kernel.git/mm/page_io.c, branch linux-5.5.y

mm/page_io.c: annotate refault stalls from swap_readpage

2019-12-01T20:59:11Z

If a block device supports rw_page operation, it doesn't submit bios so the annotation in submit_bio() for refault stall doesn't work. It happens with zram in android, especially swap read path which could consume CPU cycle for decompress. It is also a problem for zswap which uses frontswap. Annotate swap_readpage() to account the synchronous IO overhead to prevent underreport memory pressure. [akpm@linux-foundation.org: add comment, per Johannes] Link: http://lkml.kernel.org/r/20191010152134.38545-1-minchan@kernel.org Signed-off-by: Minchan Kim Acked-by: Johannes Weiner Reviewed-by: Shakeel Butt Cc: Seth Jennings Cc: Dan Streetman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm/page_io.c: do not free shared swap slots

2019-11-16T02:34:00Z

The following race is observed due to which a processes faulting on a swap entry, finds the page neither in swapcache nor swap. This causes zram to give a zero filled page that gets mapped to the process, resulting in a user space crash later. Consider parent and child processes Pa and Pb sharing the same swap slot with swap_count 2. Swap is on zram with SWP_SYNCHRONOUS_IO set. Virtual address 'VA' of Pa and Pb points to the shared swap entry. Pa Pb fault on VA fault on VA do_swap_page do_swap_page lookup_swap_cache fails lookup_swap_cache fails Pb scheduled out swapin_readahead (deletes zram entry) swap_free (makes swap_count 1) Pb scheduled in swap_readpage (swap_count == 1) Takes SWP_SYNCHRONOUS_IO path zram enrty absent zram gives a zero filled page Fix this by making sure that swap slot is freed only when swap count drops down to one. Link: http://lkml.kernel.org/r/1571743294-14285-1-git-send-email-vinmenon@codeaurora.org Fixes: aa8d22a11da9 ("mm: swap: SWP_SYNCHRONOUS_IO: skip swapcache only if swapped page has no other reference") Signed-off-by: Vinayak Menon Suggested-by: Minchan Kim Acked-by: Minchan Kim Cc: Michal Hocko Cc: Hugh Dickins Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm, swap: use rbtree for swap_extent

2019-07-12T18:05:43Z

swap_extent is used to map swap page offset to backing device's block offset. For a continuous block range, one swap_extent is used and all these swap_extents are managed in a linked list. These swap_extents are used by map_swap_entry() during swap's read and write path. To find out the backing device's block offset for a page offset, the swap_extent list will be traversed linearly, with curr_swap_extent being used as a cache to speed up the search. This works well as long as swap_extents are not huge or when the number of processes that access swap device are few, but when the swap device has many extents and there are a number of processes accessing the swap device concurrently, it can be a problem. On one of our servers, the disk's remaining size is tight: $df -h Filesystem Size Used Avail Use% Mounted on ... ... /dev/nvme0n1p1 1.8T 1.3T 504G 72% /home/t4 When creating a 80G swapfile there, there are as many as 84656 swap extents. The end result is, kernel spends abou 30% time in map_swap_entry() and swap throughput is only 70MB/s. As a comparison, when I used smaller sized swapfile, like 4G whose swap_extent dropped to 2000, swap throughput is back to 400-500MB/s and map_swap_entry() is about 3%. One downside of using rbtree for swap_extent is, 'struct rbtree' takes 24 bytes while 'struct list_head' takes 16 bytes, that's 8 bytes more for each swap_extent. For a swapfile that has 80k swap_extents, that means 625KiB more memory consumed. Test: Since it's not possible to reboot that server, I can not test this patch diretly there. Instead, I tested it on another server with NVMe disk. I created a 20G swapfile on an NVMe backed XFS fs. By default, the filesystem is quite clean and the created swapfile has only 2 extents. Testing vanilla and this patch shows no obvious performance difference when swapfile is not fragmented. To see the patch's effects, I used some tweaks to manually fragment the swapfile by breaking the extent at 1M boundary. This made the swapfile have 20K extents. nr_task=4 kernel swapout(KB/s) map_swap_entry(perf) swapin(KB/s) map_swap_entry(perf) vanilla 165191 90.77% 171798 90.21% patched 858993 +420% 2.16% 715827 +317% 0.77% nr_task=8 kernel swapout(KB/s) map_swap_entry(perf) swapin(KB/s) map_swap_entry(perf) vanilla 306783 92.19% 318145 87.76% patched 954437 +211% 2.35% 1073741 +237% 1.57% swapout: the throughput of swap out, in KB/s, higher is better 1st map_swap_entry: cpu cycles percent sampled by perf swapin: the throughput of swap in, in KB/s, higher is better. 2nd map_swap_entry: cpu cycles percent sampled by perf nr_task=1 doesn't show any difference, this is due to the curr_swap_extent can be effectively used to cache the correct swap extent for single task workload. [akpm@linux-foundation.org: s/BUG_ON(1)/BUG()/] Link: http://lkml.kernel.org/r/20190523142404.GA181@aaronlu Signed-off-by: Aaron Lu Cc: Huang Ying Cc: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

