Details Now if you press the "Details" link you will find a log of recent activity on your account, including client and ip address. You can be either delighted that *you* can monitor your account activity, or terrified... the choice is yours.
Re: [gentoo-user] want to upgrade 50 month old installation
On Tue, 4 Aug 2015, Daniel Frey wrote: > I found that I had to do it in little pieces at a time and portage got > in my way constantly. I wish there was a setting to just forcibly > compile a package and then manually deal with breakage afterward with > something like revdep-rebuild, rather than trying to solve problems it > can't deal with beforehand. It would have been a lot easier. You can do that, just go into /usr/portage and treat the whole thing like a BSD ports system. When you're in a package directory execute ebuild .ebuild merge and portage will do it's thing, only without checking dependencies. So you'll probably get configure failures, but once a package is built that way portage will treat it as though it were installed normally.
Re: [gentoo-user] Installing BTRFS on MBR with OpenRC
On Thu, 6 Aug 2015, Peter Humphrey wrote: > First, btrfs balance. I had no idea that was needed, so of course I didn't > include it in my attempts. Could that be why, on booting, the kernel couldn't > mount the file system? I don't think that balancing an empty btrfs filesystem is necessary. It should have no effect at this point and would not affect the kernel's ability to mount the btrfs volume. > Second, --xattrs on the tarball extraction. I haven't seen this recommended > before, and I wonder what extra attributes are needed for btrfs. This is part of the normal installation procedure and is not specific to btrfs. See [1]. For all I know this has always been part of the installation. > Third, why do you do your kernel compiling in /var/tmp/linux? It seems like > extra work and I can't see a reason for it. The kernel developers recommend to __never__ compile the kernel as root. You never know when something like `rm -rf /` might creep into the Makefile ;) The portage unpacked kernel sources belong to root:root so Rich's method compiles them in /var/tmp/linux which can be done as the local user. > Finally, can I assume that your procedure would work just as well installing > into, say, /dev/sd[ab]4? Yep, but as stated by Rich there isn't much to a btrfs raid 1 install aside from mkfs.btrfs, building btrfs support into the kernel, the slightly different fstab entry, and using an initramfs (dracut makes it easy). I do recommend that you try it in a virtual machine first. With an initramfs it worked out of the box for me but I wasn't able to get it to work with just the kernel command line. [1]: https://wiki.gentoo.org/wiki/Handbook:AMD64/Full/Installation#Unpacking_the_stage_tarball
Re: [gentoo-user] Hubris?
On Wed, Jul 29, 2015 at 12:53 PM, Bruce Schultz wrote: > On 29 July 2015 6:18:43 AM AEST, Neil Bothwick wrote: >> On Wed, 29 Jul 2015 05:29:18 +1000, Bruce Schultz wrote: >>> But I think you do if your btrfs is raid 1. The kernel can't mount >>> multidisk btrfs until it done a btrfs device scan in userspace, run >>> from initramfs. >> >> According to the btrfs wiki you can pass >> device=/dev/sda1,device=/dev/sdb1 on the kernel boot line. > > I'd forgotten that option. Btrfs wiki also says this though: > > "Using device is not recommended, as it is sensitive to device names > changing. You should really be using a initramfs. Most modern distributions > will do this for you automatically if you install their own btrfs-progs > package." I was wondering if *anyone* has actually seen this work. I'm referring to booting a raid1 btrfs volume without performing a user-space device scan, using only the kernel `rootflags=device` setting. I have been struggling with this in various settings and am slowly starting to believe that this scenario is simply broken.
Re: [gentoo-user] Re: Hubris?
> Paul Tobias gmail.com> writes: > > It works, but a patched kernel is needed. Take a look at > > https://forums.gentoo.org/viewtopic-p-7275724.html The patch there was > > still working on the latest kernel a while ago. The patch from the forum thread does indeed work, and just reading it makes it clear *why* it doesn't work in linus' kernel. Thanks! > > I used it on 2 of my systems, but I moved on and now using dracut > > everywhere. While I agree that dracut (/genkernel) is painless, not hard to setup and *should* be the preferred way to boot nowadays, I felt that its a matter of principle that this should work without one. I mean, they even mention in the btrfs wiki that this works (supposedly). In any case I feel like this is a bug in the kernel. Once I work up the courage I'll contact the btrfs(/kernel) mailing list and see what they have to say about this. On Mon, 10 Aug 2015, James wrote: > I not sure but the OpenSuse btrfs (non-raid1) standard install uses only > btrfs partition types, including /boot and all other partitions. I do not > think it uses dracut or others, but I'm not that use to opensuse to make > install. I happen to have a fairly recent OpenSuse vm lying around, featuring a pure btrfs set-up - boot in a subvolume etc. It's non-raid1 but OpenSuse uses dracut so they do not have the problem that I brought up anyway. In general I have not had any problems setting up similar gentoo systems running pure btrfs in raid1/single mode. Those systems, however, all have initramfs' - not that that's a bad thing but see above. Btw. testing the mentioned patch turned out to be a 5 minute thing. qemu-img create to make 2 disk images, qemu boot sysresuecd, create 2 device btrfs volume, mount, unpack stage3 and shutdown. Patch spare kernel in host, compile and boot passing the kernel and cmdline from outside the guest. Works nicely for quick testing (forgot to set root passwd? who cares, it boots!).
Re: [gentoo-user] why --noclear not set on tty1 in default /etc/inittab?
On Tue, 11 Aug 2015, Thanasis wrote: > > How do we enable logging to /var/log/rc.log ? > /etc/rc.conf -> rc_logger="YES" and rc_log_path="/var/log/rc.log"
Re: [gentoo-user] Re: Virtualbox-5.0.0 [wow!]
On Sat, 18 Jul 2015, Fernando Rodriguez wrote: > And it works fine with 4.2.0-rc2 for me. Did you remember to re-emerge app- > emulation/virtualbox-modules (after setting the /usr/src/linux symlink to the > new kernel source)? So it seems I wasn't the only one with the problem on 4.2 (linux): https://forums.virtualbox.org/viewtopic.php?f=7&t=68883 It has been acknowledged and fixed and will be released in a future version. I'll have to wait until then for a second impression of Virtualbox 5.0.
Re: [gentoo-user] Installing BTRFS on MBR with OpenRC
On Fri, 14 Aug 2015, Peter Humphrey wrote: > > > > I have now followed Rich's procedure to build the VM, and it does boot - > > once! > > > > After that, on every boot it complains "failed to read the system array" and Any idea what you did between the first boot and the second one? An internet search shows that this can happen if the btrfs module doesnt get modprobed correctly. > > So I ran dracut again with add_dracutmodules+=”btrfs” and > > add_drivers+=”btrfs” but that didn't help. Specifying > > device=/dev/sda4,device=/dev/sdb4 on the kernel command line didn't help > > either. Forget the device thing added to the kernel cmdline, that doesn't work. Can you verify that all pieces of btrfs are present in the initramfs? `lsinitrd initramfs.img | grep btrfs` should tell you if that's the case. Is btrfs built into the kernel or a module? > Actually, it turns out that the device scan isn't needed: just an exit or > control-d is enough. Check that the kernel cmdline is correct. You need to check the one that you provide and the one that is hardcoded into the dracut image. What does dracut.log say?
Re: [gentoo-user] UEFI booting
On Mon, 17 Aug 2015, Rod wrote: > Hi list, Hi > I'm trying to figure out how to make my boot partition to boot from UEFI, > I have grub2 installed, but I keep getting a error when I ask it to install > the boot information. First things first, are you installing gentoo from an UEFI booted installation media? From what I know the gentoo minimal install cd does not allow for this, and I will assume you are using that. If you're using some other installation method, check whether the directory /sys/firmware/efi/efivars has any content, try to mount efivarfs following the instructions in this link: https://wiki.gentoo.org/wiki/Efibootmgr#Configuration and then check again. > # efibootmgr > efibootmgr: EFI variables are not supported on this system. > > > # grub2-install --target=x86_64-efi /dev/sdc > Installing for x86_64-efi platform. > efibootmgr: EFI variables are not supported on this system. > efibootmgr: EFI variables are not supported on this system. > Installation finished. No error reported. In your case it seems that the system is not in an UEFI-booted state. But we can work around this by using a nice part of the UEFI specification, details below. > mount > /dev/sdc1 201633156 201478 1% /boot/efi > > I have the /boot/efi part mounted ok.. Before we go further make sure that the partition is a valid EFI boot partition: code EF00 (gdisk), partition flags boot/esp (for parted). > How can I get this UEFI be become bootable without media to make it boot > in to that mode to begin with ? It's actually much easier than it may seem, and it's outlined here: https://wiki.gentoo.org/wiki/GRUB2#Alternative:_using_the_default_UEFI_firmware_location Basically, for all sorts of removable media there must be a way to tell UEFI what to boot without having to hardcode the entries into NVRAM (like efibootmgr does). Therefore UEFI firmware is supposed to check for the first ESP partition on a drive and boot \EFI\boot\bootx64.efi from it. This also works for harddrives, and since you can access a FAT partition even when booted in bios mode you just put grub there. Now this works perfectly well for a linux kernel with efi stub and cmdline built-in, but grub may have trouble finding it's configuration files, I do not know. So I suggest you try it. Find grubx64.efi in /boot/efi and copy it to /boot/efi/EFI/BOOT/BOOTX64.EFI. Voila, should boot just fine. On my first UEFI install I did not know about this, efibootmgr did not work, but the handbook says to place the kernel at /EFI/boot/bootx64.efi and it just magically worked. Generally you will find that provided the UEFI implementation of your vendor is not complete shit (lots of them exist) UEFI makes it generally easier to handle booting. One partition and the EFI variables and you can boot anything, no more hidden sectors.
Re: [gentoo-user] installation failure
On Mon, 17 Aug 2015, jfmxl wrote: > but no joy. When I rebooted the machine I saw the grub and a bunch of > initializations whizzed by ... but the kernel failed to mount root. Sounds pretty much like grub is passing the wrong root parameter. > I opened a shell and looked at /etc/fstab, but found a /dev/ram0 and a > /proc and nothing like what I'd entered ... I'd followed the program in > the handbook. Would that be the rescue shell within the initramfs? In that case that's fine, I see the same thing in a genkernel initramfs. > When I tried to unmount the /mnt/gentoo (formerly) chroot partition at > the end the complaint was that something using it ... I couldn't > discover what that was, If you did mount --rbind on sys and dev and then tried to umount than I can tell you I always fail to unmount these properly. Rebooting (software reboot) is fine. > after waiting an retrying several times, I just > shut the vm down and restarted ... with the result above. Is there any > hope of rescuing this? Should be entirely salvageable. > Since I installed the whole system to a qemu 'partition' inside a binary file > I have no access to what I've done or failed to do. So I guess I'm just out of > luck? You can simply reboot the livecd within the VM and just go to the 'chroot` step. You could also mount the binary disk image on the host; depending on what type of image you're using the instructions are both here https://en.wikibooks.org/wiki/QEMU/Images#Mounting_an_image_on_the_host The only thing you're missing is the proper bootloader/initramfs configuration, so go back to this part. >To install an initramfs, install sys-kernel/genkernel first, then have it > generate an initramfs: >root #emerge genkernel >root #genkernel --install initramfs > > So yes, that's what I did, apparently. Without knowing the specifics of your setup, here's the steps I would take to fix it: 1. regenerate genkernel initramfs; you can try to specify a default real_root by executing genkernel --real-root, or in /etc/genkernel.conf. But it seems to me that if the correct root parameter from grub is passed then things should work. 2. grub2-mkconfig again; check the created grub.cfg to see that sane flags are being passed to the kernel. 3. try booting 4. if you don't have a freaky partition setup and the filesystem/block device controller modules are built-into the kernel, you can try to launch the kernel directly from grub-cmdline. Something like set root='hd0,2' linux /boot/vmlinuz-whatever root=/dev/sda4 boot should work (plug in the right partition numbers). This is actually what I would do, initramfs only as a last resort. if these don't work then you still have step 5: 5. generate a dracut initramfs and redo steps 2 and 3. Dracut can generate and save a kernel cmdline within itself which should fix any problems. > ... Ill give it another shot, but I won't stick around on this damn list any > longer. Seems pretty rash to run away after one email, don't you think?
