- 07 Apr, 2014 40 commits
-
-
Sergey Senozhatsky authored
While fixing lockdep spew of ->init_lock reported by Sasha Levin [1], Minchan Kim noted [2] that it's better to move compression backend allocation (using GPF_KERNEL) out of the ->init_lock lock, same way as with zram_meta_alloc(), in order to prevent the same lockdep spew. [1] https://lkml.org/lkml/2014/2/27/337 [2] https://lkml.org/lkml/2014/3/3/32Signed-off-by:
Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Reported-by:
Minchan Kim <minchan@kernel.org> Acked-by:
Jerome Marchand <jmarchan@redhat.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Sergey Senozhatsky authored
Introduce LZ4 compression backend and make it available for selection. LZ4 support is optional and requires user to set ZRAM_LZ4_COMPRESS config option. The default compression backend is LZO. TEST (x86_64, core i5, 2 cores + 2 hyperthreading, zram disk size 1G, ext4 file system, 3 compression streams) iozone -t 3 -R -r 16K -s 60M -I +Z Test LZO LZ4 ---------------------------------------------- Initial write 1642744.62 1317005.09 Rewrite 2498980.88 1800645.16 Read 3957026.38 5877043.75 Re-read 3950997.38 5861847.00 Reverse Read 2937114.56 5047384.00 Stride read 2948163.19 4929587.38 Random read 3292692.69 4880793.62 Mixed workload 1545602.62 3502940.38 Random write 2448039.75 1758786.25 Pwrite 1670051.03 1338329.69 Pread 2530682.00 5097177.62 Fwrite 3232085.62 3275942.56 Fread 6306880.25 6645271.12 So on my system LZ4 is slower in write-only tests, while it performs better in read-only and mixed (reads + writes) tests. Official LZ4 benchmarks available here http://code.google.com/p/lz4/ (linux kernel uses revision r90). Signed-off-by: -
Sergey Senozhatsky authored
Add and document `comp_algorithm' device attribute. This attribute allows to show supported compression and currently selected compression algorithms: cat /sys/block/zram0/comp_algorithm [lzo] lz4 and change selected compression algorithm: echo lzo > /sys/block/zram0/comp_algorithm Signed-off-by:
-
Sergey Senozhatsky authored
This patch allows to change max_comp_streams on initialised zcomp. Introduce zcomp set_max_streams() knob, zcomp_strm_multi_set_max_streams() and zcomp_strm_single_set_max_streams() callbacks to change streams limit for zcomp_strm_multi and zcomp_strm_single, accordingly. set_max_streams for single steam zcomp does nothing. If user has lowered the limit, then zcomp_strm_multi_set_max_streams() attempts to immediately free extra streams (as much as it can, depending on idle streams availability). Note, this patch does not allow to change stream 'policy' from single to multi stream (or vice versa) on already initialised compression backend. Signed-off-by:
-
Sergey Senozhatsky authored
Existing zram (zcomp) implementation has only one compression stream (buffer and algorithm private part), so in order to prevent data corruption only one write (compress operation) can use this compression stream, forcing all concurrent write operations to wait for stream lock to be released. This patch changes zcomp to keep a compression streams list of user-defined size (via sysfs device attr). Each write operation still exclusively holds compression stream, the difference is that we can have N write operations (depending on size of streams list) executing in parallel. See TEST section later in commit message for performance data. Introduce struct zcomp_strm_multi and a set of functions to manage zcomp_strm stream access. zcomp_strm_multi has a list of idle zcomp_strm structs, spinlock to protect idle list and wait queue, making it possible to perform parallel compressions. The following set of functions added: - zcomp_strm_multi_find()/zcomp_strm_multi_release() find and release a compression stream, implement required locking - zcomp_strm_multi_create()/zcomp_strm_multi_destroy() create and destroy zcomp_strm_multi zcomp ->strm_find() and ->strm_release() callbacks are set during initialisation to zcomp_strm_multi_find()/zcomp_strm_multi_release() correspondingly. Each time zcomp issues a zcomp_strm_multi_find() call, the following set of operations performed: - spin lock strm_lock - if idle