Run3 full-file SHA samples, captured approximately eight seconds after launch

repo-pre-7
state: RC
file offset: 1245184
kernel stack: lzma2_lzma -> erofs_xz_dec_microlzma_run -> lzma_decompress -> z_erofs_decompress -> z_erofs_do_read -> z_erofs_read_uio -> VOP_READ_APV -> vn_read
sha_after: no process row, no sha.out hash, no sha.err content
termination: TERM sent; KILL then reported No such process

repo-pre-8
state: RC
file offset: 1150976
kernel stack: lzma2_lzma -> erofs_xz_dec_microlzma_run -> lzma_decompress -> z_erofs_decompress -> z_erofs_do_read -> z_erofs_read_uio -> VOP_READ_APV -> vn_read
sha_after: no process row, no sha.out hash, no sha.err content
termination: TERM sent; KILL then reported No such process

repo-pre-9
state: RC
file offset: 1044480
kernel stack: lzma2_lzma -> erofs_xz_dec_microlzma_run -> lzma_decompress -> z_erofs_decompress -> z_erofs_do_read -> z_erofs_read_uio -> VOP_READ_APV -> vn_read
sha_after: no process row, no sha.out hash, no sha.err content
termination: TERM sent; KILL then reported No such process

Interpretation: the SHA processes were alive and executing in LZMA decode at
the sample. Their later disappearance does not prove successful completion.
No full-file hash was captured, so full-file correctness remains BLOCKED.
