This came to light when adding hard-flaot support to ARC glibc port
without hardware sqrt support causing glibc build to fail:
| ../sysdeps/ieee754/dbl-64/e_sqrt.c: In function '__ieee754_sqrt':
| ../sysdeps/ieee754/dbl-64/e_sqrt.c:58:54: error: unused variable 'ty'
[-Werror=unused-variable]
|
Reviewed-by: Adhemerval Zanella
---
sysdeps/generic/math-use-builtins.h | 3 +++
sysdeps/ieee754/dbl-64/e_sqrt.c | 6 ++
sysdeps/ieee754/flt-32/e_sqrtf.c| 16 ++--
3 files changed, 19 insertions(+), 6 deletions(-)
diff --git a/sysdeps/generic/math-use-builtins.h
b/sys
---
sysdeps/generic/math-use-builtins.h | 5 +
sysdeps/ieee754/dbl-64/s_fma.c | 6 ++
sysdeps/ieee754/dbl-64/s_fmaf.c | 6 ++
sysdeps/ieee754/float128/float128_private.h | 2 ++
sysdeps/ieee754/ldbl-128/s_fmal.c | 5 +
5 files changed, 24
Hi,
This was suggested by Adhemerval Zanella as part of ARC hard float
review.
There was testing snafu last time around (wrong branch on build server).
Although only relevant configs are aarch64 and PPC, the series passes
build-many-glibcs for following:
ARCH="aarch64-linux-gnu arm-linux-gnueabi
This is supposed to be a non functional change
Passes build-many for powerpc-linux-gnu
Reviewed-by: Adhemerval Zanella
---
sysdeps/powerpc/fpu/math-use-builtins.h | 69 +
sysdeps/powerpc/fpu/s_fma.c | 27 --
sysdeps/powerpc/fpu/s_fmaf.c| 2
introduce sysdep header math-use-builtins.h to replace aarch64
impementations with corresponding generic ones
- newly inroduced generic sqrt{,f}, fma{,f}
- existing floor{,f}, nearbyint{,f}, rint{,f}, round{,f}, trunc{,f}
- Note that generic copysign was already enabled (via generic
math-use
This is purely for review purposes to attest the interface defined
in prior patch
---
sysdeps/arc/dl-runtime.h | 42
1 file changed, 42 insertions(+)
create mode 100644 sysdeps/arc/dl-runtime.h
diff --git a/sysdeps/arc/dl-runtime.h b/sysdeps/arc/dl-runtim
Vineet Gupta (2):
dl-runtime: reloc_{offset,index} now functions arch overide'able
ARC/dl-runtime helper macros
elf/dl-runtime.c| 28 +++--
elf/dl-runtime.h| 30 ++
sysdeps/arc/dl-runtime.h| 42 +++
The existing macros are fragile and expect local variables with a
certain name. Fix this by defining them as functions with default
implementation in a new header dl-runtime.h which arches can override
if need be.
This came up during ARC port review, hence the need for argument pltgot
in reloc_ind
On Mon, Jun 1, 2020 at 10:45 PM Vineet Gupta via Libc-alpha
wrote:
>
> On 6/1/20 9:38 AM, Adhemerval Zanella via Libc-alpha wrote:
> >
> >
> > On 29/05/2020 23:00, Vineet Gupta wrote:
> >> Signed-off-by: Vineet Gupta
> >
> > LGTM, some comments below.
> >
> >> -#include
> >> -#include
> >> -
>
On 6/1/20 9:38 AM, Adhemerval Zanella via Libc-alpha wrote:
>
>
> On 29/05/2020 23:00, Vineet Gupta wrote:
>> Signed-off-by: Vineet Gupta
>
> LGTM, some comments below.
>
>> -#include
>> -#include
>> -
>> -double
>> -__ieee754_sqrt (double d)
>> -{
>> - return __builtin_sqrt (d);
>> -}
>> -
On 27/05/2020 22:36, Vineet Gupta wrote:
> On 5/27/20 12:17 PM, Adhemerval Zanella via Libc-alpha wrote:
>>
>>
>> On 22/04/2020 22:41, Vineet Gupta via Libc-alpha wrote:
>>> This includes all 4 TLS addressing models
>>>
>>> Signed-off-by: Vineet Gupta
>>
>> As prior patch we do not use DCO, but
On 01/06/2020 15:12, Vineet Gupta wrote:
> On 6/1/20 7:42 AM, Adhemerval Zanella via Libc-alpha wrote:
>> You will need to move this definitions inside the !USE_SQRTF_BUILTIN
>> to avoid defined by not used warnings. Current practice is to just
>> open code the constants and let compiler optimi
On 6/1/20 7:42 AM, Adhemerval Zanella via Libc-alpha wrote:
> You will need to move this definitions inside the !USE_SQRTF_BUILTIN
> to avoid defined by not used warnings. Current practice is to just
> open code the constants and let compiler optimize the constant pool:
Won't it be better to keep
On 6/1/20 7:18 AM, Florian Weimer wrote:
> There's a typo in commit subject: “prvoide”.
oops sorry, fixed now.
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
On 29/05/2020 23:00, Vineet Gupta wrote:
LGTM, thanks.
This new addition, along with the aarch64 one, makes me wonder if
it would be better to decompose the USE_* defined in multiple files
so the architecture adds only the required ones (instead of copy/paste
the whole file with the generic on
On 29/05/2020 23:00, Vineet Gupta wrote:
> Signed-off-by: Vineet Gupta
LGTM, some comments below.
> ---
> sysdeps/aarch64/fpu/e_sqrt.c| 27 --
> sysdeps/aarch64/fpu/e_sqrtf.c | 27 --
> sysdeps/aarch64/fpu/math-use-builtins.h | 70 +++
The commit subject contains a typo (“prvoide”).
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
On 01/06/2020 11:13, Adhemerval Zanella wrote:
>
>
> On 29/05/2020 23:00, Vineet Gupta wrote:
>
> LGTM, thanks.
>
> Reviewed-by: Adhemerval Zanella
>
>> ---
>> sysdeps/generic/math-use-builtins.h | 3 +++
>> sysdeps/ieee754/dbl-64/e_sqrt.c | 6 ++
>> sysdeps/ieee754/flt-32/e_sqrt
On 29/05/2020 23:00, Vineet Gupta wrote:
Ok with the fixes below.
Reviewed-by: Adhemerval Zanella
> ---
> sysdeps/generic/math-use-builtins.h | 4
> sysdeps/ieee754/dbl-64/s_fma.c | 6 ++
> sysdeps/ieee754/dbl-64/s_fmaf.c | 6 ++
> sysdeps/ieee
There's a typo in commit subject: “prvoide”.
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
On 29/05/2020 23:00, Vineet Gupta wrote:
LGTM, thanks.
Reviewed-by: Adhemerval Zanella
> ---
> sysdeps/generic/math-use-builtins.h | 3 +++
> sysdeps/ieee754/dbl-64/e_sqrt.c | 6 ++
> sysdeps/ieee754/flt-32/e_sqrtf.c| 6 ++
> 3 files changed, 15 insertions(+)
>
> diff --gi
22 matches
Mail list logo
