[PATCH v2 0/2] btrfs: allow mechanism to override quota

[PATCH v2 0/2] btrfs: allow mechanism to override quota

[Sargun Dhillon]

This patchset makes it so that on a per-filesystem basis one can disable
quota enforcement for users with cap_sys_resource. This patchset can
likely later be extended to per-qgroup, or a per-volume basis. I'm
thinking of extending the sysfs interface to list the qgroups and
this same interface for the qgroups themselves.

Changes since v1:
  -Rather than a separate member of btrfs_fs_info, use the existing
   flags field

Sargun Dhillon (2):
  btrfs: add quota override flag to enable quota override for
    sys_resource
  btrfs: Add quota_override knob into sysfs

 fs/btrfs/ctree.h  |  2 ++
 fs/btrfs/qgroup.c |  5 +++++
 fs/btrfs/sysfs.c  | 41 +++++++++++++++++++++++++++++++++++++++++
 3 files changed, 48 insertions(+)

-- 
2.9.3

[PATCH v2 1/2] btrfs: add quota override flag to enable quota override for sys_resource

[Sargun Dhillon]

This patch introduces the quota override flag to btrfs_fs_info, and
a change to quota limit checking code to temporarily allow for quota
to be overridden for processes with cap_sys_resource.

It's useful for administrative programs, such as log rotation,
that may need to temporarily use more disk space in order to free up
a greater amount of overall disk space without yielding more disk
space to the rest of userland.

Eventually, we may want to add the idea of an operator-specific
quota, operator reserved space, or something else to allow for
administrative override, but this is perhaps the simplest
solution.

Signed-off-by: Sargun Dhillon <sargun@sargun.me>
---
 fs/btrfs/ctree.h  | 2 ++
 fs/btrfs/qgroup.c | 5 +++++
 2 files changed, 7 insertions(+)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 643c70d..e86cb7c 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -716,6 +716,8 @@ struct btrfs_delayed_root;
 #define BTRFS_FS_BTREE_ERR			11
 #define BTRFS_FS_LOG1_ERR			12
 #define BTRFS_FS_LOG2_ERR			13
+#define BTRFS_FS_QUOTA_OVERRIDE			14
+
 /*
  * Indicate that a whole-filesystem exclusive operation is running
  * (device replace, resize, device add/delete, balance)
diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
index deffbeb..458fec0 100644
--- a/fs/btrfs/qgroup.c
+++ b/fs/btrfs/qgroup.c
@@ -2338,6 +2338,11 @@ static int qgroup_reserve(struct btrfs_root *root, u64 num_bytes, bool enforce)
 
 	if (num_bytes == 0)
 		return 0;
+
+	if (test_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags) &&
+	    capable(CAP_SYS_RESOURCE))
+		enforce = false;
+
 retry:
 	spin_lock(&fs_info->qgroup_lock);
 	quota_root = fs_info->quota_root;
-- 
2.9.3

[PATCH v2 2/2] btrfs: Add quota_override knob into sysfs

[Sargun Dhillon]

This patch adds the read-write attribute quota_override into sysfs.
Any process which has cap_sys_resource can set this flag to on, and
once it is set to true, processes with cap_sys_resource can exceed
the quota.

Signed-off-by: Sargun Dhillon <sargun@sargun.me>
---
 fs/btrfs/sysfs.c | 41 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 41 insertions(+)

diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c
index 1f157fb..c2d5f35 100644
--- a/fs/btrfs/sysfs.c
+++ b/fs/btrfs/sysfs.c
@@ -447,11 +447,52 @@ static ssize_t btrfs_clone_alignment_show(struct kobject *kobj,
 
 BTRFS_ATTR(clone_alignment, btrfs_clone_alignment_show);
 