Re: [gentoo-user] Install PreQualifying Matrix
On Mon, 17 Aug 2015, James wrote: > So another day, another borked set of installs. You are aware that for every failed install that comes to this mailing lists there are countless that go just fine, right? Planning questions are an OK-ish idea, but I surely wouldn't link to derivative distributions to answer them. We have appropriate wiki pages for all options, those that are insufficient should be improved. These could be linked to so that people know what to expect. > What I really would appreciate is some feedback on the Planning > Questions listed below, as to help folks organized their thoughts and > hardware details BEFORE actually performing an install or test-drive. It's always good to plan before doing something so *this* part of your proposal I support. > A recent discussion of the dev list showed > encouragement for pointing gentoo-noobs to some of the gentoo derivative > distros for a quick install experience. Perhaps it would be enough to extend this page https://wiki.gentoo.org/wiki/Handbook:AMD64/Installation/About and under `Troubles` mention derivative distributions (by name) with a _hint_ that their installers quickly lead to a working base system. The decisions to be made during the installation are mostly orthogonal, so I wouldn't try to break the current installation procedure which is for the most part linear. A matrix implies some form of interaction between the options, which I don't quite see. > straightforward for folks to discern the best route to their desired final > result. When new installation semantics [1] mature, the installation matrix > can be modified to include those options as links. > > Install PreQualifying Matrx::QUESTIONS > > > > Live Testdrive options before installation(usb/cd/dvd):: Pretty much already covered by https://wiki.gentoo.org/wiki/Handbook:AMD64/Installation/Media and should really be a pre-install thing which it currently already is. > Intended Usage (workstation/server/device/) > Hardware or Vitual installation:: > PC mobo or tablet/embedded/device:: > Processor/Ram characteristics:: How are any of these relevant to the installation? For virtual installations I would only add mention of the `make kvmconfig` option that quickly pulls in qemu drivers. But the other things you mentioned don't have corresponding choices that need to be made (during the install and especially by newcomers). I'd remove them. > MBR vs (u)EFI (type of mobo):: > Single or Multi or RAID disk configuration:: > File System type(s):: > Grub1 vs Grub2 or other boot-semantics:: To me these are the only real things that need to be thought about during the install. MBR vs UEFI is well explained if you ask me. Single/RAID and filesystems are strongly connected but can be chosen freely independent of the other two. Grub2 can boot pretty much anything and if you use the EFI stub kernel on the ESP with initrd then that too can handle anything. So no dependencies here. So I would ask these questions in this order, and this is actually the order in which they show up in the handbook... which makes me wonder whether there is really a need for this. > OpenRC or Systemd:: More of a post install thing if you ask me but the handbook currently links to the systemd article at just the right time. If anything I would actually go for a simplification of the install procedure, to something extremely low maintenence (for the handbook authors ofc). An ext4 single disk install with grub2 (meh) that every one can handle. Sure gentoo gives you choices but you have to be ready to handle them, so perhaps the first install is not the right one for experimenting? > as well as valid install links > like sabayon for gentoo(ish) systemd > like calculate-linus for gentoo(ish)openrc > like pentoo for gentoo-penetration systems > like zentoo for gentoo CI systems > Like funtoo as an option install > like gentooliveUSB for a gentoo + persistence experience. The goal should be to get people to come to gentoo-gentoo, not to go elsewhere. > gentoo installs, so he is one of those guys that can single-handedly > solve this crisis: I actually don't feel that there is any crisis. The only time I've ever had problems with the install was when I decided to not follow the handbook. Most people should just stick to the handbook and learn. Experiment once they know what they're dealing with. I think an einstein quote is relevant here: Everything should be made as simple as possible, but not simpler. The current install procedure is pretty much as simple as can be, once you think about it.
Re: [gentoo-user] how to get a couple of -9999 packages to behave
/usr/lib64/pkgconfig/atk-bridge.2.0.pc is part of at-spi2-atk which comes as a dependency of gtk+[X]. So you need at-spi2-atk[abi_x86_32] which should be pulled in by gtk+[abi_x86_32]. emerge @preserved-rebuild, make sure all flag changes are applied (emerge --changed-use --deep @world) and then try again. It could be that this is a recent change to packages you want (at-spi2-core/at-spi2-atk), they are live ebuilds after all, so you may need to report to the gnome overlay maintainers.
Re: [gentoo-user] how to get a couple of -9999 packages to behave
On Tue, 18 Aug 2015, cov...@ccs.covici.com wrote: > OK, I have the following in /etc/portage/package.use > app-accessibility/at-spi2-atk abi_x86_32 > app-accessibility/at-spi2-core abi_x86_32 Can you verify that these are being picked up properly by running emerge -av at-spi2-atk at-spi2-core Should show you the active use flags and hopefully abi_x86_32 will be among them. Additionally I'm going to go out on a limb here and ask: are you using a multilib profile? > But the 999 versions are not giving the /usr/lib32 items including > /usr/lib32/pkgconfig/atspi-2.pc and this seems to be the problem The gnome-overlay ebuilds do not differ too much from the main tree ebuilds and those are multlib friendly so the live ones should be too. > -- how can I fix, or how can I do the same thing outside of tree i.e. > get both the /usr/lib64 and /usr/lib32 items like the one mentioned > above. Multilib ebuilds basically copy the sources into two folders and run different configure commands in them. Something like CFLAGS=-m32 CXXFLAGS=-m32 LDFLAGS=-m32 ./configure --build=i686-pc-linux-gnu --host=i686-pc-linux-gnu --prefix=/usr/local --libdir=/usr/local/lib32 should build the 32 bit variant properly. You will also be able to use everything straight away since /usr/local/include is on the default preprocessor search path and /usr/local/lib32 (should) be in /etc/ld.so.conf by default. But I don't think it is likely an ebuild problem here, this kind of thing would be quickly caught.
Re: [gentoo-user] Re: !!!!
On Tue, 18 Aug 2015, Dale wrote: > Grant Edwards wrote: > > On 2015-08-18, Dale wrote: > > I don't have any libtinfo. on any of my systems either. By > > default, I don't think a separate libtinfo is built. One suggestion I > > saw for this problem (on a different distro) is to symlink libtinfo to > > libncurses. Ncurses can be compiled as a single library or as two (ncurses + tinfo), in either case all the symbols are present on the system it's just a question of where they are located. Many packages are not prepared to handle the seperate tinfo library, we have many bugs in the bugzilla that deal with tracking down such build failures and correcting them. But this is the first I hear of anyone having a problem with the reverse. Symlinking libtinfo to libncurses *should* work, or atleast seems like a valid rescue attempt. > Maybe we will get some replies from the OP soon. Would help, I too don't recall any serious updates lately.
Re: [gentoo-user] Re: [WAS: keyboard stops working] Recent kernels block the loading of non-GPL kernel modules
On Tue, 18 Aug 2015, Dale wrote: > I did some searching based on the config option you gave and I'm unable > to find a way to override this myself. It doesn't seem to be a setting > I can put in make.conf or package.use etc either. If this is the case, > I may wish Nvidia would switch to open source but it sort of rubs me the > wrong way that someone else is making the decision and me having no way > to exercise my decision to use it anyway. I don't care if Nvidia > doesn't show its code as long as it works and it isn't spying on me or > blowing up my house here. The issue is more of a legal one. The kernel has *always* been under the GPLv2, so derivative work should also be under a compatible license. Companies have frequently been violating the terms of the license or applying their own interpretation of what constitutes derived work. The point is the kernel devs are *not* changing anything by modifying some symbols to be exported GPL-only. Legally they have always been such, and they are just now starting to enforce what has always been their intention. And it's their right to do so. If anyone is to blame then it's nvidia and amd. Ofcourse as a user with access to the source code you can modify it in any way you like, as long as you don't redistribute it. Patch the kernel sources, patch the module sources, contact the graphics card vendor and demand a patch. But don't blame the kernel people. I have nothing against proprietary/closed source kernel modules as long as they comply with the terms of the licenses of open source software that they are using. They expect the same from you...
Re: [gentoo-user] Re: !!!!
On Tue, 18 Aug 2015, John Campbell wrote: > On 08/18/2015 06:38 PM, walt wrote: > > On Tue, 18 Aug 2015 18:03:31 -0700 > > John Campbell wrote: > > > > > I haven't really been following this closely but I haven't seen any > > > suggestion to use "emerge -1 --quiet=y smart-live-rebuild" to remove > > > the offending curses output. Hopefully emerge doesn't check/use > > > curses unless it's producing actual output. >From the initial post it seems that the binary that has a problem is bash/sh. They are both linked against tinfo/ncurses. And since portage runs using bash... you see where I'm going with this. You could maybe use a different shell (although I can't imagine a shell not using the terminal capabilities ;) ), but still wouldn't be able to use portage. > Maybe "emerge --quiet=y -a @smart-live-rebuild" Smart live rebuild only deals with live ebuilds. How would it help in this case?
Re: [gentoo-user] Re: [WAS: keyboard stops working] Recent kernels block the loading of non-GPL kernel modules
On Wed, 19 Aug 2015, Rich Freeman wrote:
> It sounds like you not only expect them to comply with the license,
> but also with the kernel devs personal interpretation of copyright
> law.
What is a license but a statement of the intent of the authors as to what
can and can't be done with their work? Well, it does have some legal force
too...
If you'd try to take away their right to decide about that next thing
they'd do is modify the license to be even larger and explicitly cover all
corner cases. But the world of licenses is complex enough so the next best
thing is saying "this is GPLv2 and we consider this to be derived work".
You can try to bypass that but will hit walls such as EXPORT_SYMBOL_GPL.
And intent does matter in copyright law, if explicitly stated as in this
case.
> I think the real issue here is what constitutes a "derived work." I
> suspect the GPU legal teams have given these practices a thumbs-up,
> and there is probably a reason that the Linux foundation hasn't tried
> to sue them over it.
That has always been the issue but I'll allow myself to quote an email
on the matter:
---
Newsgroups: fa.linux.kernel
From: Linus Torvalds
Subject: Re: Linux GPL and binary module exception clause?
Original-Message-ID:
Date: Thu, 4 Dec 2003 15:59:35 GMT
Message-ID:
On Thu, 4 Dec 2003, Jason Kingsland wrote:
> > - anything that has knowledge of and plays with fundamental internal
> >Linux behaviour is clearly a derived work. If you need to muck
> >around
> >with core code, you're derived, no question about it.
>
>
> If that is the case, why the introduction of EXPORT_SYMBOL_GPL and
> MODULE_LICENSE()?
It is really just documentation.
