- 26 Oct, 2018 40 commits
-
-
Johannes Weiner authored
The refault stats go better with the page fault stats, and are of higher interest than the stats on LRU operations. In fact they used to be grouped together; when the LRU operation stats were added later on, they were wedged in between. Move them back together. Documentation/admin-guide/cgroup-v2.rst already lists them in the right order. Link: http://lkml.kernel.org/r/20181010140239.GA2527@cmpxchg.orgSigned-off-by:
Johannes Weiner <hannes@cmpxchg.org> Cc: Rik van Riel <riel@redhat.com> Cc: Michal Hocko <mhocko@suse.com> Cc: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Johannes Weiner authored
The page cache and most shrinkable slab caches hold data that has been read from disk, but there are some caches that only cache CPU work, such as the dentry and inode caches of procfs and sysfs, as well as the subset of radix tree nodes that track non-resident page cache. Currently, all these are shrunk at the same rate: using DEFAULT_SEEKS for the shrinker's seeks setting tells the reclaim algorithm that for every two page cache pages scanned it should scan one slab object. This is a bogus setting. A virtual inode that required no IO to create is not twice as valuable as a page cache page; shadow cache entries with eviction distances beyond the size of memory aren't either. In most cases, the behavior in practice is still fine. Such virtual caches don't tend to grow and assert themselves aggressively, and usually get picked up before they cause problems. But there are scenarios where that's not true. Our database workloads suffer from two of those. For one, their file workingset is several times bigger than available memory, which has the kernel aggressively create shadow page cache entries for the non-resident parts of it. The workingset code does tell the VM that most of these are expendable, but the VM ends up balancing them 2:1 to cache pages as per the seeks setting. This is a huge waste of memory. These workloads also deal with tens of thousands of open files and use /proc for introspection, which ends up growing the proc_inode_cache to absurdly large sizes - again at the cost of valuable cache space, which isn't a reasonable trade-off, given that proc inodes can be re-created without involving the disk. This patch implements a "zero-seek" setting for shrinkers that results in a target ratio of 0:1 between their objects and IO-backed caches. This allows such virtual caches to grow when memory is available (they do cache/avoid CPU work after all), but effectively disables them as soon as IO-backed objects are under pressure. It then switches the shrinkers for procfs and sysfs metadata, as well as excess page cache shadow nodes, to the new zero-seek setting. Link: http://lkml.kernel.org/r/20181009184732.762-5-hannes@cmpxchg.orgSigned-off-by:
Domas Mituzas <dmituzas@fb.com> Reviewed-by:
Rik van Riel <riel@surriel.com> Acked-by:
Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by:
-
Johannes Weiner authored
Make it easier to catch bugs in the shadow node shrinker by adding a counter for the shadow nodes in circulation. [akpm@linux-foundation.org: assert that irqs are disabled, for __inc_lruvec_page_state()] [akpm@linux-foundation.org: s/WARN_ON_ONCE/VM_WARN_ON_ONCE/, per Johannes] Link: http://lkml.kernel.org/r/20181009184732.762-4-hannes@cmpxchg.orgSigned-off-by:
-
Johannes Weiner authored
No need to use the preemption-safe lruvec state function inside the reclaim region that has irqs disabled. Link: http://lkml.kernel.org/r/20181009184732.762-3-hannes@cmpxchg.orgSigned-off-by:
-
Johannes Weiner authored
On a system that executes multiple cgrouped jobs and independent workloads, we don't just care about the health of the overall system, but also that of individual jobs, so that we can ensure individual job health, fairness between jobs, or prioritize some jobs over others. This patch implements pressure stall tracking for cgroups. In kernels with CONFIG_PSI=y, cgroup2 groups will have cpu.pressure, memory.pressure, and io.pressure files that track aggregate pressure stall times for only the tasks inside the cgroup. Link: http://lkml.kernel.org/r/20180828172258.3185-10-hannes@cmpxchg.orgSigned-off-by:
Tejun Heo <tj@kernel.org> Acked-by:
Daniel Drake <drake@endlessm.com> Tested-by:
Suren Baghdasaryan <surenb@google.com> Cc: Christopher Lameter <cl@linux.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Johannes Weiner <jweiner@fb.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Enderborg <peter.enderborg@sony.com> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: Shakeel Butt <shakeelb@google.com> Cc: Vinayak Menon <vinmenon@codeaurora.org> Signed-off-by:
-
Johannes Weiner authored
