* [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2
@ 2023-07-31 11:16 Anand Jain
2023-07-31 11:16 ` [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid Anand Jain
` (8 more replies)
0 siblings, 9 replies; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
v2:
The new patch 1/7 peels btrfs_sb_fsid_prt() helper function from 1/4 in v1,
This function is used in patch 2/7 and patch 4/7 in v2.
Patch 6/7, adds btrfs_sb_metadata_uuid_or_null() another helper func, used
in 7/7 in volumes.c. Further, used in the upcoming fsid_changing patch.
Pushing this set now which are ready. This has passed fstests -g quick.
----- original -----
These patches are cleanups related to metadata_uuid. Please ref to the
patch for details.
Anand Jain (7):
btrfs: add a helper to read the superblock metadata_uuid
btrfs: simplify memcpy either of metadata_uuid or fsid
btrfs: fix fsid in btrfs_validate_super
btrfs: fix metadata_uuid in btrfs_validate_super
btrfs: drop redundant check to use fs_devices::metadata_uuid
btrfs: add a superblock helper to read metadata_uuid or NULL
btrfs: use btrfs_sb_metadata_uuid_or_null
fs/btrfs/disk-io.c | 28 ++++++++++------------------
fs/btrfs/volumes.c | 29 +++++++++++++++++++----------
fs/btrfs/volumes.h | 1 +
3 files changed, 30 insertions(+), 28 deletions(-)
--
2.39.2
^ permalink raw reply [flat|nested] 24+ messages in thread
* [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
@ 2023-07-31 11:16 ` Anand Jain
2023-08-01 13:16 ` Johannes Thumshirn
2023-07-31 11:16 ` [PATCH 2/7] btrfs: simplify memcpy either of metadata_uuid or fsid Anand Jain
` (7 subsequent siblings)
8 siblings, 1 reply; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
In some cases, we need to read the FSID from the superblock when the
metadata_uuid is not set, and otherwise, read the metadata_uuid. So,
add a helper.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/volumes.c | 8 ++++++++
fs/btrfs/volumes.h | 1 +
2 files changed, 9 insertions(+)
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 70d69d4b44d2..2020b1fac764 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -681,6 +681,14 @@ static int btrfs_open_one_device(struct btrfs_fs_devices *fs_devices,
return -EINVAL;
}
+u8 *btrfs_sb_fsid_ptr(struct btrfs_super_block *sb)
+{
+ bool has_metadata_uuid = (btrfs_super_incompat_flags(sb) &
+ BTRFS_FEATURE_INCOMPAT_METADATA_UUID);
+
+ return has_metadata_uuid ? sb->metadata_uuid : sb->fsid;
+}
+
/*
* Handle scanned device having its CHANGING_FSID_V2 flag set and the fs_devices
* being created with a disk that has already completed its fsid change. Such
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index b8c51f16ba86..0f87057bb575 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -749,5 +749,6 @@ int btrfs_verify_dev_extents(struct btrfs_fs_info *fs_info);
bool btrfs_repair_one_zone(struct btrfs_fs_info *fs_info, u64 logical);
bool btrfs_pinned_by_swapfile(struct btrfs_fs_info *fs_info, void *ptr);
+u8 *btrfs_sb_fsid_ptr(struct btrfs_super_block *sb);
#endif
--
2.39.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [PATCH 2/7] btrfs: simplify memcpy either of metadata_uuid or fsid
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
2023-07-31 11:16 ` [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid Anand Jain
@ 2023-07-31 11:16 ` Anand Jain
2023-08-01 13:19 ` Johannes Thumshirn
2023-07-31 11:16 ` [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super Anand Jain
` (6 subsequent siblings)
8 siblings, 1 reply; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
There is a helper which provides either metadata_uuid or fsid as per
METADATA_UUID flag. So use it.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/volumes.c | 11 ++---------
1 file changed, 2 insertions(+), 9 deletions(-)
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 2020b1fac764..2f470404ff83 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -841,15 +841,8 @@ static noinline struct btrfs_device *device_list_add(const char *path,
found_transid > fs_devices->latest_generation) {
memcpy(fs_devices->fsid, disk_super->fsid,
BTRFS_FSID_SIZE);
-
- if (has_metadata_uuid)
- memcpy(fs_devices->metadata_uuid,
- disk_super->metadata_uuid,
- BTRFS_FSID_SIZE);
- else
- memcpy(fs_devices->metadata_uuid,
- disk_super->fsid, BTRFS_FSID_SIZE);
-
+ memcpy(fs_devices->metadata_uuid,
+ btrfs_sb_fsid_ptr(disk_super), BTRFS_FSID_SIZE);
fs_devices->fsid_change = false;
}
}
--
2.39.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
2023-07-31 11:16 ` [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid Anand Jain
2023-07-31 11:16 ` [PATCH 2/7] btrfs: simplify memcpy either of metadata_uuid or fsid Anand Jain
@ 2023-07-31 11:16 ` Anand Jain
2023-08-01 13:21 ` Johannes Thumshirn
2023-08-11 15:27 ` David Sterba
2023-07-31 11:16 ` [PATCH 4/7] btrfs: fix metadata_uuid " Anand Jain
` (5 subsequent siblings)
8 siblings, 2 replies; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
The function btrfs_validate_super() should verify the fsid in the provided
superblock argument. Because, all its callers expects it to do that.
