[Bug target/27770] [4.2 Regression] wrong code in spec tests for -ftree-vectorize -maltivec

["dje at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>]

------- Comment #6 from dje at gcc dot gnu dot org  2006-06-05 21:25 -------
I still think this looks like an alignment problem.  Without section anchors
GCC generates:

        .lcomm  domain_array,13916,16
        .type   domain_array, @object
        .lcomm  N_domains,4,4
        .type   N_domains, @object
        .lcomm  N_domain_trees,4,4
        .type   N_domain_trees, @object
        .lcomm  word_links,1000,16
        .type   word_links, @object
        .lcomm  visited,1000,16
        .type   visited, @object

with section-anchors GCC generates:

        .section        ".bss"
        .align 2
        .set    .LANCHOR0,. + 0
        .type   N_domains, @object
        .size   N_domains, 4
N_domains:
        .zero   4
        .type   domain_array, @object
        .size   domain_array, 13916
domain_array:
        .zero   13916
        .type   visited, @object
        .size   visited, 1000
visited:
        .zero   1000
        .type   word_links, @object
        .size   word_links, 1000
word_links:
        .zero   1000
        .type   N_domain_trees, @object
        .size   N_domain_trees, 4
N_domain_trees:
        .zero   4

with domain_array and word_links not guaranteed the alignment of 16, if I am
reading the assembly correctly.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27770