ARM32/Thumb2: generated asm fixes#10725
Open
SparkiDev wants to merge 1 commit into
Open
Conversation
Contributor
Author
|
Code generated with PR: |
Contributor
Author
|
Jenkins: retest this please |
wolfSSL-Fenrir-bot
left a comment
wolfSSL-Fenrir-bot
left a comment
There was a problem hiding this comment.
Fenrir Automated Review — PR #10725
No scan targets match the changed files in this PR. Review skipped.
Contributor
There was a problem hiding this comment.
Pull request overview
This PR updates multiple ARM32/Thumb2 hand-written/generated assembly sources to improve correctness and simplify/improve instruction sequences, primarily by adjusting reduction/carry handling and making immediate operands more consistent.
Changes:
- Normalize many immediate operands (hex → decimal,
#0x0→#0, etc.) across Thumb2 SHA2/SHA3/ChaCha/Poly1305/ML-KEM codegen outputs. - Update ARM32 Curve25519 field ops to a simpler underflow/overflow handling approach and adjust
fe_isnegativeto avoid a reload by caching the low-bit early. - Optimize AES/GCM assembly (including
ubfxusage and arch<7 fallbacks) and remove a redundant register move in the AES key schedule path.
Reviewed changes
Copilot reviewed 18 out of 20 changed files in this pull request and generated no comments.
Show a summary per file
| File | Description |
|---|---|
| wolfcrypt/src/port/arm/thumb2-sha512-asm.S | Immediate-operand normalization in SHA-512 transform loop. |
| wolfcrypt/src/port/arm/thumb2-sha512-asm_c.c | Mirrors SHA-512 Thumb2 asm immediate changes in inline-asm C. |
| wolfcrypt/src/port/arm/thumb2-sha3-asm.S | Immediate-operand normalization in SHA-3 Thumb2 implementation. |
| wolfcrypt/src/port/arm/thumb2-sha3-asm_c.c | Mirrors SHA-3 Thumb2 asm immediate changes in inline-asm C. |
| wolfcrypt/src/port/arm/thumb2-sha256-asm.S | Immediate-operand normalization in SHA-256 Thumb2 implementation. |
| wolfcrypt/src/port/arm/thumb2-sha256-asm_c.c | Mirrors SHA-256 Thumb2 asm immediate changes in inline-asm C. |
| wolfcrypt/src/port/arm/thumb2-poly1305-asm.S | Immediate-operand normalization and small simplifications in Poly1305 Thumb2 asm. |
| wolfcrypt/src/port/arm/thumb2-poly1305-asm_c.c | Mirrors Poly1305 Thumb2 asm immediate changes in inline-asm C. |
| wolfcrypt/src/port/arm/thumb2-mlkem-asm.S | Immediate-operand normalization in ML-KEM Thumb2 asm loops. |
| wolfcrypt/src/port/arm/thumb2-mlkem-asm_c.c | Mirrors ML-KEM Thumb2 asm immediate changes in inline-asm C. |
| wolfcrypt/src/port/arm/thumb2-chacha-asm.S | Immediate-operand normalization in ChaCha Thumb2 asm. |
| wolfcrypt/src/port/arm/thumb2-chacha-asm_c.c | Mirrors ChaCha Thumb2 asm immediate changes in inline-asm C. |
| wolfcrypt/src/port/arm/thumb2-aes-asm.S | GCM nibble extraction simplification (ubfx) and immediate normalization; updates perf comment. |
| wolfcrypt/src/port/arm/armv8-32-curve25519.S | Simplifies underflow/overflow adjustment logic and tweaks fe_isnegative bit usage. |
| wolfcrypt/src/port/arm/armv8-32-curve25519_c.c | Mirrors Curve25519 asm changes; adds r12 clobber for updated fe_isnegative. |
| wolfcrypt/src/port/arm/armv8-32-aes-asm.S | Removes redundant move in AES key schedule; adds arch<7 fallback sequences for GCM nibble extraction. |
| wolfcrypt/src/port/arm/armv8-32-aes-asm_c.c | Mirrors AES/GCM asm changes in inline-asm C. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
douzzer
requested changes
Jun 18, 2026
douzzer
left a comment
douzzer
left a comment
Contributor
There was a problem hiding this comment.
don't merge as-is -- see wolfssl/scripts#590
Fix Thumb2 Curve25519 asm to do full reduce. Change ARM32 to simpler carry/overflow processing. Minor optimizations - use ubfx, no need to move register into temporary, cache value instead of loading again later. Reduce the register push and pops in Thumb2 generated code. Fix Thumb2 to have values less than 64 in decimal.
SparkiDev
force-pushed
the
aes_x25519_arm32_thumb2_fixes
branch
from
8d2c23d to
9558b0d
Compare
SparkiDev
added
Not For This Release
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Fix Thumb2 Curve25519 asm to do full reduce.
Change ARM32 to simpler carry/overflow processing. Minor optimizations - use ubfx, no need to move register into temporary, cache value instead of loading again later. Reduce the register push and pops in Thumb2 generated code. Fix Thumb2 to have values less than 64 in decimal.
Testing
ARM32 (armv7a) and Thumb2 (armv7m) assembly configurations in QEMU.