Such as in the following stack:
write_all_supers()
sb = fs_info->super_for_commit;
btrfs_validate_write_super(.., sb)
btrfs_validate_super(.., sb, ..)
scrub_one_super()
btrfs_validate_super(.., sb, ..)
And
check_dev_super()
btrfs_validate_super(.., sb, ..)
However, it currently verifies the fs_info::super_copy::fsid instead,
which does not help.
Fix this using the correct fsid in the superblock argument.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/disk-io.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index b4495d4c1533..f2279eb93370 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2373,11 +2373,10 @@ int btrfs_validate_super(struct btrfs_fs_info *fs_info,
ret = -EINVAL;
}
- if (memcmp(fs_info->fs_devices->fsid, fs_info->super_copy->fsid,
- BTRFS_FSID_SIZE)) {
+ if (memcmp(fs_info->fs_devices->fsid, sb->fsid, BTRFS_FSID_SIZE)) {
btrfs_err(fs_info,
"superblock fsid doesn't match fsid of fs_devices: %pU != %pU",
- fs_info->super_copy->fsid, fs_info->fs_devices->fsid);
+ sb->fsid, fs_info->fs_devices->fsid);
ret = -EINVAL;
}
--
2.39.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [PATCH 4/7] btrfs: fix metadata_uuid in btrfs_validate_super
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
` (2 preceding siblings ...)
2023-07-31 11:16 ` [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super Anand Jain
@ 2023-07-31 11:16 ` Anand Jain
2023-08-01 13:23 ` Johannes Thumshirn
2023-07-31 11:16 ` [PATCH 5/7] btrfs: drop redundant check to use fs_devices::metadata_uuid Anand Jain
` (4 subsequent siblings)
8 siblings, 1 reply; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
The function btrfs_validate_super() should verify the metadata_uuid in
the provided superblock argument. Because, all its callers expects it to
do that.
Such as in the following stacks:
write_all_supers()
sb = fs_info->super_for_commit;
btrfs_validate_write_super(.., sb)
btrfs_validate_super(.., sb, ..)
scrub_one_super()
btrfs_validate_super(.., sb, ..)
And
check_dev_super()
btrfs_validate_super(.., sb, ..)
However, it currently verifies the fs_info::super_copy::metadata_uuid
instead.
Fix this using the correct metadata_uuid in the superblock argument.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/disk-io.c | 8 +++-----
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index f2279eb93370..e2fb11e89279 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2380,13 +2380,11 @@ int btrfs_validate_super(struct btrfs_fs_info *fs_info,
ret = -EINVAL;
}
- if (btrfs_fs_incompat(fs_info, METADATA_UUID) &&
- memcmp(fs_info->fs_devices->metadata_uuid,
- fs_info->super_copy->metadata_uuid, BTRFS_FSID_SIZE)) {
+ if (memcmp(fs_info->fs_devices->metadata_uuid, btrfs_sb_fsid_ptr(sb),
+ BTRFS_FSID_SIZE)) {
btrfs_err(fs_info,
"superblock metadata_uuid doesn't match metadata uuid of fs_devices: %pU != %pU",
- fs_info->super_copy->metadata_uuid,
- fs_info->fs_devices->metadata_uuid);
+ btrfs_sb_fsid_ptr(sb), fs_info->fs_devices->metadata_uuid);
ret = -EINVAL;
}
--
2.39.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [PATCH 5/7] btrfs: drop redundant check to use fs_devices::metadata_uuid
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
` (3 preceding siblings ...)
2023-07-31 11:16 ` [PATCH 4/7] btrfs: fix metadata_uuid " Anand Jain
@ 2023-07-31 11:16 ` Anand Jain
2023-08-01 13:25 ` Johannes Thumshirn
2023-07-31 11:16 ` [PATCH 6/7] btrfs: add a superblock helper to read metadata_uuid or NULL Anand Jain
` (3 subsequent siblings)
8 siblings, 1 reply; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
fs_devices::metadata_uuid value is already updated based on the
super_block::METADATA_UUID flag for either fsid or metadata_uuid as
appropriate. So, fs_devices::metadata_uuid can be used directly.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/disk-io.c | 15 +++++----------
1 file changed, 5 insertions(+), 10 deletions(-)
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index e2fb11e89279..902dfc4aec53 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -313,21 +313,16 @@ static bool check_tree_block_fsid(struct extent_buffer *eb)
struct btrfs_fs_info *fs_info = eb->fs_info;
struct btrfs_fs_devices *fs_devices = fs_info->fs_devices, *seed_devs;
u8 fsid[BTRFS_FSID_SIZE];
- u8 *metadata_uuid;
read_extent_buffer(eb, fsid, offsetof(struct btrfs_header, fsid),
BTRFS_FSID_SIZE);
+
/*
- * Checking the incompat flag is only valid for the current fs. For
- * seed devices it's forbidden to have their uuid changed so reading
- * ->fsid in this case is fine
+ * alloc_fs_devices() copies the fsid into metadata_uuid if the
+ * metadata_uuid is unset in the superblock, including for a seed device.
