On 09/11/2012 09:33 PM, Vivek Goyal wrote:
> On Tue, Sep 11, 2012 at 09:31:40AM +0800, Dave Young wrote:
>> On 09/10/2012 10:40 PM, Vivek Goyal wrote:
>>
>>> On Mon, Sep 10, 2012 at 04:10:46PM +0800, Dave Young wrote:
>>>> kdump module also need to convert dev name to udev symlinks.
>>>> So better to move function get_persistent_dev() to dracut-functions.sh
>>>>
>>>> Also in this patch improvement and fix the original function:
>>>> a) use udevadm info --query=name to get the kernel name.
>>>> This will fix the issue caused by passing symbolic link of a device.
>>>> b) fix a bug to compare $_tmp instead of $i with $_dev. Really sorry,
>>>> should have tested more carefully.
>>>>
>>>> Signed-off-by: Dave Young <dyoung@xxxxxxxxxx>
>>>> ---
>>>> dracut-functions.sh | 14 ++++++++++++++
>>>> modules.d/99base/module-setup.sh | 13 -------------
>>>> 2 files changed, 14 insertions(+), 13 deletions(-)
>>>>
>>>> --- dracut.orig/dracut-functions.sh
>>>> +++ dracut/dracut-functions.sh
>>>> @@ -239,6 +239,21 @@ else
>>>> }
>>>> fi
>>>>
>>>> +get_persistent_dev() {
>>>> + local i _tmp _dev
>>>> +
>>>> + _dev=$(udevadm info --query=name --name="$1" 2>/dev/null)
>>>> + [ -z "$_dev" ] && return
>>>> +
>>>> + for i in /dev/disk/by-id/*; do
>>>> + _tmp=$(udevadm info --query=name --name="$i" 2>/dev/null)
>>>> + if [ "$_tmp" = "$_dev" ]; then
>>>> + echo $i
>>>> + return
>>>> + fi
>>>> + done
>>>> +}
>>>> +
>>>
>>> Can't we just compare the maj:min number of device and come up with udev
>>> name? I think that would be simpler and we will not have to call into
>>> udevadm.
>>
>>
>> Hi,
>>
>> /dev/disk/by-id/* are symlinks, $1 could be soft link as well, using
>> udevadm will be simpler because we do not need follow symlinks and
>> handle the string prefix.
>
> dracut function get_maj_min() uses "stat" with -L option which will
> automatically following symlinks if need be. Calling into udevadm
> for every possible device (/dev/disk/by-id/*) seems overkill to me. Anyway,
> harald has already committed this patch. So we will simplify it some
> other time.
Hi, thanks for the explaination.
Actually udevadm info is not so heavy, see below strace compare:
$ strace udevadm info --query=name --name=/dev/sda 2>a.txt
$ strace stat -L -c '$((0x%t)):$((0x%T))' /dev/sda 2>b.txt
$ wc a.txt
88 494 5997 a.txt
$ wc b.txt
61 347 4041 b.txt
Since get_maj_min works well with symlinks, converting to use majmin is
also fine to me.
>
> One advantage of converting to kernel names is that looking into
> /proc/mounts will give us the mount point. I was converting device
> name into maj:min and looking into /proc/self/mountinfo to determine
> the mount point.
>
> So converting to kernel name does seem to have this little advantage
> of being able to find device name and associated mountpoint in
> /proc/mounts.
>
> Thanks
> Vivek
--
Thanks
Dave
--
To unsubscribe from this list: send the line "unsubscribe initramfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
