summaryrefslogtreecommitdiff
path: root/README
diff options
context:
space:
mode:
authorDavid Schleef <ds@schleef.org>2004-12-02 20:28:39 +0000
committerDavid Schleef <ds@schleef.org>2004-12-02 20:28:39 +0000
commit8da0b313ff66a50c2c70dd7a916dbfd8f8b2d11e (patch)
tree4a62bb8a713a74ae483063f6c8ab68e91334702e /README
parent10a2b7ce3d808da4d1c1044ac6a80d88438c0667 (diff)
downloadliboil-8da0b313ff66a50c2c70dd7a916dbfd8f8b2d11e.tar.gz
* README: some notes
* examples/.cvsignore: update some cvsignore files * examples/md5/.cvsignore: * liboil/.cvsignore: * liboil/colorspace/.cvsignore: * liboil/copy/.cvsignore: * liboil/md5/.cvsignore: * testsuite/.cvsignore:
Diffstat (limited to 'README')
-rw-r--r--README15
1 files changed, 15 insertions, 0 deletions
diff --git a/README b/README
new file mode 100644
index 0000000..e3c5f80
--- /dev/null
+++ b/README
@@ -0,0 +1,15 @@
+
+Ideas:
+
+implementations should be flagged as to whether or not they contain
+assembly, and also whether or not they are the product of the
+alternate optimization. This way, testing code can automatically
+flag functions where the alternate optimization is faster than the
+C or asm code, indicating that the compiler output should be turned
+into assembly.
+
+
+
+Why is trans8x8_f64 so slow on powerpc (compared to trans8x8_u16)?
+
+