+ * So, we can use fs_devices->metadata_uuid.
*/
- if (btrfs_fs_incompat(fs_info, METADATA_UUID))
- metadata_uuid = fs_devices->metadata_uuid;
- else
- metadata_uuid = fs_devices->fsid;
-
- if (!memcmp(fsid, metadata_uuid, BTRFS_FSID_SIZE))
+ if (!memcmp(fsid, fs_info->fs_devices->metadata_uuid, BTRFS_FSID_SIZE))
return false;
list_for_each_entry(seed_devs, &fs_devices->seed_list, seed_list)
--
2.39.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [PATCH 6/7] btrfs: add a superblock helper to read metadata_uuid or NULL
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
` (4 preceding siblings ...)
2023-07-31 11:16 ` [PATCH 5/7] btrfs: drop redundant check to use fs_devices::metadata_uuid Anand Jain
@ 2023-07-31 11:16 ` Anand Jain
2023-08-11 15:39 ` David Sterba
2023-07-31 11:16 ` [PATCH 7/7] btrfs: use btrfs_sb_metadata_uuid_or_null Anand Jain
` (2 subsequent siblings)
8 siblings, 1 reply; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
This is preparatory patch next two patches.
In some parts of the code, when the METADATA_UUID flag is set, we need to
use metadata_uuid otherwise NULL. Add a helper function that do it.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/volumes.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 2f470404ff83..661dc69543eb 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -681,6 +681,14 @@ static int btrfs_open_one_device(struct btrfs_fs_devices *fs_devices,
return -EINVAL;
}
+static u8 *btrfs_sb_metadata_uuid_or_null(struct btrfs_super_block *sb)
+{
+ bool has_metadata_uuid = (btrfs_super_incompat_flags(sb) &
+ BTRFS_FEATURE_INCOMPAT_METADATA_UUID);
+
+ return has_metadata_uuid ? sb->metadata_uuid : NULL;
+}
+
u8 *btrfs_sb_fsid_ptr(struct btrfs_super_block *sb)
{
bool has_metadata_uuid = (btrfs_super_incompat_flags(sb) &
--
2.39.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [PATCH 7/7] btrfs: use btrfs_sb_metadata_uuid_or_null
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
` (5 preceding siblings ...)
2023-07-31 11:16 ` [PATCH 6/7] btrfs: add a superblock helper to read metadata_uuid or NULL Anand Jain
@ 2023-07-31 11:16 ` Anand Jain
2023-08-11 15:42 ` David Sterba
2023-08-03 13:32 ` [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Guilherme G. Piccoli
2023-08-11 15:45 ` David Sterba
8 siblings, 1 reply; 24+ messages in thread
From: Anand Jain @ 2023-07-31 11:16 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/volumes.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 661dc69543eb..316839d2aecc 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -821,7 +821,7 @@ static noinline struct btrfs_device *device_list_add(const char *path,
if (!fs_devices) {
fs_devices = alloc_fs_devices(disk_super->fsid,
- has_metadata_uuid ? disk_super->metadata_uuid : NULL);
+ btrfs_sb_metadata_uuid_or_null(disk_super));
if (IS_ERR(fs_devices))
return ERR_CAST(fs_devices);
--
2.39.2
^ permalink raw reply related [flat|nested] 24+ messages in thread
* Re: [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid
2023-07-31 11:16 ` [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid Anand Jain
@ 2023-08-01 13:16 ` Johannes Thumshirn
2023-08-11 15:18 ` David Sterba
0 siblings, 1 reply; 24+ messages in thread
From: Johannes Thumshirn @ 2023-08-01 13:16 UTC (permalink / raw)
To: Anand Jain, linux-btrfs@vger.kernel.org
On 31.07.23 13:18, Anand Jain wrote:
> +u8 *btrfs_sb_fsid_ptr(struct btrfs_super_block *sb)
> +{
> + bool has_metadata_uuid = (btrfs_super_incompat_flags(sb) &
> + BTRFS_FEATURE_INCOMPAT_METADATA_UUID);
> +
> + return has_metadata_uuid ? sb->metadata_uuid : sb->fsid;
> +}
Not a huge fan of the overly long variable name and first line, but the
concept looks good:
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 2/7] btrfs: simplify memcpy either of metadata_uuid or fsid
2023-07-31 11:16 ` [PATCH 2/7] btrfs: simplify memcpy either of metadata_uuid or fsid Anand Jain
@ 2023-08-01 13:19 ` Johannes Thumshirn
0 siblings, 0 replies; 24+ messages in thread
From: Johannes Thumshirn @ 2023-08-01 13:19 UTC (permalink / raw)
To: Anand Jain, linux-btrfs@vger.kernel.org
On 31.07.23 13:18, Anand Jain wrote:
> There is a helper which provides either metadata_uuid or fsid as per
> METADATA_UUID flag. So use it.
Now that we have 'btrfs_sb_fsid_ptr()' use it to simplify metadata uuid
copying in device_list_add()?
