From: Dan Williams <dan.j.williams@intel.com> To: akpm@linux-foundation.org Cc: jack@suse.cz, linux-nvdimm@lists.01.org, Matthew Wilcox <mawilcox@microsoft.com>, linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org, hch@lst.de Subject: [PATCH v4 15/18] mm, fs, dax: use page->mapping to warn if dma collides with truncate Date: Sat, 23 Dec 2017 16:57:20 -0800 [thread overview] Message-ID: <151407704063.38751.15403501871756055025.stgit@dwillia2-desk3.amr.corp.intel.com> (raw) In-Reply-To: <151407695916.38751.2866053440557472361.stgit@dwillia2-desk3.amr.corp.intel.com> Catch cases where truncate encounters pages that are still under active dma. This warning is a canary for potential data corruption as truncated blocks could be allocated to a new file while the device is still perform i/o. Here is an example of a collision that this implementation catches: WARNING: CPU: 2 PID: 1286 at fs/dax.c:343 dax_disassociate_entry+0x55/0x80 [..] Call Trace: __dax_invalidate_mapping_entry+0x6c/0xf0 dax_delete_mapping_entry+0xf/0x20 truncate_exceptional_pvec_entries.part.12+0x1af/0x200 truncate_inode_pages_range+0x268/0x970 ? tlb_gather_mmu+0x10/0x20 ? up_write+0x1c/0x40 ? unmap_mapping_range+0x73/0x140 xfs_free_file_space+0x1b6/0x5b0 [xfs] ? xfs_file_fallocate+0x7f/0x320 [xfs] ? down_write_nested+0x40/0x70 ? xfs_ilock+0x21d/0x2f0 [xfs] xfs_file_fallocate+0x162/0x320 [xfs] ? rcu_read_lock_sched_held+0x3f/0x70 ? rcu_sync_lockdep_assert+0x2a/0x50 ? __sb_start_write+0xd0/0x1b0 ? vfs_fallocate+0x20c/0x270 vfs_fallocate+0x154/0x270 SyS_fallocate+0x43/0x80 entry_SYSCALL_64_fastpath+0x1f/0x96 Cc: Jan Kara <jack@suse.cz> Cc: Jeff Moyer <jmoyer@redhat.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Matthew Wilcox <mawilcox@microsoft.com> Cc: Ross Zwisler <ross.zwisler@linux.intel.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> --- fs/dax.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 56 insertions(+) diff --git a/fs/dax.c b/fs/dax.c index fadc1b13838b..7d9fff8a1195 100644 --- a/fs/dax.c +++ b/fs/dax.c @@ -366,6 +366,56 @@ static void put_unlocked_mapping_entry(struct address_space *mapping, dax_wake_mapping_entry_waiter(mapping, index, entry, false); } +static unsigned long dax_entry_size(void *entry) +{ + if (dax_is_zero_entry(entry)) + return 0; + else if (dax_is_empty_entry(entry)) + return 0; + else if (dax_is_pmd_entry(entry)) + return HPAGE_SIZE; + else + return PAGE_SIZE; +} + +#define for_each_entry_pfn(entry, pfn, end_pfn) \ + for (pfn = dax_radix_pfn(entry), \ + end_pfn = pfn + dax_entry_size(entry) / PAGE_SIZE; \ + pfn < end_pfn; \ + pfn++) + +static void dax_associate_entry(void *entry, struct address_space *mapping) +{ + unsigned long pfn, end_pfn; + + if (IS_ENABLED(CONFIG_FS_DAX_LIMITED)) + return; + + for_each_entry_pfn(entry, pfn, end_pfn) { + struct page *page = pfn_to_page(pfn); + + WARN_ON_ONCE(page->mapping); + page->mapping = mapping; + } +} + +static void dax_disassociate_entry(void *entry, struct address_space *mapping, + bool trunc) +{ + unsigned long pfn, end_pfn; + + if (IS_ENABLED(CONFIG_FS_DAX_LIMITED)) + return; + + for_each_entry_pfn(entry, pfn, end_pfn) { + struct page *page = pfn_to_page(pfn); + + WARN_ON_ONCE(trunc && page_ref_count(page) > 1); + WARN_ON_ONCE(page->mapping && page->mapping != mapping); + page->mapping = NULL; + } +} + /* * Find radix tree entry at given index. If it points to an exceptional entry, * return it with the radix tree entry locked. If the radix tree doesn't @@ -472,6 +522,7 @@ static void *grab_mapping_entry(struct address_space *mapping, pgoff_t index, } if (pmd_downgrade) { + dax_disassociate_entry(entry, mapping, false); radix_tree_delete(&mapping->page_tree, index); mapping->nrexceptional--; dax_wake_mapping_entry_waiter(mapping, index, entry, @@ -521,6 +572,7 @@ static int __dax_invalidate_mapping_entry(struct address_space *mapping, (radix_tree_tag_get(page_tree, index, PAGECACHE_TAG_DIRTY) || radix_tree_tag_get(page_tree, index, PAGECACHE_TAG_TOWRITE))) goto out; + dax_disassociate_entry(entry, mapping, trunc); radix_tree_delete(page_tree, index); mapping->nrexceptional--; ret = 1; @@ -617,6 +669,10 @@ static void *dax_insert_mapping_entry(struct address_space *mapping, spin_lock_irq(&mapping->tree_lock); new_entry = dax_radix_locked_entry(pfn, flags); + if (dax_entry_size(entry) != dax_entry_size(new_entry)) { + dax_disassociate_entry(entry, mapping, false); + dax_associate_entry(new_entry, mapping); + } if (dax_is_zero_entry(entry) || dax_is_empty_entry(entry)) { /* _______________________________________________ Linux-nvdimm mailing list Linux-nvdimm@lists.01.org https://lists.01.org/mailman/listinfo/linux-nvdimm
WARNING: multiple messages have this Message-ID (diff)
From: Dan Williams <dan.j.williams@intel.com> To: akpm@linux-foundation.org Cc: jack@suse.cz, Matthew Wilcox <mawilcox@microsoft.com>, linux-nvdimm@lists.01.org, linux-xfs@vger.kernel.org, Jeff Moyer <jmoyer@redhat.com>, linux-fsdevel@vger.kernel.org, ross.zwisler@linux.intel.com, hch@lst.de Subject: [PATCH v4 15/18] mm, fs, dax: use page->mapping to warn if dma collides with truncate Date: Sat, 23 Dec 2017 16:57:20 -0800 [thread overview] Message-ID: <151407704063.38751.15403501871756055025.stgit@dwillia2-desk3.amr.corp.intel.com> (raw) In-Reply-To: <151407695916.38751.2866053440557472361.stgit@dwillia2-desk3.amr.corp.intel.com> Catch cases where truncate encounters pages that are still under active dma. This warning is a canary for potential data corruption as truncated blocks could be allocated to a new file while the device is still perform i/o. Here is an example of a collision that this implementation catches: WARNING: CPU: 2 PID: 1286 at fs/dax.c:343 dax_disassociate_entry+0x55/0x80 [..] Call Trace: __dax_invalidate_mapping_entry+0x6c/0xf0 dax_delete_mapping_entry+0xf/0x20 truncate_exceptional_pvec_entries.part.12+0x1af/0x200 truncate_inode_pages_range+0x268/0x970 ? tlb_gather_mmu+0x10/0x20 ? up_write+0x1c/0x40 ? unmap_mapping_range+0x73/0x140 xfs_free_file_space+0x1b6/0x5b0 [xfs] ? xfs_file_fallocate+0x7f/0x320 [xfs] ? down_write_nested+0x40/0x70 ? xfs_ilock+0x21d/0x2f0 [xfs] xfs_file_fallocate+0x162/0x320 [xfs] ? rcu_read_lock_sched_held+0x3f/0x70 ? rcu_sync_lockdep_assert+0x2a/0x50 ? __sb_start_write+0xd0/0x1b0 ? vfs_fallocate+0x20c/0x270 vfs_fallocate+0x154/0x270 SyS_fallocate+0x43/0x80 entry_SYSCALL_64_fastpath+0x1f/0x96 Cc: Jan Kara <jack@suse.cz> Cc: Jeff Moyer <jmoyer@redhat.