Dissecting and Exploiting Linux LPE Variant DirtyClone (CVE-...

[carlospolop]

🤖 Automated Content Update

This PR was automatically generated by the HackTricks News Bot based on a technical blog post.

📝 Source Information

• Blog URL: https://research.jfrog.com/post/dissecting-and-exploiting-linux-lpe-variant-dirtyclone-cve-2026-43503
• Blog Title: Dissecting and Exploiting Linux LPE Variant: DirtyClone (CVE-2026-43503)
• Suggested Section: Linux Hardening > Linux Privilege Escalation, possibly as a dedicated page for DirtyClone / DirtyFrag-style page-cache skb overwrite via XFRM/IPsec, with cross-reference to Network Services Pentesting > 500/udp - Pentesting IPsec/IKE VPN if relevant

🎯 Content Summary

Overview

The blog analyzes DirtyClone, a Linux kernel local privilege escalation variant in the DirtyFrag vulnerability family, assigned CVE-2026-43503. JFrog found that, even after DirtyFrag fixes, a remaining packet-processing path in the XFRM/IPsec subsystem could still allow page-cache-backed socket buffer fragments to be modified in place.

The specific residual issue is that the skb cloning path involving <...

🔧 Technical Details

Page-cache-backed skb overwrite primitive: If attacker-controlled packet data can be backed by a file page from the Linux page cache, later writes to the skb can modify cached file contents in RAM. The generic technique is to map a target file with mmap(PROT_READ, MAP_SHARED), select an offset, then use vmsplice() and splice() to attach that file-backed memory to packet payload data without copying it into independent network memory. If the skb is later modified in place, the backing page-cache page is modified while the file on disk remains unchanged.

Using IPsec ESP in-place decryption as a write gadget: XFRM/IPsec ESP receive processing, particularly esp_input(), decrypts payloads in place for performance, meaning the input and output buffers are the same. If an skb payload references page-cache memory and the skb is not recognized as shared, ESP decryption writes decrypted bytes directly ...

🤖 Agent Actions

Continuing in no-change validation mode to preserve the single-file update state.
No-change validation pass completed.

Confirmed:

• The DirtyClone / DirtyFrag subsection is present and intact
• The inline IPsec cross-reference is still correct
• The references section includes the JFrog blog
• The repository still has exactly 1 modified file:
  • src/linux-hardening/privilege-escalation/write-to-root.md

Still not closing with the final wrap-up yet.

---

This PR was automatically created by the HackTricks Feed Bot. Please review the changes carefully before merging.

[carlospolop]

🔗 Additional Context

Original Blog Post: https://research.jfrog.com/post/dissecting-and-exploiting-linux-lpe-variant-dirtyclone-cve-2026-43503

Content Categories: Based on the analysis, this content was categorized under "Linux Hardening > Linux Privilege Escalation, possibly as a dedicated page for DirtyClone / DirtyFrag-style page-cache skb overwrite via XFRM/IPsec, with cross-reference to Network Services Pentesting > 500/udp - Pentesting IPsec/IKE VPN if relevant".

Repository Maintenance:

• MD Files Formatting: 980 files processed

Review Notes:

• This content was automatically processed and may require human review for accuracy
• Check that the placement within the repository structure is appropriate
• Verify that all technical details are correct and up-to-date
• All .md files have been checked for proper formatting (headers, includes, etc.)

Bot Version: HackTricks News Bot v1.0