swap_readpage(): avoid blk_wake_io_task() if !synchronous

2019-07-05T02:12:07Z

swap_readpage() sets waiter = bio->bi_private even if synchronous = F, this means that the caller can get the spurious wakeup after return. This can be fatal if blk_wake_io_task() does set_current_state(TASK_RUNNING) after the caller does set_special_state(), in the worst case the kernel can crash in do_task_dead(). Link: http://lkml.kernel.org/r/20190704160301.GA5956@redhat.com Fixes: 0619317ff8baa2d ("block: add polled wakeup task helper") Signed-off-by: Oleg Nesterov Reported-by: Qian Cai Acked-by: Hugh Dickins Reviewed-by: Jens Axboe Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm, swap: fix THP swap out

2019-06-29T08:43:45Z

0-Day test system reported some OOM regressions for several THP (Transparent Huge Page) swap test cases. These regressions are bisected to 6861428921b5 ("block: always define BIO_MAX_PAGES as 256"). In the commit, BIO_MAX_PAGES is set to 256 even when THP swap is enabled. So the bio_alloc(gfp_flags, 512) in get_swap_bio() may fail when swapping out THP. That causes the OOM. As in the patch description of 6861428921b5 ("block: always define BIO_MAX_PAGES as 256"), THP swap should use multi-page bvec to write THP to swap space. So the issue is fixed via doing that in get_swap_bio(). BTW: I remember I have checked the THP swap code when 6861428921b5 ("block: always define BIO_MAX_PAGES as 256") was merged, and thought the THP swap code needn't to be changed. But apparently, I was wrong. I should have done this at that time. Link: http://lkml.kernel.org/r/20190624075515.31040-1-ying.huang@intel.com Fixes: 6861428921b5 ("block: always define BIO_MAX_PAGES as 256") Signed-off-by: "Huang, Ying" Reviewed-by: Ming Lei Cc: Michal Hocko Cc: Johannes Weiner Cc: Hugh Dickins Cc: Minchan Kim Cc: Rik van Riel Cc: Daniel Jordan Cc: Jens Axboe Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm/page_io.c: fix polled swap page in

2019-01-04T21:13:48Z

swap_readpage() wants to do polling to bring in pages if asked to, but it doesn't mark the bio as being polled. Additionally, the looping around the blk_poll() check isn't correct - if we get a zero return, we should call io_schedule(), we can't just assume that the bio has completed. The regular bio->bi_private check should be used for that. Link: http://lkml.kernel.org/r/e15243a8-2cdf-c32c-ecee-f289377c8ef9@kernel.dk Signed-off-by: Jens Axboe Reviewed-by: Andrew Morton Cc: Christoph Hellwig Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

block: don't use un-ordered __set_current_state(TASK_UNINTERRUPTIBLE)

2019-01-02T18:46:03Z

This mostly reverts commit 849a370016a5 ("block: avoid ordered task state change for polled IO"). It was wrongly claiming that the ordering wasn't necessary. The memory barrier _is_ necessary. If something is truly polling and not going to sleep, it's the whole state setting that is unnecessary, not the memory barrier. Whenever you set your state to a sleeping state, you absolutely need the memory barrier. Note that sometimes the memory barrier can be elsewhere. For example, the ordering might be provided by an external lock, or by setting the process state to sleeping before adding yourself to the wait queue list that is used for waking up (where the wait queue lock itself will guarantee that any wakeup will correctly see the sleeping state). But none of those cases were true here. NOTE! Some of the polling paths may indeed be able to drop the state setting entirely, at which point the memory barrier also goes away. (Also note that this doesn't revert the TASK_RUNNING cases: there is no race between a wakeup and setting the process state to TASK_RUNNING, since the end result doesn't depend on ordering). Cc: Jens Axboe Cc: Christoph Hellwig Signed-off-by: Linus Torvalds

blkcg: associate a blkg for pages being evicted by swap

2018-12-08T05:26:37Z

A prior patch in this series added blkg association to bios issued by cgroups. There are two other paths that we want to attribute work back to the appropriate cgroup: swap and writeback. Here we modify the way swap tags bios to include the blkg. Writeback will be tackle in the next patch. Signed-off-by: Dennis Zhou Reviewed-by: Josef Bacik Acked-by: Tejun Heo Signed-off-by: Jens Axboe

block: make blk_poll() take a parameter on whether to spin or not

2018-11-26T15:25:53Z

blk_poll() has always kept spinning until it found an IO. This is fine for SYNC polling, since we need to find one request we have pending, but in preparation for ASYNC polling it can be beneficial to just check if we have any entries available or not. Existing callers are converted to pass in 'spin == true', to retain the old behavior. Signed-off-by: Jens Axboe

block: avoid ordered task state change for polled IO

2018-11-19T15:34:49Z

For the core poll helper, the task state setting don't need to imply any atomics, as it's the current task itself that is being modified and we're not going to sleep. For IRQ driven, the wakeup path have the necessary barriers to not need us using the heavy handed version of the task state setting. Reviewed-by: Christoph Hellwig Signed-off-by: Jens Axboe