list is not empty, remove zcomp_strm from idle list, spin unlock and return zcomp stream pointer to caller - if idle list is empty, current adds itself to wait queue. it will be awaken by zcomp_strm_multi_release() caller. zcomp_strm_multi_release(): - spin lock strm_lock - add zcomp stream to idle list - spin unlock, wake up sleeper Minchan Kim reported that spinlock-based locking scheme has demonstrated a severe perfomance regression for single compression stream case, comparing to mutex-based (see https://lkml.org/lkml/2014/2/18/16) base spinlock mutex ==Initial write ==Initial write ==Initial write records: 5 records: 5 records: 5 avg: 1642424.35 avg: 699610.40 avg: 1655583.71 std: 39890.95(2.43%) std: 232014.19(33.16%) std: 52293.96 max: 1690170.94 max: 1163473.45 max: 1697164.75 min: 1568669.52 min: 573429.88 min: 1553410.23 ==Rewrite ==Rewrite ==Rewrite records: 5 records: 5 records: 5 avg: 1611775.39 avg: 501406.64 avg: 1684419.11 std: 17144.58(1.06%) std: 15354.41(3.06%) std: 18367.42 max: 1641800.95 max: 531356.78 max: 1706445.84 min: 1593515.27 min: 488817.78 min: 1655335.73 When only one compression stream available, mutex with spin on owner tends to perform much better than frequent wait_event()/wake_up(). This is why single stream implemented as a special case with mutex locking. Introduce and document zram device attribute max_comp_streams. This attr shows and stores current zcomp's max number of zcomp streams (max_strm). Extend zcomp's zcomp_create() with `max_strm' parameter. `max_strm' limits the number of zcomp_strm structs in compression backend's idle list (max_comp_streams). max_comp_streams used during initialisation as follows: -- passing to zcomp_create() max_strm equals to 1 will initialise zcomp using single compression stream zcomp_strm_single (mutex-based locking). -- passing to zcomp_create() max_strm greater than 1 will initialise zcomp using multi compression stream zcomp_strm_multi (spinlock-based locking). default max_comp_streams value is 1, meaning that zram with single stream will be initialised. Later patch will introduce configuration knob to change max_comp_streams on already initialised and used zcomp. TEST iozone -t 3 -R -r 16K -s 60M -I +Z test base 1 strm (mutex) 3 strm (spinlock) ----------------------------------------------------------------------- Initial write 589286.78 583518.39 718011.05 Rewrite 604837.97 596776.38 1515125.72 Random write 584120.11 595714.58 1388850.25 Pwrite 535731.17 541117.38 739295.27 Fwrite 1418083.88 1478612.72 1484927.06 Usage example: set max_comp_streams to 4 echo 4 > /sys/block/zram0/max_comp_streams show current max_comp_streams (default value is 1). cat /sys/block/zram0/max_comp_streams Signed-off-by:
-
Sergey Senozhatsky authored
This is preparation patch to add multi stream support to zcomp. Introduce struct zcomp_strm_single and a set of functions to manage zcomp_strm stream access. zcomp_strm_single implements single compession stream, same way as current zcomp implementation. This moves zcomp_strm stream control and locking from zcomp, so compressing backend zcomp is not aware of required locking. Single and multi streams require different locking schemes. Minchan Kim reported that spinlock-based locking scheme (which is used in multi stream implementation) has demonstrated a severe perfomance regression for single compression stream case, comparing to mutex-based. see https://lkml.org/lkml/2014/2/18/16 The following set of functions added: - zcomp_strm_single_find()/zcomp_strm_single_release() find and release a compression stream, implement required locking - zcomp_strm_single_create()/zcomp_strm_single_destroy() create and destroy zcomp_strm_single New ->strm_find() and ->strm_release() callbacks added to zcomp, which are set to zcomp_strm_single_find() and zcomp_strm_single_release() during initialisation. Instead of direct locking and zcomp_strm access from zcomp_strm_find() and zcomp_strm_release(), zcomp now calls ->strm_find() and ->strm_release() correspondingly. Signed-off-by:
-
Sergey Senozhatsky authored
Do not perform direct LZO compress/decompress calls, initialise and use zcomp LZO backend (single compression stream) instead. [akpm@linux-foundation.org: resolve conflicts with zram-delete-zram_init_device-fix.patch] Signed-off-by:
-
Sergey Senozhatsky authored
ZRAM performs direct LZO compression algorithm calls, making it the one and only option. While LZO is generally performs well, LZ4 algorithm tends to have a faster decompression (see http://code.google.com/p/lz4/ for full report) Name Ratio C.speed D.speed MB/s MB/s LZ4 (r101) 2.084 422 1820 LZO 2.06 2.106 414 600 Thus, users who have mostly read (decompress) usage scenarious or mixed workflow (writes with relatively high read ops number) will benefit from using LZ4 compression backend. Introduce compressing backend abstraction zcomp in order to support multiple compression algorithms with the following set of operations: .create .destroy .compress .decompress Schematically zram write() usually contains the following steps: 0) preparation (decompression of partioal IO, etc.) 1) lock buffer_lock mutex (protects meta compress buffers) 2) compress (using meta compress buffers) 3) alloc and map zs_pool object 4) copy compressed data (from meta compress buffers) to object allocated by 3) 5) free previous pool page, assign a new one 6) unlock buffer_lock mutex As we can see, compressing buffers must remain untouched from 1) to 4), because, otherwise, concurrent write() can overwrite data. At the same time, zram_meta must be aware of a) specific compression algorithm memory requirements and b) necessary locking to protect compression buffers. To remove requirement a) new struct zcomp_strm introduced, which contains a compress/decompress `buffer' and compression algorithm `private' part. While struct zcomp implements zcomp_strm stream handling and locking and removes requirement b) from zram meta. zcomp ->create() and ->destroy(), respectively, allocate and deallocate algorithm specific zcomp_strm `private' part. Every zcomp has zcomp stream and mutex to protect its compression stream. Stream usage semantics remains the same -- only one write can hold stream lock and use its buffers. zcomp_strm_find() turns caller into exclusive user of a stream (holding stream mutex until zram release stream), and zcomp_strm_release() makes zcomp stream available (unlock the stream mutex). Hence no concurrent write (compression) operations possible at the moment. iozone -t 3 -R -r 16K -s 60M -I +Z test base patched -------------------------------------------------- Initial write 597992.91 591660.58 Rewrite 609674.34 616054.97 Read 2404771.75 2452909.12 Re-read 2459216.81 2470074.44 Reverse Read 1652769.66 1589128.66 Stride read 2202441.81 2202173.31 Random read 2236311.47 2276565.31 Mixed workload 1423760.41 1709760.06 Random write 579584.08 615933.86 Pwrite 597550.02 594933.70 Pread 1703672.53 1718126.72 Fwrite 1330497.06 1461054.00 Fread 3922851.00 3957242.62 Usage examples: comp = zcomp_create(NAME) /* NAME e.g. "lzo" */ which initialises compressing backend if requested algorithm is supported. Compress: zstrm = zcomp_strm_find(comp) zcomp_compress(comp, zstrm, src, &dst_len) [..] /* copy compressed data */ zcomp_strm_release(comp, zstrm) Decompress: zcomp_decompress(comp, src, src_len, dst); Free compessing backend and its zcomp stream: zcomp_destroy(comp) Signed-off-by:
-
Sergey Senozhatsky authored
allocate new `zram_meta' in disksize_store() only for uninitialised zram device, saving a number of allocations and deallocations in case if disksize_store() was called on currently used device. at the same time zram_meta stack variable is not necessary, because we can set ->meta directly. there is also no need in setting QUEUE_FLAG_NONROT queue on every disksize_store(), set it once during device creation. [minchan@kernel.org: handle zram->meta alloc fail case] [minchan@kernel.org: prevent lockdep spew of init_lock] Signed-off-by:
-
Sergey Senozhatsky authored
Document `failed_reads' and `failed_writes' device attributes. Remove info about `discard' - there is no such zram attr. Signed-off-by:
-
Sergey Senozhatsky authored
Move zram warning about disksize and size of memory correlation to zram documentation. Signed-off-by:
-
Sergey Senozhatsky authored
struct table `count' member is not used. Signed-off-by:
-
Sergey Senozhatsky authored
zram accounted but did not report numbers of failed read and write queries. make these stats available as failed_reads and failed_writes attrs. Signed-off-by:
-
Sergey Senozhatsky authored
Introduce ZRAM_ATTR_RO macro that generates device_attribute and default ATTR show() function for existing atomic64_t zram stats. Signed-off-by:
-
Sergey Senozhatsky authored
This is a preparation patch for stats code duplication removal. 1) use atomic64_t for `pages_zero' and `pages_stored' zram stats. 2) `compr_size' and `pages_zero' struct zram_stats members did not follow the existing device attr naming scheme: zram_stats.ATTR has ATTR_show() function. rename them: -- compr_size -> compr_data_size -- pages_zero -> zero_pages Minchan Kim's note: If we really have trouble with atomic stat operation, we could change it with percpu_counter so that it could solve atomic overhead and unnecessary memory space by introducing unsigned long instead of 64bit atomic_t. Signed-off-by:
-
Sergey Senozhatsky authored
Remove `good' and `bad' compressed sub-requests stats. RW request may cause a number of RW sub-requests. zram used to account `good' compressed sub-queries (with compressed size less than 50% of original size), `bad' compressed sub-queries (with compressed size greater that 75% of original size), leaving sub-requests with compression size between 50% and 75% of original size not accounted and not reported. zram already accounts each sub-request's compression size so we can calculate real device compression ratio. Signed-off-by:
-
Sergey Senozhatsky authored
Do not pass rw argument down the __zram_make_request() -> zram_bvec_rw() chain, decode it in zram_bvec_rw() instead. Besides, this is the place where we distinguish READ and WRITE bio data directions, so account zram RW stats here, instead of __zram_make_request(). This also allows to account a real number of zram READ/WRITE operations, not just requests (single RW request may cause a number of zram RW ops with separate locking, compression/decompression, etc). Signed-off-by:
-
Sergey Senozhatsky authored
Introduce init_done() helper function which allows us to drop `init_done' struct zram member. init_done() uses the fact that ->init_done == 1 equals to ->meta != NULL. Signed-off-by:
-
John Hubbard authored
A new dump_page() routine was recently added, and marked EXPORT_SYMBOL_GPL. dump_page() was also added to the VM_BUG_ON_PAGE() macro, and so the end result is that non-GPL code can no longer call get_page() and a few other routines. This only happens if the kernel was compiled with CONFIG_DEBUG_VM. Change dump_page() to be EXPORT_SYMBOL. Longer explanation: Prior to commit 309381fe ("mm: dump page when hitting a VM_BUG_ON using VM_BUG_ON_PAGE") , it was possible to build MIT-licensed (non-GPL) drivers on Fedora. Fedora is semi-unique, in that it sets CONFIG_VM_DEBUG. Because Fedora sets CONFIG_VM_DEBUG, they end up pulling in dump_page(), via VM_BUG_ON_PAGE, via get_page(). As one of the authors of NVIDIA's new, open source, "UVM-Lite" kernel module, I originally choose to use the kernel's get_page() routine from within nvidia_uvm_page_cache.c, because get_page() has always seemed to be very clearly intended for use by non-GPL, driver code. So I'm hoping that making get_page() widely accessible again will not be too controversial. We did check with Fedora first, and they responded (https://bugzilla.redhat.com/show_bug.cgi?id=1074710#c3) that we should try to get upstream changed, before asking Fedora to change. Their reasoning seems beneficial to Linux: leaving CONFIG_DEBUG_VM set allows Fedora to help catch mm bugs. Signed-off-by:
John Hubbard <jhubbard@nvidia.com> Cc: Sasha Levin <sasha.levin@oracle.com> Cc: Josh Boyer <jwboyer@redhat.com> Signed-off-by:
-
Srikar Dronamraju authored
