- 07 Sep, 2017 40 commits
-
-
Huang Ying authored
Patch series "mm, THP, swap: Delay splitting THP after swapped out", v3. This is the second step of THP (Transparent Huge Page) swap optimization. In the first step, the splitting huge page is delayed from almost the first step of swapping out to after allocating the swap space for the THP and adding the THP into the swap cache. In the second step, the splitting is delayed further to after the swapping out finished. The plan is to delay splitting THP step by step, finally avoid splitting THP for the THP swapping out and swap out/in the THP as a whole. In the patchset, more operations for the anonymous THP reclaiming, such as TLB flushing, writing the THP to the swap device, removing the THP from the swap cache are batched. So that the performance of anonymous THP swapping out are improved. During the development, the following scenarios/code paths have been checked, - swap out/in - swap off - write protect page fault - madvise_free - process exit - split huge page With the patchset, the swap out throughput improves 42% (from about 5.81GB/s to about 8.25GB/s) in the vm-scalability swap-w-seq test case with 16 processes. At the same time, the IPI (reflect TLB flushing) reduced about 78.9%. The test is done on a Xeon E5 v3 system. The swap device used is a RAM simulated PMEM (persistent memory) device. To test the sequential swapping out, the test case creates 8 processes, which sequentially allocate and write to the anonymous pages until the RAM and part of the swap device is used up. Below is the part of the cover letter for the first step patchset of THP swap optimization which applies to all steps. ========================= Recently, the performance of the storage devices improved so fast that we cannot saturate the disk bandwidth with single logical CPU when do page swap out even on a high-end server machine. Because the performance of the storage device improved faster than that of single logical CPU. And it seems that the trend will not change in the near future. On the other hand, the THP becomes more and more popular because of increased memory size. So it becomes necessary to optimize THP swap performance. The advantages of the THP swap support include: - Batch the swap operations for the THP to reduce TLB flushing and lock acquiring/releasing, including allocating/freeing the swap space, adding/deleting to/from the swap cache, and writing/reading the swap space, etc. This will help improve the performance of the THP swap. - The THP swap space read/write will be 2M sequential IO. It is particularly helpful for the swap read, which are usually 4k random IO. This will improve the performance of the THP swap too. - It will help the memory fragmentation, especially when the THP is heavily used by the applications. The 2M continuous pages will be free up after THP swapping out. - It will improve the THP utilization on the system with the swap turned on. Because the speed for khugepaged to collapse the normal pages into the THP is quite slow. After the THP is split during the swapping out, it will take quite long time for the normal pages to collapse back into the THP after being swapped in. The high THP utilization helps the efficiency of the page based memory management too. There are some concerns regarding THP swap in, mainly because possible enlarged read/write IO size (for swap in/out) may put more overhead on the storage device. To deal with that, the THP swap in should be turned on only when necessary. For example, it can be selected via "always/never/madvise" logic, to be turned on globally, turned off globally, or turned on only for VMA with MADV_HUGEPAGE, etc. This patch (of 12): Previously, swapcache_free_cluster() is used only in the error path of shrink_page_list() to free the swap cluster just allocated if the THP (Transparent Huge Page) is failed to be split. In this patch, it is enhanced to clear the swap cache flag (SWAP_HAS_CACHE) for the swap cluster that holds the contents of THP swapped out. This will be used in delaying splitting THP after swapping out support. Because there is no THP swapping in as a whole support yet, after clearing the swap cache flag, the swap cluster backing the THP swapped out will be split. So that the swap slots in the swap cluster can be swapped in as normal pages later. Link: http://lkml.kernel.org/r/20170724051840.2309-2-ying.huang@intel.comSigned-off-by:
"Huang, Ying" <ying.huang@intel.com> Acked-by:
Rik van Riel <riel@redhat.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Minchan Kim <minchan@kernel.org> Cc: Hugh Dickins <hughd@google.com> Cc: Shaohua Li <shli@kernel.org> Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Michal Hocko <mhocko@kernel.org> Cc: Ross Zwisler <ross.zwisler@intel.com> [for brd.c, zram_drv.c, pmem.c] Cc: Vishal L Verma <vishal.l.verma@intel.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Matthias Kaehlcke authored
