[arch-projects] [mkinitcpio PATCH 1/3] /run/initramfs: copy the whole ramfs

[Tom Gundersen]

On Sat, Nov 26, 2011 at 4:03 AM, Gerardo Exequiel Pozzi
<vmlinuz386 at yahoo.com.ar> wrote:
> On 11/25/2011 11:46 PM, Dave Reisner wrote:
>>
>> On Fri, Nov 25, 2011 at 11:34:17PM -0300, Gerardo Exequiel Pozzi wrote:
>>>
>>> On 11/17/2011 01:08 PM, Tom Gundersen wrote:
>>>>
>>>> This is useful for the future shutdown hook, and might make debugging
>>>> easier,
>>>> as well as potentially opening up for other usecases (that I cannot yet
>>>> imagine ;-) ).
>>>>
>>>> Signed-off-by: Tom Gundersen<teg at jklm.no>
>>>> ---
>>>>  init         |    2 +-
>>>>  install/base |    1 +
>>>>  2 files changed, 2 insertions(+), 1 deletions(-)
>>>>
>>>> diff --git a/init b/init
>>>> index 7cc2d47..3c6fb81 100644
>>>> --- a/init
>>>> +++ b/init
>>>> @@ -21,7 +21,7 @@ else
>>>>    mknod /dev/mem c 1 1
>>>>  fi
>>>>  mount -t tmpfs run /run -o nosuid,nodev,mode=755,size=10M
>>>> -mkdir /run/initramfs
>>>> +cp -ax / /run/initramfs
>>>>
>>>>  # parse the kernel command line
>>>>  parse_cmdline
>>>> diff --git a/install/base b/install/base
>>>> index 0726fa2..84e7f02 100644
>>>> --- a/install/base
>>>> +++ b/install/base
>>>> @@ -10,6 +10,7 @@ build() {
>>>>      add_binary /sbin/blkid
>>>>      add_binary /bin/mount
>>>>      add_binary /sbin/switch_root
>>>> +    add_binary /bin/cp
>>>>
>>>>      add_symlink "/etc/mtab" "/proc/self/mounts"
>>>>
>>> I guess, we should start thinking incrementing size of /run, 10M is
>>> becoming small, for example I am using 56% due archiso_shutdown hook (it
>>> takes 5.3M).
>>>
>>>
>>> --
>>> Gerardo Exequiel Pozzi
>>> \cos^2\alpha + \sin^2\alpha = 1
>>>
>> Given that /run is owned by root, I don't see why we need to worry about
>> hard capping the size.
>>
>> d
>
> Sure I never understand why limiting this, also for /dev. Initial patch
> [#1], size limit added [#2]
>
> [#1]
> http://projects.archlinux.org/initscripts.git/commit/?id=e5248347fd0c00e9ea72ad3b3b3434573f08d96d
> [#2]
> http://projects.archlinux.org/initscripts.git/commit/?id=757f653b8c9ba9d671dd95bd998bc3d487ee20d1

I agree, we should probably just stop doing this. Worst case scenario
something fills up dev/run, but it sure is worse running out of space
on dev or run, than it is to potentially run out of memory... (?)

-t