On Thu, Mar 29, 2018 at 6:28 PM, Miklos Szeredi <miklos@xxxxxxxxxx> wrote:
> On Thu, Mar 29, 2018 at 4:18 PM, Amir Goldstein <amir73il@xxxxxxxxx> wrote:
>> With mount option "xino", mounter declares that there are enough
>> free high bits in underlying fs to hold the layer fsid.
>> If overlayfs does encounter underlying inodes using the high xino
>> bits reserved for layer fsid, a warning will be emitted and the original
>> inode number will be used.
>>
>> The mount option name "xino" goes after a similar meaning mount option
>> of aufs, but in overlayfs case, the mapping is stateless.
>>
>> An example for a use case of "xino" is when upper/lower is on an xfs
>> filesystem. xfs uses 64bit inode numbers, but it currently never uses the
>> upper 8bit for inode numbers exposed via stat(2) and that is not likely to
>> change in the future without user opting-in for a new xfs feature. The
>> actual number of unused upper bit is much larger and determined by the xfs
>> filesystem geometry (64 - agno_log - agblklog - inopblog). That means
>> that for all practical purpose, there are enough unused bits in xfs
>> inode numbers for more than OVL_MAX_STACK unique fsid's.
>>
>> Another example for a use case of "xino" is when upper/lower is on tmpfs.
>> tmpfs inode numbers are allocated sequentially since boot, so they will
>> practially never use the high inode number bits.
>>
>> Signed-off-by: Amir Goldstein <amir73il@xxxxxxxxx>
>> ---
>> fs/overlayfs/ovl_entry.h | 1 +
>> fs/overlayfs/super.c | 34 ++++++++++++++++++++++++++++++++--
>> 2 files changed, 33 insertions(+), 2 deletions(-)
>>
>> diff --git a/fs/overlayfs/ovl_entry.h b/fs/overlayfs/ovl_entry.h
>> index 6a077fb2a75f..e830470c77bd 100644
>> --- a/fs/overlayfs/ovl_entry.h
>> +++ b/fs/overlayfs/ovl_entry.h
>> @@ -18,6 +18,7 @@ struct ovl_config {
>> const char *redirect_mode;
>> bool index;
>> bool nfs_export;
>> + bool xino;
>> };
>>
>> struct ovl_sb {
>> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
>> index d7284444f404..26a5db244081 100644
>> --- a/fs/overlayfs/super.c
>> +++ b/fs/overlayfs/super.c
>> @@ -352,6 +352,8 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>> if (ofs->config.nfs_export != ovl_nfs_export_def)
>> seq_printf(m, ",nfs_export=%s", ofs->config.nfs_export ?
>> "on" : "off");
>> + if (ofs->config.xino)
>> + seq_puts(m, ",xino");
>> return 0;
>> }
>>
>> @@ -386,6 +388,7 @@ enum {
>> OPT_INDEX_OFF,
>> OPT_NFS_EXPORT_ON,
>> OPT_NFS_EXPORT_OFF,
>> + OPT_XINO,
>> OPT_ERR,
>> };
>>
>> @@ -399,6 +402,7 @@ static const match_table_t ovl_tokens = {
>> {OPT_INDEX_OFF, "index=off"},
>> {OPT_NFS_EXPORT_ON, "nfs_export=on"},
>> {OPT_NFS_EXPORT_OFF, "nfs_export=off"},
>> + {OPT_XINO, "xino"},
>> {OPT_ERR, NULL}
>> };
>>
>> @@ -513,6 +517,10 @@ static int ovl_parse_opt(char *opt, struct ovl_config *config)
>> config->nfs_export = false;
>> break;
>>
>> + case OPT_XINO:
>> + config->xino = true;
>> + break;
>> +
>> default:
>> pr_err("overlayfs: unrecognized mount option \"%s\" or missing value\n", p);
>> return -EINVAL;
>> @@ -1197,9 +1205,31 @@ static int ovl_get_lower_layers(struct ovl_fs *ofs, struct path *stack,
>> ofs->numlower++;
>> }
>>
>> - /* When all layers on same fs, overlay can use real inode numbers */
>> - if (!ofs->numlowerfs || (ofs->numlowerfs == 1 && !ofs->upper_mnt))
>> + /*
>> + * When all layers on same fs, overlay can use real inode numbers.
>> + * With mount option "xino", mounter declares that there are enough
>> + * free high bits in underlying fs to hold the unique fsid.
>> + * If overlayfs does encounter underlying inodes using the high xino
>> + * bits reserved for fsid, it emits a warning and uses the original
>> + * inode number.
>> + */
>> + if (!ofs->numlowerfs || (ofs->numlowerfs == 1 && !ofs->upper_mnt)) {
>> ofs->xino_bits = 0;
>> + ofs->config.xino = false;
>> + } else if (ofs->config.xino && !ofs->xino_bits) {
>> + /*
>> + * This is a roundup of number of bits needed for numlowerfs+1
>> + * (i.e. ilog2(numlowerfs+1 - 1) + 1). fsid 0 is reserved for
>> + * upper fs even with non upper overlay.
>> + */
>> + BUILD_BUG_ON(ilog2(OVL_MAX_STACK) > 31);
>> + ofs->xino_bits = ilog2(ofs->numlowerfs) + 1;
>
> Shouldn't this be
>
> ilog2(ofs->numlowerfs + (ofs->upper_mnt ? 1 : 0))
>
> ?
>
> Upper layer doesn't require a separate bit, just a separate fsid slot.
>
+1 is not for upper fs bit its for round up.
This is confusing hence the comment above.
ilog2(2^N+a) returns log2 or the "rounded down" value (i.e. N).
So for 2^N+a fsids we need N+1 bits.
The accurate expression is therefore:
ilog2(ofs->numlowerfs + (ofs->upper_mnt ? 1 : 0) - 1) + 1
However, for simplicity, if there is no upper_mnt, first fsid is still 1
so I ommitted the condition and left with
ilog2(ofs->numlowerfs + 1 - 1) + 1
I leave it to you as an exercise to see how hard it would be to get
rid of not reserving fsid 0 for upper fs (it makes reference into the
lower_fs array conditional on upper_mnt.
Maybe I just didn't try hard enough or wasn't creative enough.
Anyway, I did not think it was important not reserving 1 fsid for
non upper case.
Thanks,
Amir.
--
To unsubscribe from this list: send the line "unsubscribe linux-unionfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
