From: Zhigang Wang <zhigang.x.wang@oracle.com>
To: Ian Campbell <Ian.Campbell@eu.citrix.com>
Cc: xen-devel@lists.xensource.com, Yu Zhiguo <yuzg@cn.fujitsu.com>
Subject: Re: [PATCH] Don't attach needless options when launch pygrub
Date: Thu, 15 Jul 2010 11:44:58 +0800 [thread overview]
Message-ID: <4C3E843A.9070008@oracle.com> (raw)
In-Reply-To: <1279095028.24714.186.camel@zakaz.uk.xensource.com>
[-- Attachment #1: Type: text/plain, Size: 2542 bytes --]
On 07/14/2010 04:10 PM, Ian Campbell wrote:
> On Wed, 2010-07-14 at 08:29 +0100, Yu Zhiguo wrote:
>> Hi Ian,
>>
>> Ian Campbell wrote:
>>> On Wed, 2010-07-14 at 07:15 +0100, Yu Zhiguo wrote:
>>>> We should always run grub if bootloader is specified,
>>>> options 'kernel' and 'ramdisk' are needless.
>>>
>>> Not quite. If you specify both bootloader and kernel then this instructs
>>> pygrub to extract the specific named file from the guest file system,
>>> similarly for the ramdisk.
>>>
>>
>> Do you mean in this case, pygrub will use specified kernel
>> that lie in the filesystem of the DomU?
>
> I thought so, I looks like I was mistaken though.
>
>> I think this is good. But now pygrub's action is using the
>> specified kernel in Dom0, but not run grub.
>
> Hmm, pygrub is certainly run, regardless of having a kernel configured
> or not. What is in question is what --kernel and --ramdisk actually
> cause pygrub to do and whether that is useful.
>
> As far as I can see the --kernel and --ramdisk options end up in the
> incfg map which only used in a handful of places, most of which just
> extract incfg["args"]. The only places which do not do this are the
> calls to sniff_solaris and sniff_netware both of which appear to make
> use of incfg["kernel"] (but not incfg["ramdisk"]).
>
> So it looks like specifying the kernel option in addition to bootloader
> is infact useful if you are booting a Solaris or Netware domU but is
> harmless/ignored otherwise. I think we need to continue to support this
> use case and I don't see any particular reason to force those users to
> change their configuration file syntax for this issue (if it's even an
> issue, I still don't really see the problem).
>
> Perhaps it would be better to update pygrub so that --kernel actually
> does something consistent in the non-{Solaris,Netware} case, such as
> perhaps selecting the configuration entry with the match kernel path
> instead of defaulting to entry 0? (e.g. make "-q --kernel=/boot/FOO"
> select the entry with kernel /boot/FOO)
>
> It looks like --ramdisk (and the associated plumbing through xend) may
> in fact be useless at this time. I'd say it is harmless to plumb it
> through for consistency though -- perhaps in the future pygrub (or
> another bootloader) might want to use it.
>
Yes. Please don't remove --kernel/--ramdisk: pygrub is not the only pv guest
bootloader.
Here is a pv guest boot loader we are using which grab vmlinuz/initrd from
network. It uses --kernle/--ramdisk parameters.
Thanks,
Zhigang
[-- Attachment #2: xenpvboot-0.1.tar.gz --]
[-- Type: application/x-gzip, Size: 4063 bytes --]
[-- Attachment #3: Type: text/plain, Size: 138 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
next prev parent reply other threads:[~2010-07-15 3:44 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-07-14 6:15 [PATCH] Don't attach needless options when launch pygrub Yu Zhiguo
2010-07-14 6:47 ` Ian Campbell
2010-07-14 7:29 ` Yu Zhiguo
2010-07-14 8:10 ` Ian Campbell
2010-07-14 9:36 ` Yu Zhiguo
2010-07-14 9:46 ` Ian Campbell
2010-07-14 10:07 ` Yu Zhiguo
2010-07-14 10:21 ` Yu Zhiguo
2010-07-14 10:22 ` Ian Campbell
2010-07-14 11:01 ` Yu Zhiguo
2010-07-14 11:10 ` Ian Campbell
2010-07-14 11:21 ` Yu Zhiguo
2010-07-14 12:33 ` Ian Campbell
2010-07-15 2:37 ` [PATCH] xm: needless to check 'kernel/ramdisk' is existent or not Yu Zhiguo
2010-07-15 7:42 ` Ian Campbell
2010-07-15 7:56 ` Ian Campbell
2010-07-15 3:44 ` Zhigang Wang [this message]
2010-07-15 5:03 ` [PATCH] Don't attach needless options when launch pygrub Yu Zhiguo
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=4C3E843A.9070008@oracle.com \
--to=zhigang.x.wang@oracle.com \
--cc=Ian.Campbell@eu.citrix.com \
--cc=xen-devel@lists.xensource.com \
--cc=yuzg@cn.fujitsu.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.