Hi folks, This is the second version of the RFC I originally posted here: https://lore.kernel.org/linux-xfs/20190801021752.4986-1-david@xxxxxxxxxxxxx/ The original description of the patchset is below, the issues and approach to solving them has not changed. THere is some restructuring of the patch set - the first few patches are all the XFS fixes that can be merged regardless of the rest of the patchset, but the non-blocking reclaim is somewhat dependent of them for correct behaviour. The second set of patches are the shrinker infrastructure changes needed for the shrinkers to feed back reclaim progress to the main reclaim instructure and act on the feedback. The last set of patches are the XFS changes needed to convert inode reclaim over to a non-blocking, IO-less algorithm. Version 2: - added current_reclaim_account_pages() wrapper for reclaim_state updates - moved xfs_buf page free accounting to the page freeing code rather than the reclaim loop that drops the LRU buffer reference. - increased log CIL flush limit by 2x - moved xfs_buf GFP_NOFS reclaim changes to correct patch - renamed sc->will_defer to sc->defer_work - used min() in do_shrink_slab() appropriately - converted xfs_trans_ail_update_bulk() to use new xfs_ail_update_finish() function (fixes missing wakeups w/ xfs_ail_push_sync()) - fixed CIL size limit calculation units - was calculating min size in sectors rather than bytes. Fixes performance degradation w/ patch series. - dropped shadow buffer freeing - Brain pointed out it wasn't doing exactly what I thought it was, and the memory saving were from the massively undersized CIL limits, not early freeing of the shadow buffers. Needs rethinking. - fixed stray NULLCOMMITLSN in comments - pinned items don't track the commit LSN, just the CIL sequence number so we can't use that to push the AIL. - removed stale tracing debug from AIL push code. - fixes to memory reclaim shrinker accounting in 5.3-rc3 result in direct reclaim backoff working a whole lot better, such that it's no long necessary for the XFS inode shrinker to wait for IO to complete. Changed the LRU reclaim logic to simply push on the AIL if dirty inodes are hit, but never wait on them. See the following commits for details of the shrinker accounting changes: 0308f7cf19c9 ("mm/vmscan.c: calculate reclaimed slab caches in all reclaim paths") e5ca8071fe65 ("mm/vmscan.c: add a new member reclaim_state in struct shrink_control") - Added a patch to convert xfs_reclaim_inodes() to use xfs_ail_push_all() which gets rid of the last user of xfs_ail_push_sync(). This allows it to be removed as "temporary infrastructure for the series" rather than having to be fixed up and made robust. The optimisations and factoring it drove are retained, as they are still a net improvement overall. - fixed atomic_long vs atomic64 issues with shrinker deferral rework. - don't drop ag reclaim cursor locking any more, it gets removed when all the old reclaim code is removed. - added patch to change inode reclaim vs unreferenced XFS indoe lookup done by inode write clustering and inode cluster freeing. This gets rid of the need to cycle the ILOCK before running call_rcu() to queue the inode to be freed when the current RCU grace period expires. In doing so, the last major blocking point in XFS inode reclaim is removed. Comments, thoughts, testing all welcome.... Cheers, Dave. --- Diffstat: drivers/staging/android/ashmem.c | 8 +- fs/gfs2/glock.c | 5 +- fs/gfs2/quota.c | 6 +- fs/inode.c | 3 +- fs/nfs/dir.c | 6 +- fs/super.c | 6 +- fs/xfs/xfs_buf.c | 4 +- fs/xfs/xfs_icache.c | 628 +++++++++++---------------------------- fs/xfs/xfs_icache.h | 20 +- fs/xfs/xfs_inode.c | 131 ++++---- fs/xfs/xfs_inode.h | 8 + fs/xfs/xfs_inode_item.c | 28 +- fs/xfs/xfs_log.c | 10 +- fs/xfs/xfs_log_cil.c | 37 ++- fs/xfs/xfs_log_priv.h | 53 +++- fs/xfs/xfs_mount.c | 10 +- fs/xfs/xfs_mount.h | 6 +- fs/xfs/xfs_qm.c | 11 +- fs/xfs/xfs_super.c | 94 ++++-- fs/xfs/xfs_trace.h | 1 + fs/xfs/xfs_trans_ail.c | 88 ++++-- fs/xfs/xfs_trans_priv.h | 6 +- include/linux/shrinker.h | 9 +- include/linux/swap.h | 23 +- include/trace/events/vmscan.h | 69 ++--- mm/slab.c | 3 +- mm/slob.c | 4 +- mm/slub.c | 3 +- mm/vmscan.c | 231 +++++++++----- net/sunrpc/auth.c | 5 +- 30 files changed, 741 insertions(+), 775 deletions(-) --- [Original V1 patchset description] We've had a problem with inode reclaim for a long time - XFS is capable of caching tens of millions of inodes with ease and dirtying hundreds of thousands of those cached inodes every second. It is also capable of reclaiming more than half a million clean inodes per second per reclaim thread. The result of this is that when there is a significant change in sustained memory pressure on a system ith a large inode cache, memory reclaim rapdily frees all the clean inodes, but cannot make progress on reclaiming dirty inodes because they are rate limited by IO. However, the shrinker infrastructure in the kernel has no way to feed back rate