This is exactly so that it is more clear which cases are black-and-white,
and where people shouldn't even have to think about it for a single
second. It still doesn't make the gray area go away, but it limits it a
bit ("if you need this export, you're clearly doing something that
requires the GPL").
Note: since the kernel itself is under the GPL, clearly anybody can modify
the EXPORT_SYMBOL_GPL() line, and remove the _GPL part. That wouldn't be
against the license per se. But it doesn't make a module that needs that
symbol any less needful of the GPL - exactly because the thing is just a
big cluehint rather than anything else.
Linus
---
Apparently the Linux foundation considers some pieces to be *clearly*
GPL-only and these are marked as such. And I don't see a way for vendors
around this other than to avoid the usage of such symbols - this is what
they have done so far and this is what they will continue doing. If they
want their drivers to be useful to anyone using linux that is.
> The reason neither party talks about it openly
> is probably because they can't be 100% sure which way a court will go
> so it isn't in anybody's interest to stir things up.
Who wants to go court when the current scheme is effective. The whole
issue of "derivative work" is more applicable to binary kernel modules
anyway. In this case we have the source code, and the kernel module build
system will not allow a module that claims a certain license to use
certain symbols. They (vendors) can:
a) change their license to say "GPL" but then people would be allowed to
demand full source code from them including binary blobs.
b) apply the patch you posted - not really practical, definitely
wouldn't be popular but I don't really see how that would be illegal.
Full GPL would apply to the kernel anyway.
c) avoid the usage of such symbols.
and they will do c) because they really don't have a choice.
Re: [gentoo-user] how to get a couple of -9999 packages to behave
On Wed, 19 Aug 2015, cov...@ccs.covici.com wrote:
> Well, I would be willing to do that, but the only difference between the
> .999 ones and the regular ones is that the . ones inherit virtualx
> and they have the following
> inherit eutils gnome2 multilib-minimal virtualx
> if [[ ${PV} = ]]; then
> inherit gnome2-live
> fi
>
> So, can you give me any guidance as to why multilib is not being
> honored?
I tried emerging them like you and I found the cause. If you read
multlilib-minimal.eclass you will find the following:
# multilib-minimal should _always_ go last in inherit order!
Apply the following patch and everything builds fine as multilib. And
please report this problem along with the solution to the gnome-overlay
maintainers.
---
diff --git a/app-accessibility/at-spi2-atk/at-spi2-atk-.ebuild
b/app-accessibility/at-spi2-atk/at-spi2-atk-.ebuild
index fc52ded..3ab77f8 100644
--- a/app-accessibility/at-spi2-atk/at-spi2-atk-.ebuild
+++ b/app-accessibility/at-spi2-atk/at-spi2-atk-.ebuild
@@ -6,10 +6,11 @@ EAPI="5"
GCONF_DEBUG="no"
GNOME2_LA_PUNT="yes"
-inherit eutils gnome2 multilib-minimal virtualx
+inherit eutils gnome2 virtualx
if [[ ${PV} = ]]; then
inherit gnome2-live
fi
+inherit multilib-minimal
DESCRIPTION="Gtk module for bridging AT-SPI to Atk"
HOMEPAGE="http://live.gnome.org/Accessibility";
diff --git a/app-accessibility/at-spi2-core/at-spi2-core-.ebuild
b/app-accessibility/at-spi2-core/at-spi2-core-.ebuild
index 3d13936..4170a8e 100644
--- a/app-accessibility/at-spi2-core/at-spi2-core-.ebuild
+++ b/app-accessibility/at-spi2-core/at-spi2-core-.ebuild
@@ -6,10 +6,11 @@ EAPI="5"
GCONF_DEBUG="no"
GNOME2_LA_PUNT="yes"
-inherit eutils gnome2 multilib-minimal
+inherit eutils gnome2
if [[ ${PV} = ]]; then
inherit gnome2-live
fi
+inherit multilib-minimal
DESCRIPTION="D-Bus accessibility specifications and registration daemon"
HOMEPAGE="http://live.gnome.org/Accessibility";
Re: [gentoo-user] how to get a couple of -9999 packages to behave
On Wed, 19 Aug 2015, cov...@ccs.covici.com wrote: > So do I file a bug on bgo, or is there some other way I should contact > the gnome overlay maintainers/? > I think the easiest way would be to report throught the #gentoo-desktop IRC channel. They might tell you to go to the bugzilla.
Re: [gentoo-user] Re: !!!!
On Wed, 19 Aug 2015, Alan McKinnon wrote: > On 19/08/2015 21:58, Alec Ten Harmsel wrote: > > On Wed, Aug 19, 2015 at 08:55:16PM +0100, Mick wrote: > >> On Wednesday 19 Aug 2015 10:28:48 Jeremi Piotrowski wrote: > >>> Smart live rebuild only deals with live ebuilds. How would it help in this > >>> case? > >> > >> Anyone cares to explain what is a "live ebuild"? > >> > >> Then I may be able to understand what @smart-live-rebuild may be useful > >> for. > >> :-/ > >> > >> -- > >> Regards, > >> Mick > > > > A "live ebuild" is an ebuild that pulls the code to build straight from > > whatever version control the developers are using, so you always have > > the latest and greatest. > > > > Alec > > > > > they usually have version number - > portage has no way of knowing if the repository the package comes from has been updated without fetching the sources and this is done during the merge process. So portage has no knowledge of the state of live-ebuild packages prior to starting a merge - it doesn't know if they have been updated upstream, so it does nothing to them during normal updates. To update them you can use the set @live-rebuild. But this causes live packages to be unconditionally rebuilt even if they haven't changed. Smart-live-rebuild deals with this by updating the repositories and then only re-emerging packages that have been changed.
Re: [gentoo-user] Re: [WAS: keyboard stops working] Recent kernels block the loading of non-GPL kernel modules
On Wed, 19 Aug 2015, Rich Freeman wrote:
> On Wed, Aug 19, 2015 at 8:31 AM, Jeremi Piotrowski
> wrote:
> > On Wed, 19 Aug 2015, Rich Freeman wrote:
> >
> >> It sounds like you not only expect them to comply with the license,
> >> but also with the kernel devs personal interpretation of copyright
> >> law.
> >
> > What is a license but a statement of the intent of the authors as to what
> > can and can't be done with their work? Well, it does have some legal force
> > too...
>
> The only reason that licenses matter at all is that they have legal
> force. They only matter to the degree that they do.
This discussion has been going on ever since the kernel was first
conceived, and some matters are still not entirely clear. I have tried
reading about the subject and the discussion is always the same as the one
we're having in this thread.
What I consider a final word on the matter is pretty much written here:
https://lwn.net/Articles/154602/
briefly: the kernel devs *have* consulted lawyers and believe EXPORT_SYMBOL_GPL
has *some* legal meaning.
But I am not a lawyer, nor do I want to be so let's tune down the law
stuff and get back to technical issues.
> > If you'd try to take away their right to decide about that next thing
> > they'd do is modify the license to be even larger and explicitly cover all
> > corner cases.
>
> And that would accomplish nothing, since companies would just ignore
> any text which doesn't have a basis in law.
They could, but I was under the impression that by using licensed software
you agree to follow it's terms. And the binding nature of licenses is
codified in copyright law.
> The problem is that Linus's words mean nothing unless they're based in
> law. You don't need to quote Linus. You need to quote laws, court
> decisions, or treaties.
Copyright law talks of licenses. Linus' talks about the license under
which the kernel is licensed. Ergo, seems to me as though his words
should have some authority, but I won't argue that.
> And none of them say a word about linking creating a derived work.
The FSF says that and who knows if they're actually right, but I believe
so far everyone generally complies with their interpretation.
> And would we really want to live in a world where they did? Do you
> really want to need permission to use a product in a manner the author
> didn't originally intend?
Proprietary licenses already say that I can't do that, and it's the free
licenses that tell me I can do whatever I want as long as I release the
sources. Sounds reasonable to me.
But again, I'm speaking mostly out of common sense and opinion here.
Neither one of us is going to go around citing cases and laws as that
would be a waste of bandwidth.
>
> Why not just write API wrappers? If somebody exposes an API called
> do_this() using EXPORT_SYMBOL_GPL just write a module that is licensed
> GPL, exposes an API called do_this_nonGPL() using EXPORT_SYMBOL, and
> have do_this_nonGPL() just pass through a call to do_this().
>From what I gather online, this is actually what Nvidia does (or did).
They have an LGPL wrapper in their module, which interfaces with the
kernel, and then talks to their proprietary blob. This [1] stackoverflow
answer has a nice picture:
/.-\
| Kernel |
| |
| /\ |
| | Module | | /---\
| | (LGPL) <> proprietary code |
| \/ | \---/
\--/
[1]: http://stackoverflow.com/a/2721696
So they already go out of their way to comply with the terms of the kernel
license. But they also need to deal with the breakage on each update.
> All they have to do is have the human-readable license say non-GPL,
> and have it report GPL to the kernel, and not ship the source. The
> only recourse anybody has is to sue them, and it is doubtful that a
> court is going to force them to comply, as they clearly indicated
> their intent to not release the code as GPL.
On the other hand they would also be clearly indicating to others that
their code is GPL (if they did in fact add MODULE_LICENSE("GPL")). I may
be wrong here, but if that is not in the least bit ambiguous then shoot me.
> Also, you stated that the check is implemented in the build system. I
> don't know if that is true, but if it is, then the module can simply
> be built using a patched build system. I suspect the check is really
> in the module loader, which the module builder would not have control
> over.
Some reasearch tells me that this checking is implemented in both the build
system (consider it a warning) and in the module loader.
To me it seems like it is not hard to bypass, but it would be *atleast*
indecent.
Re: [gentoo-user] Snort compiling problems
On Mon, Aug 24, 2015 at 12:38 PM, Rod wrote: > > USE="-* ... I have not been following this thread but why do you have this setting? > # grep snort /etc/portage/package.use > net-analyser/snort active-response flexresp3 gre mpls That's a typo in the category name, it's supposed to be: net-analyzer
Re: [gentoo-user] Re: emerge world looking grim
On Mon, 24 Aug 2015, Alan McKinnon wrote: > > Or you can get a binary package from anywhere else you trust. > My personal favourite: chroot into a stage3 and quickpkg gcc. Then copy to your install and voila.
