- 15 Jan, 2016 40 commits
-
-
Daniel Cashman authored
x86: arch_mmap_rnd() uses hard-coded values, 8 for 32-bit and 28 for 64-bit, to generate the random offset for the mmap base address. This value represents a compromise between increased ASLR effectiveness and avoiding address-space fragmentation. Replace it with a Kconfig option, which is sensibly bounded, so that platform developers may choose where to place this compromise. Keep default values as new minimums. Signed-off-by:
Daniel Cashman <dcashman@google.com> Cc: Russell King <linux@arm.linux.org.uk> Acked-by:
Kees Cook <keescook@chromium.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jonathan Corbet <corbet@lwn.net> Cc: Don Zickus <dzickus@redhat.com> Cc: Eric W. Biederman <ebiederm@xmission.com> Cc: Heinrich Schuchardt <xypron.glpk@gmx.de> Cc: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Mel Gorman <mgorman@suse.de> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: David Rientjes <rientjes@google.com> Cc: Mark Salyzyn <salyzyn@android.com> Cc: Jeff Vander Stoep <jeffv@google.com> Cc: Nick Kralevich <nnk@google.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Will Deacon <will.deacon@arm.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Hector Marco-Gisbert <hecmargi@upv.es> Cc: Borislav Petkov <bp@suse.de> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Daniel Cashman authored
arm64: arch_mmap_rnd() uses STACK_RND_MASK to generate the random offset for the mmap base address. This value represents a compromise between increased ASLR effectiveness and avoiding address-space fragmentation. Replace it with a Kconfig option, which is sensibly bounded, so that platform developers may choose where to place this compromise. Keep default values as new minimums. Signed-off-by:
-
Daniel Cashman authored
arm: arch_mmap_rnd() uses a hard-code value of 8 to generate the random offset for the mmap base address. This value represents a compromise between increased ASLR effectiveness and avoiding address-space fragmentation. Replace it with a Kconfig option, which is sensibly bounded, so that platform developers may choose where to place this compromise. Keep 8 as the minimum acceptable value. [arnd@arndb.de: ARM: avoid ARCH_MMAP_RND_BITS for NOMMU] Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
-
Daniel Cashman authored
Address Space Layout Randomization (ASLR) provides a barrier to exploitation of user-space processes in the presence of security vulnerabilities by making it more difficult to find desired code/data which could help an attack. This is done by adding a random offset to the location of regions in the process address space, with a greater range of potential offset values corresponding to better protection/a larger search-space for brute force, but also to greater potential for fragmentation. The offset added to the mmap_base address, which provides the basis for the majority of the mappings for a process, is set once on process exec in arch_pick_mmap_layout() and is done via hard-coded per-arch values, which reflect, hopefully, the best compromise for all systems. The trade-off between increased entropy in the offset value generation and the corresponding increased variability in address space fragmentation is not absolute, however, and some platforms may tolerate higher amounts of entropy. This patch introduces both new Kconfig values and a sysctl interface which may be used to change the amount of entropy used for offset generation on a system. The direct motivation for this change was in response to the libstagefright vulnerabilities that affected Android, specifically to information provided by Google's project zero at: http://googleprojectzero.blogspot.com/2015/09/stagefrightened.html The attack presented therein, by Google's project zero, specifically targeted the limited randomness used to generate the offset added to the mmap_base address in order to craft a brute-force-based attack. Concretely, the attack was against the mediaserver process, which was limited to respawning every 5 seconds, on an arm device. The hard-coded 8 bits used resulted in an average expected success rate of defeating the mmap ASLR after just over 10 minutes (128 tries at 5 seconds a piece). With this patch, and an accompanying increase in the entropy value to 16 bits, the same attack would take an average expected time of over 45 hours (32768 tries), which makes it both less feasible and more likely to be noticed. The introduced Kconfig and sysctl options are limited by per-arch minimum and maximum values, the minimum of which was chosen to match the current hard-coded value and the maximum of which was chosen so as to give the greatest flexibility without generating an invalid mmap_base address, generally a 3-4 bits less than the number of bits in the user-space accessible virtual address space. When decided whether or not to change the default value, a system developer should consider that mmap_base address could be placed anywhere up to 2^(value) bits away from the non-randomized location, which would introduce variable-sized areas above and below the mmap_base address such that the maximum vm_area_struct size may be reduced, preventing very large allocations. This patch (of 4): ASLR only uses as few as 8 bits to generate the random offset for the mmap base address on 32 bit architectures. This value was chosen to prevent a poorly chosen value from dividing the address space in such a way as to prevent large allocations. This may not be an issue on all platforms. Allow the specification of a minimum number of bits so that platforms desiring greater ASLR protection may determine where to place the trade-off. Signed-off-by:
-
Piotr Kwapulinski authored
The following flag comparison in mmap_region makes no sense: if (!(vm_flags & MAP_FIXED)) return -ENOMEM; The condition is always false and thus the above "return -ENOMEM" is never executed. The vm_flags must not be compared with MAP_FIXED flag. The vm_flags may only be compared with VM_* flags. MAP_FIXED has the same value as VM_MAYREAD. Hitting the rlimit is a slow path and find_vma_intersection should realize that there is no overlapping VMA for !MAP_FIXED case pretty quickly. Signed-off-by:Piotr Kwapulinski <kwapulinski.piotr@gmail.com> Acked-by:
Michal Hocko <mhocko@suse.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Chris Metcalf <cmetcalf@ezchip.com> Reviewed-by:
Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by:
-
Michal Hocko authored
zone_reclaimable_pages counts how many pages are reclaimable in the given zone. This currently includes all pages on file lrus and anon lrus if there is an available swap storage. We do not consider NR_ISOLATED_{ANON,FILE} counters though which is not correct because these counters reflect temporarily isolated pages which are still reclaimable because they either get back to their LRU or get freed either by the page reclaim or page migration. The number of these pages might be sufficiently high to confuse users of zone_reclaimable_pages (e.g. mbind can migrate large ranges of memory at once). Signed-off-by:Johannes Weiner <hannes@cmpxchg.org> Acked-by:
Vladimir Davydov <vdavydov@virtuozzo.com> Acked-by:
David Rientjes <rientjes@google.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Mel Gorman <mgorman@suse.de> Signed-off-by:
-
Andrew Morton authored
bdev_write_page() is used by swapout and by writepage where we cannot use __GFP_FS or __GFP_IO. So it is misleading to mention GFP_KERNEL here. blk_queue_enter() only actually looks at __GFP_DIRECT_RECLAIM, so no bugs were harmed in the making of this patch. Cc: Dan Williams <dan.j.williams@intel.com> Cc: Jens Axboe <axboe@fb.com> Signed-off-by:
-
Vladimir Davydov authored
There are two bits defined for cg_proto->flags - MEMCG_SOCK_ACTIVATED and MEMCG_SOCK_ACTIVE - both are set in tcp_update_limit, but the former is never cleared while the latter can be cleared by unsetting the limit. This allows to disable tcp socket accounting for new sockets after it was enabled by writing -1 to memory.kmem.tcp.limit_in_bytes while still guaranteeing that memcg_socket_limit_enabled static key will be decremented on memcg destruction. This functionality looks dubious, because it is not clear what a use case would be. By enabling tcp accounting a user accepts the price. If they then find the performance degradation unacceptable, they can always restart their workload with tcp accounting disabled. It does not seem there is any need to flip it while the workload is running. Besides, it contradicts to how kmem accounting API works: writing whatever to memory.kmem.limit_in_bytes enables kmem accounting for the cgroup in question, after which it cannot be disabled. Therefore one might expect that writing -1 to memory.kmem.tcp.limit_in_bytes just enables socket accounting w/o limiting it, which might be useful by itself, but it isn't true. Since this API peculiarity is not documented anywhere, I propose to drop it. This will allow to simplify the code by dropping cg_proto->flags. Signed-off-by:
-
Vladimir Davydov authored
We assume there is enough inactive page cache if the size of inactive file lru is greater than the size of active file lru, in which case we force-scan file lru ignoring anonymous pages. While this logic works fine when there are plenty of page cache pages, it fails if the size of file lru is small (several MB): in this case (lru_size >> prio) will be 0 for normal scan priorities, as a result, if inactive file lru happens to be larger than active file lru, anonymous pages of a cgroup will never get evicted unless the system experiences severe memory pressure, even if there are gigabytes of unused anonymous memory there, which is unfair in respect to other cgroups, whose workloads might be page cache oriented. This patch attempts to fix this by elaborating the "enough inactive page cache" check: it makes it not only check that inactive lru size > active lru size, but also that we will scan something from the cgroup at the current scan priority. If these conditions do not hold, we proceed to SCAN_FRACT as usual. Signed-off-by:
-
David Rientjes authored
VM_VPAGES is unnecessary, it's easier to check is_vmalloc_addr() when reading /proc/vmallocinfo. [akpm@linux-foundation.org: remove VM_VPAGES reference via kvfree()] Signed-off-by:
-
Geliang Tang authored
Simplify the code with list_first_entry_or_null(). Signed-off-by:
Geliang Tang <geliangtang@163.com> Acked-by:
Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Signed-off-by:
-
Jerome Marchand authored
There are several shortcomings with the accounting of shared memory (SysV shm, shared anonymous mapping, mapping of a tmpfs file). The values in /proc/<pid>/status and <...>/statm don't allow to distinguish between shmem memory and a shared mapping to a regular file, even though theirs implication on memory usage are quite different: during reclaim, file mapping can be dropped or written back on disk, while shmem needs a place in swap. Also, to distinguish the memory occupied by anonymous and file mappings, one has to read the /proc/pid/statm file, which has a field for the file mappings (again, including shmem) and total memory occupied by these mappings (i.e. equivalent to VmRSS in the <...>/status file. Getting the value for anonymous mappings only is thus not exactly user-friendly (the statm file is intended to be rather efficiently machine-readable). To address both of these shortcomings, this patch adds a breakdown of VmRSS in /proc/<pid>/status via new fields RssAnon, RssFile and RssShmem, making use of the previous preparatory patch. These fields tell the user the memory occupied by private anonymous pages, mapped regular files and shmem, respectively. Other existing fields in /status and /statm files are left without change. The /statm file can be extended in the future, if there's a need for that. Example (part of) /proc/pid/status output including the new Rss* fields: VmPeak: 2001008 kB VmSize: 2001004 kB VmLck: 0 kB VmPin: 0 kB VmHWM: 5108 kB VmRSS: 5108 kB RssAnon: 92 kB RssFile: 1324 kB RssShmem: 3692 kB VmData: 192 kB VmStk: 136 kB VmExe: 4 kB VmLib: 1784 kB VmPTE: 3928 kB VmPMD: 20 kB VmSwap: 0 kB HugetlbPages: 0 kB [vbabka@suse.cz: forward-porting, tweak changelog] Signed-off-by:
Jerome Marchand <jmarchan@redhat.com> Signed-off-by:
Vlastimil Babka <vbabka@suse.cz> Acked-by:
Konstantin Khlebnikov <khlebnikov@yandex-team.ru> Acked-by:
Hugh Dickins <hughd@google.com> Signed-off-by:
-
Jerome Marchand authored
Currently looking at /proc/<pid>/status or statm, there is no way to distinguish shmem pages from pages mapped to a regular file (shmem pages are mapped to /dev/zero), even though their implication in actual memory use is quite different. The internal accounting currently counts shmem pages together with regular files. As a preparation to extend the userspace interfaces, this patch adds MM_SHMEMPAGES counter to mm_rss_stat to account for shmem pages separately from MM_FILEPAGES. The next patch will expose it to userspace - this patch doesn't change the exported values yet, by adding up MM_SHMEMPAGES to MM_FILEPAGES at places where MM_FILEPAGES was used before. The only user-visible change after this patch is the OOM killer message that separates the reported "shmem-rss" from "file-rss". [vbabka@suse.cz: forward-porting, tweak changelog] Signed-off-by:
-
Vlastimil Babka authored
Following the previous patch, further reduction of /proc/pid/smaps cost is possible for private writable shmem mappings with unpopulated areas where the page walk invokes the .pte_hole function. We can use radix tree iterator for each such area instead of calling find_get_entry() in a loop. This is possible at the extra maintenance cost of introducing another shmem function shmem_partial_swap_usage(). To demonstrate the diference, I have measured this on a process that creates a private writable 2GB mapping of a partially swapped out /dev/shm/file (which cannot employ the optimizations from the prvious patch) and doesn't populate it at all. I time how long does it take to cat /proc/pid/smaps of this process 100 times. Before this patch: real 0m3.831s user 0m0.180s sys 0m3.212s After this patch: real 0m1.176s user 0m0.180s sys 0m0.684s The time is similar to the case where a radix tree iterator is employed on the whole mapping. Signed-off-by:
-
Vlastimil Babka authored
The previous patch has improved swap accounting for shmem mapping, which however made /proc/pid/smaps more expensive for shmem mappings, as we consult the radix tree for each pte_none entry, so the overal complexity is O(n*log(n)). We can reduce this significantly for mappings that cannot contain COWed pages, because then we can either use the statistics tha shmem object itself tracks (if the mapping contains the whole object, or the swap usage of the whole object is zero), or use the radix tree iterator, which is much more effective than repeated find_get_entry() calls. This patch therefore introduces a function shmem_swap_usage(vma) and makes /proc/pid/smaps use it when possible. Only for writable private mappings of shmem objects (i.e. tmpfs files) with the shmem object itself (partially) swapped outwe have to resort to the find_get_entry() approach. Hopefully such mappings are relatively uncommon. To demonstrate the diference, I have measured this on a process that creates a 2GB mapping and dirties single pages with a stride of 2MB, and time how long does it take to cat /proc/pid/smaps of this process 100 times. Private writable mapping of a /dev/shm/file (the most complex case): real 0m3.831s user 0m0.180s sys 0m3.212s Shared mapping of an almost full mapping of a partially swapped /dev/shm/file (which needs to employ the radix tree iterator). real 0m1.351s user 0m0.096s sys 0m0.768s Same, but with /dev/shm/file not swapped (so no radix tree walk needed) real 0m0.935s user 0m0.128s sys 0m0.344s Private anonymous mapping: real 0m0.949s user 0m0.116s sys 0m0.348s The cost is now much closer to the private anonymous mapping case, unless the shmem mapping is private and writable. Signed-off-by:
-
Vlastimil Babka authored
Currently, /proc/pid/smaps will always show "Swap: 0 kB" for shmem-backed mappings, even if the mapped portion does contain pages that were swapped out. This is because unlike private anonymous mappings, shmem does not change pte to swap entry, but pte_none when swapping the page out. In the smaps page walk, such page thus looks like it was never faulted in. This patch changes smaps_pte_entry() to determine the swap status for such pte_none entries for shmem mappings, similarly to how mincore_page() does it. Swapped out shmem pages are thus accounted for. For private mappings of tmpfs files that COWed some of the pages, swaped out status of the original shmem pages is naturally ignored. If some of the private copies was also swapped out, they are accounted via their page table swap entries, so the resulting reported swap usage is then a sum of both swapped out private copies, and swapped out shmem pages that were not COWed. No double accounting can thus happen. The accounting is arguably still not as precise as for private anonymous mappings, since now we will count also pages that the process in question never accessed, but another process populated them and then let them become swapped out. I believe it is still less confusing and subtle than not showing any swap usage by shmem mappings at all. Swapped out counter might of interest of users who would like to prevent from future swapins during performance critical operation and pre-fault them at their convenience. Especially for larger swapped out regions the cost of swapin is much higher than a fresh page allocation. So a differentiation between pte_none vs. swapped out is important for those usecases. One downside of this patch is that it makes /proc/pid/smaps more expensive for shmem mappings, as we consult the radix tree for each pte_none entry, so the overal complexity is O(n*log(n)). I have measured this on a process that creates a 2GB mapping and dirties single pages with a stride of 2MB, and time how long does it take to cat /proc/pid/smaps of this process 100 times. Private anonymous mapping: real 0m0.949s user 0m0.116s sys 0m0.348s Mapping of a /dev/shm/file: real 0m3.831s user 0m0.180s sys 0m3.212s The difference is rather substantial, so the next patch will reduce the cost for shared or read-only mappings. In a less controlled experiment, I've gathered pids of processes on my desktop that have either '/dev/shm/*' or 'SYSV*' in smaps. This included the Chrome browser and some KDE processes. Again, I've run cat /proc/pid/smaps on each 100 times. Before this patch: real 0m9.050s user 0m0.518s sys 0m8.066s After this patch: real 0m9.221s user 0m0.541s sys 0m8.187s This suggests low impact on average systems. Note that this patch doesn't attempt to adjust the SwapPss field for shmem mappings, which would need extra work to determine who else could have the pages mapped. Thus the value stays zero except for COWed swapped out pages in a shmem mapping, which are accounted as usual. Signed-off-by:
-
Vlastimil Babka authored
This series is based on Jerome Marchand's [1] so let me quote the first paragraph from there: There are several shortcomings with the accounting of shared memory (sysV shm, shared anonymous mapping, mapping to a tmpfs file). The values in /proc/<pid>/status and statm don't allow to distinguish between shmem memory and a shared mapping to a regular file, even though their implications on memory usage are quite different: at reclaim, file mapping can be dropped or written back on disk while shmem needs a place in swap. As for shmem pages that are swapped-out or in swap cache, they aren't accounted at all. The original motivation for myself is that a customer found (IMHO rightfully) confusing that e.g. top output for process swap usage is unreliable with respect to swapped out shmem pages, which are not accounted for. The fundamental difference between private anonymous and shmem pages is that the latter has PTE's converted to pte_none, and not swapents. As such, they are not accounted to the number of swapents visible e.g. in /proc/pid/status VmSwap row. It might be theoretically possible to use swapents when swapping out shmem (without extra cost, as one has to change all mappers anyway), and on swap in only convert the swapent for the faulting process, leaving swapents in other processes until they also fault (so again no extra cost). But I don't know how many assumptions this would break, and it would be too disruptive change for a relatively small benefit. Instead, my approach is to document the limitation of VmSwap, and provide means to determine the swap usage for shmem areas for those who are interested and willing to pay the price, using /proc/pid/smaps. Because outside of ipcs, I don't think it's possible to currently to determine the usage at all. The previous patchset [1] did introduce new shmem-specific fields into smaps output, and functions to determine the values. I take a simpler approach, noting that smaps output already has a "Swap: X kB" line, where currently X == 0 always for shmem areas. I think we can just consider this a bug and provide the proper value by consulting the radix tree, as e.g. mincore_page() does. In the patch changelog I explain why this is also not perfect (and cannot be without swapents), but still arguably much better than showing a 0. The last two patches are adapted from Jerome's patchset and provide a VmRSS breakdown to RssAnon, RssFile and RssShm in /proc/pid/status. Hugh noted that this is a welcome addition, and I agree that it might help e.g. debugging process memory usage at albeit non-zero, but still rather low cost of extra per-mm counter and some page flag checks. [1] http://lwn.net/Articles/611966/ This patch (of 6): The documentation for /proc/pid/status does not mention that the value of VmSwap counts only swapped out anonymous private pages, and not swapped out pages of the underlying shmem objects (for shmem mappings). This is not obvious, so document this limitation. Signed-off-by:
-
Yaowei Bai authored
Make memmap_valid_within return bool due to this particular function only using either one or zero as its return value. No functional change. Signed-off-by:
Yaowei Bai <baiyaowei@cmss.chinamobile.com> Signed-off-by:
-
Geliang Tang authored
To make the intention clearer, use list_{next,first}_entry instead of list_entry. Signed-off-by: -
Michal Hocko authored
__alloc_pages_slowpath is looping over ALLOC_NO_WATERMARKS requests if __GFP_NOFAIL is requested. This is fragile because we are basically relying on somebody else to make the reclaim (be it the direct reclaim or OOM killer) for us. The caller might be holding resources (e.g. locks) which block other other reclaimers from making any progress for example. Remove the retry loop and rely on __alloc_pages_slowpath to invoke all allowed reclaim steps and retry logic. We have to be careful about __GFP_NOFAIL allocations from the PF_MEMALLOC context even though this is a very bad idea to begin with because no progress can be gurateed at all. We shouldn't break the __GFP_NOFAIL semantic here though. It could be argued that this is essentially GFP_NOWAIT context which we do not support but PF_MEMALLOC is much harder to check for existing users because they might happen deep down the code path performed much later after setting the flag so we cannot really rule out there is no kernel path triggering this combination. Signed-off-by:
Mel Gorman <mgorman@suse.de> Acked-by:
-
Michal Hocko authored
__alloc_pages_high_priority doesn't do anything special other than it calls get_page_from_freelist and loops around GFP_NOFAIL allocation until it succeeds. It would be better if the first part was done in __alloc_pages_slowpath where we modify the zonelist because this would be easier to read and understand. Opencoding the function into its only caller allows to simplify it a bit as well. This patch doesn't introduce any functional changes. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by:
-
Yaowei Bai authored
Hardcoding index to zonelists array in gfp_zonelist() is not a good idea, let's enumerate it to improve readability. No functional change. [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: fix CONFIG_NUMA=n build] [n-horiguchi@ah.jp.nec.com: fix warning in comparing enumerator] Signed-off-by:
-
Yaowei Bai authored
Since commit a0b8cab3 ("mm: remove lru parameter from __pagevec_lru_add and remove parts of pagevec API") there's no user of this function anymore, so remove it. Signed-off-by:
Hillf Danton <hillf.zj@alibaba-inc.com> Signed-off-by:
-
Yaowei Bai authored
Make memblock_is_memory() and memblock_is_reserved return bool to improve readability due to these particular functions only using either one or zero as their return value. No functional change. Signed-off-by:
-
Yaowei Bai authored
Make is_file_hugepages() return bool to improve readability due to this particular function only using either one or zero as its return value. This patch also removed the if condition to make is_file_hugepages return directly. No functional change. Signed-off-by:
-
yalin wang authored
Move node_id zone_idx shrink flags into trace function, so thay we don't need caculate these args if the trace is disabled, and will make this function have less arguments. Signed-off-by:
yalin wang <yalin.wang2010@gmail.com> Reviewed-by:
Steven Rostedt <rostedt@goodmis.org> Acked-by:
-
Joonsoo Kim authored
Now, we have tracepoint in test_pages_isolated() to notify pfn which cannot be isolated. But, in alloc_contig_range(), some error path doesn't call test_pages_isolated() so it's still hard to know exact pfn that causes allocation failure. This patch change this situation by calling test_pages_isolated() in almost error path. In allocation failure case, some overhead is added by this change, but, allocation failure is really rare event so it would not matter. In fatal signal pending case, we don't call test_pages_isolated() because this failure is intentional one. There was a bogus outer_start problem due to unchecked buddy order and this patch also fix it. Before this patch, it didn't matter, because end result is same thing. But, after this patch, tracepoint will report failed pfn so it should be accurate. Signed-off-by:
Joonsoo Kim <iamjoonsoo.kim@lge.com> Acked-by:
Michal Nazarewicz <mina86@mina86.com> Cc: David Rientjes <rientjes@google.com> Cc: Minchan Kim <minchan@kernel.org> Signed-off-by:
-
Joonsoo Kim authored
cma allocation should be guranteeded to succeed. But sometimes it can fail in the current implementation. To track down the problem, we need to know which page is problematic and this new tracepoint will report it. Signed-off-by:
-
Joonsoo Kim authored
This is preparation step to report test failed pfn in new tracepoint to analyze cma allocation failure problem. There is no functional change in this patch. Signed-off-by:
-
Nathan Zimmer authored
When running the SPECint_rate gcc on some very large boxes it was noticed that the system was spending lots of time in mpol_shared_policy_lookup(). The gamess benchmark can also show it and is what I mostly used to chase down the issue since the setup for that I found to be easier. To be clear the binaries were on tmpfs because of disk I/O requirements. We then used text replication to avoid icache misses and having all the copies banging on the memory where the instruction code resides. This results in us hitting a bottleneck in mpol_shared_policy_lookup() since lookup is serialised by the shared_policy lock. I have only reproduced this on very large (3k+ cores) boxes. The problem starts showing up at just a few hundred ranks getting worse until it threatens to livelock once it gets large enough. For example on the gamess benchmark at 128 ranks this area consumes only ~1% of time, at 512 ranks it consumes nearly 13%, and at 2k ranks it is over 90%. To alleviate the contention in this area I converted the spinlock to an rwlock. This allows a large number of lookups to happen simultaneously. The results were quite good reducing this consumtion at max ranks to around 2%. [akpm@linux-foundation.org: tidy up code comments] Signed-off-by:
Nathan Zimmer <nzimmer@sgi.com> Acked-by:
-
Chen Gang authored
__phys_to_pfn and __pfn_to_phys are symmetric, PHYS_PFN and PFN_PHYS are semmetric: - y = (phys_addr_t)x << PAGE_SHIFT - y >> PAGE_SHIFT = (phys_add_t)x - (unsigned long)(y >> PAGE_SHIFT) = x [akpm@linux-foundation.org: use macro arg name `x'] [arnd@arndb.de: include linux/pfn.h for PHYS_PFN definition] Signed-off-by:
Chen Gang <gang.chen.5i5j@gmail.com> Cc: Oleg Nesterov <oleg@redhat.com> Signed-off-by:
-
yalin wang authored
Move trace_reclaim_flags() into trace function, so that we don't need caculate these flags if the trace is disabled. Signed-off-by:
-
Chen Gang authored
Simplify may_expand_vm(). [akpm@linux-foundation.org: further simplification, per Naoya Horiguchi] Signed-off-by:
-
Alexey Klimov authored
Before usage page pointer initialized by NULL is reinitialized by follow_page_mask(). Drop useless init of page pointer in the beginning of loop. Signed-off-by:
Alexey Klimov <klimov.linux@gmail.com> Acked-by:
-
Vladimir Davydov authored
Mark those kmem allocations that are known to be easily triggered from userspace as __GFP_ACCOUNT/SLAB_ACCOUNT, which makes them accounted to memcg. For the list, see below: - threadinfo - task_struct - task_delay_info - pid - cred - mm_struct - vm_area_struct and vm_region (nommu) - anon_vma and anon_vma_chain - signal_struct - sighand_struct - fs_struct - files_struct - fdtable and fdtable->full_fds_bits - dentry and external_name - inode for all filesystems. This is the most tedious part, because most filesystems overwrite the alloc_inode method. The list is far from complete, so feel free to add more objects. Nevertheless, it should be close to "account everything" approach and keep most workloads within bounds. Malevolent users will be able to breach the limit, but this was possible even with the former "account everything" approach (simply because it did not account everything in fact). [akpm@linux-foundation.org: coding-style fixes] Signed-off-by:
-
Vladimir Davydov authored
Make vmalloc family functions allocate vmalloc area pages with alloc_kmem_pages so that if __GFP_ACCOUNT is set they will be accounted to memcg. This is needed, at least, to account alloc_fdmem allocations. Signed-off-by:
-
Vladimir Davydov authored
