- 13 Nov, 2013 40 commits
-
-
Bob Liu authored
Khugepaged will scan/free HPAGE_PMD_NR normal pages and replace with a hugepage which is allocated from the node of the first scanned normal page, but this policy is too rough and may end with unexpected result to upper users. The problem is the original page-balancing among all nodes will be broken after hugepaged started. Thinking about the case if the first scanned normal page is allocated from node A, most of other scanned normal pages are allocated from node B or C.. But hugepaged will always allocate hugepage from node A which will cause extra memory pressure on node A which is not the situation before khugepaged started. This patch try to fix this problem by making khugepaged allocate hugepage from the node which have max record of scaned normal pages hit, so that the effect to original page-balancing can be minimized. The other problem is if normal scanned pages are equally allocated from Node A,B and C, after khugepaged started Node A will still suffer extra memory pressure. Andrew Davidoff reported a related issue several days ago. He wanted his application interleaving among all nodes and "numactl --interleave=all ./test" was used to run the testcase, but the result wasn't not as expected. cat /proc/2814/numa_maps: 7f50bd440000 interleave:0-3 anon=51403 dirty=51403 N0=435 N1=435 N2=435 N3=50098 The end result showed that most pages are from Node3 instead of interleave among node0-3 which was unreasonable. This patch also fix this issue by allocating hugepage round robin from all nodes have the same record, after this patch the result was as expected: 7f78399c0000 interleave:0-3 anon=51403 dirty=51403 N0=12723 N1=12723 N2=13235 N3=12722 The simple testcase is like this: int main() { char *p; int i; int j; for (i=0; i < 200; i++) { p = (char *)malloc(1048576); printf("malloc done\n"); if (p == 0) { printf("Out of memory\n"); return 1; } for (j=0; j < 1048576; j++) { p[j] = 'A'; } printf("touched memory\n"); sleep(1); } printf("enter sleep\n"); while(1) { sleep(100); } } [akpm@linux-foundation.org: make last_khugepaged_target_node local to khugepaged_find_target_node()] Reported-by:Andrew Davidoff <davidoff@qedmf.net> Tested-by:
Bob Liu <bob.liu@oracle.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com> Cc: Wanpeng Li <liwanp@linux.vnet.ibm.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Bob Liu authored
Move alloc_hugepage() to a better place, no need for a seperate #ifndef CONFIG_NUMA Signed-off-by:
Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com> Acked-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andrew Davidoff <davidoff@qedmf.net> Cc: Wanpeng Li <liwanp@linux.vnet.ibm.com> Signed-off-by:
-
Christian Hesse authored
Signed-off-by:
Christian Hesse <mail@eworm.de> Acked-by:
Seth Jennings <sjenning@linux.vnet.ibm.com> Signed-off-by:
-
Wanpeng Li authored
Don't warn twice in __vmalloc_area_node and __vmalloc_node_range if __vmalloc_area_node allocation failure. This patch reverts commit 46c001a2 ("mm/vmalloc.c: emit the failure message before return"). Signed-off-by:
Wanpeng Li <liwanp@linux.vnet.ibm.com> Reviewed-by:
Zhang Yanfei <zhangyanfei@cn.fujitsu.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Mitsuo Hayasaka <mitsuo.hayasaka.hu@hitachi.com> Cc: David Rientjes <rientjes@google.com> Signed-off-by:
-
Wanpeng Li authored
The VM_UNINITIALIZED/VM_UNLIST flag introduced by f5252e00 ("mm: avoid null pointer access in vm_struct via /proc/vmallocinfo") is used to avoid accessing the pages field with unallocated page when show_numa_info() is called. This patch moves the check just before show_numa_info in order that some messages still can be dumped via /proc/vmallocinfo. This patch reverts commit d157a558 ("mm/vmalloc.c: check VM_UNINITIALIZED flag in s_show instead of show_numa_info"); Reviewed-by:
-
Wanpeng Li authored
There is a race window between vmap_area tear down and show vmap_area information. A B remove_vm_area spin_lock(&vmap_area_lock); va->vm = NULL; va->flags &= ~VM_VM_AREA; spin_unlock(&vmap_area_lock); spin_lock(&vmap_area_lock); if (va->flags & (VM_LAZY_FREE | VM_LAZY_FREEZING)) return 0; if (!(va->flags & VM_VM_AREA)) { seq_printf(m, "0x%pK-0x%pK %7ld vm_map_ram\n", (void *)va->va_start, (void *)va->va_end, va->va_end - va->va_start); return 0; } free_unmap_vmap_area(va); flush_cache_vunmap free_unmap_vmap_area_noflush unmap_vmap_area free_vmap_area_noflush va->flags |= VM_LAZY_FREE The assumption !VM_VM_AREA represents vm_map_ram allocation is introduced by d4033afd ("mm, vmalloc: iterate vmap_area_list, instead of vmlist, in vmallocinfo()"). However, !VM_VM_AREA also represents vmap_area is being tear down in race window mentioned above. This patch fix it by don't dump any information for !VM_VM_AREA case and also remove (VM_LAZY_FREE | VM_LAZY_FREEING) check since they are not possible for !VM_VM_AREA case. Suggested-by:Joonsoo Kim <iamjoonsoo.kim@lge.com> Acked-by:
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Signed-off-by:
-
Wanpeng Li authored
The caller address has already been set in set_vmalloc_vm(), there's no need to set it again in __vmalloc_area_node. Reviewed-by:
-
David Rientjes authored
mpol_to_str() should not fail. Currently, it either fails because the string buffer is too small or because a string hasn't been defined for a mempolicy mode. If a new mempolicy mode is introduced and no string is defined for it, just warn and return "unknown". If the buffer is too small, just truncate the string and return, the same behavior as snprintf(). This also fixes a bug where there was no NULL-byte termination when doing *p++ = '=' and *p++ ':' and maxlen has been reached. Signed-off-by:
David Rientjes <rientjes@google.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Chen Gang <gang.chen@asianux.com> Cc: Rik van Riel <riel@redhat.com> Cc: Dave Jones <davej@redhat.com> Signed-off-by:
-
Jianguo Wu authored
Use more appropriate NUMA_NO_NODE instead of -1 in all archs' module_alloc() Signed-off-by:
Jianguo Wu <wujianguo@huawei.com> Acked-by:
-
Naoya Horiguchi authored
Chen Gong pointed out that set/unset_migratetype_isolate() was done in different functions in mm/memory-failure.c, which makes the code less readable/maintainable. So this patch does it in soft_offline_page(). With this patch, we get to hold lock_memory_hotplug() longer but it's not a problem because races between memory hotplug and soft offline are very rare. Signed-off-by:
Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Reviewed-by:
Chen, Gong <gong.chen@linux.intel.com> Acked-by:
Andi Kleen <ak@linux.intel.com> Reviewed-by:
-
Toshi Kani authored
cpu_up() has #ifdef CONFIG_MEMORY_HOTPLUG code blocks, which call mem_online_node() to put its node online if offlined and then call build_all_zonelists() to initialize the zone list. These steps are specific to memory hotplug, and should be managed in mm/memory_hotplug.c. lock_memory_hotplug() should also be held for the whole steps. For this reason, this patch replaces mem_online_node() with try_online_node(), which performs the whole steps with lock_memory_hotplug() held. try_online_node() is named after try_offline_node() as they have similar purpose. There is no functional change in this patch. Signed-off-by:
Toshi Kani <toshi.kani@hp.com> Reviewed-by:
-
Robin Holt authored
On large memory machines it can take a few minutes to get through free_all_bootmem(). Currently, when free_all_bootmem() calls __free_pages_memory(), the number of contiguous pages that __free_pages_memory() passes to the buddy allocator is limited to BITS_PER_LONG. BITS_PER_LONG was originally chosen to keep things similar to mm/nobootmem.c. But it is more efficient to limit it to MAX_ORDER. base new change 8TB 202s 172s 30s 16TB 401s 351s 50s That is around 1%-3% improvement on total boot time. This patch was spun off from the boot time rfc Robin and I had been working on. Signed-off-by:Robin Holt <robin.m.holt@gmail.com> Signed-off-by:
Nathan Zimmer <nzimmer@sgi.com> Cc: Robin Holt <robinmholt@linux.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Mike Travis <travis@sgi.com> Cc: Yinghai Lu <yinghai@kernel.org> Cc: Mel Gorman <mgorman@suse.de> Acked-by:
Johannes Weiner <hannes@cmpxchg.org> Reviewed-by:
-
Qiang Huang authored
Use helper function to check if we need to deal with oom condition. Signed-off-by:
Qiang Huang <h.huangqiang@huawei.com> Acked-by:
-
Xishi Qiu authored
Use "pfn_to_nid(pfn)" instead of "page_to_nid(pfn_to_page(pfn))". Signed-off-by:
Xishi Qiu <qiuxishi@huawei.com> Acked-by:
-
Xishi Qiu authored
A is_memblock_offlined() return or 1 means memory block is offlined, but is_memblock_offlined_cb() returning 1 means memory block is not offlined, this will confuse somebody, so rename the function. Signed-off-by:
-
Xishi Qiu authored
Use "if (zone->present_pages)" instead of "if (zone->present_pages)". Simplify the code, no functional change. Signed-off-by:
-
Xishi Qiu authored
Use "pgdat_end_pfn()" instead of "pgdat->node_start_pfn + pgdat->node_spanned_pages". Simplify the code, no functional change. Signed-off-by:
-
Xishi Qiu authored
Use "pgdat_end_pfn()" instead of "pgdat->node_start_pfn + pgdat->node_spanned_pages". Simplify the code, no functional change. Signed-off-by:
-
Jianguo Wu authored
Since commit 13ece886 ("thp: transparent hugepage config choice"), transparent hugepage support is disabled by default, and TRANSPARENT_HUGEPAGE_ALWAYS is configured when TRANSPARENT_HUGEPAGE=y. And since commit d39d33c3 ("thp: enable direct defrag"), defrag is enable for all transparent hugepage page faults by default, not only in MADV_HUGEPAGE regions. Signed-off-by:
-
Naoya Horiguchi authored
The callers of free_pgd_range() and hugetlb_free_pgd_range() don't hold page table locks. The comments seems to be obsolete, so let's remove them. Signed-off-by:
-
Xishi Qiu authored
Use __free_reserved_page() to simplify the code in the others. Signed-off-by:
-
Xishi Qiu authored
Use __free_reserved_page() to simplify the code in arch. It used split_page() in consistent_alloc()/__dma_alloc_coherent()/dma_alloc_coherent(), so page->_count == 1, and we can free it safely. __free_reserved_page() ClearPageReserved() init_page_count() // it won't change the value __free_page() Signed-off-by:
-
Jerome Marchand authored
Since commit f40d1e42 ("mm: compaction: acquire the zone->lock as late as possible"), isolate_freepages_block() takes the zone->lock itself. The function description however still states that the zone->lock must be held. This patch removes this outdated statement. Signed-off-by:
Jerome Marchand <jmarchan@redhat.com> Cc: Mel Gorman <mel@csn.ul.ie> Signed-off-by:
-
Jianguo Wu authored
Use more appropriate "if (node == NUMA_NO_NODE)" instead of "if (node < 0)" Signed-off-by:
-
Joe Perches authored
kcalloc returns zeroed memory. There's no need to use this flag. Signed-off-by:
Joe Perches <joe@perches.com> Reviewed-by:
-
Andrew Morton authored
The callee force_page_cache_readahead() already does this and unlike do_readahead(), force_page_cache_readahead() remembers to check for ->readpages() as well. Signed-off-by:
-
Jan Kara authored
Ocfs2 doesn't do data journalling. Thus its ->invalidatepage and ->releasepage functions never get called on buffers that have journal heads attached. So just use standard variants of functions from buffer.c. Signed-off-by:
Jan Kara <jack@suse.cz> Cc: Joel Becker <jlbec@evilplan.org> Cc: Mark Fasheh <mfasheh@suse.com> Signed-off-by:
-
Joe Perches authored
This typedef is unnecessary and should just be removed. Signed-off-by:
Joel Becker <jlbec@evilplan.org> Signed-off-by:
-
Xue jiufei authored
When ocfs2_write_cluster_by_desc() failed in ocfs2_write_begin_nolock() because of ENOSPC, it goes to out_quota, freeing data_ac(meta_ac). Then it calls ocfs2_try_to_free_truncate_log() to free space. If enough space freed, it will try to write again. Unfortunately, some error happenes before ocfs2_lock_allocators(), it goes to out and free data_ac(meta_ac) again. Signed-off-by:
joyce <xuejiufei@huawei.com> Reviewed-by:
Jie Liu <jeff.liu@oracle.com> Acked-by:
-
Younger Liu authored
If the file is not regular or writeable, it should return errno(EPERM). This patch is based on 85a258b7 ("ocfs2: fix error handling in ocfs2_ioctl_move_extents()"). Signed-off-by:
Younger Liu <younger.liu@huawei.com> Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by:
-
Younger Liu authored
If group_bh is not initialized, there is no need to release. This problem does not cause anything wrong, but the patch would make the code more logical. Signed-off-by:
-
Younger Liu authored
If ocfs2_journal_access_di() fails, group->bg_next_group should rollback. Otherwise, there would be a inconsistency between group_bh and main_bm_bh. Signed-off-by:
-
Junxiao Bi authored
Signed-off-by:
Junxiao Bi <junxiao.bi@oracle.com> Signed-off-by:
-
Akinobu Mita authored
We already have find_last_bit(). So just use it as described in the comment. Signed-off-by:
Akinobu Mita <akinobu.mita@gmail.com> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by:
-
Xue jiufei authored
We trigger a bug in __dlm_lockres_reserve_ast() when we parallel umount 4 nodes. The situation is as follows: 1) Node A migrate all lockres it owned(eg. lockres A) to other nodes say node B when it umounts. 2) Receiving MIG_LOCKRES message from A, Node B masters the lockres A with DLM_LOCK_RES_MIGRATING state set. 3) Then we umount ocfs2 on node B. It also should migrate lockres A to another node, say node C. But now, DLM_LOCK_RES_MIGRATING state of lockers A is not cleared. Node B triggered the BUG on lockres with state DLM_LOCK_RES_MIGRATING. Signed-off-by:
-
Xue jiufei authored
A parallel umount on 4 nodes triggered a bug in dlm_process_recovery_date(). Here's the situation: Receiving MIG_LOCKRES message, A node processes the locks in migratable lockres. It copys lvb from migratable lockres when processing the first valid lock. If there is a lock in the blocked list with the EX level, it triggers the BUG. Since valid lvbs are set when locks are granted with EX or PR levels, locks in the blocked list cannot have valid lvbs. Therefore I think we should skip the locks in the blocked list. Signed-off-by:
-
Akinobu Mita authored
Use bitmap_weight() instead of reinventing the wheel. Signed-off-by:
-
Joel Becker authored
-EDQUOT is a user-visible error, not a logic problem. Teach mlog_errno() to ignore it like it ignores -ENOSPC, etc. Signed-off-by:
Marek Królikowski <admin@wset.edu.pl> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by:
-
Rui Xiang authored
sb_getblk() may return an err, so add a check for bh. [joseph.qi@huawei.com: also add a check after calling sb_getblk() in ocfs2_create_xattr_block()] Signed-off-by:
Rui Xiang <rui.xiang@huawei.com> Reviewed-by:
Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Cc: Joseph Qi <joseph.qi@huawei.com> Signed-off-by:
-
Rui Xiang authored
The only reason for sb_getblk() failing is if it can't allocate the buffer_head. So return ENOMEM instead when it fails. [joseph.qi@huawei.com: ocfs2_symlink_get_block() and ocfs2_read_blocks_sync() and ocfs2_read_blocks() need the same change] Signed-off-by:
-
