Fixed all issues pointed in the previous iteration.
There is now a significant change regarding how the sin(atan(x))
constant is calculated, as now it checks for which values such that
computing 1 + x*x won't overflow. There are two reasons for this
change: (1) Avoid an intermediate infinity value
On 10/9/18 5:29 PM, Giuliano Augusto Faulin Belinassi wrote:
> Fixed all issues pointed in the previous iteration.
> There is now a significant change regarding how the sin(atan(x))
> constant is calculated, as now it checks for which values such that
> computing 1 + x*x won't overflow. There are t
On Thu, 11 Oct 2018 at 23:07, Jeff Law wrote:
>
> On 10/9/18 5:29 PM, Giuliano Augusto Faulin Belinassi wrote:
> > Fixed all issues pointed in the previous iteration.
> > There is now a significant change regarding how the sin(atan(x))
> > constant is calculated, as now it checks for which values
On Fri, Oct 12, 2018 at 10:08 AM Christophe Lyon
wrote:
>
> On Thu, 11 Oct 2018 at 23:07, Jeff Law wrote:
> >
> > On 10/9/18 5:29 PM, Giuliano Augusto Faulin Belinassi wrote:
> > > Fixed all issues pointed in the previous iteration.
> > > There is now a significant change regarding how the sin(at
On Fri, 12 Oct 2018 at 13:27, Richard Biener wrote:
>
> On Fri, Oct 12, 2018 at 10:08 AM Christophe Lyon
> wrote:
> >
> > On Thu, 11 Oct 2018 at 23:07, Jeff Law wrote:
> > >
> > > On 10/9/18 5:29 PM, Giuliano Augusto Faulin Belinassi wrote:
> > > > Fixed all issues pointed in the previous iterat
On 10/12/18 8:11 AM, Christophe Lyon wrote:
> On Fri, 12 Oct 2018 at 13:27, Richard Biener
> wrote:
>>
>> On Fri, Oct 12, 2018 at 10:08 AM Christophe Lyon
>> wrote:
>>>
>>> On Thu, 11 Oct 2018 at 23:07, Jeff Law wrote:
On 10/9/18 5:29 PM, Giuliano Augusto Faulin Belinassi wrote:
>
On Fri, 12 Oct 2018 at 16:14, Jeff Law wrote:
>
> On 10/12/18 8:11 AM, Christophe Lyon wrote:
> > On Fri, 12 Oct 2018 at 13:27, Richard Biener
> > wrote:
> >>
> >> On Fri, Oct 12, 2018 at 10:08 AM Christophe Lyon
> >> wrote:
> >>>
> >>> On Thu, 11 Oct 2018 at 23:07, Jeff Law wrote:
>
> >>
Hello
What is the output of these functions on such arch? Since the
test didn't fail for the sinatan counterpart, an possible explanation
would be that the calculation of the sqrf, sqrt and sqrtl (lines
62-64) yielded a number that is far behind of what it should be.
However, I am still not su
On 10/12/18 9:51 AM, Giuliano Augusto Faulin Belinassi wrote:
> Hello
> What is the output of these functions on such arch? Since the
> test didn't fail for the sinatan counterpart, an possible explanation
> would be that the calculation of the sqrf, sqrt and sqrtl (lines
> 62-64) yielded a nu
> fc is built with:
>mov r0, #0
>movtr0, 24448
> so r0=0x5f80 (1.8446744E19) which looks ok
this is correct. My x86_64 yields the same value
> but then cosatanf is computed as (ie there's not call to cosatanf()):
>movwr3, #48430
>movtr3, 45883
>
On Fri, 12 Oct 2018 at 17:57, Jeff Law wrote:
>
> On 10/12/18 9:51 AM, Giuliano Augusto Faulin Belinassi wrote:
> > Hello
> > What is the output of these functions on such arch? Since the
> > test didn't fail for the sinatan counterpart, an possible explanation
> > would be that the calculati
On Fri, 12 Oct 2018 at 20:01, Giuliano Augusto Faulin Belinassi
wrote:
>
> > fc is built with:
> >mov r0, #0
> >movtr0, 24448
> > so r0=0x5f80 (1.8446744E19) which looks ok
>
> this is correct. My x86_64 yields the same value
>
> > but then cosatanf is computed as (ie t
Hello. Sorry for the late reply.
> but then cosatanf is computed as (ie there's not call to cosatanf()):
>movwr3, #48430
>movtr3, 45883
> so r3=0xb33bbd2e (-4.371139E-8) which is not zero.
Does this behavior is still present if we change the line 58 to:
int __attribute_
On Tue, 16 Oct 2018 at 17:15, Giuliano Augusto Faulin Belinassi
wrote:
>
> Hello. Sorry for the late reply.
>
> > but then cosatanf is computed as (ie there's not call to cosatanf()):
> >movwr3, #48430
> >movtr3, 45883
> > so r3=0xb33bbd2e (-4.371139E-8) which is not zero.
Hello, Christophe
Could you please dump the assembly of cosatanf here?
On Tue, Oct 16, 2018 at 12:23 PM Christophe Lyon
wrote:
>
> On Tue, 16 Oct 2018 at 17:15, Giuliano Augusto Faulin Belinassi
> wrote:
> >
> > Hello. Sorry for the late reply.
> >
> > > but then cosatanf is computed as (ie
On Tue, 16 Oct 2018 at 18:04, Giuliano Augusto Faulin Belinassi
wrote:
>
> Hello, Christophe
> Could you please dump the assembly of cosatanf here?
>
>
Sure:
.global cosatanf
.syntax unified
.arm
.fpu softvfp
.type cosatanf, %function
cosatanf:
.LFB3:
Hello,
> cosatanf:
> .LFB3:
> .loc 1 44 1 is_stmt 1
> .cfi_startproc
>@ args = 0, pretend = 0, frame = 0
> @ frame_needed = 0, uses_anonymous_args = 0
> .LVL50:
> .loc 1 45 5
> .loc 1 44 1 is_stmt 0
> push{r4, lr}
> .cfi_def_cfa_o
On Tue, 16 Oct 2018 at 19:33, Giuliano Augusto Faulin Belinassi
wrote:
>
> Hello,
>
> > cosatanf:
> > .LFB3:
> > .loc 1 44 1 is_stmt 1
> > .cfi_startproc
> >@ args = 0, pretend = 0, frame = 0
> > @ frame_needed = 0, uses_anonymous_args = 0
> > .LVL50:
> > .l
18 matches
Mail list logo