Re: [gentoo-user] Gentoo on a Lenovo X1 Carbon (3rd gen)
I have a T440s and would expect the two to be quite similar from an ACPI point of view, so let's see if I can help. On Tue, 25 Aug 2015, Ralf wrote: > It doesn't come back from standby. > After closing the lid, the standby LED starts breathing, opening the lid > doesn't change anything, even pressing the power button does not wake up > the system. The only option is to reset the system by holding down the > power button. This might be a long-shot but could you check cat /proc/acpi/wakeup The lid-wakeup action can be toggled there. I also have a SLPB device in that file, which could map to the power button in some cases? > As a try and quick fixI copied the Arch Kernel+Modules to my Gentoo > system and it works fine, which means to me that I probably have a > misconfigured kernel. But that's not the Gentoo way, I'd like to > compile the kernel on my own. You could try diffing your config with the arch kernel config. Should be present in /boot. Then look for suspicious differences, it's not as hard as it seems, I've done it with the fedora kernel to solve problems. > Does anyone know what I might be missing in my kernel config? Try my config (attached), I don't know how it compares to the arch one but suspend/resume works correctly here. It may be easier to pinpoint the cause with it. You'll have to enable systemd.# # Automatically generated file; DO NOT EDIT. # Linux/x86 4.1.5-gentoo Kernel Configuration # # # Gentoo Linux # CONFIG_GENTOO_LINUX=y CONFIG_GENTOO_LINUX_UDEV=y # # Support for init systems, system and service managers # CONFIG_GENTOO_LINUX_INIT_SCRIPT=y # CONFIG_GENTOO_LINUX_INIT_SYSTEMD is not set CONFIG_64BIT=y CONFIG_X86_64=y CONFIG_X86=y CONFIG_INSTRUCTION_DECODER=y CONFIG_PERF_EVENTS_INTEL_UNCORE=y CONFIG_OUTPUT_FORMAT="elf64-x86-64" CONFIG_ARCH_DEFCONFIG="arch/x86/configs/x86_64_defconfig" CONFIG_LOCKDEP_SUPPORT=y CONFIG_STACKTRACE_SUPPORT=y CONFIG_HAVE_LATENCYTOP_SUPPORT=y CONFIG_MMU=y CONFIG_NEED_DMA_MAP_STATE=y CONFIG_NEED_SG_DMA_LENGTH=y CONFIG_GENERIC_ISA_DMA=y CONFIG_GENERIC_BUG=y CONFIG_GENERIC_BUG_RELATIVE_POINTERS=y CONFIG_GENERIC_HWEIGHT=y CONFIG_ARCH_MAY_HAVE_PC_FDC=y CONFIG_RWSEM_XCHGADD_ALGORITHM=y CONFIG_GENERIC_CALIBRATE_DELAY=y CONFIG_ARCH_HAS_CPU_RELAX=y CONFIG_ARCH_HAS_CACHE_LINE_SIZE=y CONFIG_HAVE_SETUP_PER_CPU_AREA=y CONFIG_NEED_PER_CPU_EMBED_FIRST_CHUNK=y CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK=y CONFIG_ARCH_HIBERNATION_POSSIBLE=y CONFIG_ARCH_SUSPEND_POSSIBLE=y CONFIG_ARCH_WANT_HUGE_PMD_SHARE=y CONFIG_ARCH_WANT_GENERAL_HUGETLB=y CONFIG_ZONE_DMA32=y CONFIG_AUDIT_ARCH=y CONFIG_ARCH_SUPPORTS_OPTIMIZED_INLINING=y CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC=y CONFIG_HAVE_INTEL_TXT=y CONFIG_X86_64_SMP=y CONFIG_X86_HT=y CONFIG_ARCH_HWEIGHT_CFLAGS="-fcall-saved-rdi -fcall-saved-rsi -fcall-saved-rdx -fcall-saved-rcx -fcall-saved-r8 -fcall-saved-r9 -fcall-saved-r10 -fcall-saved-r11" CONFIG_ARCH_SUPPORTS_UPROBES=y CONFIG_FIX_EARLYCON_MEM=y CONFIG_PGTABLE_LEVELS=4 CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config" CONFIG_IRQ_WORK=y CONFIG_BUILDTIME_EXTABLE_SORT=y # # General setup # CONFIG_INIT_ENV_ARG_LIMIT=32 CONFIG_CROSS_COMPILE="" # CONFIG_COMPILE_TEST is not set CONFIG_LOCALVERSION="" # CONFIG_LOCALVERSION_AUTO is not set CONFIG_HAVE_KERNEL_GZIP=y CONFIG_HAVE_KERNEL_BZIP2=y CONFIG_HAVE_KERNEL_LZMA=y CONFIG_HAVE_KERNEL_XZ=y CONFIG_HAVE_KERNEL_LZO=y CONFIG_HAVE_KERNEL_LZ4=y CONFIG_KERNEL_GZIP=y # CONFIG_KERNEL_BZIP2 is not set # CONFIG_KERNEL_LZMA is not set # CONFIG_KERNEL_XZ is not set # CONFIG_KERNEL_LZO is not set # CONFIG_KERNEL_LZ4 is not set CONFIG_DEFAULT_HOSTNAME="(none)" CONFIG_SWAP=y CONFIG_SYSVIPC=y CONFIG_SYSVIPC_SYSCTL=y CONFIG_POSIX_MQUEUE=y CONFIG_POSIX_MQUEUE_SYSCTL=y CONFIG_CROSS_MEMORY_ATTACH=y # CONFIG_FHANDLE is not set CONFIG_USELIB=y CONFIG_AUDIT=y CONFIG_HAVE_ARCH_AUDITSYSCALL=y CONFIG_AUDITSYSCALL=y CONFIG_AUDIT_WATCH=y CONFIG_AUDIT_TREE=y # # IRQ subsystem # CONFIG_GENERIC_IRQ_PROBE=y CONFIG_GENERIC_IRQ_SHOW=y CONFIG_GENERIC_IRQ_LEGACY_ALLOC_HWIRQ=y CONFIG_GENERIC_PENDING_IRQ=y CONFIG_IRQ_DOMAIN=y CONFIG_GENERIC_MSI_IRQ=y # CONFIG_IRQ_DOMAIN_DEBUG is not set CONFIG_IRQ_FORCED_THREADING=y CONFIG_SPARSE_IRQ=y CONFIG_CLOCKSOURCE_WATCHDOG=y CONFIG_ARCH_CLOCKSOURCE_DATA=y CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE=y CONFIG_GENERIC_TIME_VSYSCALL=y CONFIG_GENERIC_CLOCKEVENTS=y CONFIG_GENERIC_CLOCKEVENTS_BROADCAST=y CONFIG_GENERIC_CLOCKEVENTS_MIN_ADJUST=y CONFIG_GENERIC_CMOS_UPDATE=y # # Timers subsystem # CONFIG_TICK_ONESHOT=y CONFIG_NO_HZ_COMMON=y # CONFIG_HZ_PERIODIC is not set CONFIG_NO_HZ_IDLE=y # CONFIG_NO_HZ_FULL is not set CONFIG_NO_HZ=y CONFIG_HIGH_RES_TIMERS=y # # CPU/Task time and stats accounting # CONFIG_TICK_CPU_ACCOUNTING=y # CONFIG_VIRT_CPU_ACCOUNTING_GEN is not set # CONFIG_IRQ_TIME_ACCOUNTING is not set CONFIG_BSD_PROCESS_ACCT=y # CONFIG_BSD_PROCESS_ACCT_V3 is not set CONFIG_TASKSTATS=y CONFIG_TASK_DELAY_ACCT=y CONFIG_TASK_XACCT=y CONFIG_TASK_IO_ACCOUNTING=y # # RCU Subsystem # CONFIG_TREE_RCU=y CONFIG_SRCU=y # CONFIG_TAS
Re: [gentoo-user] Re: emerge world looking grim
On Wed, Aug 26, 2015 at 12:10 AM, Harry Putnam wrote: > The only thing I can say on my own behalf is that there was once a > time when it wasn't so far fetched to start emerge -vC 'ing stuff. Many thing's can be removed with `emerge -C` and recovered from, but I doubt unmerging packages in @system was ever a well supported operation...
Re: [gentoo-user] Gentoo on a Lenovo X1 Carbon (3rd gen)
On Wed, Aug 26, 2015 at 10:06 AM, Ralf wrote: > To answer my own question: > > After a lot of diffing and searching, I was able to find out, that these > were the modules that I was missing: > > CONFIG_TCG_TPM=y > CONFIG_TCG_TIS=y Are you actually using the TPM? This shouldn't have any any effect on resume but I don't know enough about crypto (you said you use it) to be certain that there isn't any crosstalk. > CONFIG_HID_LENOVO=y This one is not necessary as it only affects _external_ lenovo USB keyboards. > CONFIG_THINKPAD_ACPI=y > CONFIG_THINKPAD_ACPI_ALSA_SUPPORT=y > CONFIG_THINKPAD_ACPI_VIDEO=y > CONFIG_THINKPAD_ACPI_HOTKEY_POLL=y I had assumed you had these active as they are pretty much mandatory for thinkpads. > CONFIG_ACPI_WMI=y This one is tricky. It has acpi in the name so it _could_ be the one that really made the difference, but I thought it was an extra device. In any case, it's great that you got to the bottom of this. Have you thought of starting a wiki page for the X1 carbon?
Re: [gentoo-user] Re: Grub1: Cant ? Re: keeping grub 1
On Wed, 26 Aug 2015, Alec Ten Harmsel wrote: > The main benefit is reduced compile times for some packages since I only > compile the 64-bit versions, less stuff on the filesystem, etc. If you > do not run any applications that use a 32-bit version of a library, that > library is taking up disk space and compile time, but is never used. The multilib profiles do not enable ABI_X86="32" by default so the default setup is to only build the 64-bit versions of everything. These profiles give you the _option_ to build both 32-bit and 64-bit things. The only things multilib by default on a multlib profile are pretty much glibc and gcc. > I also am a bit of a purist, and just run no-multilib because it is > emotionally satisfying. The above reasons would make emotional satisfaction and purity the only reasons to go down this road. Doesn't mean they're not valid :)
Re: [gentoo-user] Re: Grub1: Cant ? Re: keeping grub 1
On Thu, 27 Aug 2015, Fernando Rodriguez wrote: > I use GRUB2 because my efi firmware (like most) is really buggy. > Changing the boot order doesn't work at all (neither on the menu nor > through efibootmgr), so I have to delete and recreate the entries in the > right order. What I did is create 2 efi entries, one for my main kernel > and one for grub2 and I added entries for all my secondary kernels and > windows on the grub2 menu. I have a similar setup: an efi entry for my kernel and gummiboot instead of grub2. I would use the firmware boot selector for multibooting, but going there lengthens the boot by around 10 seconds (Lenovo notebook) which is too much for my liking. So instead I by default boot into gummiboot and from there I choose my kernel, separate grub2's from Fedora and Ubuntu, or FreeBSD. Everything is nice and fast that way. Fedora and Ubuntu manage their own grub configs on kernel updates and I don't have to mess with any boot loader configs myself. On Fri, 28 Aug 2015, Mick wrote: > Interesting ... here I had no such problems on an EUFI Asus MoBo. I > have a number of kernels (up to six last time I looked) and I can change > the boot order with 'efibootmgr -o ,,...,' I have a MSI B85-G43 and I must say that thing is the worst when it comes to managing the UEFI boot order. It completely ignores the one set using `efibootmgr -o` and instead seems to keep track of the order in which entries are added. Half of the time it even ignores me selecting the EFI boot entry during boot and instead merrily continues to load what it considers to be the _first_ entry. So I just use grub2 there and everything is fine.
Re: [gentoo-user] Re: Grub1: Cant ? Re: keeping grub 1
On Thu, 27 Aug 2015, Neil Bothwick wrote: > On Thu, 27 Aug 2015 14:19:29 + (UTC), Grant Edwards wrote: > > > For those of us with multiple Linux installations on a disk, that's a > > pretty big reason to stick with grub-legacy. > > Actually, that's a good scenario for GRUB2. grub2-mkconfig can detect > all Linux installations on a system, not just the running one, so you > only need one GRUB to boot everything. That's why distro installers are > so much better at setting up Linux dual booting these days, because GRUB2 > makes it simple for them. > It's true that grub2-mkconfig does Linux detection well but the problem with one grub and multiple distros is the need to manually regenerate the config. I give you the following scenario: Gentoo + another binary distro (say Fedora). Whichever one manages the grub config can regenerate it on updates. On gentoo you'd do that manually (post-install hooks?), Fedora would run grub2-mkconfig on kernel updates. But what happens when the other one (not responsible for the config) updates in a way that affects booting...? You end up with an inconsistant config. To regenerate you need to boot into the config-managing-distro or atleast chroot. But the worst thing is you have to review all updates to find out if the config needs changing. I much prefer chainloading and giving each distro free reign over their own boot loader. That way they can pretend they're the boss and work the way they were intended to and I can supervise things from gentoo.
