linux-toradex.git/mm/zsmalloc.c, branch v5.16-rc5

mm/zsmalloc.c: close race window between zs_pool_dec_isolated() and zs_unregister_migration()

2021-11-06T20:30:43+00:00

There is one possible race window between zs_pool_dec_isolated() and
zs_unregister_migration() because wait_for_isolated_drain() checks the
isolated count without holding class->lock and there is no order inside
zs_pool_dec_isolated().  Thus the below race window could be possible:

  zs_pool_dec_isolated		zs_unregister_migration
    check pool->destroying != 0
				  pool->destroying = true;
				  smp_mb();
				  wait_for_isolated_drain()
				    wait for pool->isolated_pages == 0
    atomic_long_dec(&pool->isolated_pages);
    atomic_long_read(&pool->isolated_pages) == 0

Since we observe the pool->destroying (false) before atomic_long_dec()
for pool->isolated_pages, waking pool->migration_wait up is missed.

Fix this by ensure checking pool->destroying happens after the
atomic_long_dec(&pool->isolated_pages).

Link: https://lkml.kernel.org/r/20210708115027.7557-1-linmiaohe@huawei.com
Fixes: 701d678599d0 ("mm/zsmalloc.c: fix race condition in zs_destroy_pool")
Signed-off-by: Miaohe Lin 
Cc: Minchan Kim 
Cc: Sergey Senozhatsky 
Cc: Henry Burns 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/zsmalloc.c: improve readability for async_free_zspage()

2021-07-01T18:06:02+00:00

The class is extracted from pool->size_class[class_idx] again before
calling __free_zspage().  It looks like class will change after we fetch
the class lock.  But this is misleading as class will stay unchanged.

Link: https://lkml.kernel.org/r/20210624123930.1769093-4-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin 
Cc: Minchan Kim 
Cc: Nitin Gupta 
Cc: Sergey Senozhatsky 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/zsmalloc.c: remove confusing code in obj_free()

2021-07-01T18:06:02+00:00

Patch series "Cleanup for zsmalloc".

This series contains cleanups to remove confusing code in obj_free(),
combine two atomic ops and improve readability for async_free_zspage().
More details can be found in the respective changelogs.

This patch (of 2):

OBJ_ALLOCATED_TAG is only set for handle to indicate allocated object.
It's irrelevant with obj.  So remove this misleading code to improve
readability.

Link: https://lkml.kernel.org/r/20210624123930.1769093-1-linmiaohe@huawei.com
Link: https://lkml.kernel.org/r/20210624123930.1769093-2-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin 
Cc: Minchan Kim 
Cc: Nitin Gupta 
Cc: Sergey Senozhatsky 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: fix typos in comments

2021-05-07T07:26:35+00:00

Fix ~94 single-word typos in locking code comments, plus a few
very obvious grammar mistakes.

Link: https://lkml.kernel.org/r/20210322212624.GA1963421@gmail.com
Link: https://lore.kernel.org/r/20210322205203.GB1959563@gmail.com
Signed-off-by: Ingo Molnar 
Reviewed-by: Matthew Wilcox (Oracle) 
Reviewed-by: Randy Dunlap 
Cc: Bhaskar Chowdhury 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: fix some typos and code style problems

2021-05-07T07:26:33+00:00

fix some typos and code style problems in mm.

gfp.h: s/MAXNODES/MAX_NUMNODES
mmzone.h: s/then/than
rmap.c: s/__vma_split()/__vma_adjust()
swap.c: s/__mod_zone_page_stat/__mod_zone_page_state, s/is is/is
swap_state.c: s/whoes/whose
z3fold.c: code style problem fix in z3fold_unregister_migration
zsmalloc.c: s/of/or, s/give/given

Link: https://lkml.kernel.org/r/20210419083057.64820-1-luoshijie1@huawei.com
Signed-off-by: Shijie Luo 
Signed-off-by: Miaohe Lin 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/zsmalloc: use BUG_ON instead of if condition followed by BUG.

2021-05-05T18:27:27+00:00

It can be optimized at compile time.

Link: https://lkml.kernel.org/r/1616727798-9110-1-git-send-email-zhouchuangao@vivo.com
Signed-off-by: zhouchuangao 
Cc: Minchan Kim 
Cc: Sergey Senozhatsky 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/zsmalloc.c: use page_private() to access page->private

2021-02-26T17:41:01+00:00

It's recommended to use helper macro page_private() to access the private
field of page.  Use such helper to eliminate direct access.

Link: https://lkml.kernel.org/r/20210203091857.20017-1-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin 
Cc: Minchan Kim 
Cc: Nitin Gupta 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

zsmalloc: account the number of compacted pages correctly

2021-02-26T17:41:01+00:00

There exists multiple path may do zram compaction concurrently.
1. auto-compaction triggered during memory reclaim
2. userspace utils write zram/compaction node

So, multiple threads may call zs_shrinker_scan/zs_compact concurrently.
But pages_compacted is a per zsmalloc pool variable and modification
of the variable is not serialized(through under class->lock).
There are two issues here:
1. the pages_compacted may not equal to total number of pages
freed(due to concurrently add).
2. zs_shrinker_scan may not return the correct number of pages
freed(issued by current shrinker).

The fix is simple:
1. account the number of pages freed in zs_compact locally.
2. use actomic variable pages_compacted to accumulate total number.

Link: https://lkml.kernel.org/r/20210202122235.26885-1-wu-yan@tcl.com
Fixes: 860c707dca155a56 ("zsmalloc: account the number of compacted pages")
Signed-off-by: Rokudo Yan 
Cc: Minchan Kim 
Cc: Sergey Senozhatsky 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/zsmalloc.c: convert to use kmem_cache_zalloc in cache_alloc_zspage()

2021-02-26T17:41:01+00:00

We always memset the zspage allocated via cache_alloc_zspage.  So it's
more convenient to use kmem_cache_zalloc in cache_alloc_zspage than caller
do it manually.

Link: https://lkml.kernel.org/r/20210114120032.25885-1-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin 
Reviewed-by: Sergey Senozhatsky 
Cc: Minchan Kim 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/zsmalloc.c: rework the list_add code in insert_zspage()

2020-12-15T20:13:46+00:00

Rework the list_add code to make it more readable and simple.

Link: https://lkml.kernel.org/r/20201015130107.65195-1-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin 
Acked-by: Minchan Kim 
Reviewed-by: Sergey Senozhatsky 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds