diff options
| author | Julian Taylor <jtaylor.debian@googlemail.com> | 2013-06-10 19:38:58 +0200 |
|---|---|---|
| committer | Julian Taylor <jtaylor.debian@googlemail.com> | 2013-06-10 23:03:06 +0200 |
| commit | aef286debfd11a62f1c337dea55624cee7fd4d9e (patch) | |
| tree | 34cd255a107668c8442984b4af5e9882635fb0f0 /numpy | |
| parent | c9bf9b0b1c8a85391695d4ded39921e98c63257d (diff) | |
| download | numpy-aef286debfd11a62f1c337dea55624cee7fd4d9e.tar.gz | |
ENH: enable unaligned loads on x86
x86 can handle unaligned load and there is no hand vectorized code in
this file. It would be a serious compiler bug if it adds vectorization
without checking for alignment.
Enables fast complex128 copies which are unaligned on 32 bit gcc unless
compiled with -malign-double.
Diffstat (limited to 'numpy')
| -rw-r--r-- | numpy/core/src/multiarray/lowlevel_strided_loops.c.src | 6 |
1 files changed, 2 insertions, 4 deletions
diff --git a/numpy/core/src/multiarray/lowlevel_strided_loops.c.src b/numpy/core/src/multiarray/lowlevel_strided_loops.c.src index 5c02c6e9f..bfa7ca224 100644 --- a/numpy/core/src/multiarray/lowlevel_strided_loops.c.src +++ b/numpy/core/src/multiarray/lowlevel_strided_loops.c.src @@ -21,12 +21,10 @@ #include "lowlevel_strided_loops.h" /* - * x86 platform may work with unaligned access, except when the - * compiler uses aligned SSE instructions, which gcc does in some - * cases. This is disabled for the time being. + * x86 platform works with unaligned access */ #if (defined(NPY_CPU_X86) || defined(NPY_CPU_AMD64)) -# define NPY_USE_UNALIGNED_ACCESS 0 +# define NPY_USE_UNALIGNED_ACCESS 1 #else # define NPY_USE_UNALIGNED_ACCESS 0 #endif |