[dgarske]

The windowed SP ECC scalar recode takes the low octet with y = (word8)(int8_t)n, which only truncates where char is 8 bits; on a CHAR_BIT == 16 target int8_t does not exist and the build fails outright, leaving WOLFSSL_SP_SMALL as the only usable path. Replaced with y = (word8)(n & 0xff) across the eight generated backends, matching the generator fix in wolfSSL/scripts#702; the files are edited in place rather than regenerated, since a regen would also pull in unrelated generator work not yet reflected here. Verified with make check under --enable-all --enable-sp --enable-sp-asm, and on a TI LAUNCHXL-F28P55X where ECDSA and ECDH over P-256 and P-384 now pass on hardware. On arm32, armthumb, cortexm and riscv64 the routine sits behind WOLFSSL_SP_NONBLOCK, so those four were checked with per-architecture cross-compiles rather than by CI.

[github-actions]

retest this please

[github-actions]

MemBrowse Memory Report

gcc-arm-cortex-m3

• FLASH: .text -4 B (-0.0%, 130,741 B / 262,144 B, total: 50% used)
  No memory changes detected for:
• gcc-arm-cortex-m0plus
• 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

[philljj]

looks good.

tests good with multi-test sp sanitizer and related.