Re: [gentoo-user] Grub1: Cant ? Re: keeping grub 1
On Fri, 28 Aug 2015, Michel Catudal wrote: > No one is asking them to do that. As mentioned before it works with some > override. A solution to the problem would be to remove the arrogance toward > people who want grub on a partition and remove the part in the installer that > refuses to install it unless you give it an override. To me they are dealing with this in the right way. As the developers they have to decide what setups they want to support as the spectrum is huge and manpower is limited. There are problems with installing grub to a partition, read [1]. Therefore it is not supported and not allowed by default, because if they don't do this people: 1. _will_ try installing to a partition 2. _will_ render their system unbootable 3. _will_ come running for help and complaining 4. _will_ get angry when you tell them `I told you so' Seems perfectly legit to want to spare yourself this trouble. > If I say write the > bootloader on the partition, that should work as requested, they can still > write a comment that they do not like us doing it but should not keep us from > doing it. If it doesn't work we will see it soon enough. > I don't get you - that _is_ exactly what they are doing. You say 'write bootloader to partition' by adding the force flag and grub2 complains but does what it is told. [1]: https://bbs.archlinux.org/viewtopic.php?pid=1229097#p1229097
Re: [gentoo-user] Grub1: Cant ? Re: keeping grub 1
> > On Thursday, August 27, 2015 9:25:01 PM Michel Catudal wrote: > > > This is nonsense. I have never had a case where it would not boot when I > > have grub correctly installed on the partition. This hasn't happened to you so it obviously means it isn't possible... On Fri, 28 Aug 2015, Michel Catudal wrote: > Le 2015-08-27 23:36, Fernando Rodriguez a écrit : > > Install grub to a partition and do something like this: > > > > su > > cd > > mv /boot/grub grub > > cp -r grub /boot > > rm -r grub > > > What is your point? same if I do that with grub1, it was even more fun with > windows 98 by deleting win.ini or renaming it "win .ini" > With grub on the partition my bootloader doesn't get wacked and I can restore > the OS if I do a stupid thing like this. > You seem to be trying really hard to _not_ get what people are saying to you. The above commands will *not* change anything visibly - they are akin to backing up your grub configuration and later restoring. It ends up at the same location on the filesystem, but this doesn't mean it is at the same location on the block device. That's the problem with installing grub to a partition: all it really does is store (in the PBR) a list of blocks where the grub core.img is located without caring about the filesystem structure on top. This is a _fragile_ set-up. So there. There are technical reasons why this is not a supported setup. No one has removed this option from grub2, so all your complaints about dictatorship are invalid.
Re: [gentoo-user] a few blockers I can't figure out
On Mon, 31 Aug 2015, cov...@ccs.covici.com wrote: [snip] > Just to let you know, most of the python entries were mandated by > portage, certainly the systemd one. > emerge --info > Portage 2.2.20.1 (python 2.7.10-final-0, > default/linux/amd64/13.0/desktop/gnome, gcc-4.9.3, glibc-2.21-r1, > 3.16.3-gentoo x86_64) I think this is your problem right here: you don't have the systemd profile selected. You're trying to splice together the use settings needed to get this to work when all of them are already gathered in the right profile. Remove most of the use settings you were forced to make because of systemd related issues, change your profile to default/linux/amd64/13.0/desktop/gnome/systemd and emerge -uDUva @world. This should remove most (if not all) of the blockers caused by conflicting use flags you currently have set.
Re: [gentoo-user] Re: a few blockers I can't figure out
On Wed, 2 Sep 2015, walt wrote: > Thank you. I've been running systemd for months and this is the first > time I've heard about systemd profiles. I'm not using either gnome or > kde, so should I use default/linux/amd64/13.0/systemd, which doesn't > seem to care if I'm running a desktop machine or a headless server? > I just had a closer look of the profile and it's definitely more suitable for servers. However do remember that this is gentoo so we can work around that. Easy way: stay on the desktop profile and copy all of the systemd profile files into /etc/portage: mkdir -p /etc/portage/profile cp /usr/portage/profiles/targets/systemd/* /etc/portage/profile A bit harder (but way more interesting!): roll your own profile. Never done this before and the documentation is a bit scarce, but this seems to do the right thing: # mkdir -p /etc/portage/repos.conf # cat >/etc/portage/repos.conf/local.confrepo_name # echo -e "amd64\tdefault/linux/amd64/13.0/desktop/systemd\texp" >profiles.desc # mkdir -p default/linux/amd64/13.0/desktop/systemd # cd default/linux/amd64/13.0/desktop/systemd # echo "gentoo:default/linux/amd64/13.0/desktop" >parent # echo "gentoo:default/linux/amd64/13.0/systemd" >>parent Then eselect your new profile and emerge -uDNva @world.
Re: [gentoo-user] Re: Portage is proposing an ncurses update and I don't understand what it means
On Wed, 2 Sep 2015, walt wrote: > qlop shows that both updates completed, but eix shows that I now have > only ncurses-5.9-r5 installed. (This is apparently the desired > result, but I'm only guessing what the desired result really is.) The desired output depends on what other packages you have installed but for me it's having both ncurses-5.9-r5 ncurses-5.9-r99. > I think every portage tool should announce very clearly whether a > package is slotted/subslotted, and exactly which slot and subslot the > package belongs in. The subject of slots is way too confusing to > withhold such information. If the devs can't explain slots to their > users then they don't understand it themselves. (Hm. That phrase > sounds familiar. Where did I get that?) > I don't know why your portage display is different than mine, I always run portage in verbose mode and it clearly shows me subslots: # emerge -uDUva @world These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild U ] sys-libs/ncurses-5.9-r5:0/5::gentoo [5.9-r3:0/0::gentoo] USE="..." ABI_X86="(64) -32 (-x32)" 2,761 KiB [ebuild NS] sys-libs/ncurses-5.9-r99:5::gentoo [5.9-r3:0::gentoo] USE="..." ABI_X86="(64) -32 (-x32)" 0 KiB
Re: [gentoo-user] Re: new computer : any advice ?
On Wed, 9 Sep 2015, Stefan G. Weichinger wrote: > On 09.09.2015 20:01, Stefan G. Weichinger wrote: > > Digging up that thread now somewhere ... > > Ah, I even participated then ;-) > > AFAI understand gcc-5 should compile faster? > And generate faster code in some cases? No, and yes. Compilation is not affected in any way and runtime performance can only be improved _if_ this stuff is explicitly used within the code. Meaning you would feel a difference in no less then 5 years when gcc-6 is widely used and accelerator support is not restricted to intel MIC and nvidia gpus. James is getting a bit ahead of himself calling this a "game changer" - yeah... not really right now. Right now this functionality is a toy for the HPC community and will stay that way. To use it you have to build a separate offloading compiler, need custom code used by few, and expensive hardware. The tree ebuild doesn't even provide a way for enabling the accelerator support. > > ... does it use this new stuff anyway, do we need a specific USE-flag > enabled (I can't spot it, looking for something like "acc" or "rdma" > ;-)), do we need specific CFLAGS .. ? > > just curious. I can't speak for RDMA (can't find any mention of it in gcc) because that's an even more exotic thing than plain old accelerator support (unless you run infiniband at home...), but the flags are: -fopenmp -foffload -fopenacc However enabling them is as useful as having CFLAGS=-fopenmp currently. It changes __nothing__ unless an application has openmp annotations, and the ones that do should already provide a means of doing so in the build system. tldr: don't buy a dedicated gpu just because you read something on a mailing list ;)
Re: [gentoo-user] clean-up root partition
On Thu, 1 Oct 2015, Philip Webb wrote: > 151001 the...@sys-concept.com wrote: > > How do you folks clean-up root partition, I have too much junk in there. > > df -h > > Filesystem Size Used Avail Use% Mounted on > > /dev/sda374G 61G 9.3G 87% / > > I've already removed all the files from /usr/portage/distfiles . > > 'du' is your friend : it has lots of options, so read the 'man'. > that will tell you what's using so much space, then you can delete stuff. I highly recommend sys-fs/ncdu, an ncurses front end to du. Definitely much better than piping du output through sort. You can even directly delete folders/files when you see that something is taking up more space than expected.
Re: [gentoo-user] kernel panick in 4.2.1 from gentoo-sources
On Wed, Oct 07, 2015 at 10:12:28PM -0400, cov...@ccs.covici.com wrote: > > Do you bother with the gentoo patches? Only the Kconfig patch that gives quick access to gentoo required configs is really useful, the rest I would call minor. If you copy your config from gentoo-sources to the git tree, the required ones will already be active you can skip genpatches altogether. > I have the kernel tree and none of the tags say longterm, The tags don't say longterm, the website (kernel.org) says which releases are longterm support ones. > do I have the wrong tree or something?The url I have is > git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git That's the development tree, if you want the stable/longterm releases use this one: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git You should be able to add it to your existing tree using something like: git remote add stable git fetch stable although it does take a while ;)
Re: [gentoo-user] Making sense of reported ZFS disk usage
On Sat, Oct 31, 2015 at 10:37:55AM +0100, Remy Blank wrote: > I'm trying to make sense of the disk usage reported by "zfs list". > Here's what I get: > > $ zfs list \ > -o name,used,avail,refer,usedbydataset,usedbychildren,usedbysnapshots \ > -t all > > NAME USED AVAIL REFER USEDDS USEDCHILD USEDSNAP > pool/data58.0G 718G 46.7G 46.7G 0 11.3G > pool/data@2015-10-03 0 - 46.5G - - - > ... > pool/data@2015-10-12 0 - 46.5G - - - > pool/data@2015-10-13 734M - 46.7G - - - > pool/data@2015-10-14 0 - 46.7G - - - > ... > pool/data@2015-10-28 0 - 46.7G - - - > pool/data@2015-10-29 755M - 46.7G - - - > pool/data@2015-10-30 757M - 46.7G - - - > pool/data@2015-10-31 0 - 46.7G - - - > > What I don't understand: I have 29 snapshots, only three of them use > ~750M, but in total they take 11.3G. Where do the excess 9.1G come from? > I'm going to go out on a limb and assume that zfs works in a similar way to btrfs here (my quick googling shows that atleast in this case that may be true). You then have to understand the numbers in the following way: USEDSNAP refers to _data_ that is not in pool/data but in the snapshots. The value for USED is _data_ that is only present in *this one* snapshot, and not in any other snapshots or in pool/data. _data_ that is shared between atleast two snapshots is not shown as USED because removing one of the snapshots would not free it (it is still referenced by another snapshot). So in your case you have 3 snapshots which each have 750 MB exclusively, and the remaining ~9 GB is in some way shared between all snapshots. So if you were to delete any one of the 3 snapshots, you would free 750 MB. If you were to delete all snapshots you would free 11.3 GB. But deleting any one snapshot can change the USED count of any other snapshot. This is one of the problems with copy-on-write filesystems - they make disk space accounting more complicated especially with snapshots. Perhaps zfs has something similar to btrfs qgroups which would allow you to group snapshots in arbitrary ways to find out how much any group of snapshots uses. Here's the example output of 'btrfs qgroup show' on my machine: qgroupid rfer excl parent child -- - 0/5 0.00GiB 0.00GiB --- --- 0/262 6.37GiB 0.03GiB --- --- 0/265 3.52GiB 2.38GiB 1/0 --- 0/270 6.38GiB 0.16GiB --- --- 0/275 0.00GiB 0.00GiB --- --- 0/276 4.38GiB 0.35GiB 1/0 --- 0/277 0.00GiB 0.00GiB --- --- 0/278 4.98GiB 0.40GiB 1/1 --- 0/279 4.62GiB 0.12GiB 1/0 --- 0/285 5.59GiB 0.01GiB 1/0 --- 0/286 5.69GiB 0.01GiB 1/0 --- 0/289 6.34GiB 0.42GiB 1/1 --- 0/290 6.35GiB 0.01GiB 1/0 --- 0/291 6.38GiB 0.15GiB 1/1 --- 1/0 10.02GiB 3.68GiB --- 0/265,0/276,0/279,0/285,0/286,0/290 1/1 7.20GiB 0.98GiB --- 0/278,0/289,0/291 0/262 is / 0/270 is /home 1/0 contains all snapshots of / 1/1 contains all snapshots of /home but I could also have grouped a subgroup of the snapshots in some other way to find out how much space they take exclusively and how much space they would free if they were to be deleted.
