[AutoPR- Security] Patch libsoup for CVE-2026-77014 [MEDIUM]

[Azure Linux Security Servicing Account (azurelinux-security)]

Auto Patch libsoup for CVE-2026-77014.

Autosec pipeline run -> https://dev.azure.com/mariner-org/mariner/_build/results?buildId=1189685&view=results

CVE-2026-77014 : Single Patch Backporter Pipeline Run -> https://dev.azure.com/mariner-org/mariner/_build/results?buildId=1189691&view=results

Merge Checklist

All boxes should be checked before merging the PR (just tick any boxes which don't apply to this PR)

• The toolchain has been rebuilt successfully (or no changes were made to it)
• The toolchain/worker package manifests are up-to-date
• Any updated packages successfully build (or no packages were changed)
• Packages depending on static components modified in this PR (Golang, *-static subpackages, etc.) have had their Release tag incremented.
• Package tests (%check section) have been verified with RUN_CHECK=y for existing SPEC files, or added to new SPEC files
• All package sources are available
• cgmanifest files are up-to-date and sorted (./cgmanifest.json, ./toolkit/scripts/toolchain/cgmanifest.json, .github/workflows/cgmanifest.json)
• LICENSE-MAP files are up-to-date (./LICENSES-AND-NOTICES/SPECS/data/licenses.json, ./LICENSES-AND-NOTICES/SPECS/LICENSES-MAP.md, ./LICENSES-AND-NOTICES/SPECS/LICENSE-EXCEPTIONS.PHOTON)
• All source files have up-to-date hashes in the *.signatures.json files
• sudo make go-tidy-all and sudo make go-test-coverage pass
• Documentation has been updated to match any changes to the build system
• Ready to merge

---

Summary

What does the PR accomplish, why was it needed?

• Auto Patch libsoup for CVE-2026-77014 (MEDIUM).

Change Log

• CVE-2026-77014

Does this affect the toolchain?

YES/NO

Associated issues

• N/A

Links to CVEs

• https://nvd.nist.gov/vuln/detail/CVE-2026-77014

Test Methodology

• Pipeline build id: Buddy Build URL

[azure-pipelines]

Azure Pipelines:
There may be pipelines that require an authorized user to comment /azp run to run.

[v-swapsahu]

Patch Analysis — CVE-2026-77014 (libsoup 3.4.4)

Backported: Yes

e82c13ba: — CWE-197 numeric truncation. sort_ranges() returned a 64-bit offset difference as a 32-bit int
For differences > INT_MAX the value wraps negative, so qsort sorts ranges backwards and the coalesce step drops a range. Impact: on resources > 2 GB, multi-range requests return the wrong bytes.

Fix — compare in full 64-bit, return -1/0/1:
Same commit also adds a MAX_RANGES (200) cap rejecting over-long range lists, plus tests /ranges/parsing, /ranges/count, /server-mem/range-overlaps.

6ece9e52 — validate Content-Range when parsing. soup_message_headers_get_content_range() parsed offsets with raw g_ascii_strtoull into signed goffset with no overflow check (values > G_MAXINT64 wrap negative; leading - accepted; no end ≥ start / end < total checks).

Fix — adds a parse_content_range_offset() helper (rejects ERANGE/> G_MAXINT64) and range validation. Ships with its upstream regression test /ranges/content-range.

546a59d - RFC 416 — 416 responses now report the resource length as Content-Range: bytes */<total> (new set_content_range_unsatisfied() helper), letting clients correct an out-of-range request — RFC 9110 §15.5.17.
request_unsatisfiable_range was omitted — sits in the apache-gated range-test.c (never runs in %check without mod_ssl), and would require grafting partial upstream test code; dropping it keeps the diff minimal.

Files changed

Backport notes:

• Minimal — no HAVE_APACHE restructuring;
• range-test is apache-gated → not run in %check (no mod_ssl in chroot); server-mem-limit-test runs. Truncation fix validated via POC.
• The new /ranges/content-range test from 6ece9e52 is self-contained (no Apache) and does run in %check, giving the CVE in-build coverage.

Verification

Patch Application:

Build and Test Results:

POC Summary — CVE-2026-77014 (range-sort truncation)

Idea: Request two byte ranges whose 64-bit start-difference overflows the old 32-bit comparator, on a >2 GB resource. Buggy lib mis-sorts and drops a range.

Vulnerability reproduced on unpatched lib

count=1 → only 3000000000-3000000100 (the 0-100 range dropped)

Fix verified on patched lib

count=2 → 0-100 then 3000000000-3000000100, sorted

[Aditya Singh (v-aaditya)]

Buddy Build has been re-triggered and it has passed.

Peer-review analysis - LGTM.