Currently, if we want to account all objects of a particular kmem cache, we have to pass __GFP_ACCOUNT to each kmem_cache_alloc call, which is inconvenient. This patch introduces SLAB_ACCOUNT flag which if passed to kmem_cache_create will force accounting for every allocation from this cache even if __GFP_ACCOUNT is not passed. This patch does not make any of the existing caches use this flag - it will be done later in the series. Note, a cache with SLAB_ACCOUNT cannot be merged with a cache w/o SLAB_ACCOUNT, because merged caches share the same kmem_cache struct and hence cannot have different sets of SLAB_* flags. Thus using this flag will probably reduce the number of merged slabs even if kmem accounting is not used (only compiled in). Signed-off-by:
Tejun Heo <tj@kernel.org> Acked-by:
-
Vladimir Davydov authored
Black-list kmem accounting policy (aka __GFP_NOACCOUNT) turned out to be fragile and difficult to maintain, because there seem to be many more allocations that should not be accounted than those that should be. Besides, false accounting an allocation might result in much worse consequences than not accounting at all, namely increased memory consumption due to pinned dead kmem caches. So this patch switches kmem accounting to the white-policy: now only those kmem allocations that are marked as __GFP_ACCOUNT are accounted to memcg. Currently, no kmem allocations are marked like this. The following patches will mark several kmem allocations that are known to be easily triggered from userspace and therefore should be accounted to memcg. Signed-off-by:
-
Vladimir Davydov authored
This reverts commit 8f4fc071 ("gfp: add __GFP_NOACCOUNT"). Black-list kmem accounting policy (aka __GFP_NOACCOUNT) turned out to be fragile and difficult to maintain, because there seem to be many more allocations that should not be accounted than those that should be. Besides, false accounting an allocation might result in much worse consequences than not accounting at all, namely increased memory consumption due to pinned dead kmem caches. So it was decided to switch to the white-list policy. This patch reverts bits introducing the black-list policy. The white-list policy will be introduced later in the series. Signed-off-by:
-
Vladimir Davydov authored
Currently, all kmem allocations (namely every kmem_cache_alloc, kmalloc, alloc_kmem_pages call) are accounted to memory cgroup automatically. Callers have to explicitly opt out if they don't want/need accounting for some reason. Such a design decision leads to several problems: - kmalloc users are highly sensitive to failures, many of them implicitly rely on the fact that kmalloc never fails, while memcg makes failures quite plausible. - A lot of objects are shared among different containers by design. Accounting such objects to one of containers is just unfair. Moreover, it might lead to pinning a dead memcg along with its kmem caches, which aren't tiny, which might result in noticeable increase in memory consumption for no apparent reason in the long run. - There are tons of short-lived objects. Accounting them to memcg will only result in slight noise and won't change the overall picture, but we still have to pay accounting overhead. For more info, see - http://lkml.kernel.org/r/20151105144002.GB15111%40dhcp22.suse.cz - http://lkml.kernel.org/r/20151106090555.GK29259@esperanza Therefore this patchset switches to the white list policy. Now kmalloc users have to explicitly opt in by passing __GFP_ACCOUNT flag. Currently, the list of accounted objects is quite limited and only includes those allocations that (1) are known to be easily triggered from userspace and (2) can fail gracefully (for the full list see patch no. 6) and it still misses many object types. However, accounting only those objects should be a satisfactory approximation of the behavior we used to have for most sane workloads. This patch (of 6): Revert 499611ed ("kernfs: do not account ino_ida allocations to memcg"). Black-list kmem accounting policy (aka __GFP_NOACCOUNT) turned out to be fragile and difficult to maintain, because there seem to be many more allocations that should not be accounted than those that should be. Besides, false accounting an allocation might result in much worse consequences than not accounting at all, namely increased memory consumption due to pinned dead kmem caches. So it was decided to switch to the white-list policy. This patch reverts bits introducing the black-list policy. The white-list policy will be introduced later in the series. Signed-off-by:
-