Re: [gentoo-user] This is what I get for trying...
On Thu, Nov 19, 2015 at 10:17:03AM -0500, Alan Grimes wrote: > Naturally, the instructions on the gentoo wiki FAIL > > https://wiki.gentoo.org/wiki/Basic_guide_to_write_Gentoo_Ebuilds > > > atg@tortoise ~/archive $ ebuild seamonkey-2.35.ebuild clean merge > Appending /home to PORTDIR_OVERLAY... > !!! Repository 'x-home' is missing masters attribute in > '/home/metadata/layout.conf' > !!! Set 'masters = gentoo' in this file for future compatibility This part tells you all that you need to know: > ebuild: /home/atg/archive/seamonkey-2.35.ebuild: does not seem to have a > valid PORTDIR structure I quote from the link that you posted: > Now, you can't just create a file /usr/portage/hello-world.ebuild and be > done with it; there are several reasons: > ... > 2. The ebuild file is not in the right directory: An ebuild has to be > located in the "package name" subdirectory of "category" directory; > so, for an ebuild to work you have to place it in a directory like > /usr/local/portage/app-misc/hello-world/. So move your seamonkey-2.35.ebuild to a subdirectory ww-client/seamonkey/ and run ebuild again.
Re: [gentoo-user] PPC64 on IBM Power8
On Mon, Nov 23, 2015 at 03:08:26PM +0100, luciano mannucci wrote: > Now I'm wondering: should I try to build everithing from there, or is > it possible to get a more recent install version for the ppc64 > architecture? The install medium has almost zero effect on what ends up in the installed system. The only purpose it serves is to partition and format your block devices - after that you chroot into the stage3 which will become your install environment. As long as your install cd/dvd supports the filesystems you want to use, you'll be fine using it. Just get the newest stage3 and you will start with the most recent gentoo has to offer.
Re: [gentoo-user] dcron problem on new install
On Wed, Nov 25, 2015 at 10:29:52PM -0500, waltd...@waltdnes.org wrote:
> I'm getting a bunch of messages like...
>
> > Subject: cron for user root root[ ! -x /etc/cron.hourly/0anacron ]
> > && { test -x /usr/sbin/run-crons && /usr/sbin/run-crons ; }
> >
> > /bin/sh: root: command not found
>
> /bin/sh does exist...
>
> [d531][waltdnes][~] ll /bin/sh
> lrwxrwxrwx 1 root root 4 Nov 24 12:10 /bin/sh -> bash
>
> /etc/ cron.hourly cron.weekly cron.monthly are empty except for a
> dummy file .keep_sys-process_cronbase-0
>
What /bin/sh actually means is "root: command not found". This is caused
by the difference between the system-wide crontab and the per-user
crontab.
The system-wide crontab /etc/crontab needs to know as which user to
execute the command so it has a column for the username right before the
command. The per-user crontabs which are at /var/spool/cron/crontabs (but
which should be edited with crontab -e) know as which user to run the
commands so they treat everything after the time as the command.
You got the two mixed up. Run crontab -e as the root user and remove
the word 'root' part right before the command. Should fix things right
away.
Re: [gentoo-user] Copying USB->USB: sync after every file
On Mon, Jan 23, 2017 at 04:50:33PM +0100, Frank Steinmetzger wrote: > > Can you give me a nudge? My sysctl.conf has only one active line besides > some netfilter stuff: > vm.swappiness = 1 > (This is an SSD, I don’t even have swap on this machine unless I know I > definitely need it temporarily) > I'd say the relevant sysctl's are /proc/sys/vm/dirty* and /proc/sys/vm/laptop_mode, along with the filesystems and mount options you are using for the USB drives. Please post these values. Also, could you get the output of /proc/meminfo while this is happening?
Re: [gentoo-user] Getting rid of gcc-4.7.3...how?
On Sat, Dec 20, 2014 at 9:04 PM, wrote: > Matti Nykyri [14-12-20 19:48]: > > > On Dec 20, 2014, at 17:56, meino.cra...@gmx.de wrote: > > > > > > Dale [14-12-20 02:47]: > > >> meino.cra...@gmx.de wrote: > > >>> Dale [14-12-19 17:08]: > > Mick wrote: > > > Meino, to avoid misunderstandings: 1. Emerge the new gcc package. > 2. > > > Use gcc-config to change to the new gcc version. 3. Run > 'env-update && > > > source /etc/profile'. 4. Run fix_libtool_files.sh, although I would > > > think that this is redundant these days. 5. Unmerge the old gcc > version. > > I don't recall ever running fix_libtool_files.sh after switching gcc > > versions. Usually when I see a gcc upgrade, I emerge it, switch to > it > > and the usual profile thing, run emerge -e world JUST to be safe, > then > > unmerge the old gcc. That's all I usually do here. I have skipped > the > > emerge -e world a time or two. > > > > Am I just lucky, not likely as some may know, or does emerge -e > world > > catch it or what? Now I'm curious. > > > > Dale > > > > :-) :-) > > >>> Hi Dale, > > >>> > > >>> I started compiling the new gcc this morning about ~7:00 AM...just a > > >>> few minutes ago stage3 finishes. Now ... before doing anything > else... > > >>> I am makeing a backup of all that, so...if anything fails...I am able > > >>> to reinstall the status quo. > > >>> > > >>> I will keep you informed, what happens to my little embedded > system... > > >>> > > >>> Best > > >>> Meino > > >> > > >> That's the thing about slow systems, you want to do it right the first > > >> time because it takes to much time to repeat something. Heck, I have > a > > >> 4 core AMD CPU with 16GBs of ram here and I still would rather do it > > >> right the first time. If you have something slow that takes days to > do > > >> something, you really want plan A to work. > > >> > > >> I'm also wondering if there have been changes to emerge that could > make > > >> a difference. I run the latest unstable non * version. I sorta > > >> like having all the new improvements. I'm just not sure if that > affects > > >> the issue here is all. > > >> > > >> Dale > > >> > > >> :-) :-) > > > > > > Hi, > > > > > > after a few more non-booting-systems and backup-reinstalls I think > > > I know whats the reason is...but by I dont know how to get out of it: > > > > > > The system becomes inaccessible if I do an env-update and reboot. > > > > > > Reason for that are binaries, in which the path to the old gcc is > > > hardcoded. With the sdcard mounted I checked that with my PC: > > > I did a > > > > > >grep -r '\/usr\/lib\/gcc\/armv7a-hardfloat-linux-gnueabi\/4.7.3' > > > > > > on ALL files of the sdcard and found "thousands" of hardcoded links > > > to the old gcc inside binaries... > > > > > > The new gcc installed but not doing env-update implies that any > > > further compilation will link to the old gcc. > > > > > > Doing env-update implies a system which will not survive the next > > > reboot. > > > > > > What now? > > > > If i understand your situation correctly, do: > > > > gcc-config "to set the new version" > > env-update > > logout > > login > > emerge --deep --update world > > emerge --depclean > > revdep-rebuild > > > > This will take a long time but will get your system working again. If > you don't wan't to do that you can of course tweak the libraries with > binary tools. That is easy if you know what you are doing. > > > > To prevent this in the future always before world update, update gcc and > glibc first if tere is a new version available. Gcc-config is crusial after > you have installed a bew version of gcc. > > > > -- > > -Matti > > > Hi Matti, > > not exactly... > The sequence you show looks like this in my case: > gcc-config "to set the new version" > env-update > reboot > logina attempt: impossible...system does not respond anymore > > The reason is, that env-update updates to the new compiler, > while some [CENSORED] system tools/application are hardcoded > to use the old compiler libraries... > > If I do a > > env-update --no-ldconfig > > then I am able to reboot successfully...but compilation > will be done against the old compiler a fear... > > I am installing the backup the fourth time today and I am > curious, when the first bad sectors of my sdcard will hit me... > > > > Best regards, > Meino > > > > Im surprised no one has suggested this, but have you tried running: emerge --oneshot libtool after gcc-config and before env-update && source /etc/profile? This is the advice that is given in the wiki [https://wiki.gentoo.org/wiki/Upgrading_GCC] and it has worked for me so far. It also says that running `fix_libtool_files.sh` manually is not necessary as it is run during the emerge automatically.
Re: [gentoo-user] No trace of installed simpleagenda
> Does anybody know how one can start > this extremely humble program Use equery f simpleagenda to find files installed by this package (equery is part of the gentoolkit package). Cheers
[gentoo-user] Re: Full system encryption on Gentoo
On Thu, Dec 31, 2015 at 07:45:29AM +1000, Hans wrote: > I can't follow Sakaki's_EFI_Install_Guide. The system will run in > VirtualBox and only have BIOS. No UEFI, EFI, USB stick as boot or key disk. You should still atleast read the guide to figure out how to get the encryption part right. You can skip the USB stuff and fallback to BIOS equivalents of EFI concepts. > I just have to find a way to get the same result using Gentoo with > OpenRC and if possible without LVM. Entering the pass phrase several > times is no problem. The steps are more or less the following: 1. cryptsetup your whole device 2. mkfs 3. chroot 4. install grub with device-mapper flag 5. install dracut and cryptsetup. 6. add GRUB_ENABLE_CRYPTODISK=y to /etc/default/grub 7. grub2-install 8. set 'hostonly="yes"' in /etc/dracut.conf OR add the output of `dracut --print-cmdline` to GRUB_CMDLINE_LINUX_DEFAULT in /etc/default/grub 9. grub2-mkconfig -o /boot/grub/grub.cfg 10. dracut --regenerate-all Somewhere between step 3 and 10 you need to build the kernel with atleast the dm_crypt module. This will lead to you having to enter the password twice - once when grub starts and once when the initramfs is setting up /. Check the arch wiki article on the topic [1] for more info, but don't blindly trust the boot loader part because that is specific to arch's initramfs generator. [1]: https://wiki.archlinux.org/index.php/Dm-crypt/Encrypting_an_entire_system
Re: [gentoo-user] Re: Full system encryption on Gentoo
On Thu, Dec 31, 2015 at 10:38:45AM +1000, Hans wrote: > I have a working VM with Gentoo on LVM on top of LUKS. Works fine in > change root, Just can't get it to boot. Probably somewhere missed > something. Will start from scratch using your 10 steps with dracut > instead of genkernel. I just tried the steps and indeed I forgot to mention a couple of things. You should generate the initramfs with dracut before you run grub2-mkconfig - that way grub will find the initramfs. The other issue is that of naming the root partition on the kernel cmdline. When you open the luks partition using `cryptsetup open` you give it a device-mapper name. In some cases grub will save this name in grub.cfg. So grub's kernel cmdline would contain e.g. root=/dev/mapper/crypto dracut will by default open the luks partition with a name of the form luks-. This mismatch will prevent root from mounting. To overcome this and guarantee a predictable name add an /etc/crypttab entry of the form UUID= then generate the initramfs with dracut again, and it will copy this file and use it to name the luks partition upon opening. Just make sure you use the same name during installation and in crypttab - this is not mandatory but it makes things easier. Howver, sometimes grub will generate a cmdline entry of the form `root=UUID=` if it finds an initramfs which will prevent this issue. Also remember that there are two things: the uuid of the encrypted luks partition (this needs to go in crypttab), and the uuid of the decrypted partition inside luks (this needs to go in fstab and the root cmdline). Just make sure everything is consistent.
