This is the mail archive of the
mailing list for the GCC project.
Re: [patch] [4.2 projects] vectorize type conversions - 1/6 (updated versus r114433)
- From: "Richard Guenther" <richard dot guenther at gmail dot com>
- To: "Victor Kaplansky" <VICTORK at il dot ibm dot com>
- Cc: gcc-patches at gcc dot gnu dot org, rth at redhat dot com, stevenb dot gcc at gmail dot com
- Date: Tue, 6 Jun 2006 18:28:46 +0200
- Subject: Re: [patch] [4.2 projects] vectorize type conversions - 1/6 (updated versus r114433)
- References: <OF94A55977.98995843-ONC2257185.00590A83-C2257185.email@example.com>
On 6/6/06, Victor Kaplansky <VICTORK@il.ibm.com> wrote:
This is updated version of "vectorize type conversions - 1/6" patch,
originally sent by Dorit in
This is a cleanup towards incorporating the new functionality:
Handling of aligned and unaligned loads can be largely unified, resulting
in a more readable code and avoids code duplication.
Bootstrapped with vectorization enabled and tested on the vectorizer
testcases on powerpc-linux.
ok for mainline?
+ /* <2> */
What are the <1>/<2>/<3> tags refering to?
if (alignment_support_cheme == dr_aligned
|| alignment_support_cheme == dr_unaligned_supported)
- /* Create:
p = initial_addr;
indx = 0;
Instead of using pseudo-code to comment the following code, you should put
an english sentence above the created code examples.
Otherwise this looks like a good cleanup (which I cannot approve, though).