| Message ID | 20210316103627.2954121-1-christian.brauner@ubuntu.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | fstests: add idmapped mounts tests | expand |
Looks good,
Reviewed-by: Christoph Hellwig <hch@lst.de>
On Tue, Mar 16, 2021 at 11:36:23AM +0100, Christian Brauner wrote: > Hey everyone, > > /* v9 */ > Rebased onto current master. > Hopefully we can merge this soon so we can have regular testing asap and > the rebasing is getting a bit tedious too. ;) Sorry for the late review.. You don't have to rebase on every release if updating the test seq number is the only thing changed. Seq number conflicts all the time, usually I can fix it on commit. Patch 2 is missing from the list, maybe it's too big to post, could you please provide a git repo so I can pull from there? Some minor comments go to specific patch. Thanks, Eryu > > ubuntu@f1-vm:~/src/git/xfstests$ sudo ./check generic/631 > FSTYP -- xfs (non-debug) > PLATFORM -- Linux/x86_64 f1-vm 5.12.0-rc3-inode-helpers #343 SMP Mon Mar 15 12:57:02 UTC 2021 > MKFS_OPTIONS -- -f -bsize=4096 /dev/loop1 > MOUNT_OPTIONS -- /dev/loop1 /mnt/scratch > > generic/631 9s ... 11s > Ran: generic/631 > Passed all 1 tests > > ubuntu@f1-vm:~/src/git/xfstests$ sudo ./check generic/632 > FSTYP -- xfs (non-debug) > PLATFORM -- Linux/x86_64 f1-vm 5.12.0-rc3-inode-helpers #343 SMP Mon Mar 15 12:57:02 UTC 2021 > MKFS_OPTIONS -- -f -bsize=4096 /dev/loop1 > MOUNT_OPTIONS -- /dev/loop1 /mnt/scratch > > generic/632 15s > Ran: generic/632 > Passed all 1 tests > > ubuntu@f1-vm:~/src/git/xfstests$ sudo ./check xfs/529 > FSTYP -- xfs (non-debug) > PLATFORM -- Linux/x86_64 f1-vm 5.12.0-rc3-inode-helpers #343 SMP Mon Mar 15 12:57:02 UTC 2021 > MKFS_OPTIONS -- -f -bsize=4096 /dev/loop1 > MOUNT_OPTIONS -- /dev/loop1 /mnt/scratch > > xfs/529 25s ... 43s > Ran: xfs/529 > Passed all 1 tests > > ubuntu@f1-vm:~/src/git/xfstests$ sudo ./check xfs/530 > FSTYP -- xfs (non-debug) > PLATFORM -- Linux/x86_64 f1-vm 5.12.0-rc3-inode-helpers #343 SMP Mon Mar 15 12:57:02 UTC 2021 > MKFS_OPTIONS -- -f -bsize=4096 /dev/loop1 > MOUNT_OPTIONS -- /dev/loop1 /mnt/scratch > > xfs/530 24s > Ran: xfs/530 > Passed all 1 tests > > /* v8 */ > This adds the testsuite for idmapped mounts. It has been expanded quite > a bit too since the last version: > - Add a testcase to verify that detached mounts in the new mount api > work correctly. > - Add a full testsuite for idmapped mounts. > - Add two quota tests verifying xfs behavior. > > Sorry that it took so long to resend this testsuite but I had to chase > and fix a bug in the new mount API (completely independent of idmapped > mounts) before I could add quota tests because they would otherwise run > danger to render the host mount namespace unuseable due to that bug > (cf. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d3110f256d126b44d34c1f662310cd295877c447) > All tests here pass for me on a kernel with idmapped mounts. The xfs > specific quota tests should go on top of the fixes that Darrick is > currently carrying for 5.12 and which he sent out a little while ago. > > Thanks! > Christian > > Christian Brauner (4): > generic/631: add test for detached mount propagation > generic/632: add fstests for idmapped mounts > xfs/529: quotas and idmapped mounts > xfs/530: quotas on idmapped mounts > > .gitignore | 3 + > common/rc | 25 + > configure.ac | 2 + > include/builddefs.in | 1 + > m4/Makefile | 1 + > m4/package_libcap.m4 | 4 + > src/Makefile | 8 +- > src/detached_mounts_propagation.c | 189 + > src/feature.c | 40 +- > src/idmapped-mounts/Makefile | 41 + > src/idmapped-mounts/idmapped-mounts.c | 8627 +++++++++++++++++++++++++ > src/idmapped-mounts/missing.h | 151 + > src/idmapped-mounts/mount-idmapped.c | 428 ++ > src/idmapped-mounts/utils.c | 134 + > src/idmapped-mounts/utils.h | 30 + > tests/generic/631 | 41 + > tests/generic/631.out | 2 + > tests/generic/632 | 42 + > tests/generic/632.out | 2 + > tests/generic/group | 2 + > tests/xfs/529 | 373 ++ > tests/xfs/529.out | 657 ++ > tests/xfs/530 | 274 + > tests/xfs/530.out | 129 + > tests/xfs/group | 2 + > 25 files changed, 11203 insertions(+), 5 deletions(-) > create mode 100644 m4/package_libcap.m4 > create mode 100644 src/detached_mounts_propagation.c > create mode 100644 src/idmapped-mounts/Makefile > create mode 100644 src/idmapped-mounts/idmapped-mounts.c > create mode 100644 src/idmapped-mounts/missing.h > create mode 100644 src/idmapped-mounts/mount-idmapped.c > create mode 100644 src/idmapped-mounts/utils.c > create mode 100644 src/idmapped-mounts/utils.h > create mode 100644 tests/generic/631 > create mode 100644 tests/generic/631.out > create mode 100644 tests/generic/632 > create mode 100644 tests/generic/632.out > create mode 100644 tests/xfs/529 > create mode 100644 tests/xfs/529.out > create mode 100644 tests/xfs/530 > create mode 100644 tests/xfs/530.out > > > base-commit: f6ddaf130d5b0817278afe441fdde52f464f321b > -- > 2.27.0