Code wise,
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super
2023-07-31 11:16 ` [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super Anand Jain
@ 2023-08-01 13:21 ` Johannes Thumshirn
2023-08-11 15:27 ` David Sterba
1 sibling, 0 replies; 24+ messages in thread
From: Johannes Thumshirn @ 2023-08-01 13:21 UTC (permalink / raw)
To: Anand Jain, linux-btrfs@vger.kernel.org
Looks good,
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 4/7] btrfs: fix metadata_uuid in btrfs_validate_super
2023-07-31 11:16 ` [PATCH 4/7] btrfs: fix metadata_uuid " Anand Jain
@ 2023-08-01 13:23 ` Johannes Thumshirn
0 siblings, 0 replies; 24+ messages in thread
From: Johannes Thumshirn @ 2023-08-01 13:23 UTC (permalink / raw)
To: Anand Jain, linux-btrfs@vger.kernel.org
Looks good,
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 5/7] btrfs: drop redundant check to use fs_devices::metadata_uuid
2023-07-31 11:16 ` [PATCH 5/7] btrfs: drop redundant check to use fs_devices::metadata_uuid Anand Jain
@ 2023-08-01 13:25 ` Johannes Thumshirn
0 siblings, 0 replies; 24+ messages in thread
From: Johannes Thumshirn @ 2023-08-01 13:25 UTC (permalink / raw)
To: Anand Jain, linux-btrfs@vger.kernel.org
Looks good,
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
` (6 preceding siblings ...)
2023-07-31 11:16 ` [PATCH 7/7] btrfs: use btrfs_sb_metadata_uuid_or_null Anand Jain
@ 2023-08-03 13:32 ` Guilherme G. Piccoli
2023-08-11 15:45 ` David Sterba
8 siblings, 0 replies; 24+ messages in thread
From: Guilherme G. Piccoli @ 2023-08-03 13:32 UTC (permalink / raw)
To: anand.jain; +Cc: linux-btrfs
Thanks for the patches Anand! Feel free to add (to the whole series):
Tested-by: Guilherme G. Piccoli <gpiccoli@igalia.com>
I'm working a feature for btrfs that relies on metadata_uuid
infrastructure, so I did rebase on top of your patches and tested, it
works fine.
Cheers,
Guilherme
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid
2023-08-01 13:16 ` Johannes Thumshirn
@ 2023-08-11 15:18 ` David Sterba
0 siblings, 0 replies; 24+ messages in thread
From: David Sterba @ 2023-08-11 15:18 UTC (permalink / raw)
To: Johannes Thumshirn; +Cc: Anand Jain, linux-btrfs@vger.kernel.org
On Tue, Aug 01, 2023 at 01:16:42PM +0000, Johannes Thumshirn wrote:
> On 31.07.23 13:18, Anand Jain wrote:
> > +u8 *btrfs_sb_fsid_ptr(struct btrfs_super_block *sb)
> > +{
> > + bool has_metadata_uuid = (btrfs_super_incompat_flags(sb) &
> > + BTRFS_FEATURE_INCOMPAT_METADATA_UUID);
> > +
> > + return has_metadata_uuid ? sb->metadata_uuid : sb->fsid;
> > +}
>
> Not a huge fan of the overly long variable name and first line, but the
> concept looks good:
For the incompat bit checks (btrfs_fs_incompat & co) we have the macros
that glue the prefix of the bit name but that's for the fs_info data.
For raw buffers of super block we use the unpreprocessed helpers/macros,
there are like 3 in total so that's probably not worth a helper.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super
2023-07-31 11:16 ` [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super Anand Jain
2023-08-01 13:21 ` Johannes Thumshirn
@ 2023-08-11 15:27 ` David Sterba
1 sibling, 0 replies; 24+ messages in thread
From: David Sterba @ 2023-08-11 15:27 UTC (permalink / raw)
To: Anand Jain; +Cc: linux-btrfs
On Mon, Jul 31, 2023 at 07:16:34PM +0800, Anand Jain wrote:
> The function btrfs_validate_super() should verify the fsid in the provided
> superblock argument. Because, all its callers expects it to do that.
>
> Such as in the following stack:
>
> write_all_supers()
> sb = fs_info->super_for_commit;
> btrfs_validate_write_super(.., sb)
> btrfs_validate_super(.., sb, ..)
>
> scrub_one_super()
> btrfs_validate_super(.., sb, ..)
>
> And
> check_dev_super()
> btrfs_validate_super(.., sb, ..)
>
> However, it currently verifies the fs_info::super_copy::fsid instead,
> which does not help.
>
> Fix this using the correct fsid in the superblock argument.
>
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> ---
> fs/btrfs/disk-io.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
> index b4495d4c1533..f2279eb93370 100644
> --- a/fs/btrfs/disk-io.c
> +++ b/fs/btrfs/disk-io.c
> @@ -2373,11 +2373,10 @@ int btrfs_validate_super(struct btrfs_fs_info *fs_info,
> ret = -EINVAL;
> }
>
> - if (memcmp(fs_info->fs_devices->fsid, fs_info->super_copy->fsid,
> - BTRFS_FSID_SIZE)) {
> + if (memcmp(fs_info->fs_devices->fsid, sb->fsid, BTRFS_FSID_SIZE)) {
For memcpy and strcpy please compare it against 0 so it follows the
semantics of "is equal" or "is not equal".
