You are not logged in.
Pages: 1
I have had Plymouth installed for months now, and it has been working well. I followed the directions in the wiki, installed the open source video drivers, etc etc.
Recently, it no longer shows a boot splash screen when starting up, or even when i shut down. I have tried changing the splash image, uninstalling, reinstalling, reconfiguring etc. I dont know what was updated or changed to cause this... any help would be much appreciated.
Regards
Offline
For me, Plymouth stopped working properly after the libpng update was available in the repositories at the beginning of this month. Before, plymouth started immediately after grub (I even don't see the fsck messages) and stays for some seconds till the boot process has finished. But after the update, it only comes up for a fraction of a second, just before the login prompt is reached. (rebuild, mkinitcpio etc. didn't solve the problem)...
Anyone else has this problem too, and could explain the reasons?
Btw: with Manjaro, I don't have this problem ![]()
Thanks
NyQuil
Offline
Just an fyi.. recently setup a new laptop with plymouth, then a desktop... both have same issue..... i am sure i read on another board the same issue, but no resolve... so i at least know it is not just me... anyone know the resolve yet?
Thanks
Offline
[...] it only comes up for a fraction of a second, just before the login prompt is reached [...]
Same Problem here. Fbsplash didn't work either, but maybe this has been my fault, haven't done depper digging here.
Knowing others is wisdom, knowing yourself is enlightenment. ~Lao Tse
Offline
Same here. It does not appear to be displaying with the fb during boot, but as soon as X starts, I see the plymouth screen flash just before my DM starts.
FBsplash does seem to work for me, but only on boot. But have removed that as it is not updated for systemd.
Cheers.
Last edited by Padfoot (2013-06-15 23:10:20)
Offline
Ok, a bit late in the game, but if you are still subscribed to this thread, I have the beginnings of a solution.
Firstly, my system, I am using EFI boot with refind, and the nvidia binary blob, so I need to use a framebuffer for plymouth. No problem I thought, refind sets native resolution and when it boots the kernel, an efifb is created, so no need for uvesafb.
Well, this didn't stop plymouth failing (well partially) as I was only briefly seeing the splash screen just before X and lightdm started.
So let's begin the troubleshooting:
# cp /usr/lib/systemd/system/plymouth-start.service /etc/systemd/system/
# touch /root/plymouth.debugChange the ExecStart line to:
ExecStart=/usr/bin/plymouthd --debug --debug-file=/root/plymouth.debug --mode=boot --pid-file=/run/plymouth.pid --attach-to-sessionAnd then:
# mkinitcpio -p linuxReboot and then open the debug file we just created in /root.
From this, I found plymouth was trying to start the X11 renderer first, which fails, then the drm renderer (I am using the nvidia blob, so this is going to fail too). Finally, it tries the framebuffer renderer, which succeeds and the splash is finally displayed. The problem is, with systemd, the boot is continuing all the while plymouth is deciding on a renderer. So by the time the framebuffer renderer is successfully started and the splash displayed, the boot is practically finished.
So, upon further investigation, I found the file /usr/lib/initcpio/install/plymouth contains the following:
add_binary ${PLYMOUTH_PLUGIN_PATH}/renderers/drm.so
#add_binary ${PLYMOUTH_PLUGIN_PATH}/renderers/x11.so
add_binary ${PLYMOUTH_PLUGIN_PATH}/renderers/frame-buffer.soNow X11 is already commented out. I also commented out drm and then rebuilt my initramfs. Unfortunately, plymouth still tries both of these renderers first.
Next step, onto the building of plymouth itself. Downloading the package from AUR and grabbing the source, I found the following configure options:
--enable-gtk
--enable-drm-rendererAdding these to the configure line in the PGKBUILD set as follows:
--enable-gtk=no --enable-drm-renderer=noand everything builds fine (you can even remove the gtk dependency with this option set), and now I don't have the X11 renderer. Unfortunately, the --enable-drm-renderer=no option fails, so the drm renderer is still present.
Installing the modified plymouth and rebuilding my initramfs, then rebooting, I now only get a delay when plymouth is trying the drm renderer and failing before it gets to the framebuffer renderer. My splash now appears earlier.