When systems are overcommitted and resources become contended, it's hard to tell exactly the impact this has on workload productivity, or how close the system is to lockups and OOM kills. In particular, when machines work multiple jobs concurrently, the impact of overcommit in terms of latency and throughput on the individual job can be enormous. In order to maximize hardware utilization without sacrificing individual job health or risk complete machine lockups, this patch implements a way to quantify resource pressure in the system. A kernel built with CONFIG_PSI=y creates files in /proc/pressure/ that expose the percentage of time the system is stalled on CPU, memory, or IO, respectively. Stall states are aggregate versions of the per-task delay accounting delays: cpu: some tasks are runnable but not executing on a CPU memory: tasks are reclaiming, or waiting for swapin or thrashing cache io: tasks are waiting for io completions These percentages of walltime can be thought of as pressure percentages, and they give a general sense of system health and productivity loss incurred by resource overcommit. They can also indicate when the system is approaching lockup scenarios and OOMs. To do this, psi keeps track of the task states associated with each CPU and samples the time they spend in stall states. Every 2 seconds, the samples are averaged across CPUs - weighted by the CPUs' non-idle time to eliminate artifacts from unused CPUs - and translated into percentages of walltime. A running average of those percentages is maintained over 10s, 1m, and 5m periods (similar to the loadaverage). [hannes@cmpxchg.org: doc fixlet, per Randy] Link: http://lkml.kernel.org/r/20180828205625.GA14030@cmpxchg.org [hannes@cmpxchg.org: code optimization] Link: http://lkml.kernel.org/r/20180907175015.GA8479@cmpxchg.org [hannes@cmpxchg.org: rename psi_clock() to psi_update_work(), per Peter] Link: http://lkml.kernel.org/r/20180907145404.GB11088@cmpxchg.org [hannes@cmpxchg.org: fix build] Link: http://lkml.kernel.org/r/20180913014222.GA2370@cmpxchg.org Link: http://lkml.kernel.org/r/20180828172258.3185-9-hannes@cmpxchg.orgSigned-off-by: -
Johannes Weiner authored
do_sched_yield() disables IRQs, looks up this_rq() and locks it. The next patch is adding another site with the same pattern, so provide a convenience function for it. Link: http://lkml.kernel.org/r/20180828172258.3185-8-hannes@cmpxchg.orgSigned-off-by:
-
Johannes Weiner authored
kernel/sched/sched.h includes "stats.h" half-way through the file. The next patch introduces users of sched.h's rq locking functions and update_rq_clock() in kernel/sched/stats.h. Move those definitions up in the file so they are available in stats.h. Link: http://lkml.kernel.org/r/20180828172258.3185-7-hannes@cmpxchg.orgSigned-off-by:
-
Johannes Weiner authored
It's going to be used in a later patch. Keep the churn separate. Link: http://lkml.kernel.org/r/20180828172258.3185-6-hannes@cmpxchg.orgSigned-off-by:
-
Johannes Weiner authored
There are several definitions of those functions/macros in places that mess with fixed-point load averages. Provide an official version. [akpm@linux-foundation.org: fix missed conversion in block/blk-iolatency.c] Link: http://lkml.kernel.org/r/20180828172258.3185-5-hannes@cmpxchg.orgSigned-off-by:
-
Johannes Weiner authored
Delay accounting already measures the time a task spends in direct reclaim and waiting for swapin, but in low memory situations tasks spend can spend a significant amount of their time waiting on thrashing page cache. This isn't tracked right now. To know the full impact of memory contention on an individual task, measure the delay when waiting for a recently evicted active cache page to read back into memory. Also update tools/accounting/getdelays.c: [hannes@computer accounting]$ sudo ./getdelays -d -p 1 print delayacct stats ON PID 1 CPU count real total virtual total delay total delay average 50318 745000000 847346785 400533713 0.008ms IO count delay total delay average 435 122601218 0ms SWAP count delay total delay average 0 0 0ms RECLAIM count delay total delay average 0 0 0ms THRASHING count delay total delay average 19 12621439 0ms Link: http://lkml.kernel.org/r/20180828172258.3185-4-hannes@cmpxchg.orgSigned-off-by: -
Johannes Weiner authored
Refaults happen during transitions between workingsets as well as in-place thrashing. Knowing the difference between the two has a range of applications, including measuring the impact of memory shortage on the system performance, as well as the ability to smarter balance pressure between the filesystem cache and the swap-backed workingset. During workingset transitions, inactive cache refaults and pushes out established active cache. When that active cache isn't stale, however, and also ends up refaulting, that's bonafide thrashing. Introduce a new page flag that tells on eviction whether the page has been active or not in its lifetime. This bit is then stored in the shadow entry, to classify refaults as transitioning or thrashing. How many page->flags does this leave us with on 32-bit? 20 bits are always page flags 21 if you have an MMU 23 with the zone bits for DMA, Normal, HighMem, Movable 29 with the sparsemem section bits 30 if PAE is enabled 31 with this patch. So on 32-bit PAE, that leaves 1 bit for distinguishing two NUMA nodes. If that's not enough, the system can switch to discontigmem and re-gain the 6 or 7 sparsemem section bits. Link: http://lkml.kernel.org/r/20180828172258.3185-3-hannes@cmpxchg.orgSigned-off-by:
-
Johannes Weiner authored