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 6/7] btrfs: add a superblock helper to read metadata_uuid or NULL
2023-07-31 11:16 ` [PATCH 6/7] btrfs: add a superblock helper to read metadata_uuid or NULL Anand Jain
@ 2023-08-11 15:39 ` David Sterba
2023-08-14 15:35 ` Anand Jain
0 siblings, 1 reply; 24+ messages in thread
From: David Sterba @ 2023-08-11 15:39 UTC (permalink / raw)
To: Anand Jain; +Cc: linux-btrfs
On Mon, Jul 31, 2023 at 07:16:37PM +0800, Anand Jain wrote:
> This is preparatory patch next two patches.
There's only one more patch 7/7 and given how short the helper is it
could be done in one patch. The split for perparatory work and actual
use applies more to bigger helpers.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 7/7] btrfs: use btrfs_sb_metadata_uuid_or_null
2023-07-31 11:16 ` [PATCH 7/7] btrfs: use btrfs_sb_metadata_uuid_or_null Anand Jain
@ 2023-08-11 15:42 ` David Sterba
2023-08-14 15:32 ` Anand Jain
0 siblings, 1 reply; 24+ messages in thread
From: David Sterba @ 2023-08-11 15:42 UTC (permalink / raw)
To: Anand Jain; +Cc: linux-btrfs
On Mon, Jul 31, 2023 at 07:16:38PM +0800, Anand Jain wrote:
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> ---
> fs/btrfs/volumes.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
> index 661dc69543eb..316839d2aecc 100644
> --- a/fs/btrfs/volumes.c
> +++ b/fs/btrfs/volumes.c
> @@ -821,7 +821,7 @@ static noinline struct btrfs_device *device_list_add(const char *path,
>
> if (!fs_devices) {
> fs_devices = alloc_fs_devices(disk_super->fsid,
> - has_metadata_uuid ? disk_super->metadata_uuid : NULL);
> + btrfs_sb_metadata_uuid_or_null(disk_super));
This is the only place that passes a non-NULL pointer to
alloc_fs_devices, so it might be better to drop the 2nd argument
completely and just set the metadata_uuid here, which is basically
copying fsid to metadata_uuid.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
` (7 preceding siblings ...)
2023-08-03 13:32 ` [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Guilherme G. Piccoli
@ 2023-08-11 15:45 ` David Sterba
2023-08-15 6:52 ` [PATCH] btrfs: simplify alloc_fs_devices() remove arg2 Anand Jain
8 siblings, 1 reply; 24+ messages in thread
From: David Sterba @ 2023-08-11 15:45 UTC (permalink / raw)
To: Anand Jain; +Cc: linux-btrfs
On Mon, Jul 31, 2023 at 07:16:31PM +0800, Anand Jain wrote:
> v2:
> The new patch 1/7 peels btrfs_sb_fsid_prt() helper function from 1/4 in v1,
> This function is used in patch 2/7 and patch 4/7 in v2.
>
> Patch 6/7, adds btrfs_sb_metadata_uuid_or_null() another helper func, used
> in 7/7 in volumes.c. Further, used in the upcoming fsid_changing patch.
> Pushing this set now which are ready. This has passed fstests -g quick.
>
> ----- original -----
> These patches are cleanups related to metadata_uuid. Please ref to the
> patch for details.
>
> Anand Jain (7):
> btrfs: add a helper to read the superblock metadata_uuid
> btrfs: simplify memcpy either of metadata_uuid or fsid
> btrfs: fix fsid in btrfs_validate_super
> btrfs: fix metadata_uuid in btrfs_validate_super
> btrfs: drop redundant check to use fs_devices::metadata_uuid
The above added to misc-next, thanks.
> btrfs: add a superblock helper to read metadata_uuid or NULL
> btrfs: use btrfs_sb_metadata_uuid_or_null
I think this can simplify the alloc_fs_devices a bit more as commented,
please resend. Thanks.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 7/7] btrfs: use btrfs_sb_metadata_uuid_or_null
2023-08-11 15:42 ` David Sterba
@ 2023-08-14 15:32 ` Anand Jain
0 siblings, 0 replies; 24+ messages in thread
From: Anand Jain @ 2023-08-14 15:32 UTC (permalink / raw)
To: dsterba; +Cc: linux-btrfs
On 11/08/2023 23:42, David Sterba wrote:
> On Mon, Jul 31, 2023 at 07:16:38PM +0800, Anand Jain wrote:
>> Signed-off-by: Anand Jain <anand.jain@oracle.com>
>> ---
>> fs/btrfs/volumes.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
>> index 661dc69543eb..316839d2aecc 100644
>> --- a/fs/btrfs/volumes.c
>> +++ b/fs/btrfs/volumes.c
>> @@ -821,7 +821,7 @@ static noinline struct btrfs_device *device_list_add(const char *path,
>>
>> if (!fs_devices) {
>> fs_devices = alloc_fs_devices(disk_super->fsid,
>> - has_metadata_uuid ? disk_super->metadata_uuid : NULL);
>> + btrfs_sb_metadata_uuid_or_null(disk_super));
>
> This is the only place that passes a non-NULL pointer to
> alloc_fs_devices, so it might be better to drop the 2nd argument
> completely and just set the metadata_uuid here, which is basically
> copying fsid to metadata_uuid.
