| Message ID | cover.1711837050.git.wqu@suse.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | btrfs-progs: header cleanups | expand |
Please drop the series. As David mentioned in the github pull request, the LSP server is pretty aggressive as it can fully understand the different configurations. We need a more comprehensive tests on all possible configuration/build targets/linked library etc. Thanks, Qu 在 2024/3/31 08:54, Qu Wenruo 写道: > [REPO] > https://github.com/adam900710/btrfs-progs/tree/headers_cleanup > > This series is focusing on cleanup the unused headers. > > This is mostly done by clangd, although it has some false alerts related > to macro usages of a header. > > But still it's pretty awesome to cleanup a lot of unnecessary headers. > Only one special touch on pretty_size() macro, to change it to a static > inline function to workaround the clangd bug. > > The first patch would do the main heavy lifting, meanwhile the second > patch is doing the BTRFS_FLAT_INCLUDES related cleanups for library-test > code. > > Unfortunately I didn't touch anything inside crypto/*, the main reason > is I'm not confident enough to verify all the optimization for different > instructions. > > (And even less motivation after the infamous recent XZ backdoor attempt, > by a possibly state-sponsored sleeping agent, ruining the trust among > open source community.) > > Qu Wenruo (2): > btrfs-progs: headers cleanup > btrfs-progs: library-test: header and BTRFS_FLAT_INCLUDES cleanups > > btrfs-corrupt-block.c | 1 - > btrfs-sb-mod.c | 2 -- > btrfs.c | 1 - > cmds/device.c | 1 - > cmds/filesystem-du.c | 1 - > cmds/filesystem-usage.c | 1 - > cmds/inspect.c | 3 --- > cmds/quota.c | 1 - > cmds/receive-dump.c | 1 - > cmds/receive.c | 1 - > cmds/reflink.c | 2 -- > cmds/restore.c | 1 - > cmds/scrub.c | 3 --- > common/device-scan.c | 1 - > common/open-utils.c | 1 - > common/path-utils.c | 1 - > common/send-stream.c | 1 - > common/send-utils.c | 1 - > common/string-utils.c | 2 -- > common/units.h | 7 ++++++- > common/utils.c | 1 - > convert/main.c | 1 - > convert/source-ext2.c | 1 - > kernel-lib/mktables.c | 5 +---- > kernel-shared/dir-item.c | 1 - > kernel-shared/extent-tree.c | 1 - > kernel-shared/file-item.c | 1 - > kernel-shared/file.c | 2 -- > kernel-shared/inode-item.c | 2 -- > kernel-shared/messages.c | 1 - > kernel-shared/root-tree.c | 1 - > kernel-shared/tree-checker.c | 1 - > kernel-shared/uuid-tree.c | 3 --- > kernel-shared/zoned.c | 1 - > libbtrfs/crc32c.c | 5 ++--- > mkfs/common.c | 1 - > mkfs/rootdir.c | 1 - > tests/fsstress.c | 25 ++++++++++++------------- > tests/ioctl-test.c | 1 - > tests/library-test.c | 22 +++------------------- > 40 files changed, 24 insertions(+), 86 deletions(-) > > -- > 2.44.0 > >
On Wed, Apr 03, 2024 at 08:35:07AM +1030, Qu Wenruo wrote: > Please drop the series. > > As David mentioned in the github pull request, the LSP server is pretty > aggressive as it can fully understand the different configurations. > > We need a more comprehensive tests on all possible configuration/build > targets/linked library etc. A bit more coverage done, the btrfs.box target is built by default everywhere and some CI image builds also configure with --experimental. The static has dependencies that may not exist everywhere but the coverage seems fine for now.