* Re: [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16
@ 2021-07-28 2:29 David Edelsohn
2021-07-29 22:03 ` Paul A. Clarke
0 siblings, 1 reply; 6+ messages in thread
From: David Edelsohn @ 2021-07-28 2:29 UTC (permalink / raw)
To: Paul A. Clarke; +Cc: GCC Patches, Bill Schmidt
> Add a naive implementation of the subject x86 intrinsic to
> ease porting.
>
> 2021-07-15 Paul A. Clarke <pc@us.ibm.com>
>
> gcc
> * config/rs6000/smmintrin.h (_mm_minpos_epu16): New.
Segher already approved this with the changes requested.
Thanks, David
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16
2021-07-28 2:29 [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16 David Edelsohn
@ 2021-07-29 22:03 ` Paul A. Clarke
0 siblings, 0 replies; 6+ messages in thread
From: Paul A. Clarke @ 2021-07-29 22:03 UTC (permalink / raw)
To: David Edelsohn; +Cc: Bill Schmidt, GCC Patches, segher
On Tue, Jul 27, 2021 at 10:29:13PM -0400, David Edelsohn via Gcc-patches wrote:
> > Add a naive implementation of the subject x86 intrinsic to
> > ease porting.
> >
> > 2021-07-15 Paul A. Clarke <pc@us.ibm.com>
> >
> > gcc
> > * config/rs6000/smmintrin.h (_mm_minpos_epu16): New.
>
> Segher already approved this with the changes requested.
Segher said:
| This does not compute the index correctly for big endian (it needs to
| walk from right to left for that). The construction of the return value
| looks wrong as well.
|
| Okay for trunk with that fixed. Thanks!
I responded:
| I'm not seeing the issue here. The values are numbered by element order,
| and the results are in the "first" (minimum value) and "second" (index of
| first encountered minimum value in element order) elements of the result.
I did not get a response, nor did I change any code. It feels like a stretch
to equate the above exchange to "approved", so I'll continue to wait for
explicit approval.
PC
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16
2021-08-02 22:29 ` Segher Boessenkool
@ 2021-08-03 13:01 ` Paul A. Clarke
0 siblings, 0 replies; 6+ messages in thread
From: Paul A. Clarke @ 2021-08-03 13:01 UTC (permalink / raw)
To: Segher Boessenkool; +Cc: wschmidt, gcc-patches
On Mon, Aug 02, 2021 at 05:29:08PM -0500, Segher Boessenkool wrote:
> On Thu, Jul 15, 2021 at 06:29:17PM -0500, Paul A. Clarke wrote:
> > Add a naive implementation of the subject x86 intrinsic to
> > ease porting.
>
> > --- a/gcc/config/rs6000/smmintrin.h
> > +++ b/gcc/config/rs6000/smmintrin.h
> > @@ -172,4 +172,31 @@ _mm_test_mix_ones_zeros (__m128i __A, __m128i __mask)
> > return any_ones * any_zeros;
> > }
> >
> > +/* Return horizontal packed word minimum and its index in bits [15:0]
> > + and bits [18:16] respectively. */
> > +__inline __m128i
> > +__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > +_mm_minpos_epu16 (__m128i __A)
> > +{
> > + union __u
> > + {
> > + __m128i __m;
> > + __v8hu __uh;
> > + };
> > + union __u __u = { .__m = __A }, __r = { .__m = {0} };
> > + unsigned short __ridx = 0;
> > + unsigned short __rmin = __u.__uh[__ridx];
> > + for (unsigned long __i = 1; __i < 8; __i++)
> > + {
> > + if (__u.__uh[__i] < __rmin)
> > + {
> > + __rmin = __u.__uh[__i];
> > + __ridx = __i;
> > + }
> > + }
> > + __r.__uh[0] = __rmin;
> > + __r.__uh[1] = __ridx;
> > + return __r.__m;
> > +}
>
> As before: does this work correctly on BE? Was it tested there?
Per the "cover letter":
| Tested on BE, LE (32 and 64bit).
> Okay for trunk if so. Thanks!
Thanks! I'll push this shortly.
PC
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16
2021-07-15 23:29 ` [PATCH v3 1/2] " Paul A. Clarke
2021-07-16 20:06 ` Bill Schmidt
@ 2021-08-02 22:29 ` Segher Boessenkool
2021-08-03 13:01 ` Paul A. Clarke
1 sibling, 1 reply; 6+ messages in thread
From: Segher Boessenkool @ 2021-08-02 22:29 UTC (permalink / raw)
To: Paul A. Clarke; +Cc: gcc-patches, wschmidt
Hi!
On Thu, Jul 15, 2021 at 06:29:17PM -0500, Paul A. Clarke wrote:
> Add a naive implementation of the subject x86 intrinsic to
> ease porting.
> --- a/gcc/config/rs6000/smmintrin.h
> +++ b/gcc/config/rs6000/smmintrin.h
> @@ -172,4 +172,31 @@ _mm_test_mix_ones_zeros (__m128i __A, __m128i __mask)
> return any_ones * any_zeros;
> }
>
> +/* Return horizontal packed word minimum and its index in bits [15:0]
> + and bits [18:16] respectively. */
> +__inline __m128i
> +__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> +_mm_minpos_epu16 (__m128i __A)
> +{
> + union __u
> + {
> + __m128i __m;
> + __v8hu __uh;
> + };
> + union __u __u = { .__m = __A }, __r = { .__m = {0} };
> + unsigned short __ridx = 0;
> + unsigned short __rmin = __u.__uh[__ridx];
> + for (unsigned long __i = 1; __i < 8; __i++)
> + {
> + if (__u.__uh[__i] < __rmin)
> + {
> + __rmin = __u.__uh[__i];
> + __ridx = __i;
> + }
> + }
> + __r.__uh[0] = __rmin;
> + __r.__uh[1] = __ridx;
> + return __r.__m;
> +}
As before: does this work correctly on BE? Was it tested there?