Oh right. I am sending v2. Thanks.
^ permalink raw reply [flat|nested] 24+ messages in thread
* Re: [PATCH 6/7] btrfs: add a superblock helper to read metadata_uuid or NULL
2023-08-11 15:39 ` David Sterba
@ 2023-08-14 15:35 ` Anand Jain
0 siblings, 0 replies; 24+ messages in thread
From: Anand Jain @ 2023-08-14 15:35 UTC (permalink / raw)
To: dsterba; +Cc: linux-btrfs
On 11/08/2023 23:39, David Sterba wrote:
> On Mon, Jul 31, 2023 at 07:16:37PM +0800, Anand Jain wrote:
>> This is preparatory patch next two patches.
>
> There's only one more patch 7/7 and given how short the helper is it
> could be done in one patch. The split for perparatory work and actual
> use applies more to bigger helpers.
There was one more patch which used this helper initially.
I'll merge this with 7/7.
Thanks.
^ permalink raw reply [flat|nested] 24+ messages in thread
* [PATCH] btrfs: simplify alloc_fs_devices() remove arg2
2023-08-11 15:45 ` David Sterba
@ 2023-08-15 6:52 ` Anand Jain
2023-08-23 14:52 ` [PATCH resend] " Anand Jain
2023-09-05 11:34 ` David Sterba
0 siblings, 2 replies; 24+ messages in thread
From: Anand Jain @ 2023-08-15 6:52 UTC (permalink / raw)
To: linux-btrfs
Among all the callers, only the device_list_add() function uses the
second argument of alloc_fs_devices(). It passes metadata_uuid when
available; otherwise, it passes NULL. And in turn, alloc_fs_devices()
is designed to copy either metadata_uuid or fsid into
fs_devices::metadata_uuid.
So eliminate the second argument in alloc_fs_devices(), and copy the
fsid alway. In the caller device_list_add() function, we will overwrite
it with metadata_uuid when it is available.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/disk-io.c | 7 ++++---
fs/btrfs/volumes.c | 25 ++++++++++++-------------
2 files changed, 16 insertions(+), 16 deletions(-)
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 0a96ea8c1d3a..9858c77b99e6 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -318,9 +318,10 @@ static bool check_tree_block_fsid(struct extent_buffer *eb)
BTRFS_FSID_SIZE);
/*
- * alloc_fs_devices() copies the fsid into metadata_uuid if the
- * metadata_uuid is unset in the superblock, including for a seed device.
- * So, we can use fs_devices->metadata_uuid.
+ * alloc_fsid_devices() copies the fsid into fs_devices::metadata_uuid.
+ * This is then overwritten by metadata_uuid if it is present in the
+ * device_list_add(). The same true for a seed device as well. So use of
+ * fs_devices::metadata_uuid is appropriate here.
*/
if (memcmp(fsid, fs_info->fs_devices->metadata_uuid, BTRFS_FSID_SIZE) == 0)
return false;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 189da583bb67..852590e06d76 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -358,20 +358,17 @@ struct list_head * __attribute_const__ btrfs_get_fs_uuids(void)
/*
* alloc_fs_devices - allocate struct btrfs_fs_devices
- * @fsid: if not NULL, copy the UUID to fs_devices::fsid
- * @metadata_fsid: if not NULL, copy the UUID to fs_devices::metadata_fsid
+ * @fsid: if not NULL, copy the UUID to fs_devices::fsid and to
+ * fs_devices::metadata_fsid
*
* Return a pointer to a new struct btrfs_fs_devices on success, or ERR_PTR().
* The returned struct is not linked onto any lists and can be destroyed with
* kfree() right away.