Re: [gentoo-user] Re: Full system encryption on Gentoo
On Thu, Dec 31, 2015 at 02:49:42PM +0100, Jeremi Piotrowski wrote: > I just tried the steps and indeed I forgot to mention a couple of things. And one more: don't format the full disk as luks, because there won't be any space for grub and grub2-install will error out. Make a single partition (default should be offset 2048 sectors from the beginning of the disk) which leaves plenty of space for grub's bootstrap, and format that as luks.
Re: [gentoo-user] emerging with distcc: What's taking so long?
On Tue, Jan 5, 2016 at 12:49 PM, lee wrote: > The gui monitor doesn't seem to exist. Recompile distcc with the gtk use flag.
Re: [gentoo-user] Re: Java 8 and remote access
On Sun, Jan 31, 2016 at 1:02 AM, Grant wrote: >> I need to run a Java 8 app remotely. Can this be done on Gentoo? > > > Actually I could run the app locally if there is a good way to run > relatively untrusted Java apps. Otherwise I should run it remotely. > What exactly do you mean by "run remotely"? Ssh to run on a different host, container/vm to run locally are all viable options but I don't quite understand what you're asking.
Re: [gentoo-user] Re: Java 8 and remote access
On Sun, Jan 31, 2016 at 12:31:51PM +0100, Jeremi Piotrowski wrote: > On Sun, Jan 31, 2016 at 1:02 AM, Grant wrote: > >> I need to run a Java 8 app remotely. Can this be done on Gentoo? > > > > > > Actually I could run the app locally if there is a good way to run > > relatively untrusted Java apps. Otherwise I should run it remotely. > > > > What exactly do you mean by "run remotely"? Ssh to run on a different > host, container/vm to run locally are all viable options but I don't > quite understand what you're asking. Heck, since java is already running on the jVM I would even consider it safe enough to just run as a different user or just in a chroot. In any case, none of this is gentoo specific.
Re: [gentoo-user] vdpau with intel video card
On Tue, Mar 01, 2016 at 09:08:45AM +0300, Ярослав wrote: > upgrading packeges all libs from /usr/lib64/vdpau got > deleted, except for libvdpau_trace.so. So I still didn't > get the libvdpau_i965.so. > There is no libvdapu_965.so. On Tue, Mar 01, 2016 at 09:52:57AM -0600, Corbin wrote: > Searching for VDPAU support on an Intel i695 GPU ... "libvdpau-va-gl" is > required. > That library may not be in the Gentoo Repository. It's part of the x11 overlay in layman layman --add x11 Let's clear this up: vdpau and vaapi are two APIs for hardware video acceleration. The package x11-libs/libvdpau-va-gl implements the VDPAU api using VAAPI+OPENGL library calls, the package x11-libs/libva-vdpau-driver implements the VAAPI api using VDPAU library calls. So these two are effectively opposites. I was once wondering if it would be possible to bring these two to recurse... But let's get back on topic: VAAPI is pretty much only implemented on intel cards, and only on newer ones supported by the i965 driver. So you will definitely need VIDEO_CARDS="i965 intel" set. Next you need to install x11-libs/libva which will also pull in x11-libs/libva-intel-driver. Check with vainfo that this works. After that you can install x11-libs/libvdpau-va-gl and x11-misc/vdpauinfo. You also need to have the environment variable VDPAU_DRIVER=va_gl set somewhere. Drop this in /etc/profile.d/vdpau.sh: export VDPAU_DRIVER=va_gl Do the usual env-update && source /etc/profile, and then check with vdpauinfo that everything works. Now you can build all the packages you need with the vdpau USE flag.
Re: [gentoo-user] Re: openssl upgrade may miss some needed rebuilds
On Wed, Mar 2, 2016 at 6:54 PM, Rich Freeman wrote: > Upstream really dropped the ball on this. When I'm updating packages > I certainly don't carefully review all their ABIs and SONAMEs. > Without some kind of automatic QA tool it would be a pretty big > undertaking. I might go see if there is such a tool though, maybe > that might be a good outcome if such a tool exists. > I recall reading about such a tool: http://ispras.linuxbase.org/index.php/ABI_compliance_checker I haven't tried it out, but I would be curious to see whether it would have caught this case.
Re: [gentoo-user] Can't get any boot method working
On Fri, Mar 25, 2016 at 04:31:11PM +, Dan Douglas wrote: > > grub2-mkconfig generates no menu entries. Do I need anything generated > by /etc/grub.d/00_header? That output looks like garbage. > You are better off not removing the header. Are you running grub2-mkconfig from the chroot? The kernel+initramfs should be located at /boot. In the case of btrfs you can make it a subvolume but at the time of running grub2-mkconfig you should be chrooted into your root subvolume and your kernel has to be at /boot. > The only real relevant configuration in /etc/default/grub should be correct. > > GRUB_CMDLINE_LINUX="root=UUID=f0373f0c-3798-4965-a845-b1b94cc14731 > rootfstype=btrfs > rootflags=rw,noatime,compress=zlib,space_cache,subvol=rootfs" >From what I can remember you don't need any of that - if anything would be relevant it's subvol=rootfs but grub should be able to figure that out. > grub2-install also fails: > > # grub2-install --target=x86_64-efi /dev/sda > Installing for x86_64-efi platform. > grub2-install: error: cannot find EFI directory. > > What EFI directory? The one I created under /boot? > /dev/sda1 is your EFI partition, formatted as vfat. You can have it mounted under /boot or under /boot/efi (I recommend the latter). Install like this # grub2-install --target=x86_64-efi --efi-directory=/boot/efi You can also use dracut to build an uefi executable that includes your kernel, the kernel commandline and the initramfs. # dracut --uefi
Re: [gentoo-user] NVMe drive and grub
On Thu, Apr 7, 2016 at 5:24 PM, wrote: > Hello list, > > I have a new box with an NVMe SSD drive attached to the PCI bus via an M.2 > interface. The drive shows up as /dev/nvme0n1, with partitions > /dev/nvme0n1p1, /dev/nvme0n1p2, ... > > After following the instructions in the handbook for a UEFI system, I get as > far as a kernel panic. Grub finds the kernel and starts it, but it seems to > be passing a null root device name. > What is in your grub.conf? Have you thought about adding an initramfs and letting it drop you to its rescue shell so that you can investigate?
Re: [gentoo-user] booting - I don't anystand how the (Linux) world works anymore
On Sat, Jun 25, 2016 at 08:33:31PM +0200, Helmut Jarausch wrote: > > So, I came up withroot=UUID=uuid_number of the root file system. > > But to my surprise I now got a kernel panic > syncing: VFS: unable to mount root fs on unknown block(0,0) > > So, please tell me what I'm missing? root=UUID lines are normally interpreted by the initramfs to find the right device to mount. If you don't use one, the kernel pretty much only supports root=/dev/ or root=PARTUUID identifiers (check kernel source file init/do_mounts.c:182,207 for details). You can find your PARTUUID using `blkid` or `lsblk -o +PARTUUID`.
Re: [gentoo-user] Gitlab experiences
On Wed, Jul 13, 2016 at 03:19:00PM +0200, Ralf wrote: > Hi James, > > On 07/13/16 14:44, James wrote: > > > > Has anyone attempted to install a self hosted gitlab on gentoo server(s)? > > A small gentoo cluster/container setup? Using a Distributed File System, > > like cephfs, orangefs or other DFS? > I recommend to deploy gitlab inside a Debian LXC/Docker container as > Gitlab guys provide and maintain precompiled .deb packages. You do not > want to compile it on your own as it comes with a load of dependencies. > And once dependencies change you really might run into trouble with > gentoo. Gitlab isn't just a tiny one-click-and-it-runs webservice, it's > a whole ecosystem. I would deploy it with docker. The gitlab guys push official images of the main gitlab app[1] and CI runners[2] to dockerhub. That should be the easiest path to getting it up and running in no time. That being said, gitlab does not really play well with clustering in general. I don't think the main part of the app does any kind of horizontal scaling (gitlab.com is hosted on a single server) so you need a fairly beefy server. And while storage should be entirely up to you (the app _should_ be indifferent to what you use) most folks appear to run with local disks or NFS. > > > > Any experiences with gitlab are most welcome for comment, good or bad. > Yes. Bad. Slow, unreactive, eats tons of resources. Doesn't scale with > large repos (except you have unlimited access to hardware resources). A > Linux kernel git mirror finally crashed it. I can second that - it's slow and load times are long even when performing the most basic operations such as opening a small file or viewing a single commit (you _will_ notice - it regulary takes several seconds). Mind you the gitlab folks are working on improving this since several releases. As a user, another issue I have with it is that the merge request/review interface is just terrible. There is _no_ merge request versioning, so either you submit your code in perfect shape at first try, or any change (amending/rebasing/merging) will cause the changelist to be duplicated many times over. You also lose track of the review history instantly - old comments are either concealed or swallowed. There is also no CLI utility to automate common review-related tasks (submitting/responding to review) so you are forced to do everything over the slow WebUI. If you care at all about the codereview aspects, I would recommend gerrit or phabricator. Both have cli utilities (git-review and arcanist) and while some claim they are ugly (heard that one especially about gerrit) they are 100x more practical. If you only care about having a repository browser then gitlab can work but there are simpler apps out there (gogs/pagure). [1]: https://hub.docker.com/r/gitlab/gitlab-ce/ [2]: https://hub.docker.com/r/gitlab/gitlab-runner/
Re: [gentoo-user] Install dkms: which package
On Sat, Jul 16, 2016 at 06:37:28AM -0300, Facundo Curti wrote:
>
> But how can I install the module without dkms? usually I make this
> automatically with dkms :S
Come on - the module has an install.sh script that installs it without
dkms.
$ cd /usr/src
$ hg clone https://bitbucket.org/Swoogan/aziokbd
$ ./install.sh
or instead of running their script just do:
$ make install
DKMS is a means of hooking into the kernel-update process in binary
distros to automatically recompile external modules. As we compile our own
kernels (most of us?), compiling modules manually is not that big of an
issue (we control when the kernel update happens).
However, as it is more practical to have portage handle the rebuilds, here
is a quickly written ebuild for you.
# Copyright 1999-2016 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Id$
EAPI=6
inherit eutils linux-info linux-mod
if [ "${PV}" = "" ]; then
inherit mercurial
EHG_REPO_URI="https://bitbucket.org/Swoogan/aziokbd";
KEYWORDS=""
else
SRC_URI="???"
