linux/mm/memory.c, branch v4.0

mm: numa: slow PTE scan rate if migration failures occur

2015-03-25T23:20:31Z

Dave Chinner reported the following on https://lkml.org/lkml/2015/3/1/226 Across the board the 4.0-rc1 numbers are much slower, and the degradation is far worse when using the large memory footprint configs. Perf points straight at the cause - this is from 4.0-rc1 on the "-o bhash=101073" config: - 56.07% 56.07% [kernel] [k] default_send_IPI_mask_sequence_phys - default_send_IPI_mask_sequence_phys - 99.99% physflat_send_IPI_mask - 99.37% native_send_call_func_ipi smp_call_function_many - native_flush_tlb_others - 99.85% flush_tlb_page ptep_clear_flush try_to_unmap_one rmap_walk try_to_unmap migrate_pages migrate_misplaced_page - handle_mm_fault - 99.73% __do_page_fault trace_do_page_fault do_async_page_fault + async_page_fault 0.63% native_send_call_func_single_ipi generic_exec_single smp_call_function_single This is showing excessive migration activity even though excessive migrations are meant to get throttled. Normally, the scan rate is tuned on a per-task basis depending on the locality of faults. However, if migrations fail for any reason then the PTE scanner may scan faster if the faults continue to be remote. This means there is higher system CPU overhead and fault trapping at exactly the time we know that migrations cannot happen. This patch tracks when migration failures occur and slows the PTE scanner. Signed-off-by: Mel Gorman Reported-by: Dave Chinner Tested-by: Dave Chinner Cc: Ingo Molnar Cc: Aneesh Kumar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: numa: preserve PTE write permissions across a NUMA hinting fault

2015-03-25T23:20:31Z

Protecting a PTE to trap a NUMA hinting fault clears the writable bit and further faults are needed after trapping a NUMA hinting fault to set the writable bit again. This patch preserves the writable bit when trapping NUMA hinting faults. The impact is obvious from the number of minor faults trapped during the basis balancing benchmark and the system CPU usage; autonumabench 4.0.0-rc4 4.0.0-rc4 baseline preserve Time System-NUMA01 107.13 ( 0.00%) 103.13 ( 3.73%) Time System-NUMA01_THEADLOCAL 131.87 ( 0.00%) 83.30 ( 36.83%) Time System-NUMA02 8.95 ( 0.00%) 10.72 (-19.78%) Time System-NUMA02_SMT 4.57 ( 0.00%) 3.99 ( 12.69%) Time Elapsed-NUMA01 515.78 ( 0.00%) 517.26 ( -0.29%) Time Elapsed-NUMA01_THEADLOCAL 384.10 ( 0.00%) 384.31 ( -0.05%) Time Elapsed-NUMA02 48.86 ( 0.00%) 48.78 ( 0.16%) Time Elapsed-NUMA02_SMT 47.98 ( 0.00%) 48.12 ( -0.29%) 4.0.0-rc4 4.0.0-rc4 baseline preserve User 44383.95 43971.89 System 252.61 201.24 Elapsed 998.68 1000.94 Minor Faults 2597249 1981230 Major Faults 365 364 There is a similar drop in system CPU usage using Dave Chinner's xfsrepair workload 4.0.0-rc4 4.0.0-rc4 baseline preserve Amean real-xfsrepair 454.14 ( 0.00%) 442.36 ( 2.60%) Amean syst-xfsrepair 277.20 ( 0.00%) 204.68 ( 26.16%) The patch looks hacky but the alternatives looked worse. The tidest was to rewalk the page tables after a hinting fault but it was more complex than this approach and the performance was worse. It's not generally safe to just mark the page writable during the fault if it's a write fault as it may have been read-only for COW so that approach was discarded. Signed-off-by: Mel Gorman Reported-by: Dave Chinner Tested-by: Dave Chinner Cc: Ingo Molnar Cc: Aneesh Kumar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: numa: group related processes based on VMA flags instead of page table flags

