This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH, ARM] Unaligned accesses for builtin memcpy [2/2]
- From: Ramana Radhakrishnan <ramana dot radhakrishnan at linaro dot org>
- To: Julian Brown <julian at codesourcery dot com>
- Cc: gcc-patches at gcc dot gnu dot org, paul at codesourcery dot com, Richard Earnshaw <rearnsha at arm dot com>
- Date: Mon, 17 Oct 2011 17:17:19 +0100
- Subject: Re: [PATCH, ARM] Unaligned accesses for builtin memcpy [2/2]
- References: <20110506141332.63b88bcb@rex.config> <CACUk7=VxTyCKLmKFFX9qG+RQd75ZTZU1H4ngkQV_ApKFah0ueA@mail.gmail.com> <20111013185305.03fc1f7e@rex.config>
Hi Julian,
There are a couple of minor formatting nits.
>+static int
>+arm_movmemqi_unaligned (rtx *operands)
....
>+ /* Inlined memcpy using ldr/str/ldrh/strh can be quite big: try to limit
>+ size of code if optimizing for size. We'll use ldm/stm if src_aligned
>+ or dst_aligned though: allow more interleaving in those cases since the
>+ resulting code can be smaller. */
Watch out the or being misaligned compared to the other text.
>+ /* Note that the loop created by arm_block_move_unaligned_loop may be
>+ subject to loop unrolling, which makes tuning this condition a little
>+ redundant. */
Same with `redundant'
On 13 October 2011 18:53, Julian Brown <julian@codesourcery.com> wrote:
> On Wed, 28 Sep 2011 14:33:17 +0100
> No, sorry, I don't have any benchmark results available at present. I
> think we'd have to have terrifically bad luck for it to be a
> performance degradation, though...
Hmmm OK - but please watch out for any bug reports or any test-suite
fallout this week with
multilibs other than what you might have tested.
>
> I re-tested the patch for good measure, in case of bitrot (and the new
> tests pass with the patch applied, of course). OK to apply now?
OK by me with those changes.
cheers
Ramana