- 14 May, 2019 40 commits
-
-
Jérôme Glisse authored
This is a all in one helper that fault pages in a range and map them to a device so that every single device driver do not have to re-implement this common pattern. This is taken from ODP RDMA in preparation of ODP RDMA convertion. It will be use by nouveau and other drivers. [jglisse@redhat.com: Was using wrong field and wrong enum] Link: http://lkml.kernel.org/r/20190409175340.26614-1-jglisse@redhat.com Link: http://lkml.kernel.org/r/20190403193318.16478-12-jglisse@redhat.comSigned-off-by:
Jérôme Glisse <jglisse@redhat.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Souptick Joarder <jrdr.linux@gmail.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Ira Weiny <ira.weiny@intel.com> Cc: Matthew Wilcox <willy@infradead.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Jérôme Glisse authored
The device driver can have kernel thread or worker doing work against a process mm and it is useful for those to test wether the mm is dead or alive to avoid doing useless work. Add an helper to test that so that driver can bail out early if a process is dying. Note that the helper does not perform any lock synchronization and thus is just a hint ie a process might be dying but the helper might still return the process as alive. All HMM functions are safe to use in that case as HMM internal properly protect itself with lock. If driver use this helper with non HMM functions it should ascertain that it is safe to do so. Link: http://lkml.kernel.org/r/20190403193318.16478-11-jglisse@redhat.comSigned-off-by:
-
Jérôme Glisse authored
HMM mirror is a device driver helpers to mirror range of virtual address. It means that the process jobs running on the device can access the same virtual address as the CPU threads of that process. This patch adds support for mirroring mapping of file that are on a DAX block device (ie range of virtual address that is an mmap of a file in a filesystem on a DAX block device). There is no reason to not support such case when mirroring virtual address on a device. Note that unlike GUP code we do not take page reference hence when we back-off we have nothing to undo. [jglisse@redhat.com: move THP and hugetlbfs code path behind #if KCONFIG] Link: http://lkml.kernel.org/r/20190422163741.13029-1-jglisse@redhat.com Link: http://lkml.kernel.org/r/20190403193318.16478-10-jglisse@redhat.comSigned-off-by:
Ralph Campbell <rcampbell@nvidia.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: John Hubbard <jhubbard@nvidia.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Ira Weiny <ira.weiny@intel.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Souptick Joarder <jrdr.linux@gmail.com> Signed-off-by:
-
Jérôme Glisse authored
HMM mirror is a device driver helpers to mirror range of virtual address. It means that the process jobs running on the device can access the same virtual address as the CPU threads of that process. This patch adds support for hugetlbfs mapping (ie range of virtual address that are mmap of a hugetlbfs). [rcampbell@nvidia.com: fix initial PFN for hugetlbfs pages] Link: http://lkml.kernel.org/r/20190419233536.8080-1-rcampbell@nvidia.com Link: http://lkml.kernel.org/r/20190403193318.16478-9-jglisse@redhat.comSigned-off-by:
Ira Weiny <ira.weiny@intel.com> Cc: John Hubbard <jhubbard@nvidia.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Balbir Singh <bsingharora@gmail.com> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Souptick Joarder <jrdr.linux@gmail.com> Signed-off-by:
-
Jérôme Glisse authored
The HMM mirror API can be use in two fashions. The first one where the HMM user coalesce multiple page faults into one request and set flags per pfns for of those faults. The second one where the HMM user want to pre-fault a range with specific flags. For the latter one it is a waste to have the user pre-fill the pfn arrays with a default flags value. This patch adds a default flags value allowing user to set them for a range without having to pre-fill the pfn array. Link: http://lkml.kernel.org/r/20190403193318.16478-8-jglisse@redhat.comSigned-off-by:
-
Jérôme Glisse authored
A common use case for HMM mirror is user trying to mirror a range and before they could program the hardware it get invalidated by some core mm event. Instead of having user re-try right away to mirror the range provide a completion mechanism for them to wait for any active invalidation affecting the range. This also changes how hmm_range_snapshot() and hmm_range_fault() works by not relying on vma so that we can drop the mmap_sem when waiting and lookup the vma again on retry. Link: http://lkml.kernel.org/r/20190403193318.16478-7-jglisse@redhat.comSigned-off-by:
-
Jérôme Glisse authored
Minor optimization around hmm_pte_need_fault(). Rename for consistency between code, comments and documentation. Also improves the comments on all the possible returns values. Improve the function by returning the number of populated entries in pfns array. Link: http://lkml.kernel.org/r/20190403193318.16478-6-jglisse@redhat.comSigned-off-by:
-
Jérôme Glisse authored
Rename for consistency between code, comments and documentation. Also improves the comments on all the possible returns values. Improve the function by returning the number of populated entries in pfns array. Link: http://lkml.kernel.org/r/20190403193318.16478-5-jglisse@redhat.comSigned-off-by:
John Hubbard <jhubbard@nvidia.com> Reviewed-by:
-
Jérôme Glisse authored
Users of HMM might be using the snapshot information to do preparatory step like dma mapping pages to a device before checking for invalidation through hmm_vma_range_done() so do not erase that information and assume users will do the right thing. Link: http://lkml.kernel.org/r/20190403193318.16478-4-jglisse@redhat.comSigned-off-by:
-
Jérôme Glisse authored
Every time I read the code to check that the HMM structure does not vanish before it should thanks to the many lock protecting its removal i get a headache. Switch to reference counting instead it is much easier to follow and harder to break. This also remove some code that is no longer needed with refcounting. Link: http://lkml.kernel.org/r/20190403193318.16478-3-jglisse@redhat.comSigned-off-by:
-
Jérôme Glisse authored
To avoid random config build issue, select mmu notifier when HMM is selected. In any cases when HMM get selected it will be by users that will also wants the mmu notifier. Link: http://lkml.kernel.org/r/20190403193318.16478-2-jglisse@redhat.comSigned-off-by:
Balbir Singh <bsingharora@gmail.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Ira Weiny <ira.weiny@intel.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Souptick Joarder <jrdr.linux@gmail.com> Signed-off-by:
-
Mike Kravetz authored
hugetlb uses a fault mutex hash table to prevent page faults of the same pages concurrently. The key for shared and private mappings is different. Shared keys off address_space and file index. Private keys off mm and virtual address. Consider a private mappings of a populated hugetlbfs file. A fault will map the page from the file and if needed do a COW to map a writable page. Hugetlbfs hole punch uses the fault mutex to prevent mappings of file pages. It uses the address_space file index key. However, private mappings will use a different key and could race with this code to map the file page. This causes problems (BUG) for the page cache remove code as it expects the page to be unmapped. A sample stack is: page dumped because: VM_BUG_ON_PAGE(page_mapped(page)) kernel BUG at mm/filemap.c:169! ... RIP: 0010:unaccount_page_cache_page+0x1b8/0x200 ... Call Trace: __delete_from_page_cache+0x39/0x220 delete_from_page_cache+0x45/0x70 remove_inode_hugepages+0x13c/0x380 ? __add_to_page_cache_locked+0x162/0x380 hugetlbfs_fallocate+0x403/0x540 ? _cond_resched+0x15/0x30 ? __inode_security_revalidate+0x5d/0x70 ? selinux_file_permission+0x100/0x130 vfs_fallocate+0x13f/0x270 ksys_fallocate+0x3c/0x80 __x64_sys_fallocate+0x1a/0x20 do_syscall_64+0x5b/0x180 entry_SYSCALL_64_after_hwframe+0x44/0xa9 There seems to be another potential COW issue/race with this approach of different private and shared keys as noted in commit 8382d914 ("mm, hugetlb: improve page-fault scalability"). Since every hugetlb mapping (even anon and private) is actually a file mapping, just use the address_space index key for all mappings. This results in potentially more hash collisions. However, this should not be the common case. Link: http://lkml.kernel.org/r/20190328234704.27083-3-mike.kravetz@oracle.com Link: http://lkml.kernel.org/r/20190412165235.t4sscoujczfhuiyt@linux-r8p5 Fixes: b5cec28d ("hugetlbfs: truncate_hugepages() takes a range of pages") Signed-off-by:
Mike Kravetz <mike.kravetz@oracle.com> Reviewed-by:
Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Reviewed-by:
Davidlohr Bueso <dbueso@suse.de> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com> Cc: Michal Hocko <mhocko@kernel.org> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Mike Kravetz authored