*/
-static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid,
- const u8 *metadata_fsid)
+static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid)
{
struct btrfs_fs_devices *fs_devs;
- ASSERT(fsid || !metadata_fsid);
-
fs_devs = kzalloc(sizeof(*fs_devs), GFP_KERNEL);
if (!fs_devs)
return ERR_PTR(-ENOMEM);
@@ -385,8 +382,7 @@ static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid,
if (fsid) {
memcpy(fs_devs->fsid, fsid, BTRFS_FSID_SIZE);
- memcpy(fs_devs->metadata_uuid,
- metadata_fsid ?: fsid, BTRFS_FSID_SIZE);
+ memcpy(fs_devs->metadata_uuid, fsid, BTRFS_FSID_SIZE);
}
return fs_devs;
@@ -812,8 +808,11 @@ static noinline struct btrfs_device *device_list_add(const char *path,
if (!fs_devices) {
- fs_devices = alloc_fs_devices(disk_super->fsid,
- has_metadata_uuid ? disk_super->metadata_uuid : NULL);
+ fs_devices = alloc_fs_devices(disk_super->fsid);
+ if (has_metadata_uuid)
+ memcpy(fs_devices->metadata_uuid,
+ disk_super->metadata_uuid, BTRFS_FSID_SIZE);
+
if (IS_ERR(fs_devices))
return ERR_CAST(fs_devices);
@@ -997,7 +996,7 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
lockdep_assert_held(&uuid_mutex);
- fs_devices = alloc_fs_devices(orig->fsid, NULL);
+ fs_devices = alloc_fs_devices(orig->fsid);
if (IS_ERR(fs_devices))
return fs_devices;
@@ -2454,7 +2453,7 @@ static struct btrfs_fs_devices *btrfs_init_sprout(struct btrfs_fs_info *fs_info)
* Private copy of the seed devices, anchored at
* fs_info->fs_devices->seed_list
*/
- seed_devices = alloc_fs_devices(NULL, NULL);
+ seed_devices = alloc_fs_devices(NULL);
if (IS_ERR(seed_devices))
return seed_devices;
@@ -6905,7 +6904,7 @@ static struct btrfs_fs_devices *open_seed_devices(struct btrfs_fs_info *fs_info,
if (!btrfs_test_opt(fs_info, DEGRADED))
return ERR_PTR(-ENOENT);
- fs_devices = alloc_fs_devices(fsid, NULL);
+ fs_devices = alloc_fs_devices(fsid);
if (IS_ERR(fs_devices))
return fs_devices;
--
2.38.1
^ permalink raw reply related [flat|nested] 24+ messages in thread
* [PATCH resend] btrfs: simplify alloc_fs_devices() remove arg2
2023-08-15 6:52 ` [PATCH] btrfs: simplify alloc_fs_devices() remove arg2 Anand Jain
@ 2023-08-23 14:52 ` Anand Jain
2023-09-05 11:34 ` David Sterba
1 sibling, 0 replies; 24+ messages in thread
From: Anand Jain @ 2023-08-23 14:52 UTC (permalink / raw)
To: linux-btrfs; +Cc: Anand Jain
Among all the callers, only the device_list_add() function uses the
second argument of alloc_fs_devices(). It passes metadata_uuid when
available; otherwise, it passes NULL. And in turn, alloc_fs_devices()
is designed to copy either metadata_uuid or fsid into
fs_devices::metadata_uuid.
So eliminate the second argument in alloc_fs_devices(), and copy the
fsid alway. In the caller device_list_add() function, we will overwrite
it with metadata_uuid when it is available.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
fs/btrfs/disk-io.c | 7 ++++---
fs/btrfs/volumes.c | 25 ++++++++++++-------------
2 files changed, 16 insertions(+), 16 deletions(-)
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 0a96ea8c1d3a..9858c77b99e6 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -318,9 +318,10 @@ static bool check_tree_block_fsid(struct extent_buffer *eb)
BTRFS_FSID_SIZE);
/*
- * alloc_fs_devices() copies the fsid into metadata_uuid if the
- * metadata_uuid is unset in the superblock, including for a seed device.
- * So, we can use fs_devices->metadata_uuid.
+ * alloc_fsid_devices() copies the fsid into fs_devices::metadata_uuid.
+ * This is then overwritten by metadata_uuid if it is present in the
+ * device_list_add(). The same true for a seed device as well. So use of
+ * fs_devices::metadata_uuid is appropriate here.
*/
if (memcmp(fsid, fs_info->fs_devices->metadata_uuid, BTRFS_FSID_SIZE) == 0)
return false;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 189da583bb67..852590e06d76 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -358,20 +358,17 @@ struct list_head * __attribute_const__ btrfs_get_fs_uuids(void)
/*
* alloc_fs_devices - allocate struct btrfs_fs_devices
- * @fsid: if not NULL, copy the UUID to fs_devices::fsid
- * @metadata_fsid: if not NULL, copy the UUID to fs_devices::metadata_fsid
+ * @fsid: if not NULL, copy the UUID to fs_devices::fsid and to
+ * fs_devices::metadata_fsid
*
* Return a pointer to a new struct btrfs_fs_devices on success, or ERR_PTR().
* The returned struct is not linked onto any lists and can be destroyed with
* kfree() right away.
