On Wed, Sep 01, 2021 at 07:46:01AM +0300, Amir Goldstein wrote:

On Wed, Sep 1, 2021 at 3:37 AM Darrick J. Wong [off-list ref] wrote:

quoted

From: Darrick J. Wong <djwong@kernel.org>

Create a file to document the purpose of each test group that is
currently defined in fstests, and add a build script to check that every
group mentioned in the tests is also mentioned in the documentation.

This is awesome and long due.
Thanks for doing that!

Minor nits about overlayfs groups below...

Heh, yeah, thanks for making corrections. :)

quoted

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
---
 doc/group-names.txt    |  136 ++++++++++++++++++++++++++++++++++++++++++++++++
 include/buildgrouplist |    1
 tools/check-groups     |   33 ++++++++++++
 3 files changed, 170 insertions(+)
 create mode 100644 doc/group-names.txt
 create mode 100755 tools/check-groups

diff --git a/doc/group-names.txt b/doc/group-names.txt
new file mode 100644
index 00000000..ae517328
--- /dev/null
+++ b/doc/group-names.txt

@@ -0,0 +1,136 @@
+======================= =======================================================
+Group Name:            Description:
+======================= =======================================================
+all                    All known tests, automatically generated by ./check at
+                       runtime
+auto                   Tests that should be run automatically.  These should
+                       not require more than ~5 minutes to run.
+quick                  Tests that should run in under 30 seconds.
+deprecated             Old tests that should not be run.
+
+acl                    Access Control Lists
+admin                  xfs_admin functionality
+aio                    general libaio async io tests
+atime                  file access time
+attr                   extended attributes
+attr2                  xfs v2 extended aributes
+balance                        btrfs tree rebalance
+bigtime                        timestamps beyond the year 2038
+blockdev               block device functionality
+broken                 broken tests
+cap                    Linux capabilities
+casefold               directory name casefolding
+ci                     ASCII case-insensitive directory name lookups
+clone                  FICLONE/FICLONERANGE ioctls
+clone_stress           stress testing FICLONE/FICLONERANGE
+collapse               fallocate collapse_range
+compress               file compression
+convert                        btrfs ext[34] conversion tool
+copy                   xfs_copy functionality
+copy_range             copy_file_range syscall
+copyup                 overlayfs copyup support

The tests in this group exercise copy up.
There is no such thing as overlayfs without "copyup support",
so guess my point is - remove the word "support"

OK.

quoted

+dangerous              dangerous test that can crash the system
+dangerous_bothrepair   fuzzers to evaluate xfs_scrub + xfs_repair repair
+dangerous_fuzzers      fuzzers that can crash your computer
+dangerous_norepair     fuzzers to evaluate kernel metadata verifiers
+dangerous_online_repair        fuzzers to evaluate xfs_scrub online repair
+dangerous_repair       fuzzers to evaluate xfs_repair offline repair
+dangerous_scrub                fuzzers to evaluate xfs_scrub checking
+data                   data loss checkers
+dax                    direct access mode for persistent memory files
+db                     xfs_db functional tests
+dedupe                 FIEDEDUPERANGE ioctl
+defrag                 filesystem defragmenters
+dir                    directory test functions
+dump                   dump and restore utilities
+eio                    IO error reporting
+encrypt                        encrypted file contents
+enospc                 ENOSPC error reporting
+exportfs               file handles
+filestreams            XFS filestreams allocator
+freeze                 filesystem freeze tests
+fsck                   general fsck tests
+fsmap                  FS_IOC_GETFSMAP ioctl
+fsr                    XFS free space reorganizer
+fuzzers                        filesystem fuzz tests
+growfs                 increasing the size of a filesystem
+hardlink               hardlinks
+health                 XFS health reporting
+idmapped               idmapped mount functionality
+inobtcount             XFS inode btree count tests
+insert                 fallocate insert_range
+ioctl                  general ioctl tests
+io_uring               general io_uring async io tests
+label                  filesystem labelling
+limit                  resource limits
+locks                  file locking
+log                    metadata logging
+logprint               xfs_logprint functional tests
+long_rw                        long-soak read write IO path exercisers
+metacopy               overlayfs metadata-only copy-up
+metadata               filesystem metadata update exercisers
+metadump               xfs_metadump/xfs_mdrestore functionality
+mkfs                   filesystem formatting tools
+mount                  mount option and functionality checks
+nested                 nested overlayfs instances
+nfs4_acl               NFSv4 access control lists
+nonsamefs              overlayfs layers on different filesystems
+online_repair          online repair functionality tests
+other                  dumping ground, do not add more tests to this group
+overlay                        using overlayfs on top of FSTYP

This description is a bit confusing, because the recommended
way to run overlayfs tests as described in README.overlay is
to set FSTYP=xfs and run ./check -overlay

I'm struggling for a better description but perhaps:
"using overlayfs regardless of ./check -overlay flag"?

Hmm.  Since I'm the author of the only test that uses this tag, I guess
I'm the authority (ha!) on what the name actually means.

That test (generic/631) is a regression test for a XFS whiteout handling
bug that can only be reproduced by layering overlayfs atop xfs.
Overlayfs is incidental to reproducing the XFS bug, but AFAIK overlayfs
is the only in-kernel user of whiteout, which is why it's critical here.

It's not right to make it "_supported_fs overlay" because we're not
testing overlayfs functionality; we're merely using it as a stick to
poke another filesystem.

How about: "regression tests that require the use of overlayfs in a
targetted configuration" ?

quoted

+pattern                        specific IO pattern tests
+perms                  access control and permission checking
+pipe                   pipe functionality
+pnfs                   PNFS
+posix                  POSIX behavior conformance
+prealloc               fallocate
+preallocrw             fallocate, then read and write
+punch                  fallocate punch_hole
+qgroup                 btrfs qgroup feature
+quota                  filesystem usage quotas
+raid                   btrfs RAID
+realtime               XFS realtime volumes
+recoveryloop           crash recovery loops
+redirect               overlayfs redirect_dir feature
+remote                 dump and restore with a remote tape
+remount                        remounting filesystems
+rename                 rename system call
+repair                 xfs_repair functional tests
+replace                        btrfs device replace
+replay                 dm-logwrites replays
+resize                 resize2fs functionality tests
+richacl                        rich ACL feature
+rmap                   XFS reverse mapping exercisers
+rotate                 overlayfs feature of some sort"

I guess that works :-D
but to be accurate, this is actually a unionmount testsuite feature -
at selected test points in the workload, a new upper layer is stacked
on to overlayfs, so maybe:
"upper layer rotate tests from the unionmount test suite"?

I've changed it to 'overlayfs upper layer rotate tests from the
unionmount test suite'.

quoted

+rw                     read/write IO tests
+samefs                 overlayfs when all layers are on the same fs
+scrub                  filesystem metadata scrubbers
+seed                   btrfs seeded filesystems
+seek                   llseek functionality
+send                   btrfs send/receive
+shrinkfs               decreasing the size of a filesystem
+shutdown               FS_IOC_SHUTDOWN ioctl
+snapshot               btrfs snapshots
+soak                   long soak tests of any kind
+spaceman               xfs_spaceman functional tests
+splice                 splice system call
+stress                 fsstress filesystem exerciser
+subvol                 btrfs subvolumes
+subvolume              btrfs subvolumes (again?)

A cleanup patch to fix this typo in btrfs/233?

Will do.

--D

Thanks,
Amir.