Re: linux-3.14 nfsd regression

[Mark Lord <mlord@pobox.com>]

Mark Lord wrote:
> On 14-04-03 12:33 PM, Mark Lord wrote:
>> This commit from linux-3.14 breaks our NFS-root clients here:
>>
>> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=6e14b46b91fee8a049b0940333ce13a820beaaa5
>>
>>
>> - *p++ = htonl((u32) stat->mode);
>> + *p++ = htonl((u32) (stat->mode & S_IALLUGO));
>>
>>
>> Reverting the one-liner above (on the server) fixes it for us,
>> as does reverting back to linux-3.13.8 on the server.
>>
>> The NFS-root clients are on PowerPC (big-endian) architecture,
>> running linux-3.12.16. The NFS server is on an Intel PC running linux-3.14.
>>
>> ACL is completely disabled on server and client,
>> and we're using NFSv2/v3.  No support for v4.
>>
>> I instrumented the function to see what other bits were being cleared
>> by the (stat->mode & S_IALLUGO) masking.  The results are attached.
> ..
> 
> 
> Looking into the (previously attached) trace, the bits that seem to be
> getting chopped most often are S_IFREG (0x8000) and S_IFDIR (0x4000).
> It appears that one/both of those are needed for mounting nfsroot.
> 
> 
> [ 2733.823753] encode_fattr: mode=0x000041ed mask=0x00000fff
> [ 2733.824217] encode_fattr: mode=0x000041ed mask=0x00000fff
> [ 2733.838769] encode_fattr: mode=0x000041ed mask=0x00000fff
> [ 2733.839175] encode_fattr: mode=0x0000a1ff mask=0x00000fff
> [ 2733.839895] encode_fattr: mode=0x000081ed mask=0x00000fff
...

Okay, the NFS root filesystem mounts and appears to work if I use this:

    *p++ = htonl((u32) (stat->mode & (S_IALLUGO | S_IFDIR | S_IFREG | S_IFCHR)));

But I suspect it may also need S_IFBLK for block device accesses as well.

Cheers