When a huge page is allocated, PagePrivate() is set if the allocation consumed a reservation. When freeing a huge page, PagePrivate is checked. If set, it indicates the reservation should be restored. PagePrivate being set at free huge page time mostly happens on error paths. When huge page reservations are created, a check is made to determine if the mapping is associated with an explicitly mounted filesystem. If so, pages are also reserved within the filesystem. The default action when freeing a huge page is to decrement the usage count in any associated explicitly mounted filesystem. However, if the reservation is to be restored the reservation/use count within the filesystem should not be decrementd. Otherwise, a subsequent page allocation and free for the same mapping location will cause the file filesystem usage to go 'negative'. Filesystem Size Used Avail Use% Mounted on nodev 4.0G -4.0M 4.1G - /opt/hugepool To fix, when freeing a huge page do not adjust filesystem usage if PagePrivate() is set to indicate the reservation should be restored. I did not cc stable as the problem has been around since reserves were added to hugetlbfs and nobody has noticed. Link: http://lkml.kernel.org/r/20190328234704.27083-2-mike.kravetz@oracle.comSigned-off-by:
-
Baoquan He authored
The input parameter 'phys_index' of memory_block_action() is actually the section number, but not the phys_index of memory_block. This is a relic from the past when one memory block could only contain one section. Rename it to start_section_nr. And also in remove_memory_section(), the 'node_id' and 'phys_device' arguments are not used by anyone. Remove them. Link: http://lkml.kernel.org/r/20190329144250.14315-2-bhe@redhat.comSigned-off-by:
Baoquan He <bhe@redhat.com> Acked-by:
Michal Hocko <mhocko@suse.com> Reviewed-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Reviewed-by:
Mukesh Ojha <mojha@codeaurora.org> Reviewed-by:
Oscar Salvador <osalvador@suse.de> Signed-off-by:
-
Baoquan He authored
The code comment above sparse_add_one_section() is obsolete and incorrect. Clean it up and write a new one. Link: http://lkml.kernel.org/r/20190329144250.14315-1-bhe@redhat.comSigned-off-by:
Mike Rapoport <rppt@linux.ibm.com> Signed-off-by:
-
David Hildenbrand authored
These are leftovers from the pre-"general non-lru movable page" era. Link: http://lkml.kernel.org/r/20190329122649.28404-1-david@redhat.comSigned-off-by:
David Hildenbrand <david@redhat.com> Reviewed-by:
Michael S. Tsirkin <mst@redhat.com> Acked-by:
Pankaj Gupta <pagupta@redhat.com> Acked-by:
Rafael Aquini <aquini@redhat.com> Signed-off-by:
-
Peng Fan authored
There is no function named munlock_vma_pages(). Correct it to munlock_vma_page(). Link: http://lkml.kernel.org/r/20190402095609.27181-1-peng.fan@nxp.comSigned-off-by:
Peng Fan <peng.fan@nxp.com> Reviewed-by:
Vlastimil Babka <vbabka@suse.cz> Signed-off-by:
-
Oscar Salvador authored
NODEMASK_ALLOC is used to allocate a nodemask bitmap, and it does it by first determining whether it should be allocated on the stack or dynamically, depending on NODES_SHIFT. Right now, it goes the dynamic path whenever the nodemask_t is above 32 bytes. Although we could bump it to a reasonable value, the largest a nodemask_t can get is 128 bytes, so since __nr_hugepages_store_common is called from a rather short stack we can just get rid of the NODEMASK_ALLOC call here. This reduces some code churn and complexity. Link: http://lkml.kernel.org/r/20190402133415.21983-1-osalvador@suse.deSigned-off-by:
-
Mike Kravetz authored
The number of node specific huge pages can be set via a file such as: /sys/devices/system/node/node1/hugepages/hugepages-2048kB/nr_hugepages When a node specific value is specified, the global number of huge pages must also be adjusted. This adjustment is calculated as the specified node specific value + (global value - current node value). If the node specific value provided by the user is large enough, this calculation could overflow an unsigned long leading to a smaller than expected number of huge pages. To fix, check the calculation for overflow. If overflow is detected, use ULONG_MAX as the requested value. This is inline with the user request to allocate as many huge pages as possible. It was also noticed that the above calculation was done outside the hugetlb_lock. Therefore, the values could be inconsistent and result in underflow. To fix, the calculation is moved within the routine set_max_huge_pages() where the lock is held. In addition, the code in __nr_hugepages_store_common() which tries to handle the case of not being able to allocate a node mask would likely result in incorrect behavior. Luckily, it is very unlikely we will ever take this path. If we do, simply return ENOMEM. Link: http://lkml.kernel.org/r/20190328220533.19884-1-mike.kravetz@oracle.comSigned-off-by:
Jing Xiangfeng <jingxiangfeng@huawei.com> Reviewed-by:
-
Linxu Fang authored