*/
-static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid,
- const u8 *metadata_fsid)
+static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid)
{
struct btrfs_fs_devices *fs_devs;
- ASSERT(fsid || !metadata_fsid);
-
fs_devs = kzalloc(sizeof(*fs_devs), GFP_KERNEL);
if (!fs_devs)
return ERR_PTR(-ENOMEM);
@@ -385,8 +382,7 @@ static struct btrfs_fs_devices *alloc_fs_devices(const u8 *fsid,
if (fsid) {
memcpy(fs_devs->fsid, fsid, BTRFS_FSID_SIZE);
- memcpy(fs_devs->metadata_uuid,
- metadata_fsid ?: fsid, BTRFS_FSID_SIZE);
+ memcpy(fs_devs->metadata_uuid, fsid, BTRFS_FSID_SIZE);
}
return fs_devs;
@@ -812,8 +808,11 @@ static noinline struct btrfs_device *device_list_add(const char *path,
if (!fs_devices) {
- fs_devices = alloc_fs_devices(disk_super->fsid,
- has_metadata_uuid ? disk_super->metadata_uuid : NULL);
+ fs_devices = alloc_fs_devices(disk_super->fsid);
+ if (has_metadata_uuid)
+ memcpy(fs_devices->metadata_uuid,
+ disk_super->metadata_uuid, BTRFS_FSID_SIZE);
+
if (IS_ERR(fs_devices))
return ERR_CAST(fs_devices);
@@ -997,7 +996,7 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
lockdep_assert_held(&uuid_mutex);
- fs_devices = alloc_fs_devices(orig->fsid, NULL);
+ fs_devices = alloc_fs_devices(orig->fsid);
if (IS_ERR(fs_devices))
return fs_devices;
@@ -2454,7 +2453,7 @@ static struct btrfs_fs_devices *btrfs_init_sprout(struct btrfs_fs_info *fs_info)
* Private copy of the seed devices, anchored at
* fs_info->fs_devices->seed_list
*/
- seed_devices = alloc_fs_devices(NULL, NULL);
+ seed_devices = alloc_fs_devices(NULL);
if (IS_ERR(seed_devices))
return seed_devices;
@@ -6905,7 +6904,7 @@ static struct btrfs_fs_devices *open_seed_devices(struct btrfs_fs_info *fs_info,
if (!btrfs_test_opt(fs_info, DEGRADED))
return ERR_PTR(-ENOENT);
- fs_devices = alloc_fs_devices(fsid, NULL);
+ fs_devices = alloc_fs_devices(fsid);
if (IS_ERR(fs_devices))
return fs_devices;
--
2.38.1
^ permalink raw reply related [flat|nested] 24+ messages in thread
* Re: [PATCH resend] btrfs: simplify alloc_fs_devices() remove arg2
2023-08-15 6:52 ` [PATCH] btrfs: simplify alloc_fs_devices() remove arg2 Anand Jain
2023-08-23 14:52 ` [PATCH resend] " Anand Jain
@ 2023-09-05 11:34 ` David Sterba
1 sibling, 0 replies; 24+ messages in thread
From: David Sterba @ 2023-09-05 11:34 UTC (permalink / raw)
To: Anand Jain; +Cc: linux-btrfs
On Wed, Aug 23, 2023 at 10:52:13PM +0800, Anand Jain wrote:
> Among all the callers, only the device_list_add() function uses the
> second argument of alloc_fs_devices(). It passes metadata_uuid when
> available; otherwise, it passes NULL. And in turn, alloc_fs_devices()
> is designed to copy either metadata_uuid or fsid into
> fs_devices::metadata_uuid.
>
> So eliminate the second argument in alloc_fs_devices(), and copy the
> fsid alway. In the caller device_list_add() function, we will overwrite
> it with metadata_uuid when it is available.
>
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
Added to misc-next, thanks.
^ permalink raw reply [flat|nested] 24+ messages in thread
end of thread, other threads:[~2023-09-05 16:03 UTC | newest]
Thread overview: 24+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-07-31 11:16 [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Anand Jain
2023-07-31 11:16 ` [PATCH 1/7] btrfs: add a helper to read the superblock metadata_uuid Anand Jain
2023-08-01 13:16 ` Johannes Thumshirn
2023-08-11 15:18 ` David Sterba
2023-07-31 11:16 ` [PATCH 2/7] btrfs: simplify memcpy either of metadata_uuid or fsid Anand Jain
2023-08-01 13:19 ` Johannes Thumshirn
2023-07-31 11:16 ` [PATCH 3/7] btrfs: fix fsid in btrfs_validate_super Anand Jain
2023-08-01 13:21 ` Johannes Thumshirn
2023-08-11 15:27 ` David Sterba
2023-07-31 11:16 ` [PATCH 4/7] btrfs: fix metadata_uuid " Anand Jain
2023-08-01 13:23 ` Johannes Thumshirn
2023-07-31 11:16 ` [PATCH 5/7] btrfs: drop redundant check to use fs_devices::metadata_uuid Anand Jain
2023-08-01 13:25 ` Johannes Thumshirn
2023-07-31 11:16 ` [PATCH 6/7] btrfs: add a superblock helper to read metadata_uuid or NULL Anand Jain
2023-08-11 15:39 ` David Sterba
2023-08-14 15:35 ` Anand Jain
2023-07-31 11:16 ` [PATCH 7/7] btrfs: use btrfs_sb_metadata_uuid_or_null Anand Jain
2023-08-11 15:42 ` David Sterba
2023-08-14 15:32 ` Anand Jain
2023-08-03 13:32 ` [PATCH 0/7 v2] metadata_uuid misc cleanup and fixes part2 Guilherme G. Piccoli
2023-08-11 15:45 ` David Sterba
2023-08-15 6:52 ` [PATCH] btrfs: simplify alloc_fs_devices() remove arg2 Anand Jain
2023-08-23 14:52 ` [PATCH resend] " Anand Jain
2023-09-05 11:34 ` David Sterba
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).