Several functions use an enum type as parameter for an event/state, but are called in some locations with an argument of a different enum type. Adjust the interface of these functions to reality by changing the parameter to int. This fixes a ton of enum-conversion warnings that are generated when building the kernel with clang. [mka@chromium.org: also change parameter type of inc/dec/mod_memcg_page_state()] Link: http://lkml.kernel.org/r/20170728213442.93823-1-mka@chromium.org Link: http://lkml.kernel.org/r/20170727211004.34435-1-mka@chromium.orgSigned-off-by:
Matthias Kaehlcke <mka@chromium.org> Acked-by:
Johannes Weiner <hannes@cmpxchg.org> Acked-by:
Michal Hocko <mhocko@suse.com> Cc: Vladimir Davydov <vdavydov.dev@gmail.com> Cc: Doug Anderson <dianders@chromium.org> Signed-off-by:
-
Arvind Yadav authored
attribute_group are not supposed to change at runtime. All functions working with attribute_group provided by <linux/sysfs.h> work with const attribute_group. So mark the non-const structs as const. Link: http://lkml.kernel.org/r/1501157260-3922-1-git-send-email-arvind.yadav.cs@gmail.comSigned-off-by:
Arvind Yadav <arvind.yadav.cs@gmail.com> Signed-off-by:
-
Arvind Yadav authored
attribute_group are not supposed to change at runtime. All functions working with attribute_group provided by <linux/sysfs.h> work with const attribute_group. So mark the non-const structs as const. Link: http://lkml.kernel.org/r/1501157240-3876-1-git-send-email-arvind.yadav.cs@gmail.comSigned-off-by:
-
Arvind Yadav authored
attribute_group are not supposed to change at runtime. All functions working with attribute_group provided by <linux/sysfs.h> work with const attribute_group. So mark the non-const structs as const. Link: http://lkml.kernel.org/r/1501157221-3832-1-git-send-email-arvind.yadav.cs@gmail.comSigned-off-by:
-
Arvind Yadav authored
attribute_group are not supposed to change at runtime. All functions working with attribute_group provided by <linux/sysfs.h> work with const attribute_group. So mark the non-const structs as const. Link: http://lkml.kernel.org/r/1501157186-3749-1-git-send-email-arvind.yadav.cs@gmail.comSigned-off-by:
Christoph Lameter <cl@linux.com> Signed-off-by:
-
Arvind Yadav authored
attribute_group are not supposed to change at runtime. All functions working with attribute_group provided by <linux/sysfs.h> work with const attribute_group. So mark the non-const structs as const. Link: http://lkml.kernel.org/r/1501157167-3706-2-git-send-email-arvind.yadav.cs@gmail.comSigned-off-by:
-
Roman Gushchin authored
Commit fa06235b ("cgroup: reset css on destruction") caused css_reset callback to be called from the offlining path. Although it solves the problem mentioned in the commit description ("For instance, memory cgroup needs to reset memory.low, otherwise pages charged to a dead cgroup might never get reclaimed."), generally speaking, it's not correct. An offline cgroup can still be a resource domain, and we shouldn't grant it more resources than it had before deletion. For instance, if an offline memory cgroup has dirty pages, we should still imply i/o limits during writeback. The css_reset callback is designed to return the cgroup state into the original state, that means reset all limits and counters. It's spomething different from the offlining, and we shouldn't use it from the offlining path. Instead, we should adjust necessary settings from the per-controller css_offline callbacks (e.g. reset memory.low). Link: http://lkml.kernel.org/r/20170727130428.28856-2-guro@fb.comSigned-off-by:
Roman Gushchin <guro@fb.com> Acked-by:
Tejun Heo <tj@kernel.org> Acked-by:
-
Roman Gushchin authored
A removed memory cgroup with a defined memory.low and some belonging pagecache has very low chances to be freed. If a cgroup has been removed, there is likely no memory pressure inside the cgroup, and the pagecache is protected from the external pressure by the defined low limit. The cgroup will be freed only after the reclaim of all belonging pages. And it will not happen until there are any reclaimable memory in the system. That means, there is a good chance, that a cold pagecache will reside in the memory for an undefined amount of time, wasting system resources. This problem was fixed earlier by fa06235b ("cgroup: reset css on destruction"), but it's not a best way to do it, as we can't really reset all limits/counters during cgroup offlining. Link: http://lkml.kernel.org/r/20170727130428.28856-1-guro@fb.comSigned-off-by:
-
Jan Kara authored
All users of pagevec_lookup() and pagevec_lookup_range() now pass PAGEVEC_SIZE as a desired number of pages. Just drop the argument. Link: http://lkml.kernel.org/r/20170726114704.7626-11-jack@suse.czSigned-off-by:
Jan Kara <jack@suse.cz> Signed-off-by:
-
Jan Kara authored
We want only pages from given range in filemap_range_has_page(), furthermore we want at most a single page. So use find_get_pages_range() instead of pagevec_lookup() and remove unnecessary code. Link: http://lkml.kernel.org/r/20170726114704.7626-10-jack@suse.czSigned-off-by:
-
Jan Kara authored
We want only pages from given range in page_cache_seek_hole_data(). Use pagevec_lookup_range() instead of pagevec_lookup() and remove unnecessary code. Note that the check for getting less pages than desired can be removed because index gets updated by pagevec_lookup_range(). Link: http://lkml.kernel.org/r/20170726114704.7626-9-jack@suse.czSigned-off-by:
-
Jan Kara authored
We want only pages from given range in remove_inode_hugepages(). Use pagevec_lookup_range() instead of pagevec_lookup(). Link: http://lkml.kernel.org/r/20170726114704.7626-8-jack@suse.czSigned-off-by:
Mike Kravetz <mike.kravetz@oracle.com> Cc: Nadia Yvette Chambers <nyc@holomorphy.com> Signed-off-by:
-
Jan Kara authored
Both occurences of pagevec_lookup() actually want only pages from a given range. Use pagevec_lookup_range() for the lookup. Link: http://lkml.kernel.org/r/20170726114704.7626-7-jack@suse.czSigned-off-by:
-
Jan Kara authored
Use pagevec_lookup_range() in ext4_find_unwritten_pgoff() since we are interested only in pages in the given range. Simplify the logic as a result of not getting pages out of range and index getting automatically advanced. Link: http://lkml.kernel.org/r/20170726114704.7626-6-jack@suse.czSigned-off-by:
-
Jan Kara authored
Commit e64855c6 ("fs: Add helper to clean bdev aliases under a bh and use it") added a wrapper for clean_bdev_aliases() that invalidates bdev aliases underlying a single buffer head. However this has caused a performance regression for bonnie++ benchmark on ext4 filesystem when delayed allocation is turned off (ext3 mode) - average of 3 runs: Hmean SeqOut Char 164787.55 ( 0.00%) 107189.06 (-34.95%) Hmean SeqOut Block 219883.89 ( 0.00%) 168870.32 (-23.20%) The reason for this regression is that clean_bdev_aliases() is slower when called for a single block because pagevec_lookup() it uses will end up iterating through the radix tree until it finds a page (which may take a while) but we are only interested whether there's a page at a particular index. Fix the problem by using pagevec_lookup_range() instead which avoids the needless iteration. Fixes: e64855c6 ("fs: Add helper to clean bdev aliases under a bh and use it") Link: http://lkml.kernel.org/r/20170726114704.7626-5-jack@suse.czSigned-off-by:
-
Jan Kara authored
Implement a variant of find_get_pages() that stops iterating at given index. This may be substantial performance gain if the mapping is sparse. See following commit for details. Furthermore lots of users of this function (through pagevec_lookup()) actually want a range lookup and all of them are currently open-coding this. Also create corresponding pagevec_lookup_range() function. Link: http://lkml.kernel.org/r/20170726114704.7626-4-jack@suse.czSigned-off-by:
-
Jan Kara authored
Make pagevec_lookup() (and underlying find_get_pages()) update index to the next page where iteration should continue. Most callers want this and also pagevec_lookup_tag() already does this. Link: http://lkml.kernel.org/r/20170726114704.7626-3-jack@suse.czSigned-off-by:
-
Jan Kara authored
Patch series "Ranged pagevec lookup", v2. In this series I make pagevec_lookup() update the index (to be consistent with pagevec_lookup_tag() and also as a preparation for ranged lookups), provide ranged variant of pagevec_lookup() and use it in places where it makes sense. This not only removes some common code but is also a measurable performance win for some use cases (see patch 4/10) where radix tree is sparse and searching & grabing of a page after the end of the range has measurable overhead. This patch (of 10): The callback doesn't ever get called. Remove it. Link: http://lkml.kernel.org/r/20170726114704.7626-2-jack@suse.czSigned-off-by:
-
Michal Hocko authored
