https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108809
Jiu Fu Guo <guojiufu at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |guojiufu at gcc dot gnu.org
--- Comment #2 from Jiu Fu Guo <guojiufu at gcc dot gnu.org> ---
(In reply to Kewen Lin from comment #1)
> It's very likely a test issue, may need to adjust some built-in function for
> endianness issue (as they have different element ordering on BE and LE).
Yeap it should be the test case issue on the difference between BE and LE:
For a buff `\001\002\003\004\005\006\a\b\t\n\v\f\r\016\017\020\021\022`,
after loading to a vector right justified, the result is different between BE
and LE.
e.g. for a vector (from 128bit view: 0x0000000000000000102030405060708),
from v16_int8 view on BE, it is v16_int8 = {0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0,
0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7, 0x8}; but from the v16_int8 view on LE,
it is v16_int8 = {0x8, 0x7, 0x6, 0x5, 0x4, 0x3, 0x2, 0x1, 0x0, 0x0, 0x0, 0x0,
0x0, 0x0, 0x0, 0x0}.
So, we would just need to update the expected result for BE for this test case.
