Jump to content

walterw

Members
  • Content Count

    39
  • Joined

  • Last visited

  • Days Won

    1

Everything posted by walterw

  1. After having setup a system with UEFI, I have a few questions: 1. Why is the 2MB BIOS partition needed as it is not formatted? Does GRUB actually install itself here? 2. The EFI partition need only contain GRUB files (including the grub.cfg which tells GRUB how to boot an OS). 3. If I were to boot a system without EFI support, the 2MB partition should kick in and GRUB would still read the grub.cfg on the EFI partition? Is that accurate? So, when installing GRUB with the EFI options, does it still setup the MBR part? Thanks, Walter
  2. If I want to build a custom debian kernel, am I correct in my understanding that I would want to use: -binary +luks for my use flags? If so, it appears the ebuild won't let me do it. If I'm not worried about building unnecessary modules and features into the kernel, then, I can just call it a day by adding the luks use flag and defaulting to binary?
  3. @drobbins Thanks for the information - I will make the switch now. Thanks, Walter
  4. Hi - has there been any update to updating gentoo-sources in 1.4? I use gentoo-sources. What is the advantage of debian-sources over gentoo-sources, stability, security? What is the advantage of gentoo-sources over debian-sources, speed? Thanks, Walter
  5. I have been using bootable USB thumbdrives for sometime and had recently attempted to boot to it to find that I cannot. I get a kernel panic: trap invalid opcode ip: ... error:0 in ld-2.29.so When I boot the same thing but from my local disk, I can boot ok. The USB thumbdrive contains a nearly identical boot: same kernel same init different grub.cfg (different UUID for device) Root is on a LUKS partition with LVM for both USB and local disk. Root is a squashfs image. I am using overlayfs to make root writable. The system I was booting on was different from my build system and I recently changed the sub-architecture I'm using. I have another configuration which specifically targets that architecture, so I need to use that instead. Walter
  6. Ah, you're right, my chromium build fails and isn't cleaned up. I will modify the build process to wipe /var/tmp for me :). I was checking after the entire build completed and /var/tmp was wiped - I forgot about failed builds not being cleaned.
  7. I am attempting to install git-annex and apparently this has not been installed for some time, but I did have git-annex installed a while back. My /var/tmp is tmpfs and has 12G allocated. When attempting to install git-annex, it requires at least 8GB free and fails: Checking for at least 8 GiB disk space at "/var/tmp/portage/dev-lang/ghc-8.0.2/temp The build will most probably fail, you should enhance the space as per failed tests. I could do a bind mount here, but why does the ebuild care if it is disk or memory for the underlying device? Is there another way around this short of doing a bind mount here? Thanks, Walter
  8. The build failed here now: /usr/include/glib-2.0/glib/gtypes.h:549:26: note: declared here 549 | typedef struct _GTimeVal GTimeVal GLIB_DEPRECATED_TYPE_IN_2_62_FOR(GDateTime); | ^~~~~~~~ cc1plus: all warnings being treated as errors cc1plus: all warnings being treated as errors cc1plus: all warnings being treated as errors > Task :graphics:ccLinuxGlassGlassgtk2 FAILED :graphics:ccLinuxGlassGlassgtk2 (Thread[Task worker for ':' Thread 2,5,main]) completed. Took 5.401 secs. FAILURE: Build failed with an exception. * What went wrong: Execution failed for task ':graphics:ccLinuxGlassGlassgtk2'. > java.util.concurrent.ExecutionException: org.gradle.process.internal.ExecException: Process 'command 'gcc'' finished with non-zero exit value 1 * Try: Run with --debug option to get more log output. Run with --scan to get full insights.
  9. Changing the ebuild to also check for that directory got my swt installation moving along ... Shall I clone the repo in git, then open an MR? Walter