+static ssize_t quota_override_show(struct kobject *kobj,
+				   struct kobj_attribute *a, char *buf)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	int quota_override;
+
+	quota_override = test_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
+	return snprintf(buf, PAGE_SIZE, "%d\n", quota_override);
+}
+
+static ssize_t quota_override_store(struct kobject *kobj,
+				    struct kobj_attribute *a,
+				    const char *buf, size_t len)
+{
+	struct btrfs_fs_info *fs_info = to_fs_info(kobj);
+	unsigned long knob;
+	int err;
+
+	if (!fs_info)
+		return -EPERM;
+
+	if (!capable(CAP_SYS_RESOURCE))
+		return -EPERM;
+
+	err = kstrtoul(buf, 10, &knob);
+	if (err)
+		return err;
+	if (knob > 1)
+		return -EINVAL;
+
+	if (knob)
+		set_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
+	else
+		clear_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
+
+	return len;
+}
+
+BTRFS_ATTR_RW(quota_override, quota_override_show, quota_override_store);
+
 static const struct attribute *btrfs_attrs[] = {
 	BTRFS_ATTR_PTR(label),
 	BTRFS_ATTR_PTR(nodesize),
 	BTRFS_ATTR_PTR(sectorsize),
 	BTRFS_ATTR_PTR(clone_alignment),
+	BTRFS_ATTR_PTR(quota_override),
 	NULL,
 };
 
-- 
2.9.3

Re: [PATCH v2 0/2] btrfs: allow mechanism to override quota

[David Sterba]

On Thu, May 11, 2017 at 09:17:01PM +0000, Sargun Dhillon wrote:
> This patchset makes it so that on a per-filesystem basis one can disable
> quota enforcement for users with cap_sys_resource. This patchset can
> likely later be extended to per-qgroup, or a per-volume basis. I'm
> thinking of extending the sysfs interface to list the qgroups and
> this same interface for the qgroups themselves.
> 
> Changes since v1:
>   -Rather than a separate member of btrfs_fs_info, use the existing
>    flags field

Looks good to me, thanks.

Re: [PATCH v2 0/2] btrfs: allow mechanism to override quota

[Sargun Dhillon]

On Fri, May 12, 2017 at 7:48 AM, David Sterba <dsterba@suse.cz> wrote:
> On Thu, May 11, 2017 at 09:17:01PM +0000, Sargun Dhillon wrote:
>> This patchset makes it so that on a per-filesystem basis one can disable
>> quota enforcement for users with cap_sys_resource. This patchset can
>> likely later be extended to per-qgroup, or a per-volume basis. I'm
>> thinking of extending the sysfs interface to list the qgroups and
>> this same interface for the qgroups themselves.
>>
>> Changes since v1:
>>   -Rather than a separate member of btrfs_fs_info, use the existing
>>    flags field
>
> Looks good to me, thanks.
I'm curious as to whether this approach is fine to get an Acked-by, or
if I need to figure out how to make it more leak-tolerant. I don't
think modifying the overridden extents inflight is a problem. I'm not
sure of a way a user would be able to create *new* chunks of data.
Alternatively, I'd be quite happy making this applicable to metadata
only, for file xattrs, creation, deletion, etc..

Opinions?

Re: [PATCH v2 0/2] btrfs: allow mechanism to override quota

[David Sterba]

On Fri, May 19, 2017 at 01:39:49AM -0700, Sargun Dhillon wrote:
> On Fri, May 12, 2017 at 7:48 AM, David Sterba <dsterba@suse.cz> wrote:
> > On Thu, May 11, 2017 at 09:17:01PM +0000, Sargun Dhillon wrote:
> >> This patchset makes it so that on a per-filesystem basis one can disable
> >> quota enforcement for users with cap_sys_resource. This patchset can
> >> likely later be extended to per-qgroup, or a per-volume basis. I'm
> >> thinking of extending the sysfs interface to list the qgroups and
> >> this same interface for the qgroups themselves.
> >>
> >> Changes since v1:
> >>   -Rather than a separate member of btrfs_fs_info, use the existing
> >>    flags field
> >
> > Looks good to me, thanks.
> I'm curious as to whether this approach is fine to get an Acked-by,

This was meant as an acked-by, the patch is now queued for 4.13, but as
it add some user-visible interface, this may need more time to see if we
haven't missed something.

> or
> if I need to figure out how to make it more leak-tolerant. I don't
> think modifying the overridden extents inflight is a problem. I'm not
> sure of a way a user would be able to create *new* chunks of data.
> Alternatively, I'd be quite happy making this applicable to metadata
> only, for file xattrs, creation, deletion, etc..

>From the interface POV, this can be set as a bitmask.  I haven't looked
if we'd be able to propagate all the changes everywhere in the code,
but sounds doable, bug I'm not sure if this level of fine-grained control
is desired.