Fix TI port AES-CTR streaming and hashCopy correctness bugs

[dgarske]

Description

Fixes two bugs in the TI (TivaWare TM4C129 AES/SHA-MD5 hardware) port. Both silently produce wrong output rather than failing, and both trace to PR #7018.

F-3078 - wc_AesCtrEncrypt drops partial streaming input (wolfcrypt/src/port/ti/ti-aes.c)

When AES-CTR is called in streaming fashion and a follow-up chunk does not complete the buffered block (aes->left + sz < WC_AES_BLOCK_SIZE), the leading carry-over branch buffered the new plaintext but never encrypted it, never wrote the caller's output, and never advanced aes->left. The caller saw uninitialized/stale output, and the next call overwrote the still-buffered bytes, silently losing plaintext. Fixed by adding the else path that mirrors the existing tail-handling block: encrypt the partial block with AES_CFG_MODE_CTR_NOCTR (no counter advance), emit the new bytes, and do aes->left += odd.

F-2646 - hashCopy discards accumulated message (wolfcrypt/src/port/ti/ti-hash.c)

hashCopy zeroed msg/used/len and copied only the never-populated hash field, so a copied context finalized to the hash of an empty message instead of the source data. Affects wc_Md5Copy, wc_ShaCopy, wc_Sha224Copy, and wc_Sha256Copy. Fixed by allocating a fresh dst->msg and copying src->len bytes (independent allocation preserves the original anti-double-free intent), matching the reference wc_Sha256Copy in devcrypto_hash.c.

[Copilot]

Pull request overview

Fixes two correctness bugs in the TI (TM4C129/TivaWare) hardware-accelerated crypto port that could silently produce incorrect output: AES-CTR streaming with partial buffered blocks, and hash context copying losing the accumulated message.

Changes:

• Fix wc_AesCtrEncrypt streaming behavior when a follow-up chunk still doesn’t complete the buffered block (encrypt/emit bytes and advance aes->left without counter advance).
• Fix TI hashCopy to deep-copy the accumulated message buffer so copied contexts finalize correctly.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
wolfcrypt/src/port/ti/ti-aes.c	Adds missing handling for the “still not a full block” carry-over path in CTR streaming.
wolfcrypt/src/port/ti/ti-hash.c	Updates TI hash context copying to include buffered message data via independent allocation.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[github-actions]

MemBrowse Memory Report

No memory changes detected for:

• gcc-arm-cortex-m0plus
• gcc-arm-cortex-m3
• gcc-arm-cortex-m4
• gcc-arm-cortex-m4-baremetal
• gcc-arm-cortex-m4-crypto-only
• gcc-arm-cortex-m4-dtls13
• gcc-arm-cortex-m4-min-ecc
• gcc-arm-cortex-m4-openssl-compat
• gcc-arm-cortex-m4-pkcs7
• gcc-arm-cortex-m4-pq
• gcc-arm-cortex-m4-rsa-only
• gcc-arm-cortex-m4-sp-math
• gcc-arm-cortex-m4-tls12
• gcc-arm-cortex-m4-tls13
• gcc-arm-cortex-m7
• gcc-arm-cortex-m7-pq
• gcc-arm-cortex-m7-tls13
• linuxkm-pie
• linuxkm-standard
• stm32-sim-stm32h753

[dgarske]

Jenkins retest this please. valgrind with openssl-all

[Frauschi]

🐺 Skoll Code Review

Overall recommendation: APPROVE
Findings: 9 total — 3 posted, 6 skipped

Posted findings

• [Medium] hashCopy leaves dst in an inconsistent state when XMALLOC fails — wolfcrypt/src/port/ti/ti-hash.c:132-138
• [Medium] hashCopy overwrites a pre-existing dst->msg without freeing it — wolfcrypt/src/port/ti/ti-hash.c:130
• [Low] wc_AesCtrEncrypt(sz == 0) with buffered bytes now performs a pointless HW AES block — wolfcrypt/src/port/ti/ti-aes.c:254

Skipped findings

• [Medium] TI port fixes are not exercised by CI; confirm on-target validation
• [Low] Copying src->len instead of src->used couples the copy to the allocator's capacity
• [Low] New CTR carry-over branch duplicates the existing tail-fragment block
• [Low] hashCopy leaves dst in an inconsistent state when the new allocation fails
• [Info] Added lines exceed the 80-column limit
• [Info] New AES-CTR carry-over branch runs a hardware AES pass for a zero-length call

Review generated by Skoll via Claude/Codex