All I need to work out now is how to disable the drm renderer, seeing as that option fails in build. I think if I get this figured out, I will again have my plymouth splash for the full boot again (it even works for me with the plymouth hook immediately after base and before udev).
If anyone is able to assist me in disabling the drm renderer, it would be greatly appreciated.
So there you go, on the way to a solution, at least for me anyway. But this should give you an idea on what you need to do with your installation to get plymouth working again.
Cheers.
Offline
Ok, I figured out, why plymouth is starting too late:
https://bbs.archlinux.org/viewtopic.php?id=167718
NyQuil ![]()
Offline
Ah, that makes sense. Something must be silently failing when mkinitcpio is run as my output indicates the hook is added in my chosen place:
# mkinitcpio -p linux
==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'default'
-> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux.img
==> Starting build: 3.9.9-1-ARCH
-> Running build hook: [base]
-> Running build hook: [plymouth]
-> Running build hook: [udev]
-> Running build hook: [autodetect]
-> Running build hook: [modconf]
-> Running build hook: [block]
-> Running build hook: [resume]
-> Running build hook: [filesystems]
-> Running build hook: [keyboard]
-> Running build hook: [fsck]
==> Generating module dependencies
==> Creating gzip initcpio image: /boot/initramfs-linux.img
==> Image generation successful
==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'fallback'
-> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-fallback.img -S autodetect
==> Starting build: 3.9.9-1-ARCH
-> Running build hook: [base]
-> Running build hook: [plymouth]
-> Running build hook: [udev]
-> Running build hook: [modconf]
-> Running build hook: [block]
-> Running build hook: [resume]
-> Running build hook: [filesystems]
-> Running build hook: [keyboard]
-> Running build hook: [fsck]
==> Generating module dependencies
==> Creating gzip initcpio image: /boot/initramfs-linux-fallback.img
==> Image generation successfulOffline
I've plymouth witch catalyst driver. The problem is as yours: after grub, some second of console some checks on arch's partition, and then start plymouth only for few seconds. After plymouth, fe seconds of black screen and then start kdm. At shutdown, sometimes plymouth does not work either.
Perhaps, at HOOK line in mkinitcpio.conf, have we to write "plymouth" at the line start to hide initial console?
ASUS P8P67 EVO | i7-2600k |RAM 16GB DDR 3 Vengeance@1600MHz | AMD 7950 Sapphire Dual-X | SSD: Samsung 830 250GB | HD: Western Digital Caviar Blue 1TB
Offline
Finally, I found the problem, why the plymouth hook isn't applied to the initrd-image. To fix this, do the following:
In the script
/usr/lib/initcpio/install/plymouthreplace the line
SCRIPT='plymouth'with
add_runscriptAfter this, generate a new initrd-image via mkinitcpio. Now, the plymouth hook is successfully applied and plymouth is starting again much earlier.
NyQuil ![]()
Offline
But isn't it an obsoleted functionality?
ASUS P8P67 EVO | i7-2600k |RAM 16GB DDR 3 Vengeance@1600MHz | AMD 7950 Sapphire Dual-X | SSD: Samsung 830 250GB | HD: Western Digital Caviar Blue 1TB
Offline
But isn't it an obsoleted functionality?
Offline
Fregtux wrote:But isn't it an obsoleted functionality?
The way I read that is that MODULES, BINARIES, FILES and SCRIPT are deprecated and add_runscript replaces them.
Yes?
Last edited by Padfoot (2013-08-08 07:04:43)
Offline
NyQuil wrote:Fregtux wrote:But isn't it an obsoleted functionality?
The way I read that is that MODULES, BINARIES, FILES and SCRIPT are deprecated and add_runscript replaces them.
Yes?
Not exactly. The variables were replaced by their corresponding add_* functions. So, add_runscript replaces (only) the SCRIPT variable. ![]()
Offline
I tried that change and it works! Thanks!
But, after boot, If I want read the messages that I saw at boot and those now I don't see because of earlier start of plymouth, what can I do? Dmesg, /var/log/boot.log, journalctl aren't what I mean.
ASUS P8P67 EVO | i7-2600k |RAM 16GB DDR 3 Vengeance@1600MHz | AMD 7950 Sapphire Dual-X | SSD: Samsung 830 250GB | HD: Western Digital Caviar Blue 1TB
Offline
Pages: 1