342332e6 ("mm/page_alloc.c: introduce kernelcore=mirror option") and later patches rewrote the calculation of node spanned pages. e506b996 ("mem-hotplug: fix node spanned pages when we have a movable node"), but the current code still has problems, When we have a node with only zone_movable and the node id is not zero, the size of node spanned pages is double added. That's because we have an empty normal zone, and zone_start_pfn or zone_end_pfn is not between arch_zone_lowest_possible_pfn and arch_zone_highest_possible_pfn, so we need to use clamp to constrain the range just like the commit <96e907d1> (bootmem: Reimplement __absent_pages_in_range() using for_each_mem_pfn_range()). e.g. Zone ranges: DMA [mem 0x0000000000001000-0x0000000000ffffff] DMA32 [mem 0x0000000001000000-0x00000000ffffffff] Normal [mem 0x0000000100000000-0x000000023fffffff] Movable zone start for each node Node 0: 0x0000000100000000 Node 1: 0x0000000140000000 Early memory node ranges node 0: [mem 0x0000000000001000-0x000000000009efff] node 0: [mem 0x0000000000100000-0x00000000bffdffff] node 0: [mem 0x0000000100000000-0x000000013fffffff] node 1: [mem 0x0000000140000000-0x000000023fffffff] node 0 DMA spanned:0xfff present:0xf9e absent:0x61 node 0 DMA32 spanned:0xff000 present:0xbefe0 absent:0x40020 node 0 Normal spanned:0 present:0 absent:0 node 0 Movable spanned:0x40000 present:0x40000 absent:0 On node 0 totalpages(node_present_pages): 1048446 node_spanned_pages:1310719 node 1 DMA spanned:0 present:0 absent:0 node 1 DMA32 spanned:0 present:0 absent:0 node 1 Normal spanned:0x100000 present:0x100000 absent:0 node 1 Movable spanned:0x100000 present:0x100000 absent:0 On node 1 totalpages(node_present_pages): 2097152 node_spanned_pages:2097152 Memory: 6967796K/12582392K available (16388K kernel code, 3686K rwdata, 4468K rodata, 2160K init, 10444K bss, 5614596K reserved, 0K cma-reserved) It shows that the current memory of node 1 is double added. After this patch, the problem is fixed. node 0 DMA spanned:0xfff present:0xf9e absent:0x61 node 0 DMA32 spanned:0xff000 present:0xbefe0 absent:0x40020 node 0 Normal spanned:0 present:0 absent:0 node 0 Movable spanned:0x40000 present:0x40000 absent:0 On node 0 totalpages(node_present_pages): 1048446 node_spanned_pages:1310719 node 1 DMA spanned:0 present:0 absent:0 node 1 DMA32 spanned:0 present:0 absent:0 node 1 Normal spanned:0 present:0 absent:0 node 1 Movable spanned:0x100000 present:0x100000 absent:0 On node 1 totalpages(node_present_pages): 1048576 node_spanned_pages:1048576 memory: 6967796K/8388088K available (16388K kernel code, 3686K rwdata, 4468K rodata, 2160K init, 10444K bss, 1420292K reserved, 0K cma-reserved) Link: http://lkml.kernel.org/r/1554178276-10372-1-git-send-email-fanglinxu@huawei.comSigned-off-by:
Linxu Fang <fanglinxu@huawei.com> Cc: Taku Izumi <izumi.taku@jp.fujitsu.com> Cc: Xishi Qiu <qiuxishi@huawei.com> Cc: Michal Hocko <mhocko@suse.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Pavel Tatashin <pavel.tatashin@microsoft.com> Cc: Oscar Salvador <osalvador@suse.de> Signed-off-by:
-
Yafang Shao authored
There are three tracepoints using this template, which are mm_vmscan_direct_reclaim_begin, mm_vmscan_memcg_reclaim_begin, mm_vmscan_memcg_softlimit_reclaim_begin. Regarding mm_vmscan_direct_reclaim_begin, sc.may_writepage is !laptop_mode, that's a static setting, and reclaim_idx is derived from gfp_mask which is already show in this tracepoint. Regarding mm_vmscan_memcg_reclaim_begin, may_writepage is !laptop_mode too, and reclaim_idx is (MAX_NR_ZONES-1), which are both static value. mm_vmscan_memcg_softlimit_reclaim_begin is the same with mm_vmscan_memcg_reclaim_begin. So we can drop them all. Link: http://lkml.kernel.org/r/1553736322-32235-1-git-send-email-laoar.shao@gmail.comSigned-off-by:
Yafang Shao <laoar.shao@gmail.com> Acked-by:
-
Aneesh Kumar K.V authored
MADV_DONTNEED is handled with mmap_sem taken in read mode. We call page_mkclean without holding mmap_sem. MADV_DONTNEED implies that pages in the region are unmapped and subsequent access to the pages in that range is handled as a new page fault. This implies that if we don't have parallel access to the region when MADV_DONTNEED is run we expect those range to be unallocated. w.r.t page_mkclean() we need to make sure that we don't break the MADV_DONTNEED semantics. MADV_DONTNEED check for pmd_none without holding pmd_lock. This implies we skip the pmd if we temporarily mark pmd none. Avoid doing that while marking the page clean. Keep the sequence same for dax too even though we don't support MADV_DONTNEED for dax mapping The bug was noticed by code review and I didn't observe any failures w.r.t test run. This is similar to commit 58ceeb6b Author: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Date: Thu Apr 13 14:56:26 2017 -0700 thp: fix MADV_DONTNEED vs. MADV_FREE race commit ced10803 Author: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Date: Thu Apr 13 14:56:20 2017 -0700 thp: fix MADV_DONTNEED vs. numa balancing race Link: http://lkml.kernel.org/r/20190321040610.14226-1-aneesh.kumar@linux.ibm.comSigned-off-by:
Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Reviewed-by:
-
John Hubbard authored
A discussion of the overall problem is below. As mentioned in patch 0001, the steps are to fix the problem are: 1) Provide put_user_page*() routines, intended to be used for releasing pages that were pinned via get_user_pages*(). 2) Convert all of the call sites for get_user_pages*(), to invoke put_user_page*(), instead of put_page(). This involves dozens of call sites, and will take some time. 3) After (2) is complete, use get_user_pages*() and put_user_page*() to implement tracking of these pages. This tracking will be separate from the existing struct page refcounting. 4) Use the tracking and identification of these pages, to implement special handling (especially in writeback paths) when the pages are backed by a filesystem. Overview ======== Some kernel components (file systems, device drivers) need to access memory that is specified via process virtual address. For a long time, the API to achieve that was get_user_pages ("GUP") and its variations. However, GUP has critical limitations that have been overlooked; in particular, GUP does not interact correctly with filesystems in all situations. That means that file-backed memory + GUP is a recipe for potential problems, some of which have already occurred in the field. GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem code to get the struct page behind a virtual address and to let storage hardware perform a direct copy to or from that page. This is a short-lived access pattern, and as such, the window for a concurrent writeback of GUP'd page was small enough that there were not (we think) any reported problems. Also, userspace was expected to understand and accept that Direct IO was not synchronized with memory-mapped access to that data, nor with any process address space changes such as munmap(), mremap(), etc. Over the years, more GUP uses have appeared (virtualization, device drivers, RDMA) that can keep the pages they get via GUP for a long period of time (seconds, minutes, hours, days, ...). This long-term pinning makes an underlying design problem more obvious. In fact, there are a number of key problems inherent to GUP: Interactions with file systems ============================== File systems expect to be able to write back data, both to reclaim pages, and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain write access to the file memory pages means that such hardware can dirty the pages, without the filesystem being aware. This can, in some cases (depending on filesystem, filesystem options, block device, block device options, and other variables), lead to data corruption, and also to kernel bugs of the form: kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899! backtrace: ext4_writepage __writepage write_cache_pages ext4_writepages do_writepages __writeback_single_inode writeback_sb_inodes __writeback_inodes_wb wb_writeback wb_workfn process_one_work worker_thread kthread ret_from_fork ...which is due to the file system asserting that there are still buffer heads attached: ({ \ BUG_ON(!PagePrivate(page)); \ ((struct buffer_head *)page_private(page)); \ }) Dave Chinner's description of this is very clear: "The fundamental issue is that ->page_mkwrite must be called on every write access to a clean file backed page, not just the first one. How long the GUP reference lasts is irrelevant, if the page is clean and you need to dirty it, you must call ->page_mkwrite before it is marked writeable and dirtied. Every. Time." This is just one symptom of the larger design problem: real filesystems that actually write to a backing device, do not actually support get_user_pages() being called on their pages, and letting hardware write directly to those pages--even though that pattern has been going on since about 2005 or so. Long term GUP ============= Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a writeable mapping is created), and the pages are file-backed. That can lead to filesystem corruption. What happens is that when a file-backed page is being written back, it is first mapped read-only in all of the CPU page tables; the file system then assumes that nobody can write to