On Sun, Mar 21, 2021 at 10:26:35PM +0800, Eryu Guan wrote: > On Tue, Mar 16, 2021 at 11:36:23AM +0100, Christian Brauner wrote: > > Hey everyone, > > > > /* v9 */ > > Rebased onto current master. > > Hopefully we can merge this soon so we can have regular testing asap and > > the rebasing is getting a bit tedious too. ;) > > Sorry for the late review.. You don't have to rebase on every release if > updating the test seq number is the only thing changed. Seq number > conflicts all the time, usually I can fix it on commit. > > Patch 2 is missing from the list, maybe it's too big to post, could you _Sigh_, this seems like another vger issue. The individual Ccs/Tos should've all gotten the patch but it's missing from the list. This is a pattern I've seen quite a few times recently with vger. There's nothing to do apart from complain about it to the vger maintainer. :) > please provide a git repo so I can pull from there? For sure. It's available in three locations: https://git.kernel.org/pub/scm/linux/kernel/git/brauner/xfstests-dev.git/log/?h=idmapped_mounts_v2 https://gitlab.com/brauner/xfstests/-/tree/idmapped_mounts_v2 https://github.com/brauner/xfstests/tree/idmapped_mounts_v2 (Are you fixing the things you pointed out in the individual patches yourself or should I resend a new version?) Thanks! Christian
On Sun, Mar 21, 2021 at 04:32:00PM +0100, Christian Brauner wrote: > On Sun, Mar 21, 2021 at 10:26:35PM +0800, Eryu Guan wrote: > > On Tue, Mar 16, 2021 at 11:36:23AM +0100, Christian Brauner wrote: > > > Hey everyone, > > > > > > /* v9 */ > > > Rebased onto current master. > > > Hopefully we can merge this soon so we can have regular testing asap and > > > the rebasing is getting a bit tedious too. ;) > > > > Sorry for the late review.. You don't have to rebase on every release if > > updating the test seq number is the only thing changed. Seq number > > conflicts all the time, usually I can fix it on commit. > > > > Patch 2 is missing from the list, maybe it's too big to post, could you > > _Sigh_, this seems like another vger issue. The individual Ccs/Tos > should've all gotten the patch but it's missing from the list. This is a > pattern I've seen quite a few times recently with vger. There's nothing > to do apart from complain about it to the vger maintainer. :) > > > please provide a git repo so I can pull from there? > > For sure. It's available in three locations: > https://git.kernel.org/pub/scm/linux/kernel/git/brauner/xfstests-dev.git/log/?h=idmapped_mounts_v2 > https://gitlab.com/brauner/xfstests/-/tree/idmapped_mounts_v2 > https://github.com/brauner/xfstests/tree/idmapped_mounts_v2 Ok, thanks! > > (Are you fixing the things you pointed out in the individual patches > yourself or should I resend a new version?) New version please, I can fix trival things like typo, renumber the test, code style, etc. Thanks, Eryu
On Mon, Mar 22, 2021 at 10:37:04AM +0800, Eryu Guan wrote: > On Sun, Mar 21, 2021 at 04:32:00PM +0100, Christian Brauner wrote: > > On Sun, Mar 21, 2021 at 10:26:35PM +0800, Eryu Guan wrote: > > > On Tue, Mar 16, 2021 at 11:36:23AM +0100, Christian Brauner wrote: > > > > Hey everyone, > > > > > > > > /* v9 */ > > > > Rebased onto current master. > > > > Hopefully we can merge this soon so we can have regular testing asap and > > > > the rebasing is getting a bit tedious too. ;) > > > > > > Sorry for the late review.. You don't have to rebase on every release if > > > updating the test seq number is the only thing changed. Seq number > > > conflicts all the time, usually I can fix it on commit. > > > > > > Patch 2 is missing from the list, maybe it's too big to post, could you > > > > _Sigh_, this seems like another vger issue. The individual Ccs/Tos > > should've all gotten the patch but it's missing from the list. This is a > > pattern I've seen quite a few times recently with vger. There's nothing > > to do apart from complain about it to the vger maintainer. :) > > > > > please provide a git repo so I can pull from there? > > > > For sure. It's available in three locations: > > https://git.kernel.org/pub/scm/linux/kernel/git/brauner/xfstests-dev.git/log/?h=idmapped_mounts_v2 > > https://gitlab.com/brauner/xfstests/-/tree/idmapped_mounts_v2 > > https://github.com/brauner/xfstests/tree/idmapped_mounts_v2 > > Ok, thanks! > > > > > (Are you fixing the things you pointed out in the individual patches > > yourself or should I resend a new version?) > > New version please, I can fix trival things like typo, renumber the I'll give you a new version hopefully today (CET). > test, code style, etc. Thanks! Christian