On Wed, Feb 24, 2016 at 09:20:11AM +0100, Clément Bœsch wrote:
> On Tue, Feb 23, 2016 at 10:40:08PM -0300, James Almer wrote:
> [...]
> > That aside, note that these runtime erros happen with every test using md5
> > and
> > don't make ubsan register them as failed. In the link above something els
Le sextidi 6 ventôse, an CCXXIV, James Almer a écrit :
> It may result in a slight speed boost, so it would be nice if someone with
> a ppc machine could test it.
I can access this:
gcc110 2TB 4x16x3.55 GHz IBM POWER7 / 64 GB RAM /
IBM Power 730 Express server / Fedora 18 ppc64
processor :
On 2/24/2016 1:15 PM, Ronald S. Bultje wrote:
> Hi,
>
> On Wed, Feb 24, 2016 at 10:47 AM, James Almer wrote:
>
>> On 2/24/2016 12:13 PM, Ronald S. Bultje wrote:
>>> Hi,
>>>
>>> On Tue, Feb 23, 2016 at 8:40 PM, James Almer wrote:
>>>
Tested on x86 and benched with no apparent speed loss
>>>
Hi,
On Wed, Feb 24, 2016 at 10:47 AM, James Almer wrote:
> On 2/24/2016 12:13 PM, Ronald S. Bultje wrote:
> > Hi,
> >
> > On Tue, Feb 23, 2016 at 8:40 PM, James Almer wrote:
> >
> >> Tested on x86 and benched with no apparent speed loss
> >
> >
> > That's because x86 supports unaligned loads.
>
On 2/24/2016 12:13 PM, Ronald S. Bultje wrote:
> Hi,
>
> On Tue, Feb 23, 2016 at 8:40 PM, James Almer wrote:
>
>> Tested on x86 and benched with no apparent speed loss
>
>
> That's because x86 supports unaligned loads.
>
> How come you get unaligned loads? Shouldn't this prevent it?
>
> -
Hi,
On Tue, Feb 23, 2016 at 8:40 PM, James Almer wrote:
> Tested on x86 and benched with no apparent speed loss
That's because x86 supports unaligned loads.
How come you get unaligned loads? Shouldn't this prevent it?
-if (HAVE_BIGENDIAN || (!HAVE_FAST_UNALIGNED && ((intptr_t)src & 3)))
On Tue, Feb 23, 2016 at 10:40:08PM -0300, James Almer wrote:
[...]
> That aside, note that these runtime erros happen with every test using md5 and
> don't make ubsan register them as failed. In the link above something else
> made
> it fail. It makes me wonder how many other ubsan errors are hidd
Signed-off-by: James Almer
---
This makes ubsan happier, removing hundreds of runtime errors as seen in
http://fatebeta.ffmpeg.org/report/x86_64-archlinux-gcc-ubsan/20160216235604
Tested on x86 and benched with no apparent speed loss, but the more people
bench it the better as it's a very importan