Patch series "psi: pressure stall information for CPU, memory, and IO", v4. Overview PSI reports the overall wallclock time in which the tasks in a system (or cgroup) wait for (contended) hardware resources. This helps users understand the resource pressure their workloads are under, which allows them to rootcause and fix throughput and latency problems caused by overcommitting, underprovisioning, suboptimal job placement in a grid; as well as anticipate major disruptions like OOM. Real-world applications We're using the data collected by PSI (and its previous incarnation, memdelay) quite extensively at Facebook, and with several success stories. One usecase is avoiding OOM hangs/livelocks. The reason these happen is because the OOM killer is triggered by reclaim not being able to free pages, but with fast flash devices there is *always* some clean and uptodate cache to reclaim; the OOM killer never kicks in, even as tasks spend 90% of the time thrashing the cache pages of their own executables. There is no situation where this ever makes sense in practice. We wrote a <100 line POC python script to monitor memory pressure and kill stuff way before such pathological thrashing leads to full system losses that would require forcible hard resets. We've since extended and deployed this code into other places to guarantee latency and throughput SLAs, since they're usually violated way before the kernel OOM killer would ever kick in. It is available here: https://github.com/facebookincubator/oomd Eventually we probably want to trigger the in-kernel OOM killer based on extreme sustained pressure as well, so that Linux can avoid memory livelocks - which technically aren't deadlocks, but to the user indistinguishable from them - out of the box. We'd continue using OOMD as the first line of defense to ensure workload health and implement complex kill policies that are beyond the scope of the kernel. We also use PSI memory pressure for loadshedding. Our batch job infrastructure used to use heuristics based on various VM stats to anticipate OOM situations, with lackluster success. We switched it to PSI and managed to anticipate and avoid OOM kills and lockups fairly reliably. The reduction of OOM outages in the worker pool raised the pool's aggregate productivity, and we were able to switch that service to smaller machines. Lastly, we use cgroups to isolate a machine's main workload from maintenance crap like package upgrades, logging, configuration, as well as to prevent multiple workloads on a machine from stepping on each others' toes. We were not able to configure this properly without the pressure metrics; we would see latency or bandwidth drops, but it would often be hard to impossible to rootcause it post-mortem. We now log and graph pressure for the containers in our fleet and can trivially link latency spikes and throughput drops to shortages of specific resources after the fact, and fix the job config/scheduling. PSI has also received testing, feedback, and feature requests from Android and EndlessOS for the purpose of low-latency OOM killing, to intervene in pressure situations before the UI starts hanging. How do you use this feature? A kernel with CONFIG_PSI=y will create a /proc/pressure directory with 3 files: cpu, memory, and io. If using cgroup2, cgroups will also have cpu.pressure, memory.pressure and io.pressure files, which simply aggregate task stalls at the cgroup level instead of system-wide. The cpu file contains one line: some avg10=2.04 avg60=0.75 avg300=0.40 total=157656722 The averages give the percentage of walltime in which one or more tasks are delayed on the runqueue while another task has the CPU. They're recent averages over 10s, 1m, 5m windows, so you can tell short term trends from long term ones, similarly to the load average. The total= value gives the absolute stall time in microseconds. This allows detecting latency spikes that might be too short to sway the running averages. It also allows custom time averaging in case the 10s/1m/5m windows aren't adequate for the usecase (or are too coarse with future hardware). What to make of this "some" metric? If CPU utilization is at 100% and CPU pressure is 0, it means the system is perfectly utilized, with one runnable thread per CPU and nobody waiting. At two or more runnable tasks per CPU, the system is 100% overcommitted and the pressure average will indicate as much. From a utilization perspective this is a great state of course: no CPU cycles are being wasted, even when 50% of the threads were to go idle (as most workloads do vary). From the perspective of the individual job it's not great, however, and they would do better with more resources. Depending on what your priority and options are, raised "some" numbers may or may not require action. The memory file contains two lines: some avg10=70.24 avg60=68.52 avg300=69.91 total=3559632828 full avg10=57.59 avg60=58.06 avg300=60.38 total=3300487258 The some line is the same as for cpu, the time in which at least one task is stalled on the resource. In the case of memory, this includes waiting on swap-in, page cache refaults and page reclaim. The full line, however, indicates time in which *nobody* is using the CPU productively due to pressure: all non-idle tasks are waiting for memory in one form or another. Significant time spent in there is a good trigger for killing things, moving jobs to other machines, or dropping incoming requests, since neither the jobs nor the machine overall are making too much headway. The io file is similar to memory. Because the block layer doesn't have a concept of hardware contention right now (how much longer is my IO request taking due to other tasks?), it reports CPU potential lost on all IO delays, not just the potential lost due to competition. FAQ Q: How is PSI's CPU component different from the load average? A: There are several quirks in the load average that make it hard to impossible to tell how overcommitted the CPU really is. 1. The load average is reported as a raw number of active tasks. You need to know how many CPUs there are in the system, how many CPUs the workload is allowed to use, then think about what the proportion between load and the number of CPUs mean for the tasks trying to run. PSI reports the percentage of wallclock time in which tasks are waiting for a CPU to run on. It doesn't matter how many CPUs are present or usable. The number always tells the quality of life of tasks in the system or in a particular cgroup. 2. The shortest averaging window is 1m, which is extremely coarse, and it's sampled in 5s intervals. A *lot* can happen on a CPU in 5 seconds. This *may* be able to identify persistent long-term trends and very clear and obvious overloads, but it's unusable for latency spikes and more subtle overutilization. PSI's shortest window is 10s. It also exports the cumulative stall times (in microseconds) of synchronously recorded events. 3. On Linux, the load average for historical reasons includes all TASK_UNINTERRUPTIBLE tasks. This gives a broader sense of how busy the system is, but on the flipside it doesn't distinguish whether tasks are likely to contend over the CPU or IO - which obviously requires very different interventions from a sys admin or a job scheduler. PSI reports independent metrics for CPU and IO. You can tell which resource is making the tasks wait, but in conjunction still see how overloaded the system is overall. Q: What's the cost / performance impact of this feature? A: PSI's primary cost is in the scheduler, in particular task wakeups and sleeps. I benchmarked this code using Facebook's two most scheduling sensitive workloads: memcache and webserver. They handle a ton of small requests - lots of wakeups and sleeps with little actual work in between - so they tend to be canaries for scheduler regressions. In the tests, the boxes were handling live traffic over the course of several hours. Half the machines, the control, ran with CONFIG_PSI=n. For memcache I used eight machines total. They're 2-socket, 14 core, 56 thread boxes. The test runs for half the test period, flips the test and control kernels on the hardware to rule out HW factors, DC location etc., then runs the other half of the test. For the webservers, I used 32 machines total. They're single socket, 16 core, 32 thread machines. During the memcache test, CPU load was nopsi=78.05% psi=78.98% in the first half and nopsi=77.52% psi=78.25%, so PSI added between 0.7 and 0.9 percentage points to the CPU load, a difference of about 1%. UPDATE: I re-ran this test with the v3 version of this patch set and the CPU utilization was equivalent between test and control. UPDATE: v4 is on par with v3. As far as end-to-end request latency from the client perspective goes, we don't sample those finely enough to capture the requests going to those particular machines during the test, but we know the p50 turnaround time in this workload is 54us, and perf bench sched pipe on those machines show nopsi=5.232666 us/op and psi=5.587347 us/op, so this doesn't add much here either. The profile for the pipe benchmark shows: 0.87% sched-pipe [kernel.vmlinux] [k] psi_group_change 0.83% perf.real [kernel.vmlinux] [k] psi_group_change 0.82% perf.real [kernel.vmlinux] [k] psi_task_change 0.58% sched-pipe [kernel.vmlinux] [k] psi_task_change The webserver load is running inside 4 nested cgroup levels. The CPU load with both nopsi and psi kernels was indistinguishable at 81%. For comparison, we had to disable the cgroup cpu controller on the webservers because it added 4 percentage points to the CPU% during this same exact test. Versions of this accounting code now run on 80% of our fleet. None of our workloads have reported regressions during the rollout. Daniel Drake said: : I just retested the latest version at : http://git.cmpxchg.org/cgit.cgi/linux-psi.git (Linux 4.18) and the results : are great. : : Test setup: : Endless OS : GeminiLake N4200 low end laptop : 2GB RAM : swap (and zram swap) disabled : : Baseline test: open a handful of large-ish apps and several website : tabs in Google Chrome. : : Results: after a couple of minutes, system is excessively thrashing, mouse : cursor can barely be moved, UI is not responding to mouse clicks, so it's : impractical to recover from this situation as an ordinary user : : Add my simple killer: : https://gist.github.com/dsd/a8988bf0b81a6163475988120fe8d9cd : : Results: when the thrashing causes the UI to become sluggish, the killer : steps in and kills something (usually a chrome tab), and the system : remains usable. I repeatedly opened more apps and more websites over a 15 : minute period but I wasn't able to get the system to a point of UI : unresponsiveness. Suren said: : Backported to 4.9 and retested on ARMv8 8 code system running Android. : Signals behave as expected reacting to memory pressure, no jumps in : "total" counters that would indicate an overflow/underflow issues. Nicely : done! This patch (of 9): If we keep just enough refault information to match the *current* page cache during reclaim time, we could lose a lot of events when there is only a temporary spike in non-cache memory consumption that pushes out all the cache. Once cache comes back, we won't see those refaults. They might not be actionable for LRU aging, but we want to know about them for measuring memory pressure. [hannes@cmpxchg.org: switch to NUMA-aware lru and slab counters] Link: http://lkml.kernel.org/r/20181009184732.762-2-hannes@cmpxchg.org Link: http://lkml.kernel.org/r/20180828172258.3185-2-hannes@cmpxchg.orgSigned-off-by:
Johannes Weiner <jweiner@fb.com> Acked-by:
-
Vlastimil Babka authored
Kmalloc cache names can get quite long for large object sizes, when the sizes are expressed in bytes. Use 'k' and 'M' prefixes to make the names as short as possible e.g. in /proc/slabinfo. This works, as we mostly use power-of-two sizes, with exceptions only below 1k. Example: 'kmalloc-4194304' becomes 'kmalloc-4M' Link: http://lkml.kernel.org/r/20180731090649.16028-7-vbabka@suse.czSuggested-by:
Matthew Wilcox <willy@infradead.org> Signed-off-by:
Vlastimil Babka <vbabka@suse.cz> Acked-by:
Mel Gorman <mgorman@techsingularity.net> Acked-by:
Christoph Lameter <cl@linux.com> Acked-by:
Roman Gushchin <guro@fb.com> Cc: David Rientjes <rientjes@google.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Laura Abbott <labbott@redhat.com> Cc: Michal Hocko <mhocko@kernel.org> Cc: Sumit Semwal <sumit.semwal@linaro.org> Cc: Vijayanand Jitta <vjitta@codeaurora.org> Signed-off-by:
-
Vlastimil Babka authored
The vmstat NR_KERNEL_MISC_RECLAIMABLE counter is for kernel non-slab allocations that can be reclaimed via shrinker. In /proc/meminfo, we can show the sum of all reclaimable kernel allocations (including slab) as "KReclaimable". Add the same counter also to per-node meminfo under /sys With this counter, users will have more complete information about kernel memory usage. Non-slab reclaimable pages (currently just the ION allocator) will not be missing from /proc/meminfo, making users wonder where part of their memory went. More precisely, they already appear in MemAvailable, but without the new counter, it's not obvious why the value in MemAvailable doesn't fully correspond with the sum of other counters participating in it. Link: http://lkml.kernel.org/r/20180731090649.16028-6-vbabka@suse.czSigned-off-by:
-
Vlastimil Babka authored
The vmstat counter NR_INDIRECTLY_RECLAIMABLE_BYTES was introduced by commit eb592546 ("mm: introduce NR_INDIRECTLY_RECLAIMABLE_BYTES") with the goal of accounting objects that can be reclaimed, but cannot be allocated via a SLAB_RECLAIM_ACCOUNT cache. This is now possible via kmalloc() with __GFP_RECLAIMABLE flag, and the dcache external names user is converted. The counter is however still useful for accounting direct page allocations (i.e. not slab) with a shrinker, such as the ION page pool. So keep it, and: - change granularity to pages to be more like other counters; sub-page allocations should be able to use kmalloc - rename the counter to NR_KERNEL_MISC_RECLAIMABLE - expose the counter again in vmstat as "nr_kernel_misc_reclaimable"; we can again remove the check for not printing "hidden" counters Link: http://lkml.kernel.org/r/20180731090649.16028-5-vbabka@suse.czSigned-off-by:
-
Vlastimil Babka authored
We can use the newly introduced kmalloc-reclaimable-X caches, to allocate external names in dcache, which will take care of the proper accounting automatically, and also improve anti-fragmentation page grouping. This effectively reverts commit f1782c9b ("dcache: account external names as indirectly reclaimable memory") and instead passes __GFP_RECLAIMABLE to kmalloc(). The accounting thus moves from NR_INDIRECTLY_RECLAIMABLE_BYTES to NR_SLAB_RECLAIMABLE, which is also considered in MemAvailable calculation and overcommit decisions. Link: http://lkml.kernel.org/r/20180731090649.16028-4-vbabka@suse.czSigned-off-by:
-
Vlastimil Babka authored
Kmem caches can be created with a SLAB_RECLAIM_ACCOUNT flag, which indicates they contain objects which can be reclaimed under memory pressure (typically through a shrinker). This makes the slab pages accounted as NR_SLAB_RECLAIMABLE in vmstat, which is reflected also the MemAvailable meminfo counter and in overcommit decisions. The slab pages are also allocated with __GFP_RECLAIMABLE, which is good for anti-fragmentation through grouping pages by mobility. The generic kmalloc-X caches are created without this flag, but sometimes are used also for objects that can be reclaimed, which due to varying size cannot have a dedicated kmem cache with SLAB_RECLAIM_ACCOUNT flag. A prominent example are dcache external names, which prompted the creation of a new, manually managed vmstat counter NR_INDIRECTLY_RECLAIMABLE_BYTES in commit f1782c9b ("dcache: account external names as indirectly reclaimable memory"). To better handle this and any other similar cases, this patch introduces SLAB_RECLAIM_ACCOUNT variants of kmalloc caches, named kmalloc-rcl-X. They are used whenever the kmalloc() call passes __GFP_RECLAIMABLE among gfp flags. They are added to the kmalloc_caches array as a new type. Allocations with both __GFP_DMA and __GFP_RECLAIMABLE will use a dma type cache. This change only applies to SLAB and SLUB, not SLOB. This is fine, since SLOB's target are tiny system and this patch does add some overhead of kmem management objects. Link: http://lkml.kernel.org/r/20180731090649.16028-3-vbabka@suse.czSigned-off-by:
-
Vlastimil Babka authored
Patch series "kmalloc-reclaimable caches", v4. As discussed at LSF/MM [1] here's a patchset that introduces kmalloc-reclaimable caches (more details in the second patch) and uses them for dcache external names. That allows us to repurpose the NR_INDIRECTLY_RECLAIMABLE_BYTES counter later in the series. With patch 3/6, dcache external names are allocated from kmalloc-rcl-* caches, eliminating the need for manual accounting. More importantly, it also ensures the reclaimable kmalloc allocations are grouped in pages separate from the regular kmalloc allocations. The need for proper accounting of dcache external names has shown it's easy for misbehaving process to allocate lots of them, causing premature OOMs. Without the added grouping, it's likely that a similar workload can interleave the dcache external names allocations with regular kmalloc allocations (note: I haven't searched myself for an example of such regular kmalloc allocation, but I would be very surprised if there wasn't some). A pathological case would be e.g. one 64byte regular allocations with 63 external dcache names in a page (64x64=4096), which means the page is not freed even after reclaiming after all dcache names, and the process can thus "steal" the whole page with single 64byte allocation. If other kmalloc users similar to dcache external names become identified, they can also benefit from the new functionality simply by adding __GFP_RECLAIMABLE to the kmalloc calls. Side benefits of the patchset (that could be also merged separately) include removed branch for detecting __GFP_DMA kmalloc(), and shortening kmalloc cache names in /proc/slabinfo output. The latter is potentially an ABI break in case there are tools parsing the names and expecting the values to be in bytes. This is how /proc/slabinfo looks like after booting in virtme: ... kmalloc-rcl-4M 0 0 4194304 1 1024 : tunables 1 1 0 : slabdata 0 0 0 ... kmalloc-rcl-96 7 32 128 32 1 : tunables 120 60 8 : slabdata 1 1 0 kmalloc-rcl-64 25 128 64 64 1 : tunables 120 60 8 : slabdata 2 2 0 kmalloc-rcl-32 0 0 32 124 1 : tunables 120 60 8 : slabdata 0 0 0 kmalloc-4M 0 0 4194304 1 1024 : tunables 1 1 0 : slabdata 0 0 0 kmalloc-2M 0 0 2097152 1 512 : tunables 1 1 0 : slabdata 0 0 0 kmalloc-1M 0 0 1048576 1 256 : tunables 1 1 0 : slabdata 0 0 0 ... /proc/vmstat with renamed nr_indirectly_reclaimable_bytes counter: ... nr_slab_reclaimable 2817 nr_slab_unreclaimable 1781 ... nr_kernel_misc_reclaimable 0 ... /proc/meminfo with new KReclaimable counter: ... Shmem: 564 kB KReclaimable: 11260 kB Slab: 18368 kB SReclaimable: 11260 kB SUnreclaim: 7108 kB KernelStack: 1248 kB ... This patch (of 6): The kmalloc caches currently mainain separate (optional) array kmalloc_dma_caches for __GFP_DMA allocations. There are tests for __GFP_DMA in the allocation hotpaths. We can avoid the branches by combining kmalloc_caches and kmalloc_dma_caches into a single two-dimensional array where the outer dimension is cache "type". This will also allow to add kmalloc-reclaimable caches as a third type. Link: http://lkml.kernel.org/r/20180731090649.16028-2-vbabka@suse.czSigned-off-by:
-
Andrea Arcangeli authored
get_mempolicy(MPOL_F_NODE|MPOL_F_ADDR) called a get_user_pages that would not be waiting for userfaults before failing and it would hit on a SIGBUS instead. Using get_user_pages_locked/unlocked instead will allow get_mempolicy to allow userfaults to resolve the fault and fill the hole, before grabbing the node id of the page. If the user calls get_mempolicy() with MPOL_F_ADDR | MPOL_F_NODE for an address inside an area managed by uffd and there is no page at that address, the page allocation from within get_mempolicy() will fail because get_user_pages() does not allow for page fault retry required for uffd; the user will get SIGBUS. With this patch, the page fault will be resolved by the uffd and the get_mempolicy() will continue normally. Background: Via code review, previously the syscall would have returned -EFAULT (vm_fault_to_errno), now it will block and wait for an userfault (if it's waken before the fault is resolved it'll still -EFAULT). This way get_mempolicy will give a chance to an "unaware" app to be compliant with userfaults. The reason this visible change is that becoming "userfault compliant" cannot regress anything: all other syscalls including read(2)/write(2) had to become "userfault compliant" long time ago (that's one of the things userfaultfd can do that PROT_NONE and trapping segfaults can't). So this is just one more syscall that become "userfault compliant" like all other major ones already were. This has been happening on virtio-bridge dpdk process which just called get_mempolicy on the guest space post live migration, but before the memory had a chance to be migrated to destination. I didn't run an strace to be able to show the -EFAULT going away, but I've the confirmation of the below debug aid information (only visible with CONFIG_DEBUG_VM=y) going away with the patch: [20116.371461] FAULT_FLAG_ALLOW_RETRY missing 0 [20116.371464] CPU: 1 PID: 13381 Comm: vhost-events Not tainted 4.17.12-200.fc28.x86_64 #1 [20116.371465] Hardware name: LENOVO 20FAS2BN0A/20FAS2BN0A, BIOS