limiting to the core memory reclaim algorithms. In fact there are no feedback mechanisms at all, and so when reclaim has freed all the clean inodes and starts hitting dirty inodes, the filesystem has no way of telling reclaim that the inode reclaim rate has dropped from 500k/s to 500/s. The result is that reclaim continues to try to free memory, and because it makes no progress freeing inodes, it puts much more pressure on the page LRUs and frees pages. When it runs out of pages, it starts swapping, and when it runs out of swap or can't get a page for swap-in it starts going on an OOM kill rampage. That does nothing to "fix" the shortage of memory caused by the slowness of dirty inode reclaim - if memory demand continues we just keep hitting the OOM killer until either something critical is killed or memory demand eases. For a long time, XFS has avoided the insane spiral of shouty OOM-killer rage death by cleaning inodes directly in the shrinker. This has the effect of throttling memory reclaim to the rate at which dirty inodes can be cleaned, and so when we get into the state when memory reclaim is dependent on inode reclaim making progress we don't ever allow LRU reclaim to run so far ahead of inode reclaim that it winds up reclaim priority and runs out of LRU pages to reclaim and/or swap. This has a downside, though. When there is a large amount of clean page cache and a small amount of inode cache that is dirty (e.g. lots of file data pressure and/or application memory demand) the inode reclaim shrinkers can run out of clean inodes to reclaim and start blocking on inode writeback. This can result in long reclaim latencies even though there is lots of memory that can be immediately reclaimed from the page cache. There are other issues, too. We have to block kswapd, too, because it will continue running until watermarks are satisfied, and that is largely the vector for shouty swappy death if it doesn't back off before priority windup from lack of progress occurs. Blocking kswapd then affects direct reclaim function, which often backs off expecting kswapd to make progress in the mean time. But if kswapd is not making progress, direct reclaim ends up in priority windup from lack of progress, too. This is especially prevalent in workloads that have a high percentage of GFP_NOFS allocations (e.g. filesystem modification workloads). The shrinkers have another problem w/ GFP_NOFS reclaim: the work that is deferred because the shrinker cannot make progress gets lumped on the first reclaim context that can do that work. That means a direct reclaimer might get lumped with scanning millions of objects during low priority scanning when it should only be scanning a couple of thousand objects. This can result in highly unpredictable and extremely long direct reclaim delays. This is most definitely sub-optimal, but it's better than random and/or premature OOM killer invocation under trivial workloads and lots of reclaimable memory still being available. This patch set aims to fix all these problems. The memory reclaim and shrinker changes involve: - a substantial rework of how the shrinker defers work, moving all the deferred work to kswapd to remove all the unpredictability from direct reclaim. Direct reclaim will only do the work the direct reclaim context determines is necesary. - deferred work is capped, and the amount of deferred work kswapd will do in each scan is increased linearly w.r.t. increasing reclaim priority. Hence when we are desparate for memory, kswapd will be running all the deferred work as quickly as possible. - The amount of deferred work and the amount of scanning that is done by the shrinkers is now tracked in the struct reclaim_state. This allows shrink_node() to see how much work is being done in comparison to both the LRU scanning and how much is being deferred to kswapd. This allows direct reclaim to back off when too much work is being deferred and hence allow kswapd to make progress on the deferred work while it waits. - A "need backoff" flag has been added to the struct reclaim_state. This allows individual shrinkers to indicate to kswapd that they need some time to finish work before being scanned again. This is basically for the same case as kswapd backs off from LRU scanning. i.e. the LRU scanning has run into the tail of the LRU and is finding dirty objects that require IO to complete before reclaim can make further progress. This is exactly the same problem we have with inode reclaim in XFS, and it is this mechanism that enables us to move to IO-less inode reclaim. The XFS changes are all over the place, and address both the reclaim blocking problems and all the other related issues I found while working on this patchest. These involve: - fixing IO priority inversion problems between metadata writeback (inodes!) and log IO caused by the block layer write throttling (more on this later). - some slab caches weren't marked as reclaimable, so were incorrectly