2015-03-25T23:20:31Z

These are three follow-on patches based on the xfsrepair workload Dave Chinner reported was problematic in 4.0-rc1 due to changes in page table management -- https://lkml.org/lkml/2015/3/1/226. Much of the problem was reduced by commit 53da3bc2ba9e ("mm: fix up numa read-only thread grouping logic") and commit ba68bc0115eb ("mm: thp: Return the correct value for change_huge_pmd"). It was known that the performance in 3.19 was still better even if is far less safe. This series aims to restore the performance without compromising on safety. For the test of this mail, I'm comparing 3.19 against 4.0-rc4 and the three patches applied on top autonumabench 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanilla vmwrite-v5r8 preserve-v5r8 slowscan-v5r8 Time System-NUMA01 124.00 ( 0.00%) 161.86 (-30.53%) 107.13 ( 13.60%) 103.13 ( 16.83%) 145.01 (-16.94%) Time System-NUMA01_THEADLOCAL 115.54 ( 0.00%) 107.64 ( 6.84%) 131.87 (-14.13%) 83.30 ( 27.90%) 92.35 ( 20.07%) Time System-NUMA02 9.35 ( 0.00%) 10.44 (-11.66%) 8.95 ( 4.28%) 10.72 (-14.65%) 8.16 ( 12.73%) Time System-NUMA02_SMT 3.87 ( 0.00%) 4.63 (-19.64%) 4.57 (-18.09%) 3.99 ( -3.10%) 3.36 ( 13.18%) Time Elapsed-NUMA01 570.06 ( 0.00%) 567.82 ( 0.39%) 515.78 ( 9.52%) 517.26 ( 9.26%) 543.80 ( 4.61%) Time Elapsed-NUMA01_THEADLOCAL 393.69 ( 0.00%) 384.83 ( 2.25%) 384.10 ( 2.44%) 384.31 ( 2.38%) 380.73 ( 3.29%) Time Elapsed-NUMA02 49.09 ( 0.00%) 49.33 ( -0.49%) 48.86 ( 0.47%) 48.78 ( 0.63%) 50.94 ( -3.77%) Time Elapsed-NUMA02_SMT 47.51 ( 0.00%) 47.15 ( 0.76%) 47.98 ( -0.99%) 48.12 ( -1.28%) 49.56 ( -4.31%) 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 User 46334.60 46391.94 44383.95 43971.89 44372.12 System 252.84 284.66 252.61 201.24 249.00 Elapsed 1062.14 1050.96 998.68 1000.94 1026.78 Overall the system CPU usage is comparable and the test is naturally a bit variable. The slowing of the scanner hurts numa01 but on this machine it is an adverse workload and patches that dramatically help it often hurt absolutely everything else. Due to patch 2, the fault activity is interesting 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 Minor Faults 2097811 2656646 2597249 1981230 1636841 Major Faults 362 450 365 364 365 Note the impact preserving the write bit across protection updates and fault reduces faults. NUMA alloc hit 1229008 1217015 1191660 1178322 1199681 NUMA alloc miss 0 0 0 0 0 NUMA interleave hit 0 0 0 0 0 NUMA alloc local 1228514 1216317 1190871 1177448 1199021 NUMA base PTE updates 245706197 240041607 238195516 244704842 115012800 NUMA huge PMD updates 479530 468448 464868 477573 224487 NUMA page range updates 491225557 479886983 