Commit Graph
4 Commits
Author SHA1 Message Date
hypermistandClaude Opus 5.5 d9a79997e1 refactor(libc): declare Dinkumware float exports in the library header
Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-25 12:11:18 +13:00
hypermistandClaude Opus 5.5 dcda232626 test(libc): pin _LFpcomp on unnormals with unequal values
Running the firmware 13.60 libSceLibcInternal.sprx and PPSA08427's
bundled libc.prx in the differential harness shows that an unnormal
operand makes the x87 compare unordered, so _LFpcomp returns 2
whatever the values, in both orders. This build already does that;
the new vectors fix it in the test.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-25 06:38:55 +13:00
hypermistandClaude Opus 5.5 af0e82ed86 fix(libc): take the guest's 64-bit long exponent in _Dscale
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01RbLgECLQ2bpDM7oe2urGZK
2026-10-21 13:20:32 +02:00
hypermistandClaude Opus 5.5 b94f257415 feat(libc): add Dinkumware float bit helpers from firmware
Add _Dint, _Dscale, _Dunscale, _Dnorm, _Fpcomp and _Poly with their
float and long double forms, matching libSceLibcInternal bit for bit.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01RbLgECLQ2bpDM7oe2urGZK
2026-10-21 13:20:32 +02:00