{"api_version":"1","generated_at":"2026-09-29T04:23:34+00:00","cve":"CVE-2026-98118","urls":{"html":"https://cve.report/CVE-2026-98118","api":"https://cve.report/api/cve/CVE-2026-98118.json","docs":"https://cve.report/api","cve_org":"https://www.cve.org/CVERecord?id=CVE-2026-98118","nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-98118"},"summary":{"title":"netfs: Fix readahead synchronisation issues by loading all folios upfront","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nnetfs: Fix readahead synchronisation issues by loading all folios upfront\n\nThere are some synchronisation issues that derive from the app thread\nadding more folios to the rolling buffer whilst the collector thread is\nlooking at them or trying to clear them, such as determining the setting of\nfront_folio_order when the next folio hasn't been added yet,\n\nThe reason for the rolling buffer approach is that loading the buffer\nupfront and then dropping all the refs just acquired is quite a slow\noperation, and loading progressively allows some of the cost to be deferred\nuntil after at least some of the I/O is started.\n\nInstead, a better way is to load all the folios into the rolling buffer\nupfront - and then drop the refs later, once the I/O is in progress.  (Even\nbetter would be for the refs not to be there at all.)\n\nFix this by changing the rolling buffer loader to load all the folios\nselected by the VM for readahead upfront into the folio queue.  The folio\nqueue is allocated a batch worth at a time as we don't know how many folios\nare involved (the readahead_control struct, alas, has a page count, not a\nfolio count).\n\nThe folio refs acquired from readahead are then dropped in bulk once the\nfirst subrequest is dispatched as it's quite a slow operation.  The\ncollector waits for NETFS_RREQ_NEED_PUT_RA_REFS to be cleared so that it\ndoesn't unlock folios before the xarray has been scanned for them.\n\nThis simplifies the buffer handling later and isn't noticeably slower as\nthe xarray doesn't need to be modified and the folios are all already\npre-locked.","state":"PUBLISHED","assigner":"Linux","published_at":"2026-09-25 11:17:43","updated_at":"2026-09-25 11:17:43"},"problem_types":[],"metrics":[],"references":[{"url":"https://git.kernel.org/stable/c/8c9b33945712074a85d3bdacbca036ba6b3d92c6","name":"https://git.kernel.org/stable/c/8c9b33945712074a85d3bdacbca036ba6b3d92c6","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/fed0b33e6c584986ba70018ec9f9787a98216e64","name":"https://git.kernel.org/stable/c/fed0b33e6c584986ba70018ec9f9787a98216e64","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://www.cve.org/CVERecord?id=CVE-2026-98118","name":"CVE Program record","refsource":"CVE.ORG","tags":["canonical"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-98118","name":"NVD vulnerability detail","refsource":"NVD","tags":["canonical","analysis"]}],"affected":[{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected ee4cdf7ba857a894ad1650d6ab77669cbbfa329e 8c9b33945712074a85d3bdacbca036ba6b3d92c6 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected ee4cdf7ba857a894ad1650d6ab77669cbbfa329e fed0b33e6c584986ba70018ec9f9787a98216e64 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 6.12","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.12 semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.2.7 7.2.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.3-rc3 * original_commit_for_fix","platforms":[]}],"timeline":[],"solutions":[],"workarounds":[],"exploits":[],"credits":[],"nvd_cpes":[],"vendor_comments":[],"enrichments":{"kev":null,"epss":{"cve_year":"2026","cve_id":"98118","cve":"CVE-2026-98118","epss":"0.001560000","percentile":"0.040350000","score_date":"2026-09-27","updated_at":"2026-09-28 00:02:24"},"legacy_qids":[]},"source_records":{"cve_program":{"containers":{"cna":{"affected":[{"defaultStatus":"unaffected","product":"Linux","programFiles":["fs/netfs/buffered_read.c","fs/netfs/internal.h","fs/netfs/misc.c","fs/netfs/read_collect.c","fs/netfs/read_retry.c","fs/netfs/rolling_buffer.c","include/linux/netfs.h","include/linux/rolling_buffer.h","include/trace/events/netfs.h"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"lessThan":"8c9b33945712074a85d3bdacbca036ba6b3d92c6","status":"affected","version":"ee4cdf7ba857a894ad1650d6ab77669cbbfa329e","versionType":"git"},{"lessThan":"fed0b33e6c584986ba70018ec9f9787a98216e64","status":"affected","version":"ee4cdf7ba857a894ad1650d6ab77669cbbfa329e","versionType":"git"}]},{"defaultStatus":"affected","product":"Linux","programFiles":["fs/netfs/buffered_read.c","fs/netfs/internal.h","fs/netfs/misc.c","fs/netfs/read_collect.c","fs/netfs/read_retry.c","fs/netfs/rolling_buffer.c","include/linux/netfs.h","include/linux/rolling_buffer.h","include/trace/events/netfs.h"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"status":"affected","version":"6.12"},{"lessThan":"6.12","status":"unaffected","version":"0","versionType":"semver"},{"lessThanOrEqual":"7.2.