476207932 489222218 229950144 NUMA hint faults 659753 656503 641678 656926 294842 NUMA hint local faults 381604 373963 360478 337585 186249 NUMA hint local percent 57 56 56 51 63 NUMA pages migrated 5412140 6374899 6266530 5277468 5755096 AutoNUMA cost 5121% 5083% 4994% 5097% 2388% Here the impact of slowing the PTE scanner on migratrion failures is obvious as "NUMA base PTE updates" and "NUMA huge PMD updates" are massively reduced even though the headline performance is very similar. As xfsrepair was the reported workload here is the impact of the series on it. xfsrepair 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanilla vmwrite-v5r8 preserve-v5r8 slowscan-v5r8 Min real-fsmark 1183.29 ( 0.00%) 1165.73 ( 1.48%) 1152.78 ( 2.58%) 1153.64 ( 2.51%) 1177.62 ( 0.48%) Min syst-fsmark 4107.85 ( 0.00%) 4027.75 ( 1.95%) 3986.74 ( 2.95%) 3979.16 ( 3.13%) 4048.76 ( 1.44%) Min real-xfsrepair 441.51 ( 0.00%) 463.96 ( -5.08%) 449.50 ( -1.81%) 440.08 ( 0.32%) 439.87 ( 0.37%) Min syst-xfsrepair 195.76 ( 0.00%) 278.47 (-42.25%) 262.34 (-34.01%) 203.70 ( -4.06%) 143.64 ( 26.62%) Amean real-fsmark 1188.30 ( 0.00%) 1177.34 ( 0.92%) 1157.97 ( 2.55%) 1158.21 ( 2.53%) 1182.22 ( 0.51%) Amean syst-fsmark 4111.37 ( 0.00%) 4055.70 ( 1.35%) 3987.19 ( 3.02%) 3998.72 ( 2.74%) 4061.69 ( 1.21%) Amean real-xfsrepair 450.88 ( 0.00%) 468.32 ( -3.87%) 454.14 ( -0.72%) 442.36 ( 1.89%) 440.59 ( 2.28%) Amean syst-xfsrepair 199.66 ( 0.00%) 290.60 (-45.55%) 277.20 (-38.84%) 204.68 ( -2.51%) 150.55 ( 24.60%) Stddev real-fsmark 4.12 ( 0.00%) 10.82 (-162.29%) 4.14 ( -0.28%) 5.98 (-45.05%) 4.60 (-11.53%) Stddev syst-fsmark 2.63 ( 0.00%) 20.32 (-671.82%) 0.37 ( 85.89%) 16.47 (-525.59%) 15.05 (-471.79%) Stddev real-xfsrepair 6.87 ( 0.00%) 4.55 ( 33.75%) 3.46 ( 49.58%) 1.78 ( 74.12%) 0.52 ( 92.50%) Stddev syst-xfsrepair 3.02 ( 0.00%) 10.30 (-241.37%) 13.17 (-336.37%) 0.71 ( 76.63%) 5.00 (-65.61%) CoeffVar real-fsmark 0.35 ( 0.00%) 0.92 (-164.73%) 0.36 ( -2.91%) 0.52 (-48.82%) 0.39 (-12.10%) CoeffVar syst-fsmark 0.06 ( 0.00%) 0.50 (-682.41%) 0.01 ( 85.45%) 0.41 (-543.22%) 0.37 (-478.78%) CoeffVar real-xfsrepair 1.52 ( 0.00%) 0.97 ( 36.21%) 0.76 ( 49.94%) 0.40 ( 73.62%) 0.12 ( 92.33%) CoeffVar syst-xfsrepair 1.51 ( 0.00%) 3.54 (-134.54%) 4.75 (-214.31%) 0.34 ( 77.20%) 3.32 (-119.63%) Max real-fsmark 1193.39 ( 0.00%) 1191.77 ( 0.14%) 1162.90 ( 2.55%) 1166.66 ( 2.24%) 1188.50 ( 0.41%) Max syst-fsmark 4114.18 ( 0.00%) 4075.45 ( 0.94%) 3987.65 ( 3.08%) 4019.45 ( 2.30%) 4082.80 ( 0.76%) Max real-xfsrepair 457.80 ( 0.00%) 474.60 ( -3.67%) 457.82 ( -0.00%) 444.42 ( 2.92%) 441.03 ( 3.66%) Max syst-xfsrepair 203.11 ( 0.00%) 303.65 (-49.50%) 294.35 (-44.92%) 