KEYWORDS="~x86 ~amd64"
fi
DESCRIPTION="Linux kernel driver for Microdia Keyboards"
HOMEPAGE="hg clone https://bitbucket.org/Swoogan/aziokbd";
LICENSE="GPL-2"
SLOT="0"
IUSE="-backslash-fix"
MODULE_NAMES="aziokbd(kernel/drivers/input/keyboard:${S})"
BUILD_TARGETS="clean default"
MODULESD_AZIOKBD_ADDITIONS=(
"options usbhid quirk=0x0c45:0x7603:0x0007"
)
src_compile() {
if use backslash-fix; then
BKSLFIX=y
else
BKSLFIX=n
fi
BUILD_PARAMS="KSRC=${KERNEL_DIR} BKSLFIX=${BKSLFIX} M=${S}"
linux-mod_src_compile
}
Re: [gentoo-user] Re: Firefox 45.2.0 segfaulting
On Thu, 14 Jul 2016, Grant Edwards wrote: > > Same behavior with 47.0.1. Starting 47 in "safe mode" avoids the > problem, but starting 47 in normal mode with all extensions disabled > still crashes. > > I downgraded to 38.8, and everyting works fine again. > Are you running all of these test with the build-from-source www-client/firefox ebuild? I used to do that but after not being able to track down the cause of some weird font-related graphics glitches, I settled for the binary www-client/firefox-bin package and haven't had problems since. I always run ~amd64 firefox-bin, currently at 47 and have no problems whatsoever. Definitely recommend you try that before going for chrome.
Re: [gentoo-user] cross-compile attempt
On Sun, Jul 31, 2016 at 07:40:37PM +0100, Mick wrote: > > * ARCH is not set... Are you missing the '/usr/i686-pc-linux- > * gnu/etc/portage/make.profile' symlink? Is the symlink correct? Is your > * portage tree complete? > === > > As far as I can tell the link is there: > > # ls -la /usr/i686-pc-linux-gnu/etc/portage/ > total 8 > drwxr-xr-x 1 root root 56 Jul 31 19:32 . > drwxr-xr-x 1 root root 20 Jul 31 18:32 .. > -rw-r--r-- 1 root root 1019 Jul 31 19:32 make.conf > lrwxrwxrwx 1 root root 30 Jul 31 17:48 make.profile -> > /usr/portage/profiles/embedded > drwxr-xr-x 1 root root 32 Jul 31 18:16 profile > > and it was created when I ran 'crossdev --stable -v -t i686-pc-linux-gnu'. > As far as I know, ARCH is one of those variables that has to be specified in a profile, not in make.conf. A quick solution is to place the ARCH=x86 line into the site-specific override .../etc/portage/profile/make.defaults Although in this case your choice of profile may simply be wrong. The embedded profile is the crossdev default that pretty much only has busybox. Choose something like default/linux/x86/13.0 or if you want a lighter libc how about default/linux/uclibc/x86 (or hardened/linux/musl/x86). That should give you a more complete @system set. > > What am I missing? How would/do you go about achieving the same objective? > Since you are doing this on an amd64 box which can natively run x86, if you want to achieve the same goal faster, start with a x86 stage3, chroot into it and emerge a couple packages that you want to add, then tar it up and load onto your 32 bit box. If you want to add packages later, emerge them with '-b' in the chroot (on the amd64 box), and then follow https://wiki.gentoo.org/wiki/Binary_package_guide to allow the 32-bit box to install them as binary packages. Ofcourse if you want to learn crossdev, then this is a great chance to do so.
Re: [gentoo-user] Re: [was cross-compile attempt] 32bit chroot
On Mon, Aug 01, 2016 at 10:21:20PM +0100, Mick wrote: > > I think libreoffice, chromium and firefox will be compiled in a chroot from > now > on and then emerged as binaries. This is the difference for libreoffice: > > Sat Aug 29 06:09:09 2015 >>> app-office/libreoffice-4.4.4.3 >merge time: 15 hours, 34 minutes and 2 seconds. > > Sun Sep 13 01:36:03 2015 >>> app-office/libreoffice-4.4.5.2 >merge time: 15 hours, 13 minutes and 17 seconds. > > Sun Nov 29 02:30:04 2015 >>> app-office/libreoffice-5.0.3.2 >merge time: 16 hours, 54 minutes and 28 seconds. > > Sun Mar 27 09:31:20 2016 >>> app-office/libreoffice-5.0.5.2 >merge time: 17 hours and 8 seconds. > > Mon Aug 1 22:17:15 2016 >>> app-office/libreoffice-5.1.4.2 >merge time: 1 minute and 31 seconds. > > (chromium takes even longer!) :-) > Does it make sense to compile your own versions of these packages and then binary merge, when portage already contains binary ebuilds for these packages? (firefox-bin/libreoffice-bin/google-chrome)
Re: [gentoo-user] Is there a reason why LLVM/Clang ebuilds don't support "mutislot"?
On Sun, Aug 28, 2016 at 11:58:03PM -0400, P Levine wrote: > Other distros like Ubuntu support the installation of multiple versions of > LLVM/Clang side by side. One of the things Clang is really good at is > support for the most recently approved upcoming features of the C++17 > standard. The best support for testing such features is with the latest > sys-devel/llvm-. However if I want to compile Mesa against a stable > version LLVM/Clang as well, I don't get that option. I'm not sure such a configuration is fully supported upstream, but the way ubuntu (and debian) does this is (was?) painfully broken. If I recall correctly they first build it in one location and then move it around and try to partially solve things through symlinks. This totally destroys llvm-config and LLVM's cmake find_package module meaning those things have hard-coded paths that have nothing to do with where the things are installed on the system. Simple example, the following CMakeLists.txt cmake_minimum_required(VERSION 3.4) project(llvm-test) find_package(LLVM) errors out on ubuntu with: CMake Error at /usr/share/llvm-3.8/cmake/LLVMConfig.cmake:178 (include): include could not find load file: /usr/share/llvm/cmake/LLVMExports.cmake Call Stack (most recent call first): CMakeLists.txt:4 (find_package) CMake Error at /usr/share/llvm-3.8/cmake/LLVMConfig.cmake:181 (include): include could not find load file: /usr/share/llvm/cmake/LLVM-Config.cmake Call Stack (most recent call first): CMakeLists.txt:4 (find_package) -- Configuring incomplete, errors occurred! Fixing this one path leads to more suffering further down the road. The way gentoo maintainers package LLVM is much saner and developer friendly.
Re: [gentoo-user] Re: help! IP blocking not working
On Tue, Sep 06, 2016 at 01:57:54PM -0700, Grant wrote: > > Hi, my site is being ravaged by an IP but dropping the IP via > > shorewall is seeming to have no effect. I'm using his IP from nginx > > logs. What you really need is to set up net-anlyzer/fail2ban and not do this kind of stuff manually. It automates parsing logs for attacks and setting up persistent iptables rules to block them. As soon as I assigned a dns domain name to my home ssh-server and made it available externally I was getting attacked by multiple IP addresses from china, and as soon as one IP was banned they came at me with another one. After I set up fail2ban and set a low preauth limit along with lifetime bans, this whole cat-and-mouse game started going more to my liking. Highly recommend you try it, it comes with lots of predefined rules/templates that you can choose from (I see nginx-botsearch and nginx-http-auth are included).
Re: [gentoo-user] Bash failed to compile
On Fri, Sep 23, 2016 at 04:14:24AM +0200, meino.cra...@gmx.de wrote: > Hi, > > may be bash is missing a lib? > > /usr/lib64/libtinfo.so.5: error adding symbols: DSO missing from command line Nope, this is a well known class of build system bugs, we have a tracker for it in our bugzilla (#457530). Search for a bash bug report, or open a new one and link it to the tracker. Briefly: libncurses can be split with some symbols (functions) landing in libncurses and others in libtinfo. This is controlled by the tinfo USE flag to sys-libs/ncurses. I don't know whose idea this was but I know that, among others, CUDA requires this split. But I'm actually starting to think that this fight is not worth fighting and we should just go back to a single library. The split just adds unnecessary complexity. Perhaps emerging sys-libs/ncurses[-tinfo] and symlinking libncurses.so to libtinfo.so would be the best solution for all.
Re: [gentoo-user] Strive for zero swap usage?
On Thu, Oct 06, 2016 at 03:31:59PM -0700, Grant wrote: > > Does this look OK? It looks to me like heavy swapping in and out with > plenty of free memory (minus buffers/cache). > Have you looked at the _units_ displayed by vmstat and munin? You're looking at **kilobytes** being swapped and calling that _heavy swapping_... In general your system shows that you have a lot of memory used as cache, and I would guess that the kernel thinks its better to keep using that memory for caching and swap out the occasional page of application data. The alternative would be dropping some cache - might be worse for performance. Swapping does not seem to be something that could be causing any issues for you. Do you even have an issue with response times of your webservers? If so (which you haven't explicitly stated here) it would be better for you to look at things from that end - what is taking long in the webserver and what is causing those things to take that long. As it stands, you're using anti-methods of performance analysis: Streetlight Anti-Method: 1. Pick observability tools there are familiar 2. Run tools 3. Look for obvious issues and Traffic Light Anti-Method 1. Open dashboard (munin) 2. All green? Assume everything is good 3. Something red? Assume that's a problem If you're really interested in this kind of stuff I recommend checking out the materials that Brendan Gregg has gathered over the years. http://www.brendangregg.com/methodology.html and other links on his page.
Re: [gentoo-user] Is there a PowerShell package?
On Sat, Nov 12, 2016 at 11:20:06PM +0100, Bertram Scharpf wrote: > Hi, > > for a special job I need to do a piece of work in Microsofts > PowerShell, but I have no Windows machine running. I know > that there is a PowerShell for Linux because I have one > installed on a Ubuntu. Yet, I do not want to shut down my > Gentoo and boot Ubuntu every time when I have to try > something in PowerShell. I would like to install it on > Gentoo. This is the perfect use case for running containers on your development machine. Just set up lxc/lxd/docker on your gentoo machine and start up an ubuntu container in any of the runtimes. Works flawlessly, and you don't have to reboot or run a heavyweight VM. Lxc/lxd are more suitable for long running containers, docker would be great for containers that you spin up and tear down whenever you need. Give it a try.
Re: [gentoo-user] OT: Can a linux vmware guest tell if its host is CPU constrained?
On Thu, Jul 30, 2020 at 2:52 AM Adam Carter wrote: >> >> > So should I run something like: date ; time > > 100%CPU for a minute> ; date ? >> >> No, date will pull from your RTC, which is usually kept up to date with an >> asynchronous >> counter. >> >> First check GNU top(1) and look in the %Cpu line for "st." That is % CPU >> time stolen. If it is >> nonzero then the guest time's accounting is probably working. It's not >> typical for the >> hypervisor to hide this information. It's really important for load >> balancing. > > > Thanks for that. I haven't seen any non-zero stolen time yet, however. > > FWIW vmstat also shows stolen time. Stolen time reporting through vmstat/top only works on xen and kvm hypervisors, it wasn't implemented for vmware. It actually looks like it was finally submitted for linux v5.7 (https://lore.kernel.org/lkml/20200331100353.ga37...@gmail.com/). If you want those numbers for older kernels, fetch this repository: https://github.com/dagwieers/vmguestlib and run vmguest-stats. You'll also need open-vm-tools installed. Have used this many times in the past and the numbers are good.