accounted. Also account for the pages xfs_buf reclaim releases. - reduced the delayed logging dirty item aggregation size (the CIL). This defines the minimum amount of memory XFS can operate in when there is heavy modifications in progress. - reduced the memory footprint of the CIL when repeated modifications to objects occur. - Added a mechanism to push the AIL to a specific LSN (metadata modification epoch) and wait for it. This forms the basis for direct inode reclaim deferring IO and waiting for some progress without issuing IO iteslf. - reworked inode reclaim to use a list_lru to track inodes in reclaim rather than a radix tree tag in the inode cache. We iterated the radix tree for reclaim because it resulted in optimal IO patterns from multiple concurrent reclaimers, but we dont' have to care about that any more because all IO comes from the AIL now. This gives us try LRU reclaim, and it allows us to effectively determine when we've run out of clean inodes to easily reclaim and provide that feedback to the higher levels via the "need backoff" flag. - direct reclaim is non-blocking while scanning, but at the end of a scan it will still block waiting for IO, but only for /some/ progress to be made and not specific individual IOs. - kswapd based reclaim is fully non-blocking. The result is that there is now enough feedback from the shrinkers into the main memory reclaim loop for it to back off in the situations where back-off is required to avoid OOM killer invocation, despite XFS now largely doing non-blocking reclaim. Testing involves at 16p/16GB machine running a fsmark workload that creates sustained heavy dirty inode cache pressure, then progressively locking 2GB of memory at time to squeeze the workload into less and less memory. A vanilla kernel runs well up to 12GB squeezed, but at 14GB squeezed performance goes to hell. With just the hacky "don't block kswapd by removing SYNC_WAIT" patch that people seem to like, OOM kills start when squeezed to 12GB. With that extended to direct reclaim, OOM kills start with squeezed to just 8GB. With the full patchset, it runs similar to a vanilla kernel up to 12GB squeezed, and vastly out-performs the vanilla kernel with 14GB squeezed. Performance only drops ~20% with a 14GB squeeze, whereas the vanilla kernel sees up to a 90% drop in performance. I also run testing with simoop, a simulated workload that Chris Mason put together to demonstrate the long tail latency and allocation stall problems the blocking in inode reclaim was causing. The vanilla kernel averaged ~5 stalls/s over a test period of 10 hours, this patch series resulted in: alloc stall rate = 0.00/sec (avg: 0.04) (p50: 0.04) (p95: 0.16) (p99: 0.32) stalls almost going away entirely. So the signs are there that this is a workable solution to the problems caused by blocking inode reclaim without re-introducing the the Death-by-OOM-killer issues the blocking avoids. Please note that I haven't full gone non-blocking on direct reclaim for a couple of reasons: 1. congestion_wait() and wait_iff_congested() are completely broken. The blkmq change-over ripped out all the block layer congestion reporting in 5.0 and didn't replace it with anything, so unless you are operating on an NFS client, Ceph, FUSE or a DVD, congestion checks and backoff aren't actually doing what they are supposed to. i.e. wait_iff_congested() never blocks, and congestion_wait() always sleeps for it's full timeout. IOWs, the whole bdi-based IO congestion feedback mechanism no longer functions as intended, and so I'm betting a lot of the memory reclaim heuristics no longer function as they were intended to... 2. The block layer write throttle is full of priority inversions. Apart from the log IO one I fixed in this series, I noticed that swap in/out has a major problem. I lost count of the number of OOM kills that occurred from the swap in path when there were several processes blocked in wbt_wait() in the block layer in the swap out path. i.e. if swap out had been making progress, swap in would not have oom killed. Hence I found it still necessary to throttle direct reclaim back in the shrinker as there wasn't a realiable way to get the core reclaim code to throttle effectively. FWIW, from the swap in/out perspective, this whole inversion problem is made worse by #1: the congestion_wait/wait_iff_congested interfaces being broken. Direct reclaim uses wait_iff_congested() to back off if kswapd has indicated that the node is congested (PGDAT_CONGESTED) and reclaim is struggling to make progress. However, this backoff never actually happens now and hence direct reclaim barrels into the swap code as hard as it can and blocks in wbt_wait() waiting behind other swap IO instead of backing off and waiting for some IO to complete and then retrying it's allocation.... So maybe if we fix the bdi congestion interfaces so they work again we can get rid of the waiting in direct reclaim, but right now I don't see any other choice....