  10. I am unable to emerge openjfx: /etc/portage/package.use/java dev-java/oracle-jdk-bin javafx gtk3 -nsplugin -webstart dev-java/openjdk-bin javafx gtk3 -nsplugin -webstart dev-java/openjdk-jre-bin javafx gtk3 -nsplugin -webstart >=dev-java/swt-4.10 cairo emerge openjfx -v My system VM is: /etc/java-config-2/current-system-vm/jre/lib/ -> /usr/lib/jvm//openjdk-bin-11 eselect java-vm list system Available Java Virtual Machines: [1] openjdk-bin-11 system-vm [2] openjdk-jre-bin-11 [3] oracle-jdk-bin-1.8 >>> Compiling source in /var/tmp/portage/dev-java/swt-4.10/work ... * libjawt.so not found in the JDK being used for compilation! * ERROR: dev-java/swt-4.10::java-kit failed (compile phase): * cannot build AWT library * * Call stack: * ebuild.sh, line 93: Called src_compile * environment, line 4508: Called die * The specific snippet of code: * die "cannot build AWT library"; * * If you need support, post the output of `emerge --info '=dev-java/swt-4.10::java-kit'`, * the complete build log and the output of `emerge -pqv '=dev-java/swt-4.10::java-kit'`. !!! When you file a bug report, please include the following information: GENTOO_VM=openjdk-bin-11 CLASSPATH="" JAVA_HOME="/opt/openjdk-bin-11.0.3_p7" JAVACFLAGS="-source 1.8 -target 1.8" COMPILER="" and of course, the output of emerge --info =swt-4.10 * The complete build log is located at '/var/tmp/portage/dev-java/swt-4.10/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/dev-java/swt-4.10/temp/environment'. * Working directory: '/var/tmp/portage/dev-java/swt-4.10/work' * S: '/var/tmp/portage/dev-java/swt-4.10/work' Now, the ebuild is checking for lib here: if [[ -f "${JAVA_HOME}/jre/lib/${AWT_ARCH}/${JAWTSO}" ]]; then export AWT_LIB_PATH="${JAVA_HOME}/jre/lib/${AWT_ARCH}" elif [[ -f "${JAVA_HOME}/jre/bin/${JAWTSO}" ]]; then export AWT_LIB_PATH="${JAVA_HOME}/jre/bin" elif [[ -f "${JAVA_HOME}/$(get_libdir)/${JAWTSO}" ]] ; then export AWT_LIB_PATH="${JAVA_HOME}/$(get_libdir)" else eerror "${JAWTSO} not found in the JDK being used for compilation!" die "cannot build AWT library" fi Now, the libawt.so is here: locate libawt.so /opt/idea-community-2019.2.4/jbr/lib/libawt.so /opt/openjdk-bin-11.0.3_p7/lib/libawt.so /opt/openjdk-jre-bin-11.0.3_p7/lib/libawt.so /opt/oracle-jdk-bin-1.8.0.202/jre/lib/amd64/libawt.so So, I think the function get_libdir isn't able to find the library in /lib/libawt.so underneath my jre home directory. Walter
  11. When I launch nestopia, I get an immediate segfault: nestopia[5264]: segfault at 10 ip 00007fe24e3e9e74 sp 00007ffc936c9e40 error 4 in libgcc_s.so.1[7fe24e3dc000+11000] [83739.944349] Code: ff ff c5 f8 29 9d 00 ff ff ff c5 f9 6f 95 80 fe ff ff c5 f9 6f 9d 90 fe ff ff c5 f8 29 a5 10 ff ff ff c5 f9 6f a5 a0 fe ff ff <49> 83 7c 24 10 00 c5 f8 29 ad 20 ff ff ff c5 f9 6f ad b0 fe ff ff I'm wondering if anyone else has it running or recommends another emulator.
  12. Thanks, for catching that. I'll make those changes and report back. My make.conf is automatically generated by a utility ... that said I am no longer writing empty values (such as VIDEO_CARD=""), but instead have this: HOST="x86_64-pc-linux-gnu" CFLAGS="-march=ivybridge -O2 -pipe" CXXFLAGS="${CFLAGS}" CPU_FLAGS_X86="aes avx mmx mmxext popcnt sse sse2 sse3 sse4_1 sse4_2 ssse3" MAKEOPTS="-j5" # patches below here # kernel USE="${USE} symlink -fortran -openmp acpi" The other changes I made were: ->=x11-base/xorg-server-1.17.4-r2 xvfb +#>=x11-base/xorg-server-1.17.4-r2 xvfb -media-libs/mesa bindist -gallium -d3d9 -llvm -x11-base/xorg-server glamor glvnd udev +#media-libs/mesa bindist -gallium -d3d9 -llvm +#x11-base/xorg-server glamor glvnd udev
  13. When playing a video, I noticed that I apparently am lacking opengl acceleration as the video plays terribly slowly. My package use flags: media-libs/mesa bindist -gallium -d3d9 -llvm x11-base/xorg-server glamor glvnd udev Part of my make.conf CHOST="x86_64-pc-linux-gnu" CFLAGS="-march=ivybridge -O2 -pipe" CXXFLAGS="${CFLAGS}" CPU_FLAGS_X86="aes avx mmx mmxext popcnt sse sse2 sse3 sse4_1 sse4_2 ssse3" MAKEOPTS="-j5" USE="opengl X xinerama dbus gtk gtk3 -gnome-keyring" VIDEO_CARDS="" SANE_BACKENDS="" LINGUAS="" I have the gfxcard-intel mix-in enabled, so my VIDEO_CARDS should be set to all intel. Additionally, I have followed the guide here to my knowledge and have installed libva-intel-driver: https://www.funtoo.org/Make.conf/VIDEO_CARDS/Mix-ins. Is there something else I'm missing? This was working fine on 1.3 and I believe I had manually set the VIDEO_CARDS setting then as there may not have been a mix-in.
  14. I must have had an old kernel built, so the kernel source directory was not configured properly.
  15. When attempting to build ZFS on release-1.4, I am getting this error: checking whether modules can be built... no configure: error: *** Unable to build an empty module. emerge zfs -vp [ebuild N ] sys-fs/zfs-kmod-0.8.1::core-kit USE="-custom-cflags -debug -rootfs" 0 KiB [ebuild N ] sys-fs/zfs-0.8.1::core-kit USE="python -custom-cflags -debug (-kernel-builtin) -rootfs -split-usr -static-libs -test-suite" PYTHON_TARGETS="python2_7 python3_7 -python3_5 -python3_6" 0 KiB Even though I have build a kernel and have already booted to it (outside of this chrooted env), I feel like there is something not quite right with my /usr/src directory possibly. Any ideas?