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Matthew Wilcox <mawilcox@microsoft.com> Cc: Ross Zwisler <ross.zwisler@linux.intel.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> --- fs/dax.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 56 insertions(+) diff --git a/fs/dax.c b/fs/dax.c index fadc1b13838b..7d9fff8a1195 100644 --- a/fs/dax.c +++ b/fs/dax.c @@ -366,6 +366,56 @@ static void put_unlocked_mapping_entry(struct address_space *mapping, dax_wake_mapping_entry_waiter(mapping, index, entry, false); } +static unsigned long dax_entry_size(void *entry) +{ + if (dax_is_zero_entry(entry)) + return 0; + else if (dax_is_empty_entry(entry)) + return 0; + else if (dax_is_pmd_entry(entry)) + return HPAGE_SIZE; + else + return PAGE_SIZE; +} + +#define for_each_entry_pfn(entry, pfn, end_pfn) \ + for (pfn = dax_radix_pfn(entry), \ + end_pfn = pfn + dax_entry_size(entry) / PAGE_SIZE; \ + pfn < end_pfn; \ + pfn++) + +static void dax_associate_entry(void *entry, struct address_space *mapping) +{ + unsigned long pfn, end_pfn; + + if (IS_ENABLED(CONFIG_FS_DAX_LIMITED)) + return; + + for_each_entry_pfn(entry, pfn, end_pfn) { + struct page *page = pfn_to_page(pfn); + + WARN_ON_ONCE(page->mapping); + page->mapping = mapping; + } +} + +static void dax_disassociate_entry(void *entry, struct address_space *mapping, + bool trunc) +{ + unsigned long pfn, end_pfn; + + if (IS_ENABLED(CONFIG_FS_DAX_LIMITED)) + return; + + for_each_entry_pfn(entry, pfn, end_pfn) { + struct page *page = pfn_to_page(pfn); + + WARN_ON_ONCE(trunc && page_ref_count(page) > 1); + WARN_ON_ONCE(page->mapping && page->mapping != mapping); + page->mapping = NULL; + } +} + /* * Find radix tree entry at given index. If it points to an exceptional entry, * return it with the radix tree entry locked. If the radix tree doesn't @@ -472,6 +522,7 @@ static void *grab_mapping_entry(struct address_space *mapping, pgoff_t index, } if (pmd_downgrade) { + dax_disassociate_entry(entry, mapping, false); radix_tree_delete(&mapping->page_tree, index); mapping->nrexceptional--; dax_wake_mapping_entry_waiter(mapping, index, entry, @@ -521,6 +572,7 @@ static int __dax_invalidate_mapping_entry(struct address_space *mapping, (radix_tree_tag_get(page_tree, index, PAGECACHE_TAG_DIRTY) || radix_tree_tag_get(page_tree, index, PAGECACHE_TAG_TOWRITE))) goto out; + dax_disassociate_entry(entry, mapping, trunc); radix_tree_delete(page_tree, index); mapping->nrexceptional--; ret = 1; @@ -617,6 +669,10 @@ static void *dax_insert_mapping_entry(struct address_space *mapping, spin_lock_irq(&mapping->tree_lock); new_entry = dax_radix_locked_entry(pfn, flags); + if (dax_entry_size(entry) != dax_entry_size(new_entry)) { + dax_disassociate_entry(entry, mapping, false); + dax_associate_entry(new_entry, mapping); + } if (dax_is_zero_entry(entry) || dax_is_empty_entry(entry)) { /*
