public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* [PATCH][AArch64] Fix PR79041
@ 2017-06-27 14:49 Wilco Dijkstra
  2017-06-27 14:55 ` Yvan Roux
  2017-07-14 14:22 ` Wilco Dijkstra
  0 siblings, 2 replies; 11+ messages in thread
From: Wilco Dijkstra @ 2017-06-27 14:49 UTC (permalink / raw)
  To: GCC Patches, James Greenhalgh; +Cc: nd, Yvan Roux

As described in PR79041, -mcmodel=large -mpc-relative-literal-loads
may be used to avoid generating ADRP/ADD or ADRP/LDR.  However both
trunk and GCC7 may still emit ADRP for some constant pool literals.
Fix this by adding a aarch64_pcrelative_literal_loads check.

OK for trunk/GCC7 backport?

ChangeLog:
2017-06-27  Wilco Dijkstra  <wdijkstr@arm.com>

	PR target/79041
	* config/aarch64/aarch64.c (aarch64_classify_symbol):
	Avoid SYMBOL_SMALL_ABSOLUTE .
	* testsuite/gcc.target/aarch64/pr79041-2.c: New test.
--
diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
index 060cd8476d2954119daac495ecb059c9be73edbe..329d244e9cf16dbdf849e5dd02b3999caf0cd5a7 100644
--- a/gcc/config/aarch64/aarch64.c
+++ b/gcc/config/aarch64/aarch64.c
@@ -10042,7 +10042,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
 	  /* This is alright even in PIC code as the constant
 	     pool reference is always PC relative and within
 	     the same translation unit.  */
-	  if (CONSTANT_POOL_ADDRESS_P (x))
+	  if (CONSTANT_POOL_ADDRESS_P (x) && !aarch64_pcrelative_literal_loads)
 	    return SYMBOL_SMALL_ABSOLUTE;
 	  else
 	    return SYMBOL_FORCE_TO_MEM;
diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
new file mode 100644
index 0000000000000000000000000000000000000000..e7899725bad2b770f8488a07f99792113275bdf2
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
@@ -0,0 +1,10 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
+
+__int128
+t (void)
+{
+  return (__int128)1 << 80;
+}
+
+/* { dg-final { scan-assembler "adr" } } */

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH][AArch64] Fix PR79041
  2017-06-27 14:49 [PATCH][AArch64] Fix PR79041 Wilco Dijkstra
@ 2017-06-27 14:55 ` Yvan Roux
  2017-06-27 15:04   ` Yvan Roux
  2017-07-14 14:22 ` Wilco Dijkstra
  1 sibling, 1 reply; 11+ messages in thread
From: Yvan Roux @ 2017-06-27 14:55 UTC (permalink / raw)
  To: Wilco Dijkstra; +Cc: GCC Patches, James Greenhalgh, nd

Hi

On 27 June 2017 at 16:49, Wilco Dijkstra <Wilco.Dijkstra@arm.com> wrote:
> As described in PR79041, -mcmodel=large -mpc-relative-literal-loads
> may be used to avoid generating ADRP/ADD or ADRP/LDR.  However both
> trunk and GCC7 may still emit ADRP for some constant pool literals.
> Fix this by adding a aarch64_pcrelative_literal_loads check.
>
> OK for trunk/GCC7 backport?

I can't approve it, but the patch is ok for me, I've built and
regtested the very same patch for aarch64-linux-gnu,
aarch64-none-elf and aarch64_be-none-elf targets :)

Yvan

> ChangeLog:
> 2017-06-27  Wilco Dijkstra  <wdijkstr@arm.com>
>
>         PR target/79041
>         * config/aarch64/aarch64.c (aarch64_classify_symbol):
>         Avoid SYMBOL_SMALL_ABSOLUTE .
>         * testsuite/gcc.target/aarch64/pr79041-2.c: New test.
> --
> diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
> index 060cd8476d2954119daac495ecb059c9be73edbe..329d244e9cf16dbdf849e5dd02b3999caf0cd5a7 100644
> --- a/gcc/config/aarch64/aarch64.c
> +++ b/gcc/config/aarch64/aarch64.c
> @@ -10042,7 +10042,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
>           /* This is alright even in PIC code as the constant
>              pool reference is always PC relative and within
>              the same translation unit.  */
> -         if (CONSTANT_POOL_ADDRESS_P (x))
> +         if (CONSTANT_POOL_ADDRESS_P (x) && !aarch64_pcrelative_literal_loads)
>             return SYMBOL_SMALL_ABSOLUTE;
>           else
>             return SYMBOL_FORCE_TO_MEM;
> diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> new file mode 100644
> index 0000000000000000000000000000000000000000..e7899725bad2b770f8488a07f99792113275bdf2
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> @@ -0,0 +1,10 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
> +
> +__int128
> +t (void)
> +{
> +  return (__int128)1 << 80;
> +}
> +
> +/* { dg-final { scan-assembler "adr" } } */

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH][AArch64] Fix PR79041
  2017-06-27 14:55 ` Yvan Roux
@ 2017-06-27 15:04   ` Yvan Roux
  0 siblings, 0 replies; 11+ messages in thread
From: Yvan Roux @ 2017-06-27 15:04 UTC (permalink / raw)
  To: Wilco Dijkstra; +Cc: GCC Patches, James Greenhalgh, nd

On 27 June 2017 at 16:55, Yvan Roux <yvan.roux@linaro.org> wrote:
> Hi
>
> On 27 June 2017 at 16:49, Wilco Dijkstra <Wilco.Dijkstra@arm.com> wrote:
>> As described in PR79041, -mcmodel=large -mpc-relative-literal-loads
>> may be used to avoid generating ADRP/ADD or ADRP/LDR.  However both
>> trunk and GCC7 may still emit ADRP for some constant pool literals.
>> Fix this by adding a aarch64_pcrelative_literal_loads check.
>>
>> OK for trunk/GCC7 backport?
>
> I can't approve it, but the patch is ok for me, I've built and
> regtested the very same patch for aarch64-linux-gnu,
> aarch64-none-elf and aarch64_be-none-elf targets :)
>
> Yvan
>
>> ChangeLog:
>> 2017-06-27  Wilco Dijkstra  <wdijkstr@arm.com>
>>
>>         PR target/79041
>>         * config/aarch64/aarch64.c (aarch64_classify_symbol):
>>         Avoid SYMBOL_SMALL_ABSOLUTE .
>>         * testsuite/gcc.target/aarch64/pr79041-2.c: New test.
>> --
>> diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
>> index 060cd8476d2954119daac495ecb059c9be73edbe..329d244e9cf16dbdf849e5dd02b3999caf0cd5a7 100644
>> --- a/gcc/config/aarch64/aarch64.c
>> +++ b/gcc/config/aarch64/aarch64.c
>> @@ -10042,7 +10042,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
>>           /* This is alright even in PIC code as the constant
>>              pool reference is always PC relative and within
>>              the same translation unit.  */
>> -         if (CONSTANT_POOL_ADDRESS_P (x))
>> +         if (CONSTANT_POOL_ADDRESS_P (x) && !aarch64_pcrelative_literal_loads)

maybe just a small note here, in my backport patch on gcc-6-branch a
kept the existing order in the condition:

-         if (nopcrelative_literal_loads
+         if (!aarch64_pcrelative_literal_loads
              && CONSTANT_POOL_ADDRESS_P (x))

Can we keep it, or maybe I should wait for your patch to be committed
on trunk, and include its backport in my patch for GCC 6.


>>             return SYMBOL_SMALL_ABSOLUTE;
>>           else
>>             return SYMBOL_FORCE_TO_MEM;
>> diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
>> new file mode 100644
>> index 0000000000000000000000000000000000000000..e7899725bad2b770f8488a07f99792113275bdf2
>> --- /dev/null
>> +++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
>> @@ -0,0 +1,10 @@
>> +/* { dg-do compile } */
>> +/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
>> +
>> +__int128
>> +t (void)
>> +{
>> +  return (__int128)1 << 80;
>> +}
>> +
>> +/* { dg-final { scan-assembler "adr" } } */

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH][AArch64] Fix PR79041
  2017-06-27 14:49 [PATCH][AArch64] Fix PR79041 Wilco Dijkstra
  2017-06-27 14:55 ` Yvan Roux
@ 2017-07-14 14:22 ` Wilco Dijkstra
  2017-07-21 11:22   ` Wilco Dijkstra
  1 sibling, 1 reply; 11+ messages in thread
From: Wilco Dijkstra @ 2017-07-14 14:22 UTC (permalink / raw)
  To: GCC Patches, James Greenhalgh; +Cc: nd, Yvan Roux

ping

    
As described in PR79041, -mcmodel=large -mpc-relative-literal-loads
may be used to avoid generating ADRP/ADD or ADRP/LDR.  However both
trunk and GCC7 may still emit ADRP for some constant pool literals.
Fix this by adding a aarch64_pcrelative_literal_loads check.

OK for trunk/GCC7 backport?

ChangeLog:
2017-06-27  Wilco Dijkstra  <wdijkstr@arm.com>

        PR target/79041
        * config/aarch64/aarch64.c (aarch64_classify_symbol):
        Avoid SYMBOL_SMALL_ABSOLUTE .
        * testsuite/gcc.target/aarch64/pr79041-2.c: New test.
--
diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
index 060cd8476d2954119daac495ecb059c9be73edbe..329d244e9cf16dbdf849e5dd02b3999caf0cd5a7 100644
--- a/gcc/config/aarch64/aarch64.c
+++ b/gcc/config/aarch64/aarch64.c
@@ -10042,7 +10042,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
           /* This is alright even in PIC code as the constant
              pool reference is always PC relative and within
              the same translation unit.  */
-         if (CONSTANT_POOL_ADDRESS_P (x))
+         if (CONSTANT_POOL_ADDRESS_P (x) && !aarch64_pcrelative_literal_loads)
             return SYMBOL_SMALL_ABSOLUTE;
           else
             return SYMBOL_FORCE_TO_MEM;
diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
new file mode 100644
index 0000000000000000000000000000000000000000..e7899725bad2b770f8488a07f99792113275bdf2
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
@@ -0,0 +1,10 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
+
+__int128
+t (void)
+{
+  return (__int128)1 << 80;
+}
+
+/* { dg-final { scan-assembler "adr" } } */
    

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH][AArch64] Fix PR79041
  2017-07-14 14:22 ` Wilco Dijkstra
@ 2017-07-21 11:22   ` Wilco Dijkstra
  2017-07-24 17:06     ` James Greenhalgh
  0 siblings, 1 reply; 11+ messages in thread
From: Wilco Dijkstra @ 2017-07-21 11:22 UTC (permalink / raw)
  To: GCC Patches, James Greenhalgh; +Cc: nd, Yvan Roux


    
ping

    
As described in PR79041, -mcmodel=large -mpc-relative-literal-loads
may be used to avoid generating ADRP/ADD or ADRP/LDR.  However both
trunk and GCC7 may still emit ADRP for some constant pool literals.
Fix this by adding a aarch64_pcrelative_literal_loads check.

OK for trunk/GCC7 backport?

ChangeLog:
2017-06-27  Wilco Dijkstra  <wdijkstr@arm.com>

        PR target/79041
        * config/aarch64/aarch64.c (aarch64_classify_symbol):
        Avoid SYMBOL_SMALL_ABSOLUTE .
        * testsuite/gcc.target/aarch64/pr79041-2.c: New test.
--
diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
index 060cd8476d2954119daac495ecb059c9be73edbe..329d244e9cf16dbdf849e5dd02b3999caf0cd5a7 100644
--- a/gcc/config/aarch64/aarch64.c
+++ b/gcc/config/aarch64/aarch64.c
@@ -10042,7 +10042,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
           /* This is alright even in PIC code as the constant
              pool reference is always PC relative and within
              the same translation unit.  */
-         if (CONSTANT_POOL_ADDRESS_P (x))
+         if (CONSTANT_POOL_ADDRESS_P (x) && !aarch64_pcrelative_literal_loads)
             return SYMBOL_SMALL_ABSOLUTE;
           else
             return SYMBOL_FORCE_TO_MEM;
diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
new file mode 100644
index 0000000000000000000000000000000000000000..e7899725bad2b770f8488a07f99792113275bdf2
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
@@ -0,0 +1,10 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
+
+__int128
+t (void)
+{
+  return (__int128)1 << 80;
+}
+
+/* { dg-final { scan-assembler "adr" } } */
        

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [PATCH][AArch64] Fix PR79041
  2017-07-21 11:22   ` Wilco Dijkstra
@ 2017-07-24 17:06     ` James Greenhalgh
  2017-07-25 12:13       ` [COMMITED][AArch64] " Wilco Dijkstra
  0 siblings, 1 reply; 11+ messages in thread
From: James Greenhalgh @ 2017-07-24 17:06 UTC (permalink / raw)
  To: Wilco Dijkstra; +Cc: GCC Patches, nd, Yvan Roux

On Fri, Jul 21, 2017 at 12:22:00PM +0100, Wilco Dijkstra wrote:
> 
>     
> ping
> 
>     
> As described in PR79041, -mcmodel=large -mpc-relative-literal-loads
> may be used to avoid generating ADRP/ADD or ADRP/LDR.  However both
> trunk and GCC7 may still emit ADRP for some constant pool literals.
> Fix this by adding a aarch64_pcrelative_literal_loads check.
> 
> OK for trunk/GCC7 backport?

OK. Either like this, or with the conditions swapped around as Yvan
suggested to make backporting easier.

Thanks,
James

> 
> ChangeLog:
> 2017-06-27  Wilco Dijkstra  <wdijkstr@arm.com>
> 
>         PR target/79041
>         * config/aarch64/aarch64.c (aarch64_classify_symbol):
>         Avoid SYMBOL_SMALL_ABSOLUTE .
>         * testsuite/gcc.target/aarch64/pr79041-2.c: New test.
> --
> diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
> index 060cd8476d2954119daac495ecb059c9be73edbe..329d244e9cf16dbdf849e5dd02b3999caf0cd5a7 100644
> --- a/gcc/config/aarch64/aarch64.c
> +++ b/gcc/config/aarch64/aarch64.c
> @@ -10042,7 +10042,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
>            /* This is alright even in PIC code as the constant
>               pool reference is always PC relative and within
>               the same translation unit.  */
> -         if (CONSTANT_POOL_ADDRESS_P (x))
> +         if (CONSTANT_POOL_ADDRESS_P (x) && !aarch64_pcrelative_literal_loads)
>              return SYMBOL_SMALL_ABSOLUTE;
>            else
>              return SYMBOL_FORCE_TO_MEM;
> diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> new file mode 100644
> index 0000000000000000000000000000000000000000..e7899725bad2b770f8488a07f99792113275bdf2
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> @@ -0,0 +1,10 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
> +
> +__int128
> +t (void)
> +{
> +  return (__int128)1 << 80;
> +}
> +
> +/* { dg-final { scan-assembler "adr" } } */
>         

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [COMMITED][AArch64] Fix PR79041
  2017-07-24 17:06     ` James Greenhalgh
@ 2017-07-25 12:13       ` Wilco Dijkstra
  2017-07-25 12:52         ` Ramana Radhakrishnan
  2017-07-25 14:15         ` Andreas Schwab
  0 siblings, 2 replies; 11+ messages in thread
From: Wilco Dijkstra @ 2017-07-25 12:13 UTC (permalink / raw)
  To: James Greenhalgh; +Cc: GCC Patches, nd, Yvan Roux

James Greenhalgh wrote:
>
> OK. Either like this, or with the conditions swapped around as Yvan
> suggested to make backporting easier.

I swapped the conditions around, not sure whether it helps...
Also I needed an additional scan-assembler, this was committed to
trunk and GCC7:

diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
index d7fab3775c06477365217e357b5754026d09752b..7713d543af788d47dcdf8ea89482a2d4f66a1344 100644
--- a/gcc/config/aarch64/aarch64.c
+++ b/gcc/config/aarch64/aarch64.c
@@ -10181,7 +10181,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
 	  /* This is alright even in PIC code as the constant
 	     pool reference is always PC relative and within
 	     the same translation unit.  */
-	  if (CONSTANT_POOL_ADDRESS_P (x))
+	  if (!aarch64_pcrelative_literal_loads && CONSTANT_POOL_ADDRESS_P (x))
 	    return SYMBOL_SMALL_ABSOLUTE;
 	  else
 	    return SYMBOL_FORCE_TO_MEM;
diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
new file mode 100644
index 0000000000000000000000000000000000000000..cd34fbab85a92d00cba7091d4146deaaf3a862a9
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
@@ -0,0 +1,11 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
+
+__int128
+t (void)
+{
+  return (__int128)1 << 80;
+}
+
+/* { dg-final { scan-assembler "adr" } } */
+/* { dg-final { scan-assembler-not "adrp" } } */

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [COMMITED][AArch64] Fix PR79041
  2017-07-25 12:13       ` [COMMITED][AArch64] " Wilco Dijkstra
@ 2017-07-25 12:52         ` Ramana Radhakrishnan
  2017-07-25 13:14           ` Wilco Dijkstra
  2017-07-25 14:15         ` Andreas Schwab
  1 sibling, 1 reply; 11+ messages in thread
From: Ramana Radhakrishnan @ 2017-07-25 12:52 UTC (permalink / raw)
  To: Wilco Dijkstra; +Cc: James Greenhalgh, GCC Patches, nd, Yvan Roux

On Tue, Jul 25, 2017 at 1:13 PM, Wilco Dijkstra <Wilco.Dijkstra@arm.com> wrote:
> James Greenhalgh wrote:
>>
>> OK. Either like this, or with the conditions swapped around as Yvan
>> suggested to make backporting easier.
>
> I swapped the conditions around, not sure whether it helps...
> Also I needed an additional scan-assembler, this was committed to
> trunk and GCC7:


BZ suggests that this affects GCC 6 but GCC 7 is fixed ? Should there
be a backport to GCC 6 as well ?

Can you please keep BZ up to date please ?


Ramana
>
> diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
> index d7fab3775c06477365217e357b5754026d09752b..7713d543af788d47dcdf8ea89482a2d4f66a1344 100644
> --- a/gcc/config/aarch64/aarch64.c
> +++ b/gcc/config/aarch64/aarch64.c
> @@ -10181,7 +10181,7 @@ aarch64_classify_symbol (rtx x, rtx offset)
>           /* This is alright even in PIC code as the constant
>              pool reference is always PC relative and within
>              the same translation unit.  */
> -         if (CONSTANT_POOL_ADDRESS_P (x))
> +         if (!aarch64_pcrelative_literal_loads && CONSTANT_POOL_ADDRESS_P (x))
>             return SYMBOL_SMALL_ABSOLUTE;
>           else
>             return SYMBOL_FORCE_TO_MEM;
> diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> new file mode 100644
> index 0000000000000000000000000000000000000000..cd34fbab85a92d00cba7091d4146deaaf3a862a9
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> @@ -0,0 +1,11 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
> +
> +__int128
> +t (void)
> +{
> +  return (__int128)1 << 80;
> +}
> +
> +/* { dg-final { scan-assembler "adr" } } */
> +/* { dg-final { scan-assembler-not "adrp" } } */

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [COMMITED][AArch64] Fix PR79041
  2017-07-25 12:52         ` Ramana Radhakrishnan
@ 2017-07-25 13:14           ` Wilco Dijkstra
  0 siblings, 0 replies; 11+ messages in thread
From: Wilco Dijkstra @ 2017-07-25 13:14 UTC (permalink / raw)
  To: Ramana Radhakrishnan; +Cc: James Greenhalgh, GCC Patches, nd, Yvan Roux

Ramana Radhakrishnan <ramana.gcc@googlemail.com>
>
> BZ suggests that this affects GCC 6 but GCC 7 is fixed ? Should there
> be a backport to GCC 6 as well ?
>
> Can you please keep BZ up to date please ?

GCC7 was affected as well, the example in PR79041 didn't trigger in GCC7,
my patch has a better example that does show the issue.

GCC6 requires a backport of multiple patches. The patch is 
https://gcc.gnu.org/ml/gcc-patches/2017-06/msg01708.html
which is still waiting for review.

Wilco

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [COMMITED][AArch64] Fix PR79041
  2017-07-25 12:13       ` [COMMITED][AArch64] " Wilco Dijkstra
  2017-07-25 12:52         ` Ramana Radhakrishnan
@ 2017-07-25 14:15         ` Andreas Schwab
  2017-07-26 12:03           ` Wilco Dijkstra
  1 sibling, 1 reply; 11+ messages in thread
From: Andreas Schwab @ 2017-07-25 14:15 UTC (permalink / raw)
  To: Wilco Dijkstra; +Cc: James Greenhalgh, GCC Patches, nd, Yvan Roux

On Jul 25 2017, Wilco Dijkstra <Wilco.Dijkstra@arm.com> wrote:

> diff --git a/gcc/testsuite/gcc.target/aarch64/pr79041-2.c b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> new file mode 100644
> index 0000000000000000000000000000000000000000..cd34fbab85a92d00cba7091d4146deaaf3a862a9
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/aarch64/pr79041-2.c
> @@ -0,0 +1,11 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -mcmodel=large -mpc-relative-literal-loads" } */
> +
> +__int128
> +t (void)
> +{
> +  return (__int128)1 << 80;
> +}
> +
> +/* { dg-final { scan-assembler "adr" } } */
> +/* { dg-final { scan-assembler-not "adrp" } } */

That fails in ILP32 mode.

Andreas.

-- 
Andreas Schwab, SUSE Labs, schwab@suse.de
GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE  1748 E4D4 88E3 0EEA B9D7
"And now for something completely different."

^ permalink raw reply	[flat|nested] 11+ messages in thread

* Re: [COMMITED][AArch64] Fix PR79041
  2017-07-25 14:15         ` Andreas Schwab
@ 2017-07-26 12:03           ` Wilco Dijkstra
  0 siblings, 0 replies; 11+ messages in thread
From: Wilco Dijkstra @ 2017-07-26 12:03 UTC (permalink / raw)
  To: Andreas Schwab; +Cc: James Greenhalgh, GCC Patches, nd, Yvan Roux

Andreas Schwab wrote:

> That fails in ILP32 mode.

Well -mabi-ilp32 and -mcmodel=large make no sense at all,
that should really give an error... I've committed a patch to 
trunk and GCC7 to disable this test with ILP32.

Wilco

^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2017-07-26 12:03 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-06-27 14:49 [PATCH][AArch64] Fix PR79041 Wilco Dijkstra
2017-06-27 14:55 ` Yvan Roux
2017-06-27 15:04   ` Yvan Roux
2017-07-14 14:22 ` Wilco Dijkstra
2017-07-21 11:22   ` Wilco Dijkstra
2017-07-24 17:06     ` James Greenhalgh
2017-07-25 12:13       ` [COMMITED][AArch64] " Wilco Dijkstra
2017-07-25 12:52         ` Ramana Radhakrishnan
2017-07-25 13:14           ` Wilco Dijkstra
2017-07-25 14:15         ` Andreas Schwab
2017-07-26 12:03           ` Wilco Dijkstra

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).