Tetsuo Handa has reported[1][2][3] that direct reclaimers might get stuck in too_many_isolated loop basically for ever because the last few pages on the LRU lists are isolated by the kswapd which is stuck on fs locks when doing the pageout or slab reclaim. This in turn means that there is nobody to actually trigger the oom killer and the system is basically unusable. too_many_isolated has been introduced by commit 35cd7815 ("vmscan: throttle direct reclaim when too many pages are isolated already") to prevent from pre-mature oom killer invocations because back then no reclaim progress could indeed trigger the OOM killer too early. But since the oom detection rework in commit 0a0337e0 ("mm, oom: rework oom detection") the allocation/reclaim retry loop considers all the reclaimable pages and throttles the allocation at that layer so we can loosen the direct reclaim throttling. Make shrink_inactive_list loop over too_many_isolated bounded and returns immediately when the situation hasn't resolved after the first sleep. Replace congestion_wait by a simple schedule_timeout_interruptible because we are not really waiting on the IO congestion in this path. Please note that this patch can theoretically cause the OOM killer to trigger earlier while there are many pages isolated for the reclaim which makes progress only very slowly. This would be obvious from the oom report as the number of isolated pages are printed there. If we ever hit this should_reclaim_retry should consider those numbers in the evaluation in one way or another. [1] http://lkml.kernel.org/r/201602092349.ACG81273.OSVtMJQHLOFOFF@I-love.SAKURA.ne.jp [2] http://lkml.kernel.org/r/201702212335.DJB30777.JOFMHSFtVLQOOF@I-love.SAKURA.ne.jp [3] http://lkml.kernel.org/r/201706300914.CEH95859.FMQOLVFHJFtOOS@I-love.SAKURA.ne.jp [mhocko@suse.com: switch to uninterruptible sleep] Link: http://lkml.kernel.org/r/20170724065048.GB25221@dhcp22.suse.cz Link: http://lkml.kernel.org/r/20170710074842.23175-1-mhocko@kernel.orgSigned-off-by:
Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Tested-by:
Mel Gorman <mgorman@suse.de> Acked-by:
Vlastimil Babka <vbabka@suse.cz> Acked-by:
-
Hui Zhu authored
Getting -EBUSY from zs_page_migrate will make migration slow (retry) or fail (zs_page_putback will schedule_work free_work, but it cannot ensure the success). I noticed this issue because my Kernel patched (https://lkml.org/lkml/2014/5/28/113) that will remove retry in __alloc_contig_migrate_range. This retry will handle the -EBUSY because it will re-isolate the page and re-call migrate_pages. Without it will make cma_alloc fail at once with -EBUSY. According to the review from Minchan Kim in https://lkml.org/lkml/2014/5/28/113, I update the patch to skip unnecessary loops but not return -EBUSY if zspage is not inuse. Following is what I got with highalloc-performance in a vbox with 2 cpu 1G memory 512 zram as swap. And the swappiness is set to 100. ori ne orig new Minor Faults 50805113 50830235 Major Faults 43918 56530 Swap Ins 42087 55680 Swap Outs 89718 104700 Allocation stalls 0 0 DMA allocs 57787 52364 DMA32 allocs 47964599 48043563 Normal allocs 0 0 Movable allocs 0 0 Direct pages scanned 45493 23167 Kswapd pages scanned 1565222 1725078 Kswapd pages reclaimed 1342222 1503037 Direct pages reclaimed 45615 25186 Kswapd efficiency 85% 87% Kswapd velocity 1897.101 1949.042 Direct efficiency 100% 108% Direct velocity 55.139 26.175 Percentage direct scans 2% 1% Zone normal velocity 1952.240 1975.217 Zone dma32 velocity 0.000 0.000 Zone dma velocity 0.000 0.000 Page writes by reclaim 89764.000 105233.000 Page writes file 46 533 Page writes anon 89718 104700 Page reclaim immediate 21457 3699 Sector Reads 3259688 3441368 Sector Writes 3667252 3754836 Page rescued immediate 0 0 Slabs scanned 1042872 1160855 Direct inode steals 8042 10089 Kswapd inode steals 54295 29170 Kswapd skipped wait 0 0 THP fault alloc 175 154 THP collapse alloc 226 289 THP splits 0 0 THP fault fallback 11 14 THP collapse fail 3 2 Compaction stalls 536 646 Compaction success 322 358 Compaction failures 214 288 Page migrate success 119608 111063 Page migrate failure 2723 2593 Compaction pages isolated 250179 232652 Compaction migrate scanned 9131832 9942306 Compaction free scanned 2093272 2613998 Compaction cost 192 189 NUMA alloc hit 47124555 47193990 NUMA alloc miss 0 0 NUMA interleave hit 0 0 NUMA alloc local 47124555 47193990 NUMA base PTE updates 0 0 NUMA huge PMD updates 0 0 NUMA page range updates 0 0 NUMA hint faults 0 0 NUMA hint local faults 0 0 NUMA hint local percent 100 100 NUMA pages migrated 0 0 AutoNUMA cost 0% 0% [akpm@linux-foundation.org: remove newline, per Minchan] Link: http://lkml.kernel.org/r/1500889535-19648-1-git-send-email-zhuhui@xiaomi.comSigned-off-by:
Hui Zhu <zhuhui@xiaomi.com> Acked-by:
Minchan Kim <minchan@kernel.org> Reviewed-by:
Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Signed-off-by:
-
Mel Gorman authored
