On Fri, 23 Dec 2016, Yunqiang Su wrote:
> > 3, kernel: the emulation when a float exception taken.
>
> The big problem is that Loongson use the same encode for (unfused)
> madd.fmt to (fused) madd.fmt. We cannot trap this in kernel.
Why is that a problem? Just add a setting like `cpu_has_fus
> 在 2016年12月22日，23:31，Richard Sandiford 写道：
>
> Matthew Fortune writes:
>> Sandra Loosemore writes:
>>> On 12/21/2016 11:54 AM, Yunqiang Su wrote:
By this patch, I add a build-time option ` --with-unfused-madd4=yes/no',
and runtime option -m(no-)unfused-madd4,
to disable generat
> 在 2016年12月23日，10:47，Paul Hua 写道：
>
> On aarch64 target the result are 1.332268e-17.
> On x86 with fma target the result are also 1.332268e-17.
>
> so, I don't think the Loongson's madd.fmt/msub.fmt is incorrect.
OMG. Will this behavior make some app wrong working abnormal?
>
> We should d
On aarch64 target the result are 1.332268e-17.
On x86 with fma target the result are also 1.332268e-17.
so, I don't think the Loongson's madd.fmt/msub.fmt is incorrect.
We should do something for usage of fused madd, the all things has
been tested an fedora21 remix for loongson(1).
1, gcc：add lo
> 在 2016年12月23日，00:18，Richard Sandiford 写道：
>
> Yunqiang Su writes:
>>> 在 2016年12月22日，23:48，Yunqiang Su 写道：
>>>
在 2016年12月22日，23:31，Richard Sandiford
写道：
Matthew Fortune writes:
> Sandra Loosemore writes:
>> On 12/21/2016 11:54 AM, Yunqiang Su wrote:
Yunqiang Su writes:
>> 在 2016年12月22日，23:48，Yunqiang Su 写道：
>>
>>>
>>> 在 2016年12月22日，23:31，Richard Sandiford
>>> 写道：
>>>
>>> Matthew Fortune writes:
Sandra Loosemore writes:
> On 12/21/2016 11:54 AM, Yunqiang Su wrote:
>> By this patch, I add a build-time option ` --with-unfused
> 在 2016年12月22日，23:48，Yunqiang Su 写道：
>
>>
>> 在 2016年12月22日，23:31，Richard Sandiford 写道：
>>
>> Matthew Fortune writes:
>>> Sandra Loosemore writes:
On 12/21/2016 11:54 AM, Yunqiang Su wrote:
> By this patch, I add a build-time option ` --with-unfused-madd4=yes/no',
> and runtime
> 在 2016年12月22日，23:31，Richard Sandiford 写道：
>
> Matthew Fortune writes:
>> Sandra Loosemore writes:
>>> On 12/21/2016 11:54 AM, Yunqiang Su wrote:
By this patch, I add a build-time option ` --with-unfused-madd4=yes/no',
and runtime option -m(no-)unfused-madd4,
to disable generat
Matthew Fortune writes:
> Sandra Loosemore writes:
>> On 12/21/2016 11:54 AM, Yunqiang Su wrote:
>> > By this patch, I add a build-time option ` --with-unfused-madd4=yes/no',
>> > and runtime option -m(no-)unfused-madd4,
>> > to disable generate madd.fmt instructions.
>>
>> Your patch also needs
Sandra Loosemore writes:
> On 12/21/2016 11:54 AM, Yunqiang Su wrote:
> > By this patch, I add a build-time option ` --with-unfused-madd4=yes/no',
> > and runtime option -m(no-)unfused-madd4,
> > to disable generate madd.fmt instructions.
>
> Your patch also needs a documentation change so that t
On 12/21/2016 11:54 AM, Yunqiang Su wrote:
By this patch, I add a build-time option ` --with-unfused-madd4=yes/no’,
and runtime option -m(no-)unfused-madd4,
to disable generate madd.fmt instructions.
Your patch also needs a documentation change so that the new
command-line option is listed in
By this patch, I add a build-time option ` --with-unfused-madd4=yes/no’,
and runtime option -m(no-)unfused-madd4,
to disable generate madd.fmt instructions.
These 2 options is needed due to madd.fmt/msub.fmt on Loongson are broken,
which may generate wrong calculator result.
diff --git a/src/gcc/
12 matches
Mail list logo
