[LU-6220] push ext4/ldiskfs patches upstream if possible

• Type: Improvement
• Resolution: Unresolved
• Priority: Minor
• Fix Version/s: Upstream
• Affects Version/s: Lustre 2.8.0
• Assignee: WC Triage
• Reporter: Andreas Dilger

Description

There are a number of ldiskfs patches that we could potentially push upstream, possibly with some cleanups. All patches should be run through checkpatch.pl first.

• ext4-print-inum-in-htree-warning.patch: could be pushed upstream after moving string to the next line so it fits inside 80 columns
• ext4-disable-mb-cache.patch: I was involved in some discussion upstream about removing the mbcache (http://lwn.net/Articles/564802/), but that patch was never accepted upstream. (upstream since v4.12-rc2-47-gcdb7ee4c63)
• ext4-inode-version.patch: It might be possible to submit a patch upstream that moves dir->i_version++ into ext4_mark_iloc_dirty() like if (IS_I_VERSION(inode) || S_ISDIR(dir)). Then, we can add a mount flag or patch that makes this check a no-op so we only need a very small patch. More importantly, it will reduce the size of ext4_inode_info because we don't need a separate i_fs_version and reduce the memory usage on the MDS.
• ext4-journal-path-opt.patch: I think this can be pushed upstream without any significant changes (upstream since v3.11-rc2-34-gad4eec6135)
• ext4-kill-dx-root.patch: could be submitted upstream as a cleanup with some style fixes
• ext4-data-in-dirent.patch: tracked via LU-17423
• ext4-htree-lock.patch: Only provides locking functionality. Needs ext4-pdirop.patch to land as well.
• ext4-pdirop.patch: Neil wasn't a big fan of this, but at minimum needs VFS support for parallel dirops before landing
• ext4-max-dir-size.patch: we should print out a deprecation warning in newer kernels that use the "max_dir_size" mount option or sysfs parameter to use "max_dir_size_kb" instead, and then we can eventually remove it? The mount option has existed since kernel 3.6, but it would be good to print a deprecation warning before removing the option completely or some filesystems may not mount.
• ext4-symlink-xattr-inode-fix.patch: fixes bug in fast symlink handling (upstream since v6.16-rc4-42-gb4cc4a4077)
• ext4-large-eas.patch: this is described in LU-908 (upstream since v4.12-rc2-16-ge50e5129f3)
• -ext4-add-periodic-superblock-update.patch: tracked via LU-16298 (upstream since v6.5-rc3-53-gff0722de89)
• ext4-mballoc-for-hybrid.patch LU-16750 ldiskfs: optimize metadata allocation for hybrid LUNs

Issue Links

• is duplicated by: LU-6141 Merge ldiskfs patches into upstream ext4 linux sources
• is related to: LU-19857 ldiskfs patch ext4-dont-check-before-replay.patch broken in 6.0+ kernels
• is related to: LU-17423 Upstream dir_data feature
• is related to: LU-9724 update ext4-large-eas.patch to match upstream ext4 feature
• is related to: LU-19856 remove ldiskfs patch series for rhel7.9 kernel
• is related to: LU-12511 Prepare lustre for adoption into the linux kernel
• is related to: LU-19847 ldiskfs: integrate ext4-data-in-dirent.patch with fscrypt+case_insensitive feature
• is related to: LU-16750 optimize ldiskfs internal metadata allocation for hybrid storage LUNs
• is related to: LU-13054 MDS kernel BUG at ldiskfs/htree_lock.c:429!
• is related to: LU-908 multi-block xattr support
• is related to: LU-6030 cleanup for ldiskfs patches
• is related to: LU-16742 e2fsck/ldiskfs/ZFS to allow multiple FIDs per dirent