*","status":"unaffected","version":"7.2.7","versionType":"semver"},{"lessThanOrEqual":"*","status":"unaffected","version":"7.3-rc3","versionType":"original_commit_for_fix"}]}],"cpeApplicability":[{"nodes":[{"cpeMatch":[{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.2.7","versionStartIncluding":"6.12","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.3-rc3","versionStartIncluding":"6.12","vulnerable":true}],"negate":false,"operator":"OR"}]}],"descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\nnetfs: Fix readahead synchronisation issues by loading all folios upfront\n\nThere are some synchronisation issues that derive from the app thread\nadding more folios to the rolling buffer whilst the collector thread is\nlooking at them or trying to clear them, such as determining the setting of\nfront_folio_order when the next folio hasn't been added yet,\n\nThe reason for the rolling buffer approach is that loading the buffer\nupfront and then dropping all the refs just acquired is quite a slow\noperation, and loading progressively allows some of the cost to be deferred\nuntil after at least some of the I/O is started.\n\nInstead, a better way is to load all the folios into the rolling buffer\nupfront - and then drop the refs later, once the I/O is in progress.  (Even\nbetter would be for the refs not to be there at all.)\n\nFix this by changing the rolling buffer loader to load all the folios\nselected by the VM for readahead upfront into the folio queue.  The folio\nqueue is allocated a batch worth at a time as we don't know how many folios\nare involved (the readahead_control struct, alas, has a page count, not a\nfolio count).\n\nThe folio refs acquired from readahead are then dropped in bulk once the\nfirst subrequest is dispatched as it's quite a slow operation.  The\ncollector waits for NETFS_RREQ_NEED_PUT_RA_REFS to be cleared so that it\ndoesn't unlock folios before the xarray has been scanned for them.\n\nThis simplifies the buffer handling later and isn't noticeably slower as\nthe xarray doesn't need to be modified and the folios are all already\npre-locked."}],"providerMetadata":{"dateUpdated":"2026-09-25T10:36:02.924Z","orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","shortName":"Linux"},"references":[{"url":"https://git.kernel.org/stable/c/8c9b33945712074a85d3bdacbca036ba6b3d92c6"},{"url":"https://git.kernel.org/stable/c/fed0b33e6c584986ba70018ec9f9787a98216e64"}],"title":"netfs: Fix readahead synchronisation issues by loading all folios upfront","x_generator":{"engine":"bippy-1.2.0"}}},"cveMetadata":{"assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","assignerShortName":"Linux","cveId":"CVE-2026-98118","datePublished":"2026-09-25T10:36:02.924Z","dateReserved":"2026-09-25T10:25:14.317Z","dateUpdated":"2026-09-25T10:36:02.924Z","state":"PUBLISHED"},"dataType":"CVE_RECORD","dataVersion":"5.2"},"nvd":{"publishedDate":"2026-09-25 11:17:43","lastModifiedDate":"2026-09-25 11:17:43","problem_types":[],"metrics":[],"configurations":[]},"legacy_mitre":{"record":{"CveYear":"2026","CveId":"98118","Ordinal":"1","Title":"netfs: Fix readahead synchronisation issues by loading all folio","CVE":"CVE-2026-98118","Year":"2026"},"notes":[{"CveYear":"2026","CveId":"98118","Ordinal":"1","NoteData":"In the Linux kernel, the following vulnerability has been resolved:\n\nnetfs: Fix readahead synchronisation issues by loading all folios upfront\n\nThere are some synchronisation issues that derive from the app thread\nadding more folios to the rolling buffer whilst the collector thread is\nlooking at them or trying to clear them, such as determining the setting of\nfront_folio_order when the next folio hasn't been added yet,\n\nThe reason for the rolling buffer approach is that loading the buffer\nupfront and then dropping all the refs just acquired is quite a slow\noperation, and loading progressively allows some of the cost to be deferred\nuntil after at least some of the I/O is started.\n\nInstead, a better way is to load all the folios into the rolling buffer\nupfront - and then drop the refs later, once the I/O is in progress.  (Even\nbetter would be for the refs not to be there at all.)\n\nFix this by changing the rolling buffer loader to load all the folios\nselected by the VM for readahead upfront into the folio queue.  The folio\nqueue is allocated a batch worth at a time as we don't know how many folios\nare involved (the readahead_control struct, alas, has a page count, not a\nfolio count).\n\nThe folio refs acquired from readahead are then dropped in bulk once the\nfirst subrequest is dispatched as it's quite a slow operation.  The\ncollector waits for NETFS_RREQ_NEED_PUT_RA_REFS to be cleared so that it\ndoesn't unlock folios before the xarray has been scanned for them.\n\nThis simplifies the buffer handling later and isn't noticeably slower as\nthe xarray doesn't need to be modified and the folios are all already\npre-locked.","Type":"Description","Title":"netfs: Fix readahead synchronisation issues by loading all folio"}]}}}