On May 4, 2012, at 8:08 AM, Scott Robertson wrote:

>> Maybe I'm missing something, but I think that entire paragraph could be
>> removed.  The book gives the command to mount the LFS partition as:
> 
>> mount -pv $LFS -v -t ext3 /dev/<xxx> $LFS
> 
>> And then warns, effectively, that if you did something like:
> 
>> mount -pv $LFS -v -t ext3 -o nosuid,noudev,noatime /dev/<xxx> $LFS
> 
>> that you should remount it.  Well, yes, obviously.  But we didn't ever
>> tell anyone to mount it with those options.  There are a myriad of other
>> ways that readers can fail to follow the instructions, but we
>> don't/shouldn't/can't protect/advise against all of them, can we?
> 
>> Ta.
> 
>> Matt.
> 
> 
> Hmmm, I followed the directions and typed it just as it is in the book:
> 
> mkdir -pv $LFS
> mount -v -t ext3 /dev/<xxx> $LFS
> 
> But when I ran the mount command without any parameters to see what options 
> were set,
> nosuid, nodev  were set by default.  So in my case at least, I did have to 
> explicitly remount.

This is, IMO, not a matter of reading the instructions correctly or not.

Some kernels in some hosts can be compiled to use different filesystem defaults 
(e.g., the epic noatime/relatime battle ages ago).

When the books talks about mounting as if you were using parameters like 
noatime, I would imagine it's referring to a situation where the host kernel is 
doing so on your behalf, without you necessarily having typed it in.  
Obviously, if you simply mount /dev/whatever onto /mnt/lfs with a bare mount 
command, it's just going to choose the kernel/FS defaults.  So, I'm guessing 
these "clarifying paragraphs" are probably not to say that people intentionally 
have (or unwittingly may have) screwed up.  Those parameters are there simply 
as proxies for what the kernel/FS defaults might be.

It's probably important to point this out, since we know little about host 
kernels.  I imagine that's the same rationale as is behind asking people not to 
mount /mnt/lfs with noatime b/c it causes some of the tests to blow up.

        Q


-- 
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page

Reply via email to