Keyboard input / init + questions
jxm at risingtidesystems.com
Mon Sep 27 09:13:06 PDT 2010
OK, I removed my patches to get access to detail mode again and the tty is
filling with "couldn't write a log entry; log text too long" messages, but
there is also some other output that scrolls past too fast for me to get it.
On Mon, Sep 27, 2010 at 5:54 PM, Jerome Martin <jxm at risingtidesystems.com>wrote:
> Mmmhh. I am not familiar with these options so I blindly copied them, but
> it does not work. Only the kernel logs in dmesg, and same as before
> /dev/vcs1. Here is my full commandline (tried both with and without
> plymouth:debug as I suspect it might be redundant with plymouth.debug= ). An
> echo in /dev/kmsg does show up in dmesg ok. I also tried moving the options
> on the line, still no cigar.
> [ 0.000000] Command line: initrd=/live/initrd.img boot=live noprompt
> nouser splash vga=0x311 plymouth:debug plymouth.debug=file:/dev/kmsg
> printk.time log_bug_len=1M BOOT_IMAGE=/live/vmlinuz
> On Mon, Sep 27, 2010 at 4:18 PM, Ray Strode <halfline at gmail.com> wrote:
>> On Mon, Sep 27, 2010 at 3:13 AM, Jerome Martin
>> <jxm at risingtidesystems.com> wrote:
>> > Fred here is using a patch very similar to my own, basically moving
>> > the TIOCSCTTY of init out of plymouth's way.
>> > As for the logs, I use a much stripped down system environment, with a
>> > root and a COW on top of it that does not get mounted right away, so
>> > might explain the log issues.
>> Oh, okay. So, if the file system might be read-only at the time
>> plymouth quits, then you should be able to boot with
>> plymouth.debug=file:/dev/kmsg printk.time log_bug_len=1M
>> on your kernel command line and the debug output will go to "dmesg"
>> along wth nice timestamps. Then you can just
>> dmesg > output.txt
>> and post it here.
> Jérôme Martin
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the plymouth