  16. I do not use KDE, so I do not know the answer to your question off-hand; however, would you be able to do: 1. post emerge --info 2. post the stack trace / log message According to this: https://packages.gentoo.org/packages/sys-apps/accountsservice You are right that systemd should be an option. Additionally, exploring the ebuild from funtoo on github: https://github.com/funtoo/nokit/blob/master/sys-apps/accountsservice/accountsservice-0.6.50-r1.ebuild REQUIRED_USE="?? ( elogind systemd )" !systemd? ( !elogind? ( sys-auth/consolekit ) ) I think if I'm interpreting that correctly, when systemd is disabled, elogind must also be disabled and you need consolekit.
  17. Yes, this is very true, and well-said. What are your thoughts on the new Linux phones coming out such as the Librem5?
  18. I was running unbound in conjunction with DNSSEC to basically ensure that responses aren't tampered with (because it is plaintext, if there is a malicious party in between me and the DNS server, it could modify the response without me knowing). Now, if I run unbound by itself in this manner, basically, my DNS queries are sent out in the open, plaintext and then unbound will do all of that magic for me, ensuring that the IP address for google.com is indeed what it should be. Now, not all zones to my knowledge are signed. Now, if I'm worried about someone seeing what my DNS traffic is, then they're going to be able to see my IP traffic too, so I don't see how much benefit there is to hide my DNS traffic. Yes, the IP address might be used by many domains, but they might all be fairly closely related. Now, if I were using forwarding DNS queries to another provider, I am basically delegating that responsibility to them and may or may not be able to validate the result. Is that an accurate assessment? And, since they may not provide DNSSEC, if the DNS provider isn't offering DoH or DNSCrypt, then I have no guarantee that the IP address returned is accurate?
  19. Hi, Is 1.1.1.1 just no logging or does it also filter out ads / spam as well - I think it supports DoH so at least if you don't have DNSSEC, then you can be sure your DNS isn't tampered with. My concerns would be: 1. is this DNS record valid, how can I be sure 2. is this DNS record for ads / malware / spam / etc. - use a blacklist / blocklist (uBlock)
  20. I do have liberation-fonts installed. I built terminator for python 3.6, I see it is looking for python 2.7. Also, on a possibly related note, I was attempting to watch videos in VLC and on my new image, the video doesn't play very well at all. I suspect there may be an issue with my opengl setup? Perhaps it is related? It does report file not found for libGldispatch.so.0 and lib64/charset.alias.
  21. Agreed, so what you need me to do is to switch back to the default and then I think I have a utility installed that lists all the open files (strace?) during execution and see which ones are reported as file does not exist? Then, that should be very clear which font is missing? I would be happy to do that, just need a teeny, tiny bit of guidance (if the above approach sounds reasonable) :).
  22. Okay, I think you're right. So, I think I thought I must have selected a monospaced font, but did not. When I selected the Deja Vu Sans Book Mono, it looks good, but when I select just a Deja Vu Sans (non-mono), I have the problem. Okay, that explains it.
  23. I installed xfce4-terminal and st and both display just fine. I am leaning toward migrating to xfce4-terminal since it is fairly lightweight, written in c versus python (I've had some annoyances in the past where terminator would hangup when I was doing a fair amount of I/O in the terminal), and still has tabs. I use a tiling WM, so terminator doesn't provide anything I cannot do with my WM with better performance. I am curious as to what happened that is causing this issue, but in the end, I am planning to migrate back toward xfce4-terminal. Walter
  24. Thanks - I don't have an .Xresources and I did try playing with those font settings. Also, my old system does not have a .Xresources either. On a random note, I noticed that between 1.3 and 1.4, the path to libraries has changed, I think the 64 was dropped and it is just /lib or something. I wonder if that might be related ... Also, I'm not sure if it helps or not, but the fonts all look perfectly fine when previewing them in terminator's configuration, but once I choose them, that is when the appear squished like above. Walter
×
×
  • Create New...