cglm's test suite crashes at mat2x3s_zero_init
Measured 2026-09-27 with the compiler that fixed bug-c-an-array-of-pointers-to-arrays-has-no-pointee-shape-so-deref-loads-instead-of-decaying:
P=compiler/pascal26; L=library_candidates; K=/home/neo/pxx-website-patches/v425-restamp
$P -Ilib/crtl/include -Ilib/crtl/src -I$L/cglm/include -I$L/cglm/test/include \
-I$L/cglm/test/src $K/drv/cglm_all.c /tmp/cglm && /tmp/cglm
rc=139 after glmc_ivec4_abs. The next entry in test/tests.h is
mat2x3s_zero_init (test/src/test_struct.c:10). Reduce it against gcc before
guessing the mechanism.
2026-09-27 (frankD): the same bug as the struct-field ticket, fixed there
mat2x3s is a union whose raw member is typed by the array typedef
mat2x3; the struct builder laid it out as one float, so .raw passed on as
a mat2x3 read past it. Fixed with
bug-c-a-struct-field-that-points-at-an-array-typedef-has-no-pointee-dims.
cglm's whole suite now runs: 1131 ran, 1131 passed, 0 failed, rc=0.
Log
- 2026-09-27 — resolved; this names the commit that carried the resolve, which is not always the one that carried the change — commit 84ea0d4fef.