205.33 ( -1.09%) 155.28 ( 23.55%) The really relevant lines as syst-xfsrepair which is the system CPU usage when running xfsrepair. Note that on my machine the overhead was 45% higher on 4.0-rc4 which may be part of what Dave is seeing. Once we preserve the write bit across faults, it's only 2.51% higher on average. With the full series applied, system CPU usage is 24.6% lower on average. Again, the impact of preserving the write bit on minor faults is obvious and the impact of slowing scanning after migration failures is obvious on the PTE updates. Note also that the number of pages migrated is much reduced even though the headline performance is comparable. 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 Minor Faults 153466827 254507978 249163829 153501373 105737890 Major Faults 610 702 690 649 724 NUMA base PTE updates 217735049 210756527 217729596 216937111 144344993 NUMA huge PMD updates 129294 85044 106921 127246 79887 NUMA pages migrated 21938995 29705270 28594162 22687324 16258075 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 Mean sdb-avgqusz 13.47 2.54 2.55 2.47 2.49 Mean sdb-avgrqsz 202.32 140.22 139.50 139.02 138.12 Mean sdb-await 25.92 5.09 5.33 5.02 5.22 Mean sdb-r_await 4.71 0.19 0.83 0.51 0.11 Mean sdb-w_await 104.13 5.21 5.38 5.05 5.32 Mean sdb-svctm 0.59 0.13 0.14 0.13 0.14 Mean sdb-rrqm 0.16 0.00 0.00 0.00 0.00 Mean sdb-wrqm 3.59 1799.43 1826.84 1812.21 1785.67 Max sdb-avgqusz 111.06 12.13 14.05 11.66 15.60 Max sdb-avgrqsz 255.60 190.34 190.01 187.33 191.78 Max sdb-await 168.24 39.28 49.22 44.64 65.62 Max sdb-r_await 660.00 52.00 280.00 76.00 12.00 Max sdb-w_await 7804.00 39.28 49.22 44.64 65.62 Max sdb-svctm 4.00 2.82 2.86 1.98 2.84 Max sdb-rrqm 8.30 0.00 0.00 0.00 0.00 Max sdb-wrqm 34.20 5372.80 5278.60 5386.60 5546.15 FWIW, I also checked SPECjbb in different configurations but it's similar observations -- minor faults lower, PTE update activity lower and performance is roughly comparable against 3.19. This patch (of 3): Threads that share writable data within pages are grouped together as related tasks. This decision is based on whether the PTE is marked dirty which is subject to timing races between the PTE scanner update and when the application writes the page. If the page is file-backed, then background flushes and sync also affect placement. This is unpredictable behaviour which is impossible to reason about so this patch makes grouping decisions based on the VMA flags. Signed-off-by: Mel Gorman Reported-by: Dave Chinner Tested-by: Dave Chinner Cc: Ingo Molnar Cc: Aneesh Kumar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: fix up numa read-only thread grouping logic

