GHSA-H82G-VJ5R-M4JV
Vulnerability from github – Published: 2026-09-17 18:32 – Updated: 2026-09-17 18:32In the Linux kernel, the following vulnerability has been resolved:
ext4: drain in-flight DIO before buffered write fallback
generic/746 started failing intermittently on ext3 (no-extent inodes). The test triggers 'Page cache invalidation failure on direct I/O' warnings and subsequent fsync returns -EIO. Adding a 50ms delay between ext4_buffered_write_iter() and filemap_write_and_wait_range() in ext4_dio_write_iter() makes the race almost always reproducible.
On no-extent inodes, DIO writes to holes cannot use unwritten extents, so ext4_iomap_alloc() leaves m_flags=0 and ext4_map_blocks() returns 0. The iomap layer then returns -ENOTBLK, causing fallback to buffered I/O.
The fallback path in ext4_dio_write_iter() calls ext4_buffered_write_iter() which dirties pages, then does flush and invalidate. However, there's an unprotected window between ext4_buffered_write_iter() returning (with inode lock released) and the subsequent flush+invalidate.
Concurrent async DIO completions from other threads can run kiocb_invalidate_post_direct_write() during this window. If pages have been re-dirtied, post-invalidation finds dirty pages and triggers the warning, setting -EIO in the error sequence.
Consider a file with two 4k extents: [hole][written]. Thread A does DIO to the written extent, while thread B does DIO spanning both:
kworker A (4k DIO, allocated block) kworker B (8k DIO, fallback) ----------------------------------- ---------------------------- inode_lock_shared() inode_lock_shared() iomap_dio_rw(): iomap_dio_rw(): kiocb_invalidate_pages -> clean iomap_begin -> -ENOTBLK submit_bio (async) dio->size = 0 inode_unlock_shared() inode_unlock_shared()
[bio pending in block layer] / fallback: lock released / ext4_buffered_write_iter() inode_lock(exclusive) generic_perform_write() -> dirty pages [0, 8k] inode_unlock(exclusive)
/* pages dirty, no lock */
[bio completes] filemap_write_and_wait_range() iomap_dio_complete() -> flush dirty pages kiocb_invalidate_post_direct_write() invalidate_mapping_pages() invalidate_inode_pages2_range() -> finds dirty page! -> dio_warn_stale_pagecache() -> errseq_set(-EIO)
This issue can be triggered through normal I/O paths, not just intentionally overlapping DIO writes from userspace. For example, generic/746 uses a loop device where multiple kworkers issue concurrent I/O to the backing file. Additionally, when block_size < folio_size, non-overlapping DIO writes that share a large folio can also trigger the race.
Add inode_dio_wait() in ext4_buffered_write_iter() before ext4_write_checks() to drain all in-flight DIO. This ensures that all DIO clears existing pages before submitting IO (via kiocb_invalidate_pages()), all BIO waits for all DIO to complete (via inode_dio_wait()), and ext4_write_checks() observes the inode size after all completed DIO so that ext4_block_zero_eof() does not race with in-flight DIO, thus eliminating the race.
{
"affected": [],
"aliases": [
"CVE-2026-92501"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-09-17T17:17:52Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\next4: drain in-flight DIO before buffered write fallback\n\ngeneric/746 started failing intermittently on ext3 (no-extent inodes).\nThe test triggers \u0027Page cache invalidation failure on direct I/O\u0027\nwarnings and subsequent fsync returns -EIO. Adding a 50ms delay\nbetween ext4_buffered_write_iter() and filemap_write_and_wait_range()\nin ext4_dio_write_iter() makes the race almost always reproducible.\n\nOn no-extent inodes, DIO writes to holes cannot use unwritten extents,\nso ext4_iomap_alloc() leaves m_flags=0 and ext4_map_blocks() returns 0.\nThe iomap layer then returns -ENOTBLK, causing fallback to buffered I/O.\n\nThe fallback path in ext4_dio_write_iter() calls\next4_buffered_write_iter() which dirties pages, then does flush and\ninvalidate. However, there\u0027s an unprotected window between\next4_buffered_write_iter() returning (with inode lock released) and\nthe subsequent flush+invalidate.\n\nConcurrent async DIO completions from other threads can run\nkiocb_invalidate_post_direct_write() during this window. If pages have\nbeen re-dirtied, post-invalidation finds dirty pages and triggers the\nwarning, setting -EIO in the error sequence.\n\nConsider a file with two 4k extents: [hole][written]. Thread A does\nDIO to the written extent, while thread B does DIO spanning both:\n\n kworker A (4k DIO, allocated block) kworker B (8k DIO, fallback)\n ----------------------------------- ----------------------------\n inode_lock_shared() inode_lock_shared()\n iomap_dio_rw(): iomap_dio_rw():\n kiocb_invalidate_pages -\u003e clean iomap_begin -\u003e -ENOTBLK\n submit_bio (async) dio-\u003esize = 0\n inode_unlock_shared() inode_unlock_shared()\n\n [bio pending in block layer] /* fallback: lock released */\n ext4_buffered_write_iter()\n inode_lock(exclusive)\n generic_perform_write()\n -\u003e dirty pages [0, 8k]\n inode_unlock(exclusive)\n\n /* pages dirty, no lock */\n [bio completes] filemap_write_and_wait_range()\n iomap_dio_complete() -\u003e flush dirty pages\n kiocb_invalidate_post_direct_write() invalidate_mapping_pages()\n invalidate_inode_pages2_range()\n -\u003e finds dirty page!\n -\u003e dio_warn_stale_pagecache()\n -\u003e errseq_set(-EIO)\n\nThis issue can be triggered through normal I/O paths, not just\nintentionally overlapping DIO writes from userspace. For example,\ngeneric/746 uses a loop device where multiple kworkers issue concurrent\nI/O to the backing file. Additionally, when block_size \u003c folio_size,\nnon-overlapping DIO writes that share a large folio can also trigger\nthe race.\n\nAdd inode_dio_wait() in ext4_buffered_write_iter() before\next4_write_checks() to drain all in-flight DIO. This ensures that\nall DIO clears existing pages before submitting IO (via\nkiocb_invalidate_pages()), all BIO waits for all DIO to complete\n(via inode_dio_wait()), and ext4_write_checks() observes the inode\nsize after all completed DIO so that ext4_block_zero_eof() does not\nrace with in-flight DIO, thus eliminating the race.",
"id": "GHSA-h82g-vj5r-m4jv",
"modified": "2026-09-17T18:32:06Z",
"published": "2026-09-17T18:32:06Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-92501"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/15cdefd0c0522f9d5e12d947fa04f4c11649b699"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/4e4e3eec506247c8f8bd8aaa1eb25e67016681a5"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/7341e234927ff215f1d5d0bcfe04b74f53af378d"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/74eee4ff9698a65b2e6e15dac0e50d6526ad5f20"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/9fd3ffc3c51c9deaba99bd7b338fff2d08f52416"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/d47cdadd6e49023f7ee248048463807f1214f1ee"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/f0af3ae09fb72382da1a5371bf6761b0264668e4"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/fd7e0dab20837b9ea1eeef7c26f78ace8ac8258c"
}
],
"schema_version": "1.4.0",
"severity": []
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.