N1CET54W (1.22 ) 02/10/2017 [20116.371466] Call Trace: [20116.371473] dump_stack+0x5c/0x80 [20116.371476] handle_userfault.cold.37+0x1b/0x22 [20116.371479] ? remove_wait_queue+0x20/0x60 [20116.371481] ? poll_freewait+0x45/0xa0 [20116.371483] ? do_sys_poll+0x31c/0x520 [20116.371485] ? radix_tree_lookup_slot+0x1e/0x50 [20116.371488] shmem_getpage_gfp+0xce7/0xe50 [20116.371491] ? page_add_file_rmap+0x1a/0x2c0 [20116.371493] shmem_fault+0x78/0x1e0 [20116.371495] ? filemap_map_pages+0x3a1/0x450 [20116.371498] __do_fault+0x1f/0xc0 [20116.371500] __handle_mm_fault+0xe2e/0x12f0 [20116.371502] handle_mm_fault+0xda/0x200 [20116.371504] __get_user_pages+0x238/0x790 [20116.371506] get_user_pages+0x3e/0x50 [20116.371510] kernel_get_mempolicy+0x40b/0x700 [20116.371512] ? vfs_write+0x170/0x1a0 [20116.371515] __x64_sys_get_mempolicy+0x21/0x30 [20116.371517] do_syscall_64+0x5b/0x160 [20116.371520] entry_SYSCALL_64_after_hwframe+0x44/0xa9 The above harmless debug message (not a kernel crash, just a dump_stack()) is shown with CONFIG_DEBUG_VM=y to more quickly identify and improve kernel spots that may have to become "userfaultfd compliant" like this one (without having to run an strace and search for syscall misbehavior). Spots like the above are more closer to a kernel bug for the non-cooperative usages that Mike focuses on, than for for dpdk qemu-cooperative usages that reproduced it, but it's still nicer to get this fixed for dpdk too. The part of the patch that caused me to think is only the implementation issue of mpol_get, but it looks like it should work safe no matter the kind of mempolicy structure that is (the default static policy also starts at 1 so it'll go to 2 and back to 1 without crashing everything at 0). [rppt@linux.vnet.ibm.com: changelog addition] http://lkml.kernel.org/r/20180904073718.GA26916@rapoport-lnx Link: http://lkml.kernel.org/r/20180831214848.23676-1-aarcange@redhat.comSigned-off-by:Andrea Arcangeli <aarcange@redhat.com> Reported-by:
Maxime Coquelin <maxime.coquelin@redhat.com> Tested-by:
Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by:
Mike Rapoport <rppt@linux.vnet.ibm.com> Signed-off-by:
-
Mike Rapoport authored
Replace bootmem allocator with memblock and enable use of NO_BOOTMEM like on most other architectures. Alpha gets the description of the physical memory from the firmware as an array of memory clusters. Each cluster that is not reserved by the firmware is added to memblock.memory. Once the memblock.memory is set up, we reserve the kernel and initrd pages with memblock reserve. Since we don't need the bootmem bitmap anymore, the code that finds an appropriate place is removed. The conversion does not take care of NUMA support which is marked broken for more than 10 years now. Link: http://lkml.kernel.org/r/1535952894-10967-1-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:
-
Mike Rapoport authored
The unicore32 architecture already supports memblock and uses it for some early memory reservations, e.g initrd and the page tables. At some point unicore32 allocates the bootmem bitmap from the memblock and then hands over the memory reservations from memblock to bootmem. This patch removes the bootmem initialization and leaves memblock as the only boot time memory manager for unicore32. Link: http://lkml.kernel.org/r/1533326330-31677-8-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:
Guan Xuetao <gxt@pku.edu.cn> Cc: Ley Foon Tan <ley.foon.tan@intel.com> Cc: Richard Kuo <rkuo@codeaurora.org> Cc: Richard Weinberger <richard@nod.at> Cc: Rob Herring <robh@kernel.org> Cc: Michal Hocko <mhocko@kernel.org> Signed-off-by:
-
Mike Rapoport authored
Replace bootmem initialization with memblock_add and memblock_reserve calls and explicit initialization of {min,max}_low_pfn. Link: http://lkml.kernel.org/r/1533326330-31677-7-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:Richard Weinberger <richard@nod.at> Cc: Guan Xuetao <gxt@pku.edu.cn> Cc: Ley Foon Tan <ley.foon.tan@intel.com> Cc: Richard Kuo <rkuo@codeaurora.org> Cc: Rob Herring <robh@kernel.org> Cc: Michal Hocko <mhocko@kernel.org> Signed-off-by:
-
Mike Rapoport authored
The setup_physmem() function receives uml_physmem and uml_reserved as parameters and still used these global variables. Replace such usage with local variables. Link: http://lkml.kernel.org/r/1533326330-31677-6-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:
-
Mike Rapoport authored
Remove bootmem bitmap initialization and replace reserve_bootmem() with memblock_reserve(). Link: http://lkml.kernel.org/r/1533326330-31677-5-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:
Ley Foon Tan <ley.foon.tan@intel.com> Cc: Guan Xuetao <gxt@pku.edu.cn> Cc: Richard Kuo <rkuo@codeaurora.org> Cc: Richard Weinberger <richard@nod.at> Cc: Rob Herring <robh@kernel.org> Cc: Michal Hocko <mhocko@kernel.org> Signed-off-by:
-
Mike Rapoport authored
All we have to do is to enable memblock, the generic FDT code will take care of the rest. Link: http://lkml.kernel.org/r/1533326330-31677-4-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:
-
Mike Rapoport authored
Memory region size is rounded down to page boundary and with sub-page region it becomes 0 and there is no point to add an empty region. Moreover, when the base is less than PAGE_SIZE we get a bogus size as (base + size - 1) evaluates to -1. 8cccffc5 ("of: check for size < 0 after rounding in early_init_dt_add_memory_arch") introduced a test for wrap around for the case when base is not page aligned, the same test can be used to ignore sub-page region sizes. Link: http://lkml.kernel.org/r/1533326330-31677-3-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:
Rob Herring <robh@kernel.org> Cc: Guan Xuetao <gxt@pku.edu.cn> Cc: Ley Foon Tan <ley.foon.tan@intel.com> Cc: Richard Kuo <rkuo@codeaurora.org> Cc: Richard Weinberger <richard@nod.at> Cc: Michal Hocko <mhocko@kernel.org> Signed-off-by:
-
Mike Rapoport authored
Patch series "switch several architectures NO_BOOTMEM". These patches perform conversion to NO_BOOTMEM of hexagon, nios2, uml and unicore32. This patch (of 7): Add registration of the system memory with memblock, eliminate bootmem initialization and convert early memory reservations from bootmem to memblock. Link: http://lkml.kernel.org/r/1533326330-31677-2-git-send-email-rppt@linux.vnet.ibm.comSigned-off-by:
Richard Kuo <rkuo@codeaurora.org> Cc: Guan Xuetao <gxt@pku.edu.cn> Cc: Ley Foon Tan <ley.foon.tan@intel.com> Cc: Richard Weinberger <richard@nod.at> Cc: Rob Herring <robh@kernel.org> Cc: Michal Hocko <mhocko@kernel.org> Signed-off-by:
-
Matthew Wilcox authored
All callers convert its errno into a vm_fault_t, so convert it to return a vm_fault_t directly. Link: http://lkml.kernel.org/r/20180828145728.11873-11-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox authored
Both of its callers currently convert its errno return into a vm_fault_t, so move the conversion into __vm_insert_mixed(). Link: http://lkml.kernel.org/r/20180828145728.11873-10-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox authored
vm_insert_pfn_prot() is only called from vmf_insert_pfn_prot(), so inline it and convert some of the errnos into vm_fault codes earlier. Link: http://lkml.kernel.org/r/20180828145728.11873-9-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox authored
All callers are now converted to vmf_insert_pfn() so convert vmf_insert_pfn() from being a compatibility wrapper around vm_insert_pfn() to being a compatibility wrapper around vmf_insert_pfn_prot(). Link: http://lkml.kernel.org/r/20180828145728.11873-8-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox authored
Documentation and comments. Link: http://lkml.kernel.org/r/20180828145728.11873-7-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox authored
Now this is no longer used outside mm/memory.c, make it static. Link: http://lkml.kernel.org/r/20180828145728.11873-6-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox authored
Return vm_fault_t codes directly from the appropriate mm routines instead of converting from errnos ourselves. Fixes a minor bug where we'd return SIGBUS instead of the correct OOM code if we ran out of memory allocating page tables. Link: http://lkml.kernel.org/r/20180828145728.11873-5-willy@infradead.orgSigned-off-by:
Thomas Gleixner <tglx@linutronix.de> Reviewed-by:
-
Matthew Wilcox authored
Like vm_insert_pfn_prot(), but returns a vm_fault_t instead of an errno. Also unexport vm_insert_pfn_prot as it has no modular users. Link: http://lkml.kernel.org/r/20180828145728.11873-4-willy@infradead.orgSigned-off-by:
-
Matthew Wilcox authored
All callers are now converted to vmf_insert_mixed() so convert vmf_insert_mixed() from being a compatibility wrapper into the real function. Link: http://lkml.kernel.org/r/20180828145728.11873-3-willy@infradead.orgSigned-off-by:
-
Nicolas Pitre authored
cramfs is the only remaining user of vm_insert_mixed() and should be converted to vmf_insert_mixed(). Based on a previous patch from Matthew Wilcox. Link: http://lkml.kernel.org/r/nycvar.YSQ.7.76.1808290945450.10215@knanqh.ubzrSigned-off-by:
Nicolas Pitre <nico@linaro.org> Reviewed-by:
-
Souptick Joarder authored
As part of vm_fault_t conversion filemap_page_mkwrite() for the NOMMU case was missed. Now converted. Link: http://lkml.kernel.org/r/20180828174952.GA29229@jordon-HP-15-Notebook-PCSigned-off-by:
Souptick Joarder <jrdr.linux@gmail.com> Reviewed-by:
-
Oscar Salvador authored
check_for_memory() looks a bit confusing. First of all, we have this: if (N_MEMORY == N_NORMAL_MEMORY) return; Checking the ENUM declaration, looks like N_MEMORY canot be equal to N_NORMAL_MEMORY. I could not find where N_MEMORY is set to N_NORMAL_MEMORY, or the other way around either, so unless I am missing something, this condition will never evaluate to true. It makes sense to get rid of it. Moving forward, the operations within the loop look a bit confusing as well. We set N_HIGH_MEMORY unconditionally, and then we set N_NORMAL_MEMORY in case we have CONFIG_HIGHMEM (N_NORMAL_MEMORY != N_HIGH_MEMORY) and zone <= ZONE_NORMAL. (N_HIGH_MEMORY falls back to N_NORMAL_MEMORY on !CONFIG_HIGHMEM systems, and that is why we can just go ahead and set N_HIGH_MEMORY unconditionally) Although this works, it is a bit subtle. I think that this could be easier to follow: First, we should only set N_HIGH_MEMORY in case we have CONFIG_HIGHMEM. And then we should set N_NORMAL_MEMORY in case zone <= ZONE_NORMAL, without further checking whether we have CONFIG_HIGHMEM or not. Link: http://lkml.kernel.org/r/20180828210158.4617-1-osalvador@techadventures.netSigned-off-by:
Oscar Salvador <osalvador@suse.de> Reviewed-by:
-