LAST_CPUPID_MASK is calculated using LAST_CPUPID_WIDTH. However LAST_CPUPID_WIDTH itself can be 0. (when LAST_CPUPID_NOT_IN_PAGE_FLAGS is set). In such a case LAST_CPUPID_MASK turns out to be 0. But with recent commit 1ae71d03: (mm: numa: bugfix for LAST_CPUPID_NOT_IN_PAGE_FLAGS) if LAST_CPUPID_MASK is 0, page_cpupid_xchg_last() and page_cpupid_reset_last() causes page->_last_cpupid to be set to 0. This causes performance regression. Its almost as if numa_balancing is off. Fix LAST_CPUPID_MASK by using LAST_CPUPID_SHIFT instead of LAST_CPUPID_WIDTH. Some performance numbers and perf stats with and without the fix. (3.14-rc6) ---------- numa01 Performance counter stats for '/usr/bin/time -f %e %S %U %c %w -o start_bench.out -a ./numa01': 12,27,462 cs [100.00%] 2,41,957 migrations [100.00%] 1,68,01,713 faults [100.00%] 7,99,35,29,041 cache-misses 98,808 migrate:mm_migrate_pages [100.00%] 1407.690148814 seconds time elapsed numa02 Performance counter stats for '/usr/bin/time -f %e %S %U %c %w -o start_bench.out -a ./numa02': 63,065 cs [100.00%] 14,364 migrations [100.00%] 2,08,118 faults [100.00%] 25,32,59,404 cache-misses 12 migrate:mm_migrate_pages [100.00%] 63.840827219 seconds time elapsed (3.14-rc6 with fix) ------------------- numa01 Performance counter stats for '/usr/bin/time -f %e %S %U %c %w -o start_bench.out -a ./numa01': 9,68,911 cs [100.00%] 1,01,414 migrations [100.00%] 88,38,697 faults [100.00%] 4,42,92,51,042 cache-misses 4,25,060 migrate:mm_migrate_pages [100.00%] 685.965331189 seconds time elapsed numa02 Performance counter stats for '/usr/bin/time -f %e %S %U %c %w -o start_bench.out -a ./numa02': 17,543 cs [100.00%] 2,962 migrations [100.00%] 1,17,843 faults [100.00%] 11,80,61,644 cache-misses 12,358 migrate:mm_migrate_pages [100.00%] 20.380132343 seconds time elapsed Signed-off-by:
Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Liu Ping Fan <pingfank@linux.vnet.ibm.com> Reviewed-by:
Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Mel Gorman <mel@csn.ul.ie> Signed-off-by:
-
Zhang Yanfei authored
s/MADV_NODUMP/MADV_DONTDUMP/ Signed-off-by:
Zhang Yanfei <zhangyanfei@cn.fujitsu.com> Signed-off-by:
-
Fabian Frederick authored
Commit f9acc8c7 ("readahead: sanify file_ra_state names") left ra_submit with a single function call. Move ra_submit to internal.h and inline it to save some stack. Thanks to Andrew Morton for commenting different versions. Signed-off-by:
Fabian Frederick <fabf@skynet.be> Suggested-by:
-
Mizuma, Masayoshi authored
When I decrease the value of nr_hugepage in procfs a lot, softlockup happens. It is because there is no chance of context switch during this process. On the other hand, when I allocate a large number of hugepages, there is some chance of context switch. Hence softlockup doesn't happen during this process. So it's necessary to add the context switch in the freeing process as same as allocating process to avoid softlockup. When I freed 12 TB hugapages with kernel-2.6.32-358.el6, the freeing process occupied a CPU over 150 seconds and following softlockup message appeared twice or more. $ echo 6000000 > /proc/sys/vm/nr_hugepages $ cat /proc/sys/vm/nr_hugepages 6000000 $ grep ^Huge /proc/meminfo HugePages_Total: 6000000 HugePages_Free: 6000000 HugePages_Rsvd: 0 HugePages_Surp: 0 Hugepagesize: 2048 kB $ echo 0 > /proc/sys/vm/nr_hugepages BUG: soft lockup - CPU#16 stuck for 67s! [sh:12883] ... Pid: 12883, comm: sh Not tainted 2.6.32-358.el6.x86_64 #1 Call Trace: free_pool_huge_page+0xb8/0xd0 set_max_huge_pages+0x128/0x190 hugetlb_sysctl_handler_common+0x113/0x140 hugetlb_sysctl_handler+0x1e/0x20 proc_sys_call_handler+0x97/0xd0 proc_sys_write+0x14/0x20 vfs_write+0xb8/0x1a0 sys_write+0x51/0x90 __audit_syscall_exit+0x265/0x290 system_call_fastpath+0x16/0x1b I have not confirmed this problem with upstream kernels because I am not able to prepare the machine equipped with 12TB memory now. However I confirmed that the amount of decreasing hugepages was directly proportional to the amount of required time. I measured required times on a smaller machine. It showed 130-145 hugepages decreased in a millisecond. Amount of decreasing Required time Decreasing rate hugepages (msec) (pages/msec) ------------------------------------------------------------ 10,000 pages == 20GB 70 - 74 135-142 30,000 pages == 60GB 208 - 229 131-144 It means decrement of 6TB hugepages will trigger softlockup with the default threshold 20sec, in this decreasing rate. Signed-off-by:
Masayoshi Mizuma <m.mizuma@jp.fujitsu.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Michal Hocko <mhocko@suse.cz> Cc: Wanpeng Li <liwanp@linux.vnet.ibm.com> Cc: Aneesh Kumar <aneesh.kumar@linux.vnet.ibm.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Fabian Frederick authored
Replace ((phys_addr_t)(x) << PAGE_SHIFT) by pfn macro. Signed-off-by:
-
Emil Medve authored
This is a small cleanup. Signed-off-by:
Emil Medve <Emilian.Medve@Freescale.com> Signed-off-by:
-
Miklos Szeredi authored
There's only one caller of set_page_dirty_balance() and that will call it with page_mkwrite == 0. The page_mkwrite argument was unused since commit b827e496 "mm: close page_mkwrite races". Signed-off-by:
Miklos Szeredi <mszeredi@suse.cz> Signed-off-by:
-
Vlastimil Babka authored
A BUG_ON(!PageLocked) was triggered in mlock_vma_page() by Sasha Levin fuzzing with trinity. The call site try_to_unmap_cluster() does not lock the pages other than its check_page parameter (which is already locked). The BUG_ON in mlock_vma_page() is not documented and its purpose is somewhat unclear, but apparently it serializes against page migration, which could otherwise fail to transfer the PG_mlocked flag. This would not be fatal, as the page would be eventually encountered again, but NR_MLOCK accounting would become distorted nevertheless. This patch adds a comment to the BUG_ON in mlock_vma_page() and munlock_vma_page() to that effect. The call site try_to_unmap_cluster() is fixed so that for page != check_page, trylock_page() is attempted (to avoid possible deadlocks as we already have check_page locked) and mlock_vma_page() is performed only upon success. If the page lock cannot be obtained, the page is left without PG_mlocked, which is again not a problem in the whole unevictable memory design. Signed-off-by:
Vlastimil Babka <vbabka@suse.cz> Signed-off-by:
Bob Liu <bob.liu@oracle.com> Reported-by:
Sasha Levin <sasha.levin@oracle.com> Cc: Wanpeng Li <liwanp@linux.vnet.ibm.com> Cc: Michel Lespinasse <walken@google.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Acked-by:
Rik van Riel <riel@redhat.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Hugh Dickins <hughd@google.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Johannes Weiner authored
On NUMA systems, a node may start thrashing cache or even swap anonymous pages while there are still free pages on remote nodes. This is a result of commits 81c0a2bb ("mm: page_alloc: fair zone allocator policy") and fff4068c ("mm: page_alloc: revert NUMA aspect of fair allocation policy"). Before those changes, the allocator would first try all allowed zones, including those on remote nodes, before waking any kswapds. But now, the allocator fastpath doubles as the fairness pass, which in turn can only consider the local node to prevent remote spilling based on exhausted fairness batches alone. Remote nodes are only considered in the slowpath, after the kswapds are woken up. But if remote nodes still have free memory, kswapd should not be woken to rebalance the local node or it may thrash cash or swap prematurely. Fix this by adding one more unfair pass over the zonelist that is allowed to spill to remote nodes after the local fairness pass fails but before entering the slowpath and waking the kswapds. This also gets rid of the GFP_THISNODE exemption from the fairness protocol because the unfair pass is no longer tied to kswapd, which GFP_THISNODE is not allowed to wake up. However, because remote spills can be more frequent now - we prefer them over local kswapd reclaim - the allocation batches on remote nodes could underflow more heavily. When resetting the batches, use atomic_long_read() directly instead of zone_page_state() to calculate the delta as the latter filters negative counter values. Signed-off-by:
Johannes Weiner <hannes@cmpxchg.org> Acked-by:
Mel Gorman <mgorman@suse.de> Cc: <stable@kernel.org> [3.12+] Signed-off-by:
-
Michal Hocko authored
mem_cgroup_newpage_charge is used only for charging anonymous memory so it is better to rename it to mem_cgroup_charge_anon. mem_cgroup_cache_charge is used for file backed memory so rename it to mem_cgroup_charge_file. Signed-off-by:
Michal Hocko <mhocko@suse.cz> Cc: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by:
-
Johannes Weiner authored
Some callsites pass a memcg directly, some callsites pass an mm that then has to be translated to a memcg. This makes for a terrible function interface. Just push the mm-to-memcg translation into the respective callsites and always pass a memcg to mem_cgroup_try_charge(). [mhocko@suse.cz: add charge mm helper] Signed-off-by:
-
Michal Hocko authored
__mem_cgroup_try_charge duplicates get_mem_cgroup_from_mm for charges which came without a memcg. The only reason seems to be a tiny optimization when css_tryget is not called if the charge can be consumed from the stock. Nevertheless css_tryget is very cheap since it has been reworked to use per-cpu counting so this optimization doesn't give us anything these days. So let's drop the code duplication so that the code is more readable. Signed-off-by:
-
Johannes Weiner authored
Instead of returning NULL from try_get_mem_cgroup_from_mm() when the mm owner is exiting, just return root_mem_cgroup. This makes sense for all callsites and gets rid of some of them having to fallback manually. [fengguang.wu@intel.com: fix warnings] Signed-off-by:
Fengguang Wu <fengguang.wu@intel.com> Acked-by:
-
Johannes Weiner authored
Users pass either a mm that has been established under task lock, or use a verified current->mm, which means the task can't be exiting. Signed-off-by:
-
Johannes Weiner authored
Only page cache charges can happen without an mm context, so push this special case out of the inner core and into the cache charge function. An ancient comment explains that the mm can also be NULL in case the task is currently being migrated, but that is not actually true with the current case, so just remove it. Signed-off-by:
-
Johannes Weiner authored
mem_cgroup_charge_common() is used by both cache and anon pages, but most of its body only applies to anon pages and the remainder is not worth having in a separate function. Signed-off-by:
-
Johannes Weiner authored
It used to disable preemption and run sanity checks but now it's only taking a number out of one percpu counter and putting it into another. Do this directly in the callsite and save the indirection. Signed-off-by:
-
Johannes Weiner authored
lock_page_cgroup() disables preemption, remove explicit preemption disabling for code paths holding this lock. Signed-off-by:
-
Kirill A. Shutemov authored
I tried to use 'dump_page(page, __func__)' for debugging, but it triggers warning: warning: passing argument 2 of `dump_page' discards `const' qualifier from pointer target type [enabled by default] Let's convert 'reason' to 'const char *' in dump_page() and friends: we shouldn't modify it anyway. Signed-off-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by:
-
Gioh Kim authored
vm_map_ram() has a fragmentation problem when it cannot purge a chunk(ie, 4M address space) if there is a pinning object in that addresss space. So it could consume all VMALLOC address space easily. We can fix the fragmentation problem by using vmap instead of vm_map_ram() but vmap() is known to be slow compared to vm_map_ram(). Minchan said vm_map_ram is 5 times faster than vmap in his tests. So I thought we should fix fragment problem of vm_map_ram because our proprietary GPU driver has used it heavily. On second thought, it's not an easy because we should reuse freed space for solving the problem and it could make more IPI and bitmap operation for searching hole. It could mitigate API's goal which is very fast mapping. And even fragmentation problem wouldn't show in 64 bit machine. Another option is that the user should separate long-life and short-life object and use vmap for long-life but vm_map_ram for short-life. If we inform the user about the characteristic of vm_map_ram the user can choose one according to the page lifetime. Let's add some notice messages to user. [akpm@linux-foundation.org: tweak comment text] Signed-off-by:
Gioh Kim <gioh.kim@lge.com> Reviewed-by:
-
Choi Gi-yong authored
Signed-off-by:
Choi Gi-yong <yong@gnoy.org> Signed-off-by:
-