next prev parent reply other threads:[~2017-12-24 1:00 UTC|newest] Thread overview: 136+ messages / expand[flat|nested] mbox.gz Atom feed top 2017-12-24 0:56 [PATCH v4 00/18] dax: fix dma vs truncate/hole-punch Dan Williams 2017-12-24 0:56 ` Dan Williams 2017-12-24 0:56 ` Dan Williams 2017-12-24 0:56 ` [PATCH v4 01/18] mm, dax: introduce pfn_t_special() Dan Williams 2017-12-24 0:56 ` Dan Williams 2018-01-04 8:16 ` Christoph Hellwig 2018-01-04 8:16 ` Christoph Hellwig 2017-12-24 0:56 ` [PATCH v4 02/18] ext4: auto disable dax instead of failing mount Dan Williams 2017-12-24 0:56 ` Dan Williams 2018-01-03 14:20 ` Jan Kara 2018-01-03 14:20 ` Jan Kara 2017-12-24 0:56 ` [PATCH v4 03/18] ext2: " Dan Williams 2017-12-24 0:56 ` Dan Williams 2018-01-03 14:21 ` Jan Kara 2018-01-03 14:21 ` Jan Kara 2017-12-24 0:56 ` [PATCH v4 04/18] dax: require 'struct page' by default for filesystem dax Dan Williams 2017-12-24 0:56 ` Dan Williams 2018-01-03 15:29 ` Jan Kara 2018-01-03 15:29 ` Jan Kara 2018-01-04 8:16 ` Christoph Hellwig 2018-01-04 8:16 ` Christoph Hellwig 2018-01-08 11:58 ` Gerald Schaefer 2018-01-08 11:58 ` Gerald Schaefer 2017-12-24 0:56 ` [PATCH v4 05/18] dax: stop using VM_MIXEDMAP for dax Dan Williams 2017-12-24 0:56 ` Dan Williams 2018-01-03 15:27 ` Jan Kara 2018-01-03 15:27 ` Jan Kara 2017-12-24 0:56 ` [PATCH v4 06/18] dax: stop using VM_HUGEPAGE " Dan Williams 2017-12-24 0:56 ` Dan Williams 2017-12-24 0:56 ` [PATCH v4 07/18] dax: store pfns in the radix Dan Williams 2017-12-24 0:56 ` Dan Williams 2017-12-27 0:17 ` Ross Zwisler 2017-12-27 0:17 ` Ross Zwisler 2018-01-02 20:15 ` Dan Williams 2018-01-02 20:15 ` Dan Williams 2018-01-03 15:39 ` Jan Kara 2018-01-03 15:39 ` Jan Kara 2017-12-24 0:56 ` [PATCH v4 08/18] tools/testing/nvdimm: add 'bio_delay' mechanism Dan Williams 2017-12-24 0:56 ` Dan Williams 2017-12-27 18:08 ` Ross Zwisler 2017-12-27 18:08 ` Ross Zwisler 2018-01-02 20:35 ` Dan Williams 2018-01-02 20:35 ` Dan Williams 2018-01-02 21:44 ` Dave Chinner 2018-01-02 21:44 ` Dave Chinner 2018-01-02 21:51 ` Dan Williams 2018-01-02 21:51 ` Dan Williams 2018-01-03 15:46 ` Jan Kara 2018-01-03 15:46 ` Jan Kara 2018-01-03 20:37 ` Jeff Moyer 2018-01-03 20:37 ` Jeff Moyer 2017-12-24 0:56 ` [PATCH v4 09/18] mm, dax: enable filesystems to trigger dev_pagemap ->page_free callbacks Dan Williams 2017-12-24 0:56 ` Dan Williams 2018-01-04 8:20 ` Christoph Hellwig 2018-01-04 8:20 ` Christoph Hellwig 2017-12-24 0:56 ` [PATCH v4 10/18] mm, dev_pagemap: introduce CONFIG_DEV_PAGEMAP_OPS Dan Williams 2017-12-24 0:56 ` Dan Williams 2018-01-04 8:25 ` Christoph Hellwig 2018-01-04 8:25 ` Christoph Hellwig 2017-12-24 0:56 ` [PATCH v4 11/18] fs, dax: introduce DEFINE_FSDAX_AOPS Dan Williams 2017-12-24 0:56 ` Dan Williams 2017-12-27 5:29 ` Matthew Wilcox 2017-12-27 5:29 ` Matthew Wilcox 2018-01-02 20:21 ` Dan Williams 2018-01-02 20:21 ` Dan Williams 2018-01-03 16:05 ` Jan Kara 2018-01-03 16:05 ` Jan Kara 2018-01-04 8:27 ` Christoph Hellwig 2018-01-04 8:27 ` Christoph Hellwig 2018-01-02 