Nadav Amit report zap_page_range only specifies that the caller protect the VMA list but does not specify whether it is held for read or write with callers using either. madvise holds mmap_sem for read meaning that a parallel zap operation can unmap PTEs which are then potentially skipped by madvise which potentially returns with stale TLB entries present. While the API could be extended, it would be a difficult API to use. This patch causes zap_page_range() to always consider flushing the full affected range. For small ranges or sparsely populated mappings, this may result in one additional spurious TLB flush. For larger ranges, it is possible that the TLB has already been flushed and the overhead is negligible. Either way, this approach is safer overall and avoids stale entries being present when madvise returns. This can be illustrated with the following program provided by Nadav Amit and slightly modified. With the patch applied, it has an exit code of 0 indicating a stale TLB entry did not leak to userspace. ---8<--- volatile int sync_step = 0; volatile char *p; static inline unsigned long rdtsc() { unsigned long hi, lo; __asm__ __volatile__ ("rdtsc" : "=a"(lo), "=d"(hi)); return lo | (hi << 32); } static inline void wait_rdtsc(unsigned long cycles) { unsigned long tsc = rdtsc(); while (rdtsc() - tsc < cycles); } void *big_madvise_thread(void *ign) { sync_step = 1; while (sync_step != 2); madvise((void*)p, PAGE_SIZE * N_PAGES, MADV_DONTNEED); } int main(void) { pthread_t aux_thread; p = mmap(0, PAGE_SIZE * N_PAGES, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0); memset((void*)p, 8, PAGE_SIZE * N_PAGES); pthread_create(&aux_thread, NULL, big_madvise_thread, NULL); while (sync_step != 1); *p = 8; // Cache in TLB sync_step = 2; wait_rdtsc(100000); madvise((void*)p, PAGE_SIZE, MADV_DONTNEED); printf("data: %d (%s)\n", *p, (*p == 8 ? "stale, broken" : "cleared, fine")); return *p == 8 ? -1 : 0; } ---8<--- Link: http://lkml.kernel.org/r/20170725101230.5v7gvnjmcnkzzql3@techsingularity.netSigned-off-by:Nadav Amit <nadav.amit@gmail.com> Cc: Andy Lutomirski <luto@kernel.org> Signed-off-by:
-
Punit Agrawal authored
When walking the page tables to resolve an address that points to !p*d_present() entry, huge_pte_offset() returns inconsistent values depending on the level of page table (PUD or PMD). It returns NULL in the case of a PUD entry while in the case of a PMD entry, it returns a pointer to the page table entry. A similar inconsitency exists when handling swap entries - returns NULL for a PUD entry while a pointer to the pte_t is retured for the PMD entry. Update huge_pte_offset() to make the behaviour consistent - return a pointer to the pte_t for hugepage or swap entries. Only return NULL in instances where we have a p*d_none() entry and the size parameter doesn't match the hugepage size at this level of the page table. Document the behaviour to clarify the expected behaviour of this function. This is to set clear semantics for architecture specific implementations of huge_pte_offset(). Discussions on the arm64 implementation of huge_pte_offset() (http://www.spinics.net/lists/linux-mm/msg133699.html) showed that there is benefit from returning a pte_t* in the case of p*d_none(). The fault handling code in hugetlb_fault() can handle p*d_none() entries and saves an extra round trip to huge_pte_alloc(). Other callers of huge_pte_offset() should be ok as well. [punit.agrawal@arm.com: v2] Link: http://lkml.kernel.org/r/20170725154114.24131-2-punit.agrawal@arm.comSigned-off-by:
Punit Agrawal <punit.agrawal@arm.com> Reviewed-by:
Catalin Marinas <catalin.marinas@arm.com> Reviewed-by:
-
Oliver O'Halloran authored
These functions are the only bits of generic code that use {pud,pmd}_pfn() without checking for CONFIG_TRANSPARENT_HUGEPAGE. This works fine on x86, the only arch with devmap support, since the *_pfn() functions are always defined there, but this isn't true for every architecture. Link: http://lkml.kernel.org/r/20170626063833.11094-1-oohall@gmail.comSigned-off-by:Oliver O'Halloran <oohall@gmail.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Signed-off-by:
-
Mike Kravetz authored
mremap will attempt to create a 'duplicate' mapping if old_size == 0 is specified. In the case of private mappings, mremap will actually create a fresh separate private mapping unrelated to the original. This does not fit with the design semantics of mremap as the intention is to create a new mapping based on the original. Therefore, return EINVAL in the case where an attempt is made to duplicate a private mapping. Also, print a warning message (once) if such an attempt is made. Link: http://lkml.kernel.org/r/cb9d9f6a-7095-582f-15a5-62643d65c736@oracle.comSigned-off-by:
-
Vlastimil Babka authored