Okay for trunk if so. Thanks!
Segher
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16
2021-07-15 23:29 ` [PATCH v3 1/2] " Paul A. Clarke
@ 2021-07-16 20:06 ` Bill Schmidt
2021-08-02 22:29 ` Segher Boessenkool
1 sibling, 0 replies; 6+ messages in thread
From: Bill Schmidt @ 2021-07-16 20:06 UTC (permalink / raw)
To: Paul A. Clarke, gcc-patches; +Cc: segher
Hi Paul,
LGTM. Recommend maintainers approve.
Thanks for the cleanups,
Bill
On 7/15/21 6:29 PM, Paul A. Clarke wrote:
> Add a naive implementation of the subject x86 intrinsic to
> ease porting.
>
> 2021-07-15 Paul A. Clarke <pc@us.ibm.com>
>
> gcc
> * config/rs6000/smmintrin.h (_mm_minpos_epu16): New.
> ---
> v3: Minor formatting changes per review from Bill.
> v2: Minor formatting changes per review from Segher.
>
> gcc/config/rs6000/smmintrin.h | 27 +++++++++++++++++++++++++++
> 1 file changed, 27 insertions(+)
>
> diff --git a/gcc/config/rs6000/smmintrin.h b/gcc/config/rs6000/smmintrin.h
> index 16fd34d836ff..6a010fdbb96f 100644
> --- a/gcc/config/rs6000/smmintrin.h
> +++ b/gcc/config/rs6000/smmintrin.h
> @@ -172,4 +172,31 @@ _mm_test_mix_ones_zeros (__m128i __A, __m128i __mask)
> return any_ones * any_zeros;
> }
>
> +/* Return horizontal packed word minimum and its index in bits [15:0]
> + and bits [18:16] respectively. */
> +__inline __m128i
> +__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> +_mm_minpos_epu16 (__m128i __A)
> +{
> + union __u
> + {
> + __m128i __m;
> + __v8hu __uh;
> + };
> + union __u __u = { .__m = __A }, __r = { .__m = {0} };
> + unsigned short __ridx = 0;
> + unsigned short __rmin = __u.__uh[__ridx];
> + for (unsigned long __i = 1; __i < 8; __i++)
> + {
> + if (__u.__uh[__i] < __rmin)
> + {
> + __rmin = __u.__uh[__i];
> + __ridx = __i;
> + }
> + }
> + __r.__uh[0] = __rmin;
> + __r.__uh[1] = __ridx;
> + return __r.__m;
> +}
> +
> #endif
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16
2021-07-15 23:29 [PATCH v3 0/2] " Paul A. Clarke
@ 2021-07-15 23:29 ` Paul A. Clarke
2021-07-16 20:06 ` Bill Schmidt
2021-08-02 22:29 ` Segher Boessenkool
0 siblings, 2 replies; 6+ messages in thread
From: Paul A. Clarke @ 2021-07-15 23:29 UTC (permalink / raw)
To: gcc-patches; +Cc: segher, wschmidt
Add a naive implementation of the subject x86 intrinsic to
ease porting.
2021-07-15 Paul A. Clarke <pc@us.ibm.com>
gcc
* config/rs6000/smmintrin.h (_mm_minpos_epu16): New.
---
v3: Minor formatting changes per review from Bill.
v2: Minor formatting changes per review from Segher.
gcc/config/rs6000/smmintrin.h | 27 +++++++++++++++++++++++++++
1 file changed, 27 insertions(+)
diff --git a/gcc/config/rs6000/smmintrin.h b/gcc/config/rs6000/smmintrin.h
index 16fd34d836ff..6a010fdbb96f 100644
--- a/gcc/config/rs6000/smmintrin.h
+++ b/gcc/config/rs6000/smmintrin.h
@@ -172,4 +172,31 @@ _mm_test_mix_ones_zeros (__m128i __A, __m128i __mask)
return any_ones * any_zeros;
}
+/* Return horizontal packed word minimum and its index in bits [15:0]
+ and bits [18:16] respectively. */
+__inline __m128i
+__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
+_mm_minpos_epu16 (__m128i __A)
+{
+ union __u
+ {
+ __m128i __m;
+ __v8hu __uh;
+ };
+ union __u __u = { .__m = __A }, __r = { .__m = {0} };
+ unsigned short __ridx = 0;
+ unsigned short __rmin = __u.__uh[__ridx];
+ for (unsigned long __i = 1; __i < 8; __i++)
+ {
+ if (__u.__uh[__i] < __rmin)
+ {
+ __rmin = __u.__uh[__i];
+ __ridx = __i;
+ }
+ }
+ __r.__uh[0] = __rmin;
+ __r.__uh[1] = __ridx;
+ return __r.__m;
+}
+
#endif
--
2.27.0
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2021-08-03 13:01 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-07-28 2:29 [PATCH v3 1/2] rs6000: Add support for _mm_minpos_epu16 David Edelsohn
2021-07-29 22:03 ` Paul A. Clarke
-- strict thread matches above, loose matches on Subject: below --
2021-07-15 23:29 [PATCH v3 0/2] " Paul A. Clarke
2021-07-15 23:29 ` [PATCH v3 1/2] " Paul A. Clarke
2021-07-16 20:06 ` Bill Schmidt
2021-08-02 22:29 ` Segher Boessenkool
2021-08-03 13:01 ` Paul A. Clarke
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).