21:41 ` Dave Chinner 2018-01-02 21:41 ` Dave Chinner 2017-12-24 0:57 ` [PATCH v4 12/18] xfs: use DEFINE_FSDAX_AOPS Dan Williams 2017-12-24 0:57 ` Dan Williams 2018-01-02 21:15 ` Darrick J. Wong 2018-01-02 21:15 ` Darrick J. Wong 2018-01-02 21:40 ` Dan Williams 2018-01-02 21:40 ` Dan Williams 2018-01-03 16:09 ` Jan Kara 2018-01-03 16:09 ` Jan Kara 2018-01-04 8:28 ` Christoph Hellwig 2018-01-04 8:28 ` Christoph Hellwig 2017-12-24 0:57 ` [PATCH v4 13/18] ext4: " Dan Williams 2017-12-24 0:57 ` Dan Williams 2017-12-24 0:57 ` Dan Williams 2018-01-04 8:29 ` Christoph Hellwig 2018-01-04 8:29 ` Christoph Hellwig 2018-01-04 8:29 ` Christoph Hellwig 2017-12-24 0:57 ` [PATCH v4 14/18] ext2: " Dan Williams 2017-12-24 0:57 ` Dan Williams 2018-01-04 8:29 ` Christoph Hellwig 2018-01-04 8:29 ` Christoph Hellwig 2017-12-24 0:57 ` Dan Williams [this message] 2017-12-24 0:57 ` [PATCH v4 15/18] mm, fs, dax: use page->mapping to warn if dma collides with truncate Dan Williams 2018-01-04 8:30 ` Christoph Hellwig 2018-01-04 8:30 ` Christoph Hellwig 2018-01-04 9:39 ` Jan Kara 2018-01-04 9:39 ` Jan Kara 2017-12-24 0:57 ` [PATCH v4 16/18] wait_bit: introduce {wait_on,wake_up}_atomic_one Dan Williams 2017-12-24 0:57 ` Dan Williams 2018-01-04 8:30 ` Christoph Hellwig 2018-01-04 8:30 ` Christoph Hellwig 2017-12-24 0:57 ` [PATCH v4 17/18] mm, fs, dax: dax_flush_dma, handle dma vs block-map-change collisions Dan Williams 2017-12-24 0:57 ` Dan Williams 2018-01-04 8:31 ` Christoph Hellwig 2018-01-04 8:31 ` Christoph Hellwig 2018-01-04 11:12 ` Jan Kara 2018-01-04 11:12 ` Jan Kara 2018-01-07 21:58 ` Dan Williams 2018-01-07 21:58 ` Dan Williams 2018-01-08 13:50 ` Jan Kara 2018-01-08 13:50 ` Jan Kara 2018-03-08 17:02 ` Dan Williams 2018-03-08 17:02 ` Dan Williams 2018-03-09 12:56 ` Jan Kara 2018-03-09 12:56 ` Jan Kara 2018-03-09 16:15 ` Dan Williams 2018-03-09 16:15 ` Dan Williams 2018-03-09 17:26 ` Dan Williams 2018-03-09 17:26 ` Dan Williams 2017-12-24 0:57 ` [PATCH v4 18/18] xfs, dax: wire up dax_flush_dma support via a new xfs_sync_dma helper Dan Williams 2017-12-24 0:57 ` Dan Williams 2018-01-02 21:07 ` Darrick J. Wong 2018-01-02 21:07 ` Darrick J. Wong 2018-01-02 23:00 ` Dave Chinner 2018-01-02 23:00 ` Dave Chinner 2018-01-03 2:21 ` Dan Williams 2018-01-03 2:21 ` Dan Williams 2018-01-03 7:51 ` Dave Chinner 2018-01-03 7:51 ` Dave Chinner 2018-01-04 8:34 ` Christoph Hellwig 2018-01-04 8:34 ` Christoph Hellwig 2018-01-04 8:33 ` Christoph Hellwig 2018-01-04 8:33 ` Christoph Hellwig 2018-01-04 8:17 ` [PATCH v4 00/18] dax: fix dma vs truncate/hole-punch Christoph Hellwig 2018-01-04 8:17 ` Christoph Hellwig 2018-01-04 8:17 ` Christoph Hellwig
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=151407704063.38751.15403501871756055025.stgit@dwillia2-desk3.amr.corp.intel.com \ --to=dan.j.williams@intel.com \ --cc=akpm@linux-foundation.org \ --cc=hch@lst.de \ --cc=jack@suse.cz \ --cc=linux-fsdevel@vger.kernel.org \ --cc=linux-nvdimm@lists.01.org \ --cc=linux-xfs@vger.kernel.org \ --cc=mawilcox@microsoft.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.