2015-03-12T15:45:46Z

Dave Chinner reported that commit 4d9424669946 ("mm: convert p[te|md]_mknonnuma and remaining page table manipulations") slowed down his xfsrepair test enormously. In particular, it was using more system time due to extra TLB flushing. The ultimate reason turns out to be how the change to use the regular page table accessor functions broke the NUMA grouping logic. The old special mknuma/mknonnuma code accessed the page table present bit and the magic NUMA bit directly, while the new code just changes the page protections using PROT_NONE and the regular vma protections. That sounds equivalent, and from a fault standpoint it really is, but a subtle side effect is that the *other* protection bits of the page table entries also change. And the code to decide how to group the NUMA entries together used the writable bit to decide whether a particular page was likely to be shared read-only or not. And with the change to make the NUMA handling use the regular permission setting functions, that writable bit was basically always cleared for private mappings due to COW. So even if the page actually ends up being written to in the end, the NUMA balancing would act as if it was always shared RO. This code is a heuristic anyway, so the fix - at least for now - is to instead check whether the page is dirty rather than writable. The bit doesn't change with protection changes. NOTE! This also adds a FIXME comment to revisit this issue, Not only should we probably re-visit the whole "is this a shared read-only page" heuristic (we might want to take the vma permissions into account and base this more on those than the per-page ones, and also look at whether the particular access that triggers it is a write or not), but the whole COW issue shows that we should think about the NUMA fault handling some more. For example, maybe we should do the early-COW thing that a regular fault does. Or maybe we should accept that while using the same bits as PROTNONE was a good thing (and got rid of the specual NUMA bit), we might still want to just preseve the other protection bits across NUMA faulting. Those are bigger questions, left for later. This just fixes up the heuristic so that it at least approximates working again. More analysis and work needed. Reported-by: Dave Chinner Tested-by: Mel Gorman Cc: Andrew Morton Cc: Aneesh Kumar Cc: Ingo Molnar , Signed-off-by: Linus Torvalds

mm: allow page fault handlers to perform the COW

2015-02-17T01:56:03Z

Currently COW of an XIP file is done by first bringing in a read-only mapping, then retrying the fault and copying the page. It is much more efficient to tell the fault handler that a COW is being attempted (by passing in the pre-allocated page in the vm_fault structure), and allow the handler to perform the COW operation itself. The handler cannot insert the page itself if there is already a read-only mapping at that address, so allow the handler to return VM_FAULT_LOCKED and set the fault_page to be NULL. This indicates to the MM code that the i_mmap_lock is held instead of the page lock. Signed-off-by: Matthew Wilcox Acked-by: Kirill A. Shutemov Cc: Andreas Dilger Cc: Boaz Harrosh Cc: Christoph Hellwig Cc: Dave Chinner Cc: Jan Kara Cc: Jens Axboe Cc: Mathieu Desnoyers Cc: Randy Dunlap Cc: Ross Zwisler Cc: Theodore Ts'o Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: fix XIP fault vs truncate race

2015-02-17T01:56:02Z

DAX is a replacement for the variation of XIP currently supported by the ext2 filesystem. We have three different things in the tree called 'XIP', and the new focus is on access to data rather than executables, so a name change was in order. DAX stands for Direct Access. The X is for eXciting. The new focus on data access has resulted in more careful attention to races that exist in the current XIP code, but are not hit by the use-case that it was designed for. XIP's architecture worked fine for ext2, but DAX is architected to work with modern filsystems such as ext4 and XFS. DAX is not intended for use with btrfs; the value that btrfs adds relies on manipulating data and writing data to different locations, while DAX's value is for write-in-place and keeping the kernel from touching the data. DAX was developed in order to support NV-DIMMs, but it's become clear that its usefuless extends beyond NV-DIMMs and there are several potential customers including the tracing machinery. Other people want to place the kernel log in an area of memory, as long as they have a BIOS that does not clear DRAM on reboot. Patch 1 is a bug fix, probably worth including in 3.18. Patches 2 & 3 are infrastructure for DAX. Patches 4-8 replace the XIP code with its DAX equivalents, transforming ext2 to use the DAX code as we go. Note that patch 10 is the Documentation patch. Patches 9-15 clean up after the XIP code, removing the infrastructure that is no longer needed and renaming various XIP things to DAX. Most of these patches were added after Jan found things he didn't like in an earlier version of the ext4 patch ... that had been copied from ext2. So ext2 i being transformed to do things the same way that ext4 will later. The ability to mount ext2 filesystems with the 'xip' option is retained, although the 'dax' option is now preferred. Patch 16 adds some DAX infrastructure to support ext4. Patch 17 adds DAX support to ext4. It is broadly similar to ext2's DAX support, but it is more efficient than ext4's due to its support for unwritten extents. Patch 18 is another cleanup patch renaming XIP to DAX. My thanks to Mathieu Desnoyers for his reviews of the v11 patchset. Most of the changes below were based on his feedback. This patch (of 18): Pagecache faults recheck i_size after taking the page lock to ensure that the fault didn't race against a truncate. We don't have a page to lock in the XIP case, so use i_mmap_lock_read() instead. It is locked in the truncate path in unmap_mapping_range() after updating i_size. So while we hold it in the fault path, we are guaranteed that either i_size has already been updated in the truncate path, or that the truncate will subsequently call zap_page_range_single() and so remove the mapping we have just inserted. There is a window of time in which i_size has been reduced and the thread has a mapping to a page which will be removed from the file, but this is harmless as the page will not be allocated to a different purpose before the thread's access to it is revoked. [akpm@linux-foundation.org: switch to i_mmap_lock_read(), add comment in unmap_single_vma()] Signed-off-by: Matthew Wilcox Reviewed-by: Jan Kara Acked-by: Kirill A. Shutemov Reviewed-by: Mathieu Desnoyers Cc: Andreas Dilger Cc: Boaz Harrosh Cc: Christoph Hellwig Cc: Dave Chinner Cc: Jens Axboe Cc: Randy Dunlap Cc: Ross Zwisler Cc: Theodore Ts'o Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm/memory.c: actually remap enough memory

