From: Jack <ostroffjh@users.sourceforge.net>
To: gentoo-user@lists.gentoo.org
Subject: Re: [gentoo-user] new genkernel problem
Date: Sat, 6 Jun 2020 19:55:58 -0400 [thread overview]
Message-ID: <c9f0b43e-38fe-18f1-5924-2008585ee15e@users.sourceforge.net> (raw)
In-Reply-To: <83c03ffc-4371-8168-5363-2e33d85cff5d@gmail.com>
[-- Attachment #1: Type: text/plain, Size: 4028 bytes --]
On 6/6/20 6:34 PM, Dale wrote:
> Neil Bothwick wrote:
>> On Sat, 6 Jun 2020 11:34:28 -0500, Dale wrote:
>>>>> From what I've read, I like my way better. I did have to change the
>>>>> names from bzimage* to kernel* but other than that, I can use the
>>>>> naming method I've used for years and keep the good kernels I want.
>>>> make install names the kernels vmlinuz-$VERSION, and updates a
>>>> symlink to vmlinuz if one exists.
>>> But sometimes I have more than one of the same version. I add a -1, -2,
>>> -3 etc to them as I go. From my understanding, make install doesn't do
>>> that. I do and with good reason. It seems make install won't accomplish
>>> what I do by hand.
>> There's a kernel option that does that automatically, LOCALVERSION I
>> think, but it's years since I've needed to touch it.
>
> But if it changes how it works and I don't know it, that could cause
> problems. It could even remove something I don't want removed. If I
> do it myself, I know what is done and don't remove anything I don't
> want to lose. Even the make commands change between kernels at times.
> I suspect make install has changes too. I don't want to have to track
> kernel changes just to know if I can use make install and still get
> what I have now.
LOCALVERSION is a kernel setting in .config. All the kernel make
commands handle it consistently, and have not changed in a long time.
They just append that string to files they create (kernel and symbolmap)
and the /lib/modules directory. (For the init thingy, I assume you
would give dracut that name - the exact name of the kernel to boot in
/boot, and it makes an init thingy with a matching name. It was
genkernel where I was talking about things changing - and yes, genkernel
changed it's basic kernel naming conventions a few months back, but that
was clearly a big change, and there was a news item about it, if I
remember. If you are only or mostly recompiling the kernel after
changing config settings, it's trivial to change LOCALVERSION from -01
to -02 at the same time.
>>>>> My biggest problem, getting the dracut command options right. If I
>>>>> didn't need dracut, I'd be in heaven.
>>>> If you have a plain setup, dracut shouldn't need any options.
>>> I don't have a plain setup tho nor do I really want that setup.
>> You don't boot from an encrypted drive (yet) or use unusual hardware,
>> that's what I meant by a plain system. Dracut handles booting from a a
>> btrfs root on a LUKS encrypted block device here with no fancy
>> configuration. It really is impressive the way it figures so much out for
>> itself.
>
> What I meant was, I don't use make install because I do things
> differently than make install does. It's not "plain" in that regard.
>
>
>>> I like
>>> having backup kernels and my own numbering system. It has worked for me
>>> for decades, ever since I started using Gentoo and building my own
>>> kernels. I don't see any point in changing what works unless I can
>>> streamline what I'm already getting with the results I expect. If I
>>> could get rid of the init thingy, I would have zero issues with my
>>> method. It's dracut that causes the issues. We all know how much I
>>> dislike init thingys tho. ;-) That said, dracut hasn't failed me in a
>>> while. If it can't build correctly, it does spit out it failed. It's
>>> been a while since the init thingy it creates has failed as well. So,
>>> at least there is that.
>> I use a shell script to compile and install the kernel, build the
>> initramfs and then update the bootloader. It means the process is always
>> executed consistently and I only have to remember one command :)
>>
>>
>
>
> Well, we know my scripting ability isn't anywhere near that advanced.
> Claiming I can do a basic script is a stretch. ROFL When I do it
> manually, I get consistent results to and I know what steps were taken
> to get there. My only problem is recalling the options for dracut.
>
> Dale
>
> :-) :-)
[-- Attachment #2: Type: text/html, Size: 5854 bytes --]
next prev parent reply other threads:[~2020-06-06 23:56 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-05 19:01 [gentoo-user] new genkernel problem Jack
2020-06-05 20:24 ` Andrew Udvare
2020-06-05 20:48 ` Jack
2020-06-05 22:46 ` Neil Bothwick
2020-06-05 20:57 ` Dale
2020-06-05 22:51 ` Neil Bothwick
2020-06-06 15:13 ` Dale
2020-06-06 16:13 ` Neil Bothwick
2020-06-06 16:34 ` Dale
2020-06-06 17:20 ` Jack
2020-06-06 22:11 ` Neil Bothwick
2020-06-06 22:34 ` Dale
2020-06-06 23:55 ` Jack [this message]
2020-06-11 14:18 ` antlists
2020-06-11 17:04 ` madscientistatlarge
2020-06-11 22:08 ` madscientistatlarge
2020-06-12 17:19 ` antlists
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=c9f0b43e-38fe-18f1-5924-2008585ee15e@users.sourceforge.net \
--to=ostroffjh@users.sourceforge.net \
--cc=gentoo-user@lists.gentoo.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox