- 29 Apr, 2013 40 commits
-
-
Yasuaki Ishimatsu authored
When hot removing memory presented at boot time, following messages are shown: kernel BUG at mm/slub.c:3409! invalid opcode: 0000 [#1] SMP Modules linked in: ebtable_nat ebtables xt_CHECKSUM iptable_mangle bridge stp llc ipmi_devintf ipmi_msghandler sunrpc ipt_REJECT nf_conntrack_ipv4 nf_defrag_ipv4 iptable_filter ip_tables ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables binfmt_misc vfat fat dm_mirror dm_region_hash dm_log dm_mod vhost_net macvtap macvlan tun uinput iTCO_wdt iTCO_vendor_support coretemp kvm_intel kvm crc32c_intel ghash_clmulni_intel microcode pcspkr sg i2c_i801 lpc_ich mfd_core igb i2c_algo_bit i2c_core e1000e ptp pps_core tpm_infineon ioatdma dca sr_mod cdrom sd_mod crc_t10dif usb_storage megaraid_sas lpfc scsi_transport_fc scsi_tgt scsi_mod CPU 0 Pid: 5091, comm: kworker/0:2 Tainted: G W 3.9.0-rc6+ #15 RIP: kfree+0x232/0x240 Process kworker/0:2 (pid: 5091, threadinfo ffff88084678c000, task ffff88083928ca80) Call Trace: __release_region+0xd4/0xe0 __remove_pages+0x52/0x110 arch_remove_memory+0x89/0xd0 remove_memory+0xc4/0x100 acpi_memory_device_remove+0x6d/0xb1 acpi_device_remove+0x89/0xab __device_release_driver+0x7c/0xf0 device_release_driver+0x2f/0x50 acpi_bus_device_detach+0x6c/0x70 acpi_ns_walk_namespace+0x11a/0x250 acpi_walk_namespace+0xee/0x137 acpi_bus_trim+0x33/0x7a acpi_bus_hot_remove_device+0xc4/0x1a1 acpi_os_execute_deferred+0x27/0x34 process_one_work+0x1f7/0x590 worker_thread+0x11a/0x370 kthread+0xee/0x100 ret_from_fork+0x7c/0xb0 RIP [<ffffffff811c41d2>] kfree+0x232/0x240 RSP <ffff88084678d968> The reason why the messages are shown is to release a resource structure, allocated by bootmem, by kfree(). So when we release a resource structure, we should check whether it is allocated by bootmem or not. But even if we know a resource structure is allocated by bootmem, we cannot release it since SLxB cannot treat it. So for reusing a resource structure, this patch remembers it by using bootmem_resource as follows: When releasing a resource structure by free_resource(), free_resource() checks whether the resource structure is allocated by bootmem or not. If it is allocated by bootmem, free_resource() adds it to bootmem_resource. If it is not allocated by bootmem, free_resource() release it by kfree(). And when getting a new resource structure by get_resource(), get_resource() checks whether bootmem_resource has released resource structures or not. If there is a released resource structure, get_resource() returns it. If there is not a releaed resource structure, get_resource() returns new resource structure allocated by kzalloc(). [akpm@linux-foundation.org: s/get_resource/alloc_resource/] Signed-off-by:Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com> Reviewed-by:
Toshi Kani <toshi.kani@hp.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Ram Pai <linuxram@us.ibm.com> Cc: David Rientjes <rientjes@google.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Vinayak Menon authored
There are times when HIGHMEM is enabled, but we don't prefer CONFIG_BOUNCE to be enabled. CONFIG_BOUNCE can reduce the block device throughput, and this is not ideal for machines where we don't gain much by enabling it. So provide an option to deselect CONFIG_BOUNCE. The observation was made while measuring eMMC throughput using iozone on an ARM device with 1GB RAM. Signed-off-by:
Vinayak Menon <vinayakm.list@gmail.com> Cc: David Rientjes <rientjes@google.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: Russell King <rmk@arm.linux.org.uk> Signed-off-by:
-
Joonsoo Kim authored
Currently, we do memset() before reserving the area. This may not cause any problem, but it is somewhat weird. So change execution order. Signed-off-by:
Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Yinghai Lu <yinghai@kernel.org> Acked-by:
Johannes Weiner <hannes@cmpxchg.org> Cc: Jiang Liu <liuj97@gmail.com> Signed-off-by:
-
Joonsoo Kim authored
Remove unused argument and make function static, because there is no user outside of nobootmem.c Signed-off-by:
-
majianpeng authored
bh allocation uses kmem_cache_zalloc() so we needn't call 'init_buffer(bh, NULL, NULL)' and perform other set-zero-operations. Signed-off-by:
Jianpeng Ma <majianpeng@gmail.com> Cc: Jan Kara <jack@suse.cz> Cc: Theodore Ts'o <tytso@mit.edu> Signed-off-by:
-
Yasuaki Ishimatsu authored
When booting x86 system contains memoryless node, node numbers of CPUs on memoryless node were changed to nearest online node number by init_cpu_to_node() because the node is not online. In my system, node numbers of cpu#30-44 and 75-89 were changed from 2 to 0 as follows: $ numactl --hardware available: 2 nodes (0-1) node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 node 0 size: 32394 MB node 0 free: 27898 MB node 1 cpus: 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 node 1 size: 32768 MB node 1 free: 30335 MB If we hot add memory to memoryless node and offine/online all CPUs on the node, node numbers of these CPUs are changed to correct node numbers by srat_detect_node() because the node become online. In this case, node numbers of cpu#30-44 and 75-89 were changed from 0 to 2 in my system as follows: $ numactl --hardware available: 3 nodes (0-2) node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 node 0 size: 32394 MB node 0 free: 27218 MB node 1 cpus: 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 node 1 size: 32768 MB node 1 free: 30014 MB node 2 cpus: 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 node 2 size: 16384 MB node 2 free: 16384 MB But "cpu to node" and "node to cpu" links were not changed as follows: $ ls /sys/devices/system/cpu/cpu30/|grep node node0 $ ls /sys/devices/system/node/node0/|grep cpu30 cpu30 "numactl --hardware" shows that cpu30 belongs to node 2. But sysfs links does not change. This patch changes "cpu to node" and "node to cpu" links when node number changed by onlining CPU. Signed-off-by:
-
Randy Dunlap authored
PFN_PHYS() is a phys_addr_t, which can be u32 or u64. Fix the build warning when phys_addr_t is u32. mm/memory_hotplug.c: warning: format '%llx' expects argument of type 'long long unsigned int', but argument 2 has type 'unsigned int' [-Wformat]: => 1685:3 mm/memory_hotplug.c: warning: format '%llx' expects argument of type 'long long unsigned int', but argument 3 has type 'unsigned int' [-Wformat]: => 1685:3 Signed-off-by:
Randy Dunlap <rdunlap@infradead.org> Reported-by:
Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by:
-
Mel Gorman authored
As pointed out by Andrew Morton, the swap-over-NFS writeback is not setting PageWriteback before it is queued for direct IO. While swap pages do not participate in BDI or process dirty accounting and the IO is synchronous, the writeback bit is still required and not setting it in this case was an oversight. swapoff depends on the page writeback to synchronoise all pending writes on a swap page before it is reused. Swapcache freeing and reuse depend on checking the PageWriteback under lock to ensure the page is safe to reuse. Direct IO handlers and the direct IO handler for NFS do not deal with PageWriteback as they are synchronous writes. In the case of NFS, it schedules pages (or a page in the case of swap) for IO and then waits synchronously for IO to complete in nfs_direct_write(). It is recognised that this is a slowdown from normal swap handling which is asynchronous and uses a completion handler. Shoving PageWriteback handling down into direct IO handlers looks like a bad fit to handle the swap case although it may have to be dealt with some day if swap is converted to use direct IO in general and bmap is finally done away with. At that point it will be necessary to refit asynchronous direct IO with completion handlers onto the swap subsystem. As swapcache currently depends on PageWriteback to protect against races, this patch sets PageWriteback under the page lock before queueing it for direct IO. It is cleared when the direct IO handler returns. IO errors are treated similarly to the direct-to-bio case except PageError is not set as in the case of swap-over-NFS, it is likely to be a transient error. It was asked what prevents such a page being reclaimed in parallel. With this patch applied, such a page will now be skipped (most of the time) or blocked until the writeback completes. Reclaim checks PageWriteback under the page lock before calling try_to_free_swap and the page lock should prevent the page being requeued for IO before it is freed. This and Jerome's related patch should considered for -stable as far back as 3.6 when swap-over-NFS was introduced. [akpm@linux-foundation.org: use pr_err_ratelimited()] [akpm@linux-foundation.org: remove hopefully-unneeded cast in printk] Signed-off-by:
Mel Gorman <mgorman@suse.de> Cc: Jerome Marchand <jmarchan@redhat.com> Cc: Hugh Dickins <hughd@google.com> Cc: <stable@vger.kernel.org> [3.6+] Signed-off-by:
-
Jerome Marchand authored
Since commit 62c230bc ("mm: add support for a filesystem to activate swap files and use direct_IO for writing swap pages"), swap_writepage() calls direct_IO on swap files. However, in that case the page isn't redirtied if I/O fails, and is therefore handled afterwards as if it has been successfully written to the swap file, leading to memory corruption when the page is eventually swapped back in. This patch sets the page dirty when direct_IO() fails. It fixes a memory corruption that happened while using swap-over-NFS. Signed-off-by:
Jerome Marchand <jmarchan@redhat.com> Acked-by:
-
David Rientjes authored
A memcg may livelock when oom if the process that grabs the hierarchy's oom lock is never the first process with PF_EXITING set in the memcg's task iteration. The oom killer, both global and memcg, will defer if it finds an eligible process that is in the process of exiting and it is not being ptraced. The idea is to allow it to exit without using memory reserves before needlessly killing another process. This normally works fine except in the memcg case with a large number of threads attached to the oom memcg. In this case, the memcg oom killer only gets called for the process that grabs the hierarchy's oom lock; all others end up blocked on the memcg's oom waitqueue. Thus, if the process that grabs the hierarchy's oom lock is never the first PF_EXITING process in the memcg's task iteration, the oom killer is constantly deferred without anything making progress. The fix is to give PF_EXITING processes access to memory reserves so that we've marked them as oom killed without any iteration. This allows __mem_cgroup_try_charge() to succeed so that the process may exit. This makes the memcg oom killer exemption for TIF_MEMDIE tasks, now immediately granted for processes with pending SIGKILLs and those in the exit path, to be equivalent to what is done for the global oom killer. Signed-off-by:
David Rientjes <rientjes@google.com> Acked-by:
Michal Hocko <mhocko@suse.cz> Acked-by:
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Acked-by:
-
Kirill A. Shutemov authored
Current implementation of huge zero page uses pfn value 0 to indicate that the page hasn't allocated yet. It assumes that buddy page allocator can't return page with pfn == 0. Let's rework the code to store 'struct page *' of huge zero page, not its pfn. This way we can avoid the weak assumption. [akpm@linux-foundation.org: fix sparse warning] Signed-off-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reported-by:
Minchan Kim <minchan@kernel.org> Acked-by:
Andrea Arcangeli <aarcange@redhat.com> Acked-by:
-
Li Zefan authored
This might cause a use-after-free bug. Signed-off-by:
Li Zefan <lizefan@huawei.com> Cc: Glauber Costa <glommer@parallels.com> Acked-by:
-
Dmitry Monakhov authored
There are two convenient ways to report errors to userspace 1) retun error to original syscall for example write(2) 2) mark mapping with error flag and return it on later fsync(2) Second one is broken if (mapping->nrpages == 0) This is real-life situation because after error pages are likey to be truncated or invalidated. We have to return an error regardless to number of pages in the mapping. #Original testcase: git@github.com:dmonakhov/xfstests.git MOUNT_OPTIONS="-b1024" ./check shared/305 Signed-off-by:
Dmitry Monakhov <dmonakhov@openvz.org> Reviewed-by:
Jan Kara <jack@suse.cz> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by:
-
Tang Chen authored
There is no comment for parameter nid of memblock_insert_region(). This patch adds comment for it. Signed-off-by:
Tang Chen <tangchen@cn.fujitsu.com> Signed-off-by:
-
Tang Chen authored
nr_pages is not used in pages_correctly_reserved(). So remove it. Signed-off-by:
Wang Shilong <wangsl-fnst@cn.fujitsu.com> Reviewed-by:
Wen Congyang <wency@cn.fujitsu.com> Acked-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by:
-
Yasuaki Ishimatsu authored
When hot removing memory, a firmware_map_entry which has memory range of the memory is released by release_firmware_map_entry(). If the entry is allocated by bootmem, release_firmware_map_entry() adds the entry to map_entires_bootmem list when firmware_map_find_entry() finds the entry from map_entries list. But firmware_map_find_entry never find the entry sicne map_entires list does not have the entry. So the entry just leaks. Here are steps of leaking firmware_map_entry: firmware_map_remove() -> firmware_map_find_entry() Find released entry from map_entries list -> firmware_map_remove_entry() Delete the entry from map_entries list -> remove_sysfs_fw_map_entry() ... -> release_firmware_map_entry() -> firmware_map_find_entry() Find the entry from map_entries list but the entry has been deleted from map_entries list. So the entry is not added to map_entries_bootmem. Thus the entry leaks release_firmware_map_entry() should not call firmware_map_find_entry() since releaed entry has been deleted from map_entries list. So the patch delete firmware_map_find_entry() from releae_firmware_map_entry() Signed-off-by:Wanpeng Li <liwanp@linux.vnet.ibm.com> Reviewed-by:
-
Cody P Schafer authored
Signed-off-by:
Cody P Schafer <cody@linux.vnet.ibm.com> Signed-off-by:
-
Shaohua Li authored
In page reclaim, huge page is split. split_huge_page() adds tail pages to LRU list. Since we are reclaiming a huge page, it's better we reclaim all subpages of the huge page instead of just the head page. This patch adds split tail pages to shrink page list so the tail pages can be reclaimed soon. Before this patch, run a swap workload: thp_fault_alloc 3492 thp_fault_fallback 608 thp_collapse_alloc 6 thp_collapse_alloc_failed 0 thp_split 916 With this patch: thp_fault_alloc 4085 thp_fault_fallback 16 thp_collapse_alloc 90 thp_collapse_alloc_failed 0 thp_split 1272 fallback allocation is reduced a lot. [akpm@linux-foundation.org: fix CONFIG_SWAP=n build] Signed-off-by:
Shaohua Li <shli@fusionio.com> Acked-by:
Rik van Riel <riel@redhat.com> Acked-by:
-
Seth Jennings authored
To prevent flooding the swap device with writebacks, frontswap backends need to count and limit the number of outstanding writebacks. The incrementing of the counter can be done before the call to __swap_writepage(). However, the caller must receive a notification when the writeback completes in order to decrement the counter. To achieve this functionality, this patch modifies __swap_writepage() to take the bio completion callback function as an argument. end_swap_bio_write(), the normal bio completion function, is also made non-static so that code doing the accounting can call it after the accounting is done. There should be no behavioural change to existing code. Signed-off-by:
Seth Jennings <sjenning@linux.vnet.ibm.com> Signed-off-by:
Bob Liu <bob.liu@oracle.com> Acked-by:
Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Signed-off-by:
-
Seth Jennings authored
swap_writepage() is currently where frontswap hooks into the swap write path to capture pages with the frontswap_store() function. However, if a frontswap backend wants to "resume" the writeback of a page to the swap device, it can't call swap_writepage() as the page will simply reenter the backend. This patch separates swap_writepage() into a top and bottom half, the bottom half named __swap_writepage() to allow a frontswap backend, like zswap, to resume writeback beyond the frontswap_store() hook. __add_to_swap_cache() is also made non-static so that the page for which writeback is to be resumed can be added to the swap cache. Signed-off-by:
-
Cyril Hrubis authored
Fix a corner case for MAP_FIXED when requested mapping length is larger than rlimit for virtual memory. In such case any overlapping mappings are unmapped before we check for the limit and return ENOMEM. The check is moved before the loop that unmaps overlapping parts of existing mappings. When we are about to hit the limit (currently mapped pages + len > limit) we scan for overlapping pages and check again accounting for them. This fixes situation when userspace program expects that the previous mappings are preserved after the mmap() syscall has returned with error. (POSIX clearly states that successfull mapping shall replace any previous mappings.) This corner case was found and can be tested with LTP testcase: testcases/open_posix_testsuite/conformance/interfaces/mmap/24-2.c In this case the mmap, which is clearly over current limit, unmaps dynamic libraries and the testcase segfaults right after returning into userspace. I've also looked at the second instance of the unmapping loop in the do_brk(). The do_brk() is called from brk() syscall and from vm_brk(). The brk() syscall checks for overlapping mappings and bails out when there are any (so it can't be triggered from the brk syscall). The vm_brk() is called only from binmft handlers so it shouldn't be triggered unless binmft handler created overlapping mappings. Signed-off-by:
Cyril Hrubis <chrubis@suse.cz> Reviewed-by:
-
Anton Vorontsov authored
With this patch userland applications that want to maintain the interactivity/memory allocation cost can use the pressure level notifications. The levels are defined like this: The "low" level means that the system is reclaiming memory for new allocations. Monitoring this reclaiming activity might be useful for maintaining cache level. Upon notification, the program (typically "Activity Manager") might analyze vmstat and act in advance (i.e. prematurely shutdown unimportant services). The "medium" level means that the system is experiencing medium memory pressure, the system might be making swap, paging out active file caches, etc. Upon this event applications may decide to further analyze vmstat/zoneinfo/memcg or internal memory usage statistics and free any resources that can be easily reconstructed or re-read from a disk. The "critical" level means that the system is actively thrashing, it is about to out of memory (OOM) or even the in-kernel OOM killer is on its way to trigger. Applications should do whatever they can to help the system. It might be too late to consult with vmstat or any other statistics, so it's advisable to take an immediate action. The events are propagated upward until the event is handled, i.e. the events are not pass-through. Here is what this means: for example you have three cgroups: A->B->C. Now you set up an event listener on cgroups A, B and C, and suppose group C experiences some pressure. In this situation, only group C will receive the notification, i.e. groups A and B will not receive it. This is done to avoid excessive "broadcasting" of messages, which disturbs the system and which is especially bad if we are low on memory or thrashing. So, organize the cgroups wisely, or propagate the events manually (or, ask us to implement the pass-through events, explaining why would you need them.) Performance wise, the memory pressure notifications feature itself is lightweight and does not require much of bookkeeping, in contrast to the rest of memcg features. Unfortunately, as of current memcg implementation, pages accounting is an inseparable part and cannot be turned off. The good news is that there are some efforts[1] to improve the situation; plus, implementing the same, fully API-compatible[2] interface for CONFIG_MEMCG=n case (e.g. embedded) is also a viable option, so it will not require any changes on the userland side. [1] http://permalink.gmane.org/gmane.linux.kernel.cgroups/6291 [2] http://lkml.org/lkml/2013/2/21/454 [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: fix CONFIG_CGROPUPS=n warnings] Signed-off-by:
Anton Vorontsov <anton.vorontsov@linaro.org> Acked-by:
Kirill A. Shutemov <kirill@shutemov.name> Acked-by:
-
Rasmus Villemoes authored
In madvise(), there doesn't seem to be any reason for taking the ¤t->mm->mmap_sem before start and len_in have been validated. Incidentally, this removes the need for the out: label. [akpm@linux-foundation.org: s/out_plug/out/, per David] Signed-off-by:
Rasmus Villemoes <linux@rasmusvillemoes.dk> Acked-by:
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Acked-by:
-
David Rientjes authored
__remove_pages() is only necessary for CONFIG_MEMORY_HOTREMOVE. PowerPC pseries will return -EOPNOTSUPP if unsupported. Adding an #ifdef causes several other functions it depends on to also become unnecessary, which saves in .text when disabled (it's disabled in most defconfigs besides powerpc, including x86). remove_memory_block() becomes static since it is not referenced outside of drivers/base/memory.c. Build tested on x86 and powerpc with CONFIG_MEMORY_HOTREMOVE both enabled and disabled. Signed-off-by:
-
Toshi Kani authored
Change __remove_pages() to call release_mem_region_adjustable(). This allows a requested memory range to be released from the iomem_resource table even if it does not match exactly to an resource entry but still fits into. The resource entries initialized at bootup usually cover the whole contiguous memory ranges and may not necessarily match with the size of memory hot-delete requests. If release_mem_region_adjustable() failed, __remove_pages() emits a warning message and continues to proceed as it was the case with release_mem_region(). release_mem_region(), which is defined to __release_region(), emits a warning message and returns no error since a void function. Signed-off-by:
-
Toshi Kani authored
Add release_mem_region_adjustable(), which releases a requested region from a currently busy memory resource. This interface adjusts the matched memory resource accordingly even if the requested region does not match exactly but still fits into. This new interface is intended for memory hot-delete. During bootup, memory resources are inserted from the boot descriptor table, such as EFI Memory Table and e820. Each memory resource entry usually covers the whole contigous memory range. Memory hot-delete request, on the other hand, may target to a particular range of memory resource, and its size can be much smaller than the whole contiguous memory. Since the existing release interfaces like __release_region() require a requested region to be exactly matched to a resource entry, they do not allow a partial resource to be released. This new interface is restrictive (i.e. release under certain conditions), which is consistent with other release interfaces, __release_region() and __release_resource(). Additional release conditions, such as an overlapping region to a resource entry, can be supported after they are confirmed as valid cases. There is no change to the existing interfaces since their restriction is valid for I/O resources. [akpm@linux-foundation.org: use GFP_ATOMIC under write_lock()] [akpm@linux-foundation.org: switch back to GFP_KERNEL, less buggily] [akpm@linux-foundation.org: remove unneeded and wrong kfree(), per Toshi] Signed-off-by:
Ram Pai <linuxram@us.ibm.com> Cc: T Makphaibulchoke <tmac@hp.com> Cc: Wen Congyang <wency@cn.fujitsu.com> Cc: Tang Chen <tangchen@cn.fujitsu.com> Cc: Jiang Liu <jiang.liu@huawei.com> Signed-off-by:
-
Toshi Kani authored
Add __adjust_resource(), which is called by adjust_resource() internally after the resource_lock is held. There is no interface change to adjust_resource(). This change allows other functions to call __adjust_resource() internally while the resource_lock is held. Signed-off-by:
-
Srivatsa S. Bhat authored
The comment over migrate_pages() looks quite weird, and makes it hard to grasp what it is trying to say. Rewrite it more comprehensibly. Signed-off-by:
Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Acked-by:
Christoph Lameter <cl@linux.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Hugh Dickins <hughd@google.com> Signed-off-by:
-
Minchan Kim authored
Currently the memory barrier in __do_huge_pmd_anonymous_page doesn't work. Because lru_cache_add_lru uses pagevec so it could miss spinlock easily so above rule was broken so user might see inconsistent data. I was not first person who pointed out the problem. Mel and Peter pointed out a few months ago and Peter pointed out further that even spin_lock/unlock can't make sure of it: http://marc.info/?t=134333512700004 In particular: *A = a; LOCK UNLOCK *B = b; may occur as: LOCK, STORE *B, STORE *A, UNLOCK At last, Hugh pointed out that even we don't need memory barrier in there because __SetPageUpdate already have done it from Nick's commit 0ed361de ("mm: fix PageUptodate data race") explicitly. So this patch fixes comment on THP and adds same comment for do_anonymous_page, too because everybody except Hugh was missing that. It means we need a comment about that. Signed-off-by:
-
Yijing Wang authored
CONFIG_HOTPLUG is going away as an option, cleanup CONFIG_HOTPLUG ifdefs in mm files. Signed-off-by:
Yijing Wang <wangyijing@huawei.com> Acked-by:
-
Michel Lespinasse authored
Just a trivial issue I stumbled on while doing something else... Signed-off-by:
Michel Lespinasse <walken@google.com> Signed-off-by:
-
Andrew Shewmaker authored
Alter the admin and user reserves of the previous patches in this series when memory is added or removed. If memory is added and the reserves have been eliminated or increased above the default max, then we'll trust the admin. If memory is removed and there isn't enough free memory, then we need to reset the reserves. Otherwise keep the reserve set by the admin. The reserve reset code is the same as the reserve initialization code. I tested hot addition and removal by triggering it via sysfs. The reserves shrunk when they were set high and memory was removed. They were reset higher when memory was added again. [akpm@linux-foundation.org: use register_hotmemory_notifier()] [akpm@linux-foundation.org: init_user_reserve() and init_admin_reserve can no longer be __meminit] [fengguang.wu@intel.com: make init_reserve_notifier() static] [akpm@linux-foundation.org: coding-style fixes] Signed-off-by:
Andrew Shewmaker <agshew@gmail.com> Signed-off-by:
Fengguang Wu <fengguang.wu@intel.com> Signed-off-by:
-
Andrew Shewmaker authored
Add an admin_reserve_kbytes knob to allow admins to change the hardcoded memory reserve to something other than 3%, which may be multiple gigabytes on large memory systems. Only about 8MB is necessary to enable recovery in the default mode, and only a few hundred MB are required even when overcommit is disabled. This affects OVERCOMMIT_GUESS and OVERCOMMIT_NEVER. admin_reserve_kbytes is initialized to min(3% free pages, 8MB) I arrived at 8MB by summing the RSS of sshd or login, bash, and top. Please see first patch in this series for full background, motivation, testing, and full changelog. [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: make init_admin_reserve() static] Signed-off-by:
-
Andrew Shewmaker authored
Add user_reserve_kbytes knob. Limit the growth of the memory reserved for other user processes to min(3% current process size, user_reserve_pages). Only about 8MB is necessary to enable recovery in the default mode, and only a few hundred MB are required even when overcommit is disabled. user_reserve_pages defaults to min(3% free pages, 128MB) I arrived at 128MB by taking the max VSZ of sshd, login, bash, and top ... then adding the RSS of each. This only affects OVERCOMMIT_NEVER mode. Background 1. user reserve __vm_enough_memory reserves a hardcoded 3% of the current process size for other applications when overcommit is disabled. This was done so that a user could recover if they launched a memory hogging process. Without the reserve, a user would easily run into a message such as: bash: fork: Cannot allocate memory 2. admin reserve Additionally, a hardcoded 3% of free memory is reserved for root in both overcommit 'guess' and 'never' modes. This was intended to prevent a scenario where root-cant-log-in and perform recovery operations. Note that this reserve shrinks, and doesn't guarantee a useful reserve. Motivation The two hardcoded memory reserves should be updated to account for current memory sizes. Also, the admin reserve would be more useful if it didn't shrink too much. When the current code was originally written, 1GB was considered "enterprise". Now the 3% reserve can grow to multiple GB on large memory systems, and it only needs to be a few hundred MB at most to enable a user or admin to recover a system with an unwanted memory hogging process. I've found that reducing these reserves is especially beneficial for a specific type of application load: * single application system * one or few processes (e.g. one per core) * allocating all available memory * not initializing every page immediately * long running I've run scientific clusters with this sort of load. A long running job sometimes failed many hours (weeks of CPU time) into a calculation. They weren't initializing all of their memory immediately, and they weren't using calloc, so I put systems into overcommit 'never' mode. These clusters run diskless and have no swap. However, with the current reserves, a user wishing to allocate as much memory as possible to one process may be prevented from using, for example, almost 2GB out of 32GB. The effect is less, but still significant when a user starts a job with one process per core. I have repeatedly seen a set of processes requesting the same amount of memory fail because one of them could not allocate the amount of memory a user would expect to be able to allocate. For example, Message Passing Interfce (MPI) processes, one per core. And it is similar for other parallel programming frameworks. Changing this reserve code will make the overcommit never mode more useful by allowing applications to allocate nearly all of the available memory. Also, the new admin_reserve_kbytes will be safer than the current behavior since the hardcoded 3% of available memory reserve can shrink to something useless in the case where applications have grabbed all available memory. Risks * "bash: fork: Cannot allocate memory" The downside of the first patch-- which creates a tunable user reserve that is only used in overcommit 'never' mode--is that an admin can set it so low that a user may not be able to kill their process, even if they already have a shell prompt. Of course, a user can get in the same predicament with the current 3% reserve--they just have to launch processes until 3% becomes negligible. * root-cant-log-in problem The second patch, adding the tunable rootuser_reserve_pages, allows the admin to shoot themselves in the foot by setting it too small. They can easily get the system into a state where root-can't-log-in. However, the new admin_reserve_kbytes will be safer than the current behavior since the hardcoded 3% of available memory reserve can shrink to something useless in the case where applications have grabbed all available memory. Alternatives * Memory cgroups provide a more flexible way to limit application memory. Not everyone wants to set up cgroups or deal with their overhead. * We could create a fourth overcommit mode which provides smaller reserves. The size of useful reserves may be drastically different depending on the whether the system is embedded or enterprise. * Force users to initialize all of their memory or use calloc. Some users don't want/expect the system to overcommit when they malloc. Overcommit 'never' mode is for this scenario, and it should work well. The new user and admin reserve tunables are simple to use, with low overhead compared to cgroups. The patches preserve current behavior where 3% of memory is less than 128MB, except that the admin reserve doesn't shrink to an unusable size under pressure. The code allows admins to tune for embedded and enterprise usage. FAQ * How is the root-cant-login problem addressed? What happens if admin_reserve_pages is set to 0? Root is free to shoot themselves in the foot by setting admin_reserve_kbytes too low. On x86_64, the minimum useful reserve is: 8MB for overcommit 'guess' 128MB for overcommit 'never' admin_reserve_pages defaults to min(3% free memory, 8MB) So, anyone switching to 'never' mode needs to adjust admin_reserve_pages. * How do you calculate a minimum useful reserve? A user or the admin needs enough memory to login and perform recovery operations, which includes, at a minimum: sshd or login + bash (or some other shell) + top (or ps, kill, etc.) For overcommit 'guess', we can sum resident set sizes (RSS) because we only need enough memory to handle what the recovery programs will typically use. On x86_64 this is about 8MB. For overcommit 'never', we can take the max of their virtual sizes (VSZ) and add the sum of their RSS. We use VSZ instead of RSS because mode forces us to ensure we can fulfill all of the requested memory allocations-- even if the programs only use a fraction of what they ask for. On x86_64 this is about 128MB. When swap is enabled, reserves are useful even when they are as small as 10MB, regardless of overcommit mode. When both swap and overcommit are disabled, then the admin should tune the reserves higher to be absolutley safe. Over 230MB each was safest in my testing. * What happens if user_reserve_pages is set to 0? Note, this only affects overcomitt 'never' mode. Then a user will be able to allocate all available memory minus admin_reserve_kbytes. However, they will easily see a message such as: "bash: fork: Cannot allocate memory" And they won't be able to recover/kill their application. The admin should be able to recover the system if admin_reserve_kbytes is set appropriately. * What's the difference between overcommit 'guess' and 'never'? "Guess" allows an allocation if there are enough free + reclaimable pages. It has a hardcoded 3% of free pages reserved for root. "Never" allows an allocation if there is enough swap + a configurable percentage (default is 50) of physical RAM. It has a hardcoded 3% of free pages reserved for root, like "Guess" mode. It also has a hardcoded 3% of the current process size reserved for additional applications. * Why is overcommit 'guess' not suitable even when an app eventually writes to every page? It takes free pages, file pages, available swap pages, reclaimable slab pages into consideration. In other words, these are all pages available, then why isn't overcommit suitable? Because it only looks at the present state of the system. It does not take into account the memory that other applications have malloced, but haven't initialized yet. It overcommits the system. Test Summary There was little change in behavior in the default overcommit 'guess' mode with swap enabled before and after the patch. This was expected. Systems run most predictably (i.e. no oom kills) in overcommit 'never' mode with swap enabled. This also allowed the most memory to be allocated to a user application. Overcommit 'guess' mode without swap is a bad idea. It is easy to crash the system. None of the other tested combinations crashed. This matches my experience on the Roadrunner supercomputer. Without the tunable user reserve, a system in overcommit 'never' mode and without swap does not allow the admin to recover, although the admin can. With the new tunable reserves, a system in overcommit 'never' mode and without swap can be configured to: 1. maximize user-allocatable memory, running close to the edge of recoverability 2. maximize recoverability, sacrificing allocatable memory to ensure that a user cannot take down a system Test Description Fedora 18 VM - 4 x86_64 cores, 5725MB RAM, 4GB Swap System is booted into multiuser console mode, with unnecessary services turned off. Caches were dropped before each test. Hogs are user memtester processes that attempt to allocate all free memory as reported by /proc/meminfo In overcommit 'never' mode, memory_ratio=100 Test Results 3.9.0-rc1-mm1 Overcommit | Swap | Hogs | MB Got/Wanted | OOMs | User Recovery | Admin Recovery ---------- ---- ---- ------------- ---- ------------- -------------- guess yes 1 5432/5432 no yes yes guess yes 4 5444/5444 1 yes yes guess no 1 5302/5449 no yes yes guess no 4 - crash no no never yes 1 5460/5460 1 yes yes never yes 4 5460/5460 1 yes yes never no 1 5218/5432 no no yes never no 4 5203/5448 no no yes 3.9.0-rc1-mm1-tunablereserves User and Admin Recovery show their respective reserves, if applicable. Overcommit | Swap | Hogs | MB Got/Wanted | OOMs | User Recovery | Admin Recovery ---------- ---- ---- ------------- ---- ------------- -------------- guess yes 1 5419/5419 no - yes 8MB yes guess yes 4 5436/5436 1 - yes 8MB yes guess no 1 5440/5440 * - yes 8MB yes guess no 4 - crash - no 8MB no * process would successfully mlock, then the oom killer would pick it never yes 1 5446/5446 no 10MB yes 20MB yes never yes 4 5456/5456 no 10MB yes 20MB yes never no 1 5387/5429 no 128MB no 8MB barely never no 1 5323/5428 no 226MB barely 8MB barely never no 1 5323/5428 no 226MB barely 8MB barely never no 1 5359/5448 no 10MB no 10MB barely never no 1 5323/5428 no 0MB no 10MB barely never no 1 5332/5428 no 0MB no 50MB yes never no 1 5293/5429 no 0MB no 90MB yes never no 1 5001/5427 no 230MB yes 338MB yes never no 4* 4998/5424 no 230MB yes 338MB yes * more memtesters were launched, able to allocate approximately another 100MB Future Work - Test larger memory systems. - Test an embedded image. - Test other architectures. - Time malloc microbenchmarks. - Would it be useful to be able to set overcommit policy for each memory cgroup? - Some lines are slightly above 80 chars. Perhaps define a macro to convert between pages and kb? Other places in the kernel do this. [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: make init_user_reserve() static] Signed-off-by: -
Andrew Morton authored
Use the new interface, remove one ifdef. No code size changes. We could/should have been using __meminit/__meminitdata here but there's now no point in doing that because all this code is elided at compile time. Cc: Li Zefan <lizefan@huawei.com> Cc: Tejun Heo <tj@kernel.org> Signed-off-by:
-
Andrew Morton authored
Saves an ifdef, no code size changes Signed-off-by:
-
Andrew Morton authored
Squishes a warning which my change to hotplug_memory_notifier() added. I want to keep that warning, because it is punishment for failnig to check the hotplug_memory_notifier() return value. Cc: Greg KH <greg@kroah.com> Signed-off-by:
-
Andrew Morton authored
Squishes a statement-with-no-effect warning, removes some ifdefs and shrinks .text by 2 bytes. Note that this code fails to check for blocking_notifier_chain_register() failures. Cc: Pekka Enberg <penberg@kernel.org> Signed-off-by:
-
Andrew Morton authored
Squishes a statement-with-no-effect warning, removes some ifdefs and shrinks .text by one byte! Note that this code fails to check for blocking_notifier_chain_register() failures. Signed-off-by:
-
Andrew Morton authored
When CONFIG_MEMORY_HOTPLUG=n, we don't want the memory-hotplug notifier handlers to be included in the .o files, for space reasons. The existing hotplug_memory_notifier() tries to handle this but testing with gcc-4.4.4 shows that it doesn't work - the hotplug functions are still present in the .o files. So implement a new register_hotmemory_notifier() which is a copy of register_hotcpu_notifier(), and which actually works as desired. hotplug_memory_notifier() and register_memory_notifier() callsites should be converted to use this new register_hotmemory_notifier(). While we're there, let's repair the existing hotplug_memory_notifier(): it simply stomps on the register_memory_notifier() return value, so well-behaved code cannot check for errors. Apparently non of the existing callers were well-behaved :( Cc: Andrew Shewmaker <agshew@gmail.com> Signed-off-by:
-