2015-02-13T02:54:11Z

For whatever reason, generic_access_phys() only remaps one page, but actually allows to access arbitrary size. It's quite easy to trigger large reads, like printing out large structure with gdb, which leads to a crash. Fix it by remapping correct size. Fixes: 28b2ee20c7cb ("access_process_vm device memory infrastructure") Signed-off-by: Grazvydas Ignotas Cc: Rik van Riel Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: numa: add paranoid check around pte_protnone_numa

2015-02-13T02:54:08Z

pte_protnone_numa is only safe to use after VMA checks for PROT_NONE are complete. Treating a real PROT_NONE PTE as a NUMA hinting fault is going to result in strangeness so add a check for it. BUG_ON looks like overkill but if this is hit then it's a serious bug that could result in corruption so do not even try recovering. It would have been more comprehensive to check VMA flags in pte_protnone_numa but it would have made the API ugly just for a debugging check. Signed-off-by: Mel Gorman Cc: Aneesh Kumar K.V Cc: Benjamin Herrenschmidt Cc: Dave Jones Cc: Hugh Dickins Cc: Ingo Molnar Cc: Kirill Shutemov Cc: Linus Torvalds Cc: Paul Mackerras Cc: Rik van Riel Cc: Sasha Levin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: numa: do not trap faults on the huge zero page

2015-02-13T02:54:08Z

Faults on the huge zero page are pointless and there is a BUG_ON to catch them during fault time. This patch reintroduces a check that avoids marking the zero page PAGE_NONE. Signed-off-by: Mel Gorman Cc: Aneesh Kumar K.V Cc: Benjamin Herrenschmidt Cc: Dave Jones Cc: Hugh Dickins Cc: Ingo Molnar Cc: Kirill Shutemov Cc: Linus Torvalds Cc: Paul Mackerras Cc: Rik van Riel Cc: Sasha Levin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds

mm: convert p[te|md]_mknonnuma and remaining page table manipulations

2015-02-13T02:54:08Z

With PROT_NONE, the traditional page table manipulation functions are sufficient. [andre.przywara@arm.com: fix compiler warning in pmdp_invalidate()] [akpm@linux-foundation.org: fix build with STRICT_MM_TYPECHECKS] Signed-off-by: Mel Gorman Acked-by: Linus Torvalds Acked-by: Aneesh Kumar Tested-by: Sasha Levin Cc: Benjamin Herrenschmidt Cc: Dave Jones Cc: Hugh Dickins Cc: Ingo Molnar Cc: Kirill Shutemov Cc: Paul Mackerras Cc: Rik van Riel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds