Re: [PATCH 5 3/4] nfsd_open(): rename 'int access' to 'int may_flags' in nfsd_open()

[Bernd Schubert <bernd.schubert@fastmail.fm>]

On 03/13/2012 09:34 PM, J. Bruce Fields wrote:
> On Tue, Mar 13, 2012 at 09:03:51PM +0100, Bernd Schubert wrote:
>> On 03/13/2012 09:01 PM, J. Bruce Fields wrote:
>>> On Mon, Mar 12, 2012 at 06:22:50PM -0400, bfields wrote:
>>>> On Mon, Mar 12, 2012 at 11:49:21AM -0400, J. Bruce Fields wrote:
>>>>> On Mon, Mar 12, 2012 at 11:09:12AM -0400, Ted Ts'o wrote:
>>>>>> On Fri, Mar 09, 2012 at 03:51:48PM -0500, Ted Ts'o wrote:
>>>>>>>
>>>>>>> Linus in general doesn't like cross tree merges (or any extraneous
>>>>>>> merges) unless they are absolutely necessary; but then, he trusts git
>>>>>>> merges more than many of us do.  :-)
>>>>>>>
>>>>>>> I've put the the patch series on a separate patch, with a signed tag, at:
>>>>>>>
>>>>>>>       git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4.git nfs-ext4-premerge
>>>>>>>
>>>>>>> Can you confirm that you've pulled it into your tree and it's landed
>>>>>>> in linux-next?  I probably won't bother pulling it in mine unless
>>>>>>> there is definitely a merge conflict that I need to resolve.  (I've
>>>>>>> checked and there do not appear to be any against linux-next as of
>>>>>>> last night.)
>>>>>>
>>>>>> Ping?
>>>>>
>>>>> Whoops, sorry.
>>>>>
>>>>> Looks perfect, thanks for handling these!
>>>>>
>>>>> I'm running them through my usual regression tests
>>>>
>>>> Urgh, I'm seeing a failure on the telldir test (part of the "special"
>>>> connectathon tests).  I haven't looked at what it's trying to do yet.
>>>> Reproduceable even just on a local filesystem without NFS involved.  If
>>>> someone wants to look at it, you can just do:
>>>>
>>>> 	git clone git://linux-nfs.org/~bfields/cthon04.git
>>>> 	cd cthon04
>>>> 	make
>>>> 	NFS_TESTDIR=/somewhere_on_an_ext4_fs/TMP ./runtests -s
>>>>
>>>> (Or after running that, more specifically,
>>>>
>>>> 	cd /somewhere_on_an_ext4_fs/TMP/
>>>> 	./telldir
>>>>
>>>> )
>>>>
>>>> I'll look at it tomorrow.
>>>
>>> Looks like seekdir is just not accepting an offset returned from
>>> getdents:
>>>
>>> 	openat(AT_FDCWD, "telldir-test", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 3
>>> 	getdents(3, {...{d_ino=20386, d_off=5728083968307607285, d_reclen=24, d_name="192"}...} = 4848
>>> 	lseek(3, 5728083968307607285, SEEK_SET) = -1 EINVAL (Invalid argument)
>>>
>>> Still investigating.
>>
>> Thanks! I'm also just going to work on it (just noticed your mail
>> while on vacation ...).
>
> OK, good.  You can probably figure this out faster than me....
>

Hmm, there must have gone something wrong on merging, my own test also 
fails

http://www.pci.uni-heidelberg.de/tc/usr/bernd/downloads/test_seekdir/

(Sorry, it does not say 'failure', but one needs to compare the file 
names and telldir-offset numbers)

I think I will continue in the morning as its already 1 a.m. here.


Cheers,
Bernd