init_pages_in_zone() is run under zone->lock, which means a long lock time and disabled interrupts on large machines. This is currently not an issue since it runs early in boot, but a later patch will change that. However, like other pfn scanners, we don't actually need zone->lock even when other cpus are running. The only potentially dangerous operation here is reading bogus buddy page owner due to race, and we already know how to handle that. The worst that can happen is that we skip some early allocated pages, which should not affect the debugging power of page_owner noticeably. Link: http://lkml.kernel.org/r/20170720134029.25268-4-vbabka@suse.czSigned-off-by:
-
Vlastimil Babka authored
page_ext_init() can take long on large machines, so add a cond_resched() point after each section is processed. This will allow moving the init to a later point at boot without triggering lockup reports. Link: http://lkml.kernel.org/r/20170720134029.25268-3-vbabka@suse.czSigned-off-by:
-
Vlastimil Babka authored
In init_pages_in_zone() we currently use the generic set_page_owner() function to initialize page_owner info for early allocated pages. This means we needlessly do lookup_page_ext() twice for each page, and more importantly save_stack(), which has to unwind the stack and find the corresponding stack depot handle. Because the stack is always the same for the initialization, unwind it once in init_pages_in_zone() and reuse the handle. Also avoid the repeated lookup_page_ext(). This can significantly reduce boot times with page_owner=on on large machines, especially for kernels built without frame pointer, where the stack unwinding is noticeably slower. [vbabka@suse.cz: don't duplicate code of __set_page_owner(), per Michal Hocko] [akpm@linux-foundation.org: coding-style fixes] [vbabka@suse.cz: create statically allocated fake stack trace for early allocated pages, per Michal] Link: http://lkml.kernel.org/r/45813564-2342-fc8d-d31a-f4b68a724325@suse.cz Link: http://lkml.kernel.org/r/20170720134029.25268-2-vbabka@suse.czSigned-off-by:
-
Michal Hocko authored
Commit f52407ce ("memory hotplug: alloc page from other node in memory online") has introduced N_HIGH_MEMORY checks to only use NUMA aware allocations when there is some memory present because the respective node might not have any memory yet at the time and so it could fail or even OOM. Things have changed since then though. Zonelists are now always initialized before we do any allocations even for hotplug (see 959ecc48 ("mm/memory_hotplug.c: fix building of node hotplug zonelist")). Therefore these checks are not really needed. In fact caller of the allocator should never care about whether the node is populated because that might change at any time. Link: http://lkml.kernel.org/r/20170721143915.14161-10-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
zonelists_mutex was introduced by commit 4eaf3f64 ("mem-hotplug: fix potential race while building zonelist for new populated zone") to protect zonelist building from races. This is no longer needed though because both memory online and offline are fully serialized. New users have grown since then. Notably setup_per_zone_wmarks wants to prevent from races between memory hotplug, khugepaged setup and manual min_free_kbytes update via sysctl (see cfd3da1e ("mm: Serialize access to min_free_kbytes"). Let's add a private lock for that purpose. This will not prevent from seeing halfway through memory hotplug operation but that shouldn't be a big deal becuse memory hotplug will update watermarks explicitly so we will eventually get a full picture. The lock just makes sure we won't race when updating watermarks leading to weird results. Also __build_all_zonelists manipulates global data so add a private lock for it as well. This doesn't seem to be necessary today but it is more robust to have a lock there. While we are at it make sure we document that memory online/offline depends on a full serialization either via mem_hotplug_begin() or device_lock. Link: http://lkml.kernel.org/r/20170721143915.14161-9-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
build_all_zonelists has been (ab)using stop_machine to make sure that zonelists do not change while somebody is looking at them. This is is just a gross hack because a) it complicates the context from which we can call build_all_zonelists (see 3f906ba2 ("mm/memory-hotplug: switch locking to a percpu rwsem")) and b) is is not really necessary especially after "mm, page_alloc: simplify zonelist initialization" and c) it doesn't really provide the protection it claims (see below). Updates of the zonelists happen very seldom, basically only when a zone becomes populated during memory online or when it loses all the memory during offline. A racing iteration over zonelists could either miss a zone or try to work on one zone twice. Both of these are something we can live with occasionally because there will always be at least one zone visible so we are not likely to fail allocation too easily for example. Please note that the original stop_machine approach doesn't really provide a better exclusion because the iteration might be interrupted half way (unless the whole iteration is preempt disabled which is not the case in most cases) so the some zones could still be seen twice or a zone missed. I have run the pathological online/offline of the single memblock in the movable zone while stressing the same small node with some memory pressure. Node 1, zone DMA pages free 0 min 0 low 0 high 0 spanned 0 present 0 managed 0 protection: (0, 943, 943, 943) Node 1, zone DMA32 pages free 227310 min 8294 low 10367 high 12440 spanned 262112 present 262112 managed 241436 protection: (0, 0, 0, 0) Node 1, zone Normal pages free 0 min 0 low 0 high 0 spanned 0 present 0 managed 0 protection: (0, 0, 0, 1024) Node 1, zone Movable pages free 32722 min 85 low 117 high 149 spanned 32768 present 32768 managed 32768 protection: (0, 0, 0, 0) root@test1:/sys/devices/system/node/node1# while true do echo offline > memory34/state echo online_movable > memory34/state done root@test1:/mnt/data/test/linux-3.7-rc5# numactl --preferred=1 make -j4 and it survived without any unexpected behavior. While this is not really a great testing coverage it should exercise the allocation path quite a lot. Link: http://lkml.kernel.org/r/20170721143915.14161-8-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
build_zonelists gradually builds zonelists from the nearest to the most distant node. As we do not know how many populated zones we will have in each node we rely on the _zoneref to terminate initialized part of the zonelist by a NULL zone. While this is functionally correct it is quite suboptimal because we cannot allow updaters to race with zonelists users because they could see an empty zonelist and fail the allocation or hit the OOM killer in the worst case. We can do much better, though. We can store the node ordering into an already existing node_order array and then give this array to build_zonelists_in_node_order and do the whole initialization at once. zonelists consumers still might see halfway initialized state but that should be much more tolerateable because the list will not be empty and they would either see some zone twice or skip over some zone(s) in the worst case which shouldn't lead to immediate failures. While at it let's simplify build_zonelists_node which is rather confusing now. It gets an index into the zoneref array and returns the updated index for the next iteration. Let's rename the function to build_zonerefs_node to better reflect its purpose and give it zoneref array to update. The function doesn't the index anymore. It just returns the number of added zones so that the caller can advance the zonered array start for the next update. This patch alone doesn't introduce any functional change yet, though, it is merely a preparatory work for later changes. Link: http://lkml.kernel.org/r/20170721143915.14161-7-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
try_online_node calls hotadd_new_pgdat which already calls build_all_zonelists. So the additional call is redundant. Even though hotadd_new_pgdat will only initialize zonelists of the new node this is the right thing to do because such a node doesn't have any memory so other zonelists would ignore all the zones from this node anyway. Link: http://lkml.kernel.org/r/20170721143915.14161-6-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
build_all_zonelists gets a zone parameter to initialize zone's pagesets. There is only a single user which gives a non-NULL zone parameter and that one doesn't really need the rest of the build_all_zonelists (see commit 6dcd73d7 ("memory-hotplug: allocate zone's pcp before onlining pages")). Therefore remove setup_zone_pageset from build_all_zonelists and call it from its only user directly. This will also remove a pointless zonlists rebuilding which is always good. Link: http://lkml.kernel.org/r/20170721143915.14161-5-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
__build_all_zonelists reinitializes each online cpu local node for CONFIG_HAVE_MEMORYLESS_NODES. This makes sense because previously memory less nodes could gain some memory during memory hotplug and so the local node should be changed for CPUs close to such a node. It makes less sense to do that unconditionally for a newly creaded NUMA node which is still offline and without any memory. Let's also simplify the cpu loop and use for_each_online_cpu instead of an explicit cpu_online check for all possible cpus. Link: http://lkml.kernel.org/r/20170721143915.14161-4-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
boot_pageset is a boot time hack which gets superseded by normal pagesets later in the boot process. It makes zero sense to reinitialize it again and again during memory hotplug. Link: http://lkml.kernel.org/r/20170721143915.14161-3-mhocko@kernel.orgSigned-off-by:
-
Michal Hocko authored
Patch series "cleanup zonelists initialization", v1. This is aimed at cleaning up the zonelists initialization code we have but the primary motivation was bug report [2] which got resolved but the usage of stop_machine is just too ugly to live. Most patches are straightforward but 3 of them need a special consideration. Patch 1 removes zone ordered zonelists completely. I am CCing linux-api because this is a user visible change. As I argue in the patch description I do not think we have a strong usecase for it these days. I have kept sysctl in place and warn into the log if somebody tries to configure zone lists ordering. If somebody has a real usecase for it we can revert this patch but I do not expect anybody will actually notice runtime differences. This patch is not strictly needed for the rest but it made patch 6 easier to implement. Patch 7 removes stop_machine from build_all_zonelists without adding any special synchronization between iterators and updater which I _believe_ is acceptable as explained in the changelog. I hope I am not missing anything. Patch 8 then removes zonelists_mutex which is kind of ugly as well and not really needed AFAICS but a care should be taken when double checking my thinking. This patch (of 9): Supporting zone ordered zonelists costs us just a lot of code while the usefulness is arguable if existent at all. Mel has already made node ordering default on 64b systems. 32b systems are still using ZONELIST_ORDER_ZONE because it is considered better to fallback to a different NUMA node rather than consume precious lowmem zones. This argument is, however, weaken by the fact that the memory reclaim has been reworked to be node rather than zone oriented. This means that lowmem requests have to skip over all highmem pages on LRUs already and so zone ordering doesn't save the reclaim time much. So the only advantage of the zone ordering is under a light memory pressure when highmem requests do not ever hit into lowmem zones and the lowmem pressure doesn't need to reclaim. Considering that 32b NUMA systems are rather suboptimal already and it is generally advisable to use 64b kernel on such a HW I believe we should rather care about the code maintainability and just get rid of ZONELIST_ORDER_ZONE altogether. Keep systcl in place and warn if somebody tries to set zone ordering either from kernel command line or the sysctl. [mhocko@suse.com: reading vm.numa_zonelist_order will never terminate] Link: http://lkml.kernel.org/r/20170721143915.14161-2-mhocko@kernel.orgSigned-off-by:
-
Minchan Kim authored
This patch adds document and kconfig for using of writeback feature. Link: http://lkml.kernel.org/r/1498459987-24562-10-git-send-email-minchan@kernel.orgSigned-off-by:
-
Minchan Kim authored
This patch enables read IO from backing device. For the feature, it implements two IO read functions to transfer data from backing storage. One is asynchronous IO function and other is synchronous one. A reason I need synchrnous IO is due to partial write which need to complete read IO before the overwriting partial data. We can make the partial IO's case asynchronous, too but at the moment, I don't feel adding more complexity to support such rare use cases so want to go with simple. [xieyisheng1@huawei.com: read_from_bdev_async(): return 1 to avoid call page_endio() in zram_rw_page()] Link: http://lkml.kernel.org/r/1502707447-6944-1-git-send-email-xieyisheng1@huawei.com Link: http://lkml.kernel.org/r/1498459987-24562-9-git-send-email-minchan@kernel.orgSigned-off-by:
Yisheng Xie <xieyisheng1@huawei.com> Cc: Juneho Choi <juno.choi@lge.com> Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Signed-off-by:
-
Minchan Kim authored
This patch enables write IO to transfer data to backing device. For that, it implements write_to_bdev function which creates new bio and chaining with parent bio to make the parent bio asynchrnous. For rw_page which don't have parent bio, it submit owned bio and handle IO completion by zram_page_end_io. Also, this patch defines new flag ZRAM_WB to mark written page for later read IO. [xieyisheng1@huawei.com: fix typo in comment] Link: http://lkml.kernel.org/r/1502707447-6944-2-git-send-email-xieyisheng1@huawei.com Link: http://lkml.kernel.org/r/1498459987-24562-8-git-send-email-minchan@kernel.orgSigned-off-by:
-