the page, and that the page content is therefore stable. Unfortunately, the GUP callers generally do not monitor changes to the CPU pages tables; they instead assume that the following pattern is safe (it's not): get_user_pages() Hardware can keep a reference to those pages for a very long time, and write to it at any time. Because "hardware" here means "devices that are not a CPU", this activity occurs without any interaction with the kernel's file system code. for each page set_page_dirty put_page() In fact, the GUP documentation even recommends that pattern. Anyway, the file system assumes that the page is stable (nothing is writing to the page), and that is a problem: stable page content is necessary for many filesystem actions during writeback, such as checksum, encryption, RAID striping, etc. Furthermore, filesystem features like COW (copy on write) or snapshot also rely on being able to use a new page for as memory for that memory range inside the file. Corruption during write back is clearly possible here. To solve that, one idea is to identify pages that have active GUP, so that we can use a bounce page to write stable data to the filesystem. The filesystem would work on the bounce page, while any of the active GUP might write to the original page. This would avoid the stable page violation problem, but note that it is only part of the overall solution, because other problems remain. Other filesystem features that need to replace the page with a new one can be inhibited for pages that are GUP-pinned. This will, however, alter and limit some of those filesystem features. The only fix for that would be to require GUP users to monitor and respond to CPU page table updates. Subsystems such as ODP and HMM do this, for example. This aspect of the problem is still under discussion. Direct IO ========= Direct IO can cause corruption, if userspace does Direct-IO that writes to a range of virtual addresses that are mmap'd to a file. The pages written to are file-backed pages that can be under write back, while the Direct IO is taking place. Here, Direct IO races with a write back: it calls GUP before page_mkclean() has replaced the CPU pte with a read-only entry. The race window is pretty small, which is probably why years have gone by before we noticed this problem: Direct IO is generally very quick, and tends to finish up before the filesystem gets around to do anything with the page contents. However, it's still a real problem. The solution is to never let GUP return pages that are under write back, but instead, force GUP to take a write fault on those pages. That way, GUP will properly synchronize with the active write back. This does not change the required GUP behavior, it just avoids that race. Details ======= Introduces put_user_page(), which simply calls put_page(). This provides a way to update all get_user_pages*() callers, so that they call put_user_page(), instead of put_page(). Also introduces put_user_pages(), and a few dirty/locked variations, as a replacement for release_pages(), and also as a replacement for open-coded loops that release multiple pages. These may be used for subsequent performance improvements, via batching of pages to be released. This is the first step of fixing a problem (also described in [1] and [2]) with interactions between get_user_pages ("gup") and filesystems. Problem description: let's start with a bug report. Below, is what happens sometimes, under memory pressure, when a driver pins some pages via gup, and then marks those pages dirty, and releases them. Note that the gup documentation actually recommends that pattern. The problem is that the filesystem may do a writeback while the pages were gup-pinned, and then the filesystem believes that the pages are clean. So, when the driver later marks the pages as dirty, that conflicts with the filesystem's page tracking and results in a BUG(), like this one that I experienced: kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899! backtrace: ext4_writepage __writepage write_cache_pages ext4_writepages do_writepages __writeback_single_inode writeback_sb_inodes __writeback_inodes_wb wb_writeback wb_workfn process_one_work worker_thread kthread ret_from_fork ...which is due to the file system asserting that there are still buffer heads attached: ({ \ BUG_ON(!PagePrivate(page)); \ ((struct buffer_head *)page_private(page)); \ }) Dave Chinner's description of this is very clear: "The fundamental issue is that ->page_mkwrite must be called on every write access to a clean file backed page, not just the first one. How long the GUP reference lasts is irrelevant, if the page is clean and you need to dirty it, you must call ->page_mkwrite before it is marked writeable and dirtied. Every. Time." This is just one symptom of the larger design problem: real filesystems that actually write to a backing device, do not actually support get_user_pages() being called on their pages, and letting hardware write directly to those pages--even though that pattern has been going on since about 2005 or so. The steps are to fix it are: 1) (This patch): provide put_user_page*() routines, intended to be used for releasing pages that were pinned via get_user_pages*(). 