diff mbox series

rs6000: Add cntlzdm and cnttzdm

Message ID 20200508131718.24294-1-wschmidt@linux.ibm.com
State New
Headers show
Series rs6000: Add cntlzdm and cnttzdm | expand

Commit Message

Bill Schmidt May 8, 2020, 1:17 p.m. UTC 
From: Kelvin Nilsen <kelvin@gcc.gnu.org>

Add support for new scalar instructions for counting leading or
trailing zeros under control of a bitmask.

Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no
regressions.  Is this okay for master?

Thanks,
Bill

[gcc]

2020-05-08  Kelvin Nilsen  <kelvin@gcc.gnu.org>

	* config/rs6000/rs6000-builtin.def (__builtin_cntlzdm): New
	built-in function definition.
	(__builtin_cnttzdm): Likewise.,
	* config/rs6000/rs6000.md (UNSPEC_CNTLZDM): New constant.
	(UNSPEC_CNTTZDM): Likewise.
	(cntlzdm): New insn.
	(cnttzdm): Likewise.
	* doc/extend.texi (Basic PowerPC Built-in Functions available for
	a Future Architecture): Add descriptions of __builtin_cntlzdm and
	__builtin_cnttzdm functions.

[gcc/testsuite]

2020-05-08  Kelvin Nilsen  <kelvin@gcc.gnu.org>

	* gcc.target/powerpc/cntlzdm-0.c: New test.
	* gcc.target/powerpc/cntlzdm-1.c: New test.
	* gcc.target/powerpc/cnttzdm-0.c: New test.
	* gcc.target/powerpc/cnttzdm-1.c: New test.
---
 gcc/config/rs6000/rs6000-builtin.def         |  2 +
 gcc/config/rs6000/rs6000.md                  | 20 +++++++
 gcc/doc/extend.texi                          | 16 ++++++
 gcc/testsuite/gcc.target/powerpc/cntlzdm-0.c | 57 ++++++++++++++++++++
 gcc/testsuite/gcc.target/powerpc/cntlzdm-1.c | 56 +++++++++++++++++++
 gcc/testsuite/gcc.target/powerpc/cnttzdm-0.c | 53 ++++++++++++++++++
 gcc/testsuite/gcc.target/powerpc/cnttzdm-1.c | 53 ++++++++++++++++++
 7 files changed, 257 insertions(+)
 create mode 100644 gcc/testsuite/gcc.target/powerpc/cntlzdm-0.c
 create mode 100644 gcc/testsuite/gcc.target/powerpc/cntlzdm-1.c
 create mode 100644 gcc/testsuite/gcc.target/powerpc/cnttzdm-0.c
 create mode 100644 gcc/testsuite/gcc.target/powerpc/cnttzdm-1.c

Comments

Segher Boessenkool May 8, 2020, 11:51 p.m. UTC | #1 
On Fri, May 08, 2020 at 08:17:18AM -0500, Bill Schmidt wrote:
> From: Kelvin Nilsen <kelvin@gcc.gnu.org>
> 
> Add support for new scalar instructions for counting leading or
> trailing zeros under control of a bitmask.
> 
> Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no
> regressions.  Is this okay for master?

Ooh, I found problems!

> 	* config/rs6000/rs6000-builtin.def (__builtin_cntlzdm): New
> 	built-in function definition.
> 	(__builtin_cnttzdm): Likewise.,

Stray comma.

> +(define_insn "cntlzdm"
> +  [(set (match_operand:DI 0 "gpc_reg_operand" "=r")
> +	(unspec:DI [(match_operand:DI 1 "gpc_reg_operand" "r")
> +		    (match_operand:DI 2 "gpc_reg_operand" "r")]
> +	 UNSPEC_CNTLZDM))]
> +   "TARGET_FUTURE && TARGET_64BIT"
> +   "cntlzdm %0,%1,%2"
> +   [(set_attr "type" "integer")])

TARGET_POWERPC64.

> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/powerpc/cntlzdm-0.c
> @@ -0,0 +1,57 @@
> +/* { dg-do compile } */
> +/* { dg-require-effective-target lp64 } */

And powerpc64 here as well then.

Not sure if this is a bigger problem than the comma thing though.


Segher
Li, Pan2 via Gcc-patches May 11, 2020, 4:45 p.m. UTC | #2 
On 5/8/20 6:51 PM, Segher Boessenkool wrote:
> On Fri, May 08, 2020 at 08:17:18AM -0500, Bill Schmidt wrote:
>> From: Kelvin Nilsen <kelvin@gcc.gnu.org>
>>
>> Add support for new scalar instructions for counting leading or
>> trailing zeros under control of a bitmask.
>>
>> Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no
>> regressions.  Is this okay for master?
> Ooh, I found problems!

Thanks for catching these!  Okay with them fixed?

Thanks,
Bill

>
>> 	* config/rs6000/rs6000-builtin.def (__builtin_cntlzdm): New
>> 	built-in function definition.
>> 	(__builtin_cnttzdm): Likewise.,
> Stray comma.
>
>> +(define_insn "cntlzdm"
>> +  [(set (match_operand:DI 0 "gpc_reg_operand" "=r")
>> +	(unspec:DI [(match_operand:DI 1 "gpc_reg_operand" "r")
>> +		    (match_operand:DI 2 "gpc_reg_operand" "r")]
>> +	 UNSPEC_CNTLZDM))]
>> +   "TARGET_FUTURE && TARGET_64BIT"
>> +   "cntlzdm %0,%1,%2"
>> +   [(set_attr "type" "integer")])
> TARGET_POWERPC64.
>
>> --- /dev/null
>> +++ b/gcc/testsuite/gcc.target/powerpc/cntlzdm-0.c
>> @@ -0,0 +1,57 @@
>> +/* { dg-do compile } */
>> +/* { dg-require-effective-target lp64 } */
> And powerpc64 here as well then.
>
> Not sure if this is a bigger problem than the comma thing though.
>
>
> Segher
Segher Boessenkool May 12, 2020, 1:33 p.m. UTC | #3 
On Mon, May 11, 2020 at 11:45:09AM -0500, Bill Schmidt wrote:
> On 5/8/20 6:51 PM, Segher Boessenkool wrote:
> >On Fri, May 08, 2020 at 08:17:18AM -0500, Bill Schmidt wrote:
> >>From: Kelvin Nilsen <kelvin@gcc.gnu.org>
> >>
> >>Add support for new scalar instructions for counting leading or
> >>trailing zeros under control of a bitmask.
> >>
> >>Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no
> >>regressions.  Is this okay for master?
> >Ooh, I found problems!
> 
> Thanks for catching these!  Okay with them fixed?

Erm yes, sorry I forgot to say -- I don't see any othger problems.


Segher
diff mbox series

Patch

diff --git a/gcc/config/rs6000/rs6000-builtin.def b/gcc/config/rs6000/rs6000-builtin.def
index 9d80d03fe50..c05d9f53d54 100644
--- a/gcc/config/rs6000/rs6000-builtin.def
+++ b/gcc/config/rs6000/rs6000-builtin.def
@@ -2575,6 +2575,8 @@  BU_P9_OVERLOAD_2 (CMPEQB,	"byte_in_set")
 
 /* Future architecture scalar built-ins.  */
 BU_FUTURE_MISC_2 (CFUGED, "cfuged", CONST, cfuged)
+BU_FUTURE_MISC_2 (CNTLZDM, "cntlzdm", CONST, cntlzdm)
+BU_FUTURE_MISC_2 (CNTTZDM, "cnttzdm", CONST, cnttzdm)
 
 /* Future architecture vector built-ins.  */
 BU_FUTURE_V_2 (VCFUGED, "vcfuged", CONST, vcfuged)
diff --git a/gcc/config/rs6000/rs6000.md b/gcc/config/rs6000/rs6000.md
index 001ec27f403..19b3d7263d3 100644
--- a/gcc/config/rs6000/rs6000.md
+++ b/gcc/config/rs6000/rs6000.md
@@ -149,6 +149,8 @@  (define_c_enum "unspec"
    UNSPEC_PLTSEQ
    UNSPEC_PLT16_HA
    UNSPEC_CFUGED
+   UNSPEC_CNTLZDM
+   UNSPEC_CNTTZDM
   ])
 
 ;;
@@ -2463,6 +2465,24 @@  (define_insn "cfuged"
    "cfuged %0,%1,%2"
    [(set_attr "type" "integer")])
 
+(define_insn "cntlzdm"
+  [(set (match_operand:DI 0 "gpc_reg_operand" "=r")
+	(unspec:DI [(match_operand:DI 1 "gpc_reg_operand" "r")
+		    (match_operand:DI 2 "gpc_reg_operand" "r")]
+	 UNSPEC_CNTLZDM))]
+   "TARGET_FUTURE && TARGET_64BIT"
+   "cntlzdm %0,%1,%2"
+   [(set_attr "type" "integer")])
+
+(define_insn "cnttzdm"
+  [(set (match_operand:DI 0 "gpc_reg_operand" "=r")
+	(unspec:DI [(match_operand:DI 1 "gpc_reg_operand" "r")
+		    (match_operand:DI 2 "gpc_reg_operand" "r")]
+	 UNSPEC_CNTTZDM))]
+   "TARGET_FUTURE && TARGET_64BIT"
+   "cnttzdm %0,%1,%2"
+   [(set_attr "type" "integer")])
+
 (define_insn "cmpb<mode>3"
   [(set (match_operand:GPR 0 "gpc_reg_operand" "=r")
 	(unspec:GPR [(match_operand:GPR 1 "gpc_reg_operand" "r")
diff --git a/gcc/doc/extend.texi b/gcc/doc/extend.texi
index f7b30c7ead7..8c833d39fec 100644
--- a/gcc/doc/extend.texi
+++ b/gcc/doc/extend.texi
@@ -17556,6 +17556,22 @@  Perform a 64-bit centrifuge operation, as if implemented by the Future
 @code{cfuged} instruction.
 @findex __builtin_cfuged
 
+@smallexample
+@exdent unsigned long long int
+@exdent __builtin_cntlzdm (unsigned long long int, unsigned long long int)
+@end smallexample
+Perform a 64-bit count leading zeros operation under mask, as if
+implemented by the future @code{cntlzdm} instruction.
+@findex __builtin_cntlzdm
+
+@smallexample
+@exdent unsigned long long int
+@exdent __builtin_cnttzdm (unsigned long long int, unsigned long long int)
+@end smallexample
+Perform a 64-bit count trailing zeros operation under mask, as if
+implemented by the future @code{cnttzdm} instruction.
+@findex __builtin_cnttzdm
+
 @node PowerPC AltiVec/VSX Built-in Functions
 @subsection PowerPC AltiVec/VSX Built-in Functions
 
diff --git a/gcc/testsuite/gcc.target/powerpc/cntlzdm-0.c b/gcc/testsuite/gcc.target/powerpc/cntlzdm-0.c
new file mode 100644
index 00000000000..e3ce63c3447
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/cntlzdm-0.c
@@ -0,0 +1,57 @@ 
+/* { dg-do compile } */
+/* { dg-require-effective-target lp64 } */
+/* { dg-options "-mdejagnu-cpu=future" } */
+
+extern void abort (void);
+
+unsigned long long int
+do_cntlzdm (unsigned long long int source, unsigned long long int mask)
+{
+  return __builtin_cntlzdm (source, mask);
+}
+
+int main (int argc, char *argv [])
+{
+  unsigned long long int sources [4], masks [4];
+  unsigned long long int intermediates [4][4] = {
+    /* sources[0] with each of masks [0 .. 3] */
+    { 0x0000a5f0ll, 0x00007e3cll, 0x000050ecll, 0x0000af73ll },
+    /* sources[1] with each of masks [0 .. 3] */
+    { 0x00007e3cll, 0x0000a5f0ll, 0x0000ec50ll, 0x000073afll },
+    /* sources[2] with each of masks [0 .. 3] */
+    { 0x00003ca5ll, 0x0000f07ell, 0x0000c50ell, 0x00003af7ll },
+    /* sources[3] with each of masks [0 .. 3] */
+    { 0x00005a0fll, 0x0000e7c3ll, 0x0000af73ll, 0x000050ecll },
+  };
+  unsigned long long int results [4][4] = {
+    { 0, 1, 1, 0 },
+    { 1, 0, 0, 1 },
+    { 2, 0, 0, 2 },
+    { 1, 0, 0, 1 },
+  };
+
+  sources[0] = 0xa5f07e3cll;
+  sources[1] = 0x7e3ca5f0ll;
+  sources[2] = 0x3ca5f07ell;
+  sources[3] = 0x5a0fe7c3ll;
+
+  masks[0] = 0xffff0000ll;
+  masks[1] = 0x0000ffffll;
+  masks[2] = 0x0f0f0f0fll;
+  masks[3] = 0xf0f0f0f0ll;
+
+  unsigned long long int result;
+
+  for (int i = 0; i < 4; i++)
+    {
+      for (int j = 0; j < 4; j++)
+	{
+	  if (do_cntlzdm (sources[i], masks[j]) != results [i][j])
+	    abort ();
+	}
+    }
+
+  return 0;
+}
+
+/* { dg-final { scan-assembler {\mcntlzdm\M} } } */
diff --git a/gcc/testsuite/gcc.target/powerpc/cntlzdm-1.c b/gcc/testsuite/gcc.target/powerpc/cntlzdm-1.c
new file mode 100644
index 00000000000..3239c386bb7
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/cntlzdm-1.c
@@ -0,0 +1,56 @@ 
+/* { dg-do run } */
+/* { dg-require-effective-target powerpc_future_hw } */
+/* { dg-require-effective-target lp64 } */
+/* { dg-options "-mdejagnu-cpu=future" } */
+
+extern void abort (void);
+
+unsigned long long int
+do_cntlzdm (unsigned long long int source, unsigned long long int mask)
+{
+  return __builtin_cntlzdm (source, mask);
+}
+
+int main (int argc, char *argv [])
+{
+  unsigned long long int sources [4], masks [4];
+  unsigned long long int intermediates [4][4] = {
+    /* sources[0] with each of masks [0 .. 3] */
+    { 0x0000a5f0ll, 0x00007e3cll, 0x000050ecll, 0x0000af73ll },
+    /* sources[1] with each of masks [0 .. 3] */
+    { 0x00007e3cll, 0x0000a5f0ll, 0x0000ec50ll, 0x000073afll },
+    /* sources[2] with each of masks [0 .. 3] */
+    { 0x00003ca5ll, 0x0000f07ell, 0x0000c50ell, 0x00003af7ll },
+    /* sources[3] with each of masks [0 .. 3] */
+    { 0x00005a0fll, 0x0000e7c3ll, 0x0000af73ll, 0x000050ecll },
+  };
+  unsigned long long int results [4][4] = {
+    { 0, 1, 1, 0 },
+    { 1, 0, 0, 1 },
+    { 2, 0, 0, 2 },
+    { 1, 0, 0, 1 },
+  };
+
+  sources[0] = 0xa5f07e3cll;
+  sources[1] = 0x7e3ca5f0ll;
+  sources[2] = 0x3ca5f07ell;
+  sources[3] = 0x5a0fe7c3ll;
+
+  masks[0] = 0xffff0000ll;
+  masks[1] = 0x0000ffffll;
+  masks[2] = 0x0f0f0f0fll;
+  masks[3] = 0xf0f0f0f0ll;
+
+  unsigned long long int result;
+
+  for (int i = 0; i < 4; i++)
+    {
+      for (int j = 0; j < 4; j++)
+	{
+	  if (do_cntlzdm (sources[i], masks[j]) != results [i][j])
+	    abort ();
+	}
+    }
+
+  return 0;
+}
diff --git a/gcc/testsuite/gcc.target/powerpc/cnttzdm-0.c b/gcc/testsuite/gcc.target/powerpc/cnttzdm-0.c
new file mode 100644
index 00000000000..6601f1f6471
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/cnttzdm-0.c
@@ -0,0 +1,53 @@ 
+/* { dg-do compile } */
+/* { dg-options "-mdejagnu-cpu=future" } */
+
+extern void abort (void);
+
+unsigned long long int
+do_cnttzdm (unsigned long long int source, unsigned long long int mask) {
+  return __builtin_cnttzdm (source, mask);
+}
+
+int main (int argc, char *argv [])
+{
+  unsigned long long int sources [4], masks [4];
+  unsigned long long int intermediates [4][4] = {
+    /* sources[0] with each of masks [0 .. 3] */
+    { 0x0000a5f0ll, 0x00007e3cll, 0x000050ecll, 0x0000af73ll },
+    /* sources[1] with each of masks [0 .. 3] */
+    { 0x00007e3cll, 0x0000a5f0ll, 0x0000ec50ll, 0x000073afll },
+    /* sources[2] with each of masks [0 .. 3] */
+    { 0x00003ca5ll, 0x0000f07ell, 0x0000c50ell, 0x00003af7ll },
+    /* sources[3] with each of masks [0 .. 3] */
+    { 0x00005a0fll, 0x0000e7c3ll, 0x0000af73ll, 0x000050ecll },
+  };
+  unsigned long long int results [4][4] = {
+    { 4, 2, 2, 0 },
+    { 2, 4, 4, 0 },
+    { 0, 1, 1, 0 },
+    { 0, 0, 0, 2 },
+  };
+
+  sources[0] = 0xa5f07e3cll;
+  sources[1] = 0x7e3ca5f0ll;
+  sources[2] = 0x3ca5f07ell;
+  sources[3] = 0x5a0fe7c3ll;
+
+  masks[0] = 0xffff0000ll;
+  masks[1] = 0x0000ffffll;
+  masks[2] = 0x0f0f0f0fll;
+  masks[3] = 0xf0f0f0f0ll;
+
+  for (int i = 0; i < 4; i++)
+    {
+      for (int j = 0; j < 4; j++)
+	{
+	  if (do_cnttzdm (sources[i], masks[j]) != results [i][j])
+	    abort ();
+	}
+    }
+
+  return 0;
+}
+
+/* { dg-final { scan-assembler {\mcnttzdm\M} } } */
diff --git a/gcc/testsuite/gcc.target/powerpc/cnttzdm-1.c b/gcc/testsuite/gcc.target/powerpc/cnttzdm-1.c
new file mode 100644
index 00000000000..d04b1192d15
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/cnttzdm-1.c
@@ -0,0 +1,53 @@ 
+/* { dg-do run } */
+/* { dg-require-effective-target powerpc_future_hw } */
+/* { dg-require-effective-target lp64 } */
+/* { dg-options "-mdejagnu-cpu=future" } */
+
+extern void abort (void);
+
+unsigned long long int
+do_cnttzdm (unsigned long long int source, unsigned long long int mask) {
+  return __builtin_cnttzdm (source, mask);
+}
+
+int main (int argc, char *argv [])
+{
+  unsigned long long int sources [4], masks [4];
+  unsigned long long int intermediates [4][4] = {
+    /* sources[0] with each of masks [0 .. 3] */
+    { 0x0000a5f0ll, 0x00007e3cll, 0x000050ecll, 0x0000af73ll },
+    /* sources[1] with each of masks [0 .. 3] */
+    { 0x00007e3cll, 0x0000a5f0ll, 0x0000ec50ll, 0x000073afll },
+    /* sources[2] with each of masks [0 .. 3] */
+    { 0x00003ca5ll, 0x0000f07ell, 0x0000c50ell, 0x00003af7ll },
+    /* sources[3] with each of masks [0 .. 3] */
+    { 0x00005a0fll, 0x0000e7c3ll, 0x0000af73ll, 0x000050ecll },
+  };
+  unsigned long long int results [4][4] = {
+    { 4, 2, 2, 0 },
+    { 2, 4, 4, 0 },
+    { 0, 1, 1, 0 },
+    { 0, 0, 0, 2 },
+  };
+
+  sources[0] = 0xa5f07e3cll;
+  sources[1] = 0x7e3ca5f0ll;
+  sources[2] = 0x3ca5f07ell;
+  sources[3] = 0x5a0fe7c3ll;
+
+  masks[0] = 0xffff0000ll;
+  masks[1] = 0x0000ffffll;
+  masks[2] = 0x0f0f0f0fll;
+  masks[3] = 0xf0f0f0f0ll;
+
+  for (int i = 0; i < 4; i++)
+    {
+      for (int j = 0; j < 4; j++)
+	{
+	  if (do_cnttzdm (sources[i], masks[j]) != results [i][j])
+	    abort ();
+	}
+    }
+
+  return 0;
+}