diff options
| author | Torbjorn Granlund <tege@gmplib.org> | 2009-11-15 11:43:55 +0100 |
|---|---|---|
| committer | Torbjorn Granlund <tege@gmplib.org> | 2009-11-15 11:43:55 +0100 |
| commit | 2fb438b0ac9abb72f2cbd85f624ad0603fdefd4b (patch) | |
| tree | 8289ec0264c4386ca44c3e9d38e209f696c842ed /mpn/cray | |
| parent | 442c9f32476abf86274f76f91a3a725c9fefce8d (diff) | |
| download | gmp-2fb438b0ac9abb72f2cbd85f624ad0603fdefd4b.tar.gz | |
More spelling fixes.
Diffstat (limited to 'mpn/cray')
| -rw-r--r-- | mpn/cray/README | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/mpn/cray/README b/mpn/cray/README index ab3b03270..ccd743997 100644 --- a/mpn/cray/README +++ b/mpn/cray/README @@ -34,7 +34,7 @@ systems with cfp floating point, the main obstacle is the forming of 128-bit products. For IEEE systems, adding, and in particular computing carry is the main issue. There are no vectorizing unsigned-less-than instructions, and the sequence that implement that -opetration is very long. +operation is very long. Shifting is the only operation that is simple to make fast. All Cray systems have a bitblt instructions (Vi Vj,Vj<Ak and Vi Vj,Vj>Ak) that |