2) Convert all of the call sites for get_user_pages*(), to invoke put_user_page*(), instead of put_page(). This involves dozens of call sites, and will take some time. 3) After (2) is complete, use get_user_pages*() and put_user_page*() to implement tracking of these pages. This tracking will be separate from the existing struct page refcounting. 4) Use the tracking and identification of these pages, to implement special handling (especially in writeback paths) when the pages are backed by a filesystem. [1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()" [2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()" Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.comSigned-off-by:Jan Kara <jack@suse.cz> Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs] Reviewed-by:
Christoph Lameter <cl@linux.com> Tested-by:
-
Alexandre Ghiti authored
On systems without CONTIG_ALLOC activated but that support gigantic pages, boottime reserved gigantic pages can not be freed at all. This patch simply enables the possibility to hand back those pages to memory allocator. Link: http://lkml.kernel.org/r/20190327063626.18421-5-alex@ghiti.frSigned-off-by:
Alexandre Ghiti <alex@ghiti.fr> Acked-by: David S. Miller <davem@davemloft.net> [sparc] Reviewed-by:
-
Alexandre Ghiti authored
This condition allows to define alloc_contig_range, so simplify it into a more accurate naming. Link: http://lkml.kernel.org/r/20190327063626.18421-4-alex@ghiti.frSigned-off-by:
-
Alexandre Ghiti authored
sparc actually supports gigantic pages and selecting ARCH_HAS_GIGANTIC_PAGE allows it to allocate and free gigantic pages at runtime. sparc allows configuration such as huge pages of 16GB, pages of 8KB and MAX_ORDER = 13 (default): HPAGE_SHIFT (34) - PAGE_SHIFT (13) = 21 >= MAX_ORDER (13) Link: http://lkml.kernel.org/r/20190327063626.18421-3-alex@ghiti.frSigned-off-by:
David S. Miller <davem@davemloft.net> Cc: Andy Lutomirsky <luto@kernel.org> Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Borislav Petkov <bp@alien8.de> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: "H . Peter Anvin" <hpa@zytor.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Mike Kravetz <mike.kravetz@oracle.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rich Felker <dalias@libc.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Will Deacon <will.deacon@arm.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Signed-off-by:
-
Alexandre Ghiti authored
Patch series "Fix free/allocation of runtime gigantic pages", v8. This series fixes sh and sparc that did not advertise their gigantic page support and then were not able to allocate and free those pages at runtime. It renames MEMORY_ISOLATION && COMPACTION || CMA condition into the more accurate CONTIG_ALLOC, since it allows the definition of alloc_contig_range function. Finally, it then fixes the wrong definition of ARCH_HAS_GIGANTIC_PAGE config that, without MEMORY_ISOLATION && COMPACTION || CMA defined, did not allow architectures to free boottime allocated gigantic pages although unrelated. This patch (of 4): sh actually supports gigantic pages and selecting ARCH_HAS_GIGANTIC_PAGE allows it to allocate and free gigantic pages at runtime. At least sdk7786_defconfig exposes such a configuration with huge pages of 64MB, pages of 4KB and MAX_ORDER = 11: HPAGE_SHIFT (26) - PAGE_SHIFT (12) = 14 >= MAX_ORDER (11) Link: http://lkml.kernel.org/r/20190327063626.18421-2-alex@ghiti.frSigned-off-by:
-
Mike Rapoport authored
The riscv version of free_initmem() differs from the generic one only in that it sets the freed memory to zero. Make ricsv use the generic version and poison the freed memory. Link: http://lkml.kernel.org/r/1550515285-17446-5-git-send-email-rppt@linux.ibm.comSigned-off-by:
Palmer Dabbelt <palmer@sifive.com> Reviewed-by:
-
Mike Rapoport authored
Various architectures including x86 poison the freed init memory. Do the same in the generic free_initmem implementation and switch sparc32 architecture that is identical to the generic code over to it now. Link: http://lkml.kernel.org/r/1550515285-17446-4-git-send-email-rppt@linux.ibm.comSigned-off-by:
-
Mike Rapoport authored
hexagon implementation of free_initmem() is currently empty and marked with comment * Todo: free pages between __init_begin and __init_end; possibly * some devtree related stuff as well. Switch it to the generic implementation. Link: http://lkml.kernel.org/r/1550515285-17446-3-git-send-email-rppt@linux.ibm.comSigned-off-by:
-
Mike Rapoport authored
Patch series "provide a generic free_initmem implementation", v2. Many architectures implement free_initmem() in exactly the same or very similar way: they wrap the call to free_initmem_default() with sometimes different 'poison' parameter. These patches switch those architectures to use a generic implementation that does free_initmem_default(POISON_FREE_INITMEM). This was inspired by Christoph's patches for free_initrd_mem [1] and I shamelessly copied changelog entries from his patches :) [1] https://lore.kernel.org/lkml/20190213174621.29297-1-hch@lst.de/ This patch (of 2): For most architectures free_initmem just a wrapper for the same free_initmem_default(-1) call. Provide that as a generic implementation marked __weak. Link: http://lkml.kernel.org/r/1550515285-17446-2-git-send-email-rppt@linux.ibm.comSigned-off-by:
-
Christoph Hellwig authored
Various architectures including x86 poison the freed initrd memory. Do the same in the generic free_initrd_mem implementation and switch a few more architectures that are identical to the generic code over to it now. Link: http://lkml.kernel.org/r/20190213174621.29297-9-hch@lst.deSigned-off-by:
Christoph Hellwig <hch@lst.de> Acked-by:
-
Christoph Hellwig authored
For most architectures free_initrd_mem just expands to the same free_reserved_area call. Provide that as a generic implementation marked __weak. Link: http://lkml.kernel.org/r/20190213174621.29297-8-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
No need to handle the freeing disable in arch code when we already have a core hook (and a different name for the option) for it. Link: http://lkml.kernel.org/r/20190213174621.29297-7-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
The code for kernels that support ramdisks or not is mostly the same. Unify it by using an IS_ENABLED for the info message, and moving the error message into a stub for populate_initrd_image. [cai@lca.pw: fix a compilation error] Link: http://lkml.kernel.org/r/20190328014806.36375-1-cai@lca.pw Link: http://lkml.kernel.org/r/20190213174621.29297-6-hch@lst.deSigned-off-by:
Qian Cai <cai@lca.pw> Acked-by:
-
Christoph Hellwig authored
This will allow for cleaner code sharing in the caller. Link: http://lkml.kernel.org/r/20190213174621.29297-5-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Factor the kexec logic into a separate helper, and then inline the rest of free_initrd into the only caller. Link: http://lkml.kernel.org/r/20190213174621.29297-4-hch@lst.deSigned-off-by:
-
Christoph Hellwig authored
Patch series "initramfs tidyups". I've spent some time chasing down behavior in initramfs and found plenty of opportunity to improve the code. A first stab on that is contained in this series. This patch (of 7): We free the initrd memory for all successful or error cases except for the case where opening /initrd.image fails, which looks like an oversight. Steven said: : This also changes the behaviour when CONFIG_INITRAMFS_FORCE is enabled : - specifically it means that the initrd is freed (previously it was : ignored and never freed). But that seems like reasonable behaviour and : the previous behaviour looks like another oversight. Link: http://lkml.kernel.org/r/20190213174621.29297-3-hch@lst.deSigned-off-by:
Steven Price <steven.price@arm.com> Acked-by:
-
Yue Hu authored
f022d8cb ("mm: cma: Don't crash on allocation if CMA area can't be activated") fixes the crash issue when activation fails via setting cma->count as 0, same logic exists if bitmap allocation fails. Link: http://lkml.kernel.org/r/20190325081309.6004-1-zbestahu@gmail.comSigned-off-by:
Yue Hu <huyue2@yulong.com> Reviewed-by:
Anshuman Khandual <anshuman.khandual@arm.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Laura Abbott <labbott@redhat.com> Cc: Mike Rapoport <rppt@linux.vnet.ibm.com> Cc: Randy Dunlap <rdunlap@infradead.org> Signed-off-by:
-
Johannes Weiner authored
Only memcg_numa_stat_show() uses those wrappers and the lru bitmasks, group them together. Link: http://lkml.kernel.org/r/20190228163020.24100-7-hannes@cmpxchg.orgSigned-off-by:
Johannes Weiner <hannes@cmpxchg.org> Reviewed-by:
Roman Gushchin <guro@fb.com> Cc: Michal Hocko <mhocko@kernel.org> Cc: Tejun Heo <tj@kernel.org> Signed-off-by:
-
