20260306-wc_Hash-refactor

[douzzer]

refactor wc_Hash* so that known wc_HashType values are unconditionally defined in enum wc_HashType, and always either succeed if used properly, or return HASH_TYPE_E if gated out or used improperly; add detailed error code tracing in wolfcrypt/src/hash.c.

wolfssl/wolfcrypt/error-crypt.h: add do-nothing WC_ERR_TRACE() fallthrough definition to make WC_ERR_TRACE() safe to use ungated in code.

tested with wolfssl-multi-test.sh ... super-quick-check

[douzzer]

retest this please

[douzzer]

retest this please

[dgarske]

@douzzer please resolve merge conflicts and review CI errors

[douzzer]

retest this please
(Jenkins controller reboot)

[douzzer]

retest this please

FAIL: scripts/crl-gen-openssl

[douzzer]

Note the PRB-libwolfssl-size-change failure is a false positive likely caused by another already-merged PR, or by a faulty test script. There is no significant footprint change from this PR:

$ diff /tmp/wolfcrypt-sizes.pre.fixed /tmp/wolfcrypt-sizes.post
11c11
< -rw-r--r-- 1 douzzer users  19440 Mar 23 20:50 src_libwolfssl_la-hash.o
---
> -rw-r--r-- 1 douzzer users  20040 Mar 23 20:50 src_libwolfssl_la-hash.o

(using --enable-opensslextra=small as in testing/Jenkins/PRB-libwolfssl-size-change/PRB.Jenkinsfile)

[douzzer]

Note also the multi-test failure:

[all-c89-clang-tidy] [1 of 1] [wolfssl]
    autogen.sh wolfssl...   real 0m12.954s  user 0m11.739s  sys 0m0.272s
    configure...   real 0m13.987s  user 0m10.522s  sys 0m4.052s
    build...   real 13m4.139s  user 12m46.441s  sys 0m17.848s
    check...FAIL: scripts/ocsp-stapling2.test
FAIL: scripts/ocsp-stapling_tls13multi.test

is also false positives unrelated to this PR.

[dgarske]

Jenkins retest this please. FIPS 140-3 agent offline