[PATCH] testsuite: Fix up lra effective target

[PATCH] testsuite: Fix up lra effective target

[Jakub Jelinek]

Hi!

Given the recent discussions on IRC started with Andrew P. mentioning that
an asm goto outputs test should have { target lra } and the lra effective
target in GCC 11/12 only returning 0 for PA and in 13/14 for PA/AVR, while
we clearly have 14 other targets which don't support LRA and a couple of
further ones which have an -mlra/-mno-lra switch (whatever default they
have), seems to me the effective target is quite broken.

The following patch rewrites it, such that it has a fast path for heavily
used targets which are for years known to use only LRA (just an
optimization) plus determines whether it is a LRA target or reload target
by scanning the -fdump-rtl-reload-details dump on an empty function,
LRA has quite a few always emitted messages in that case while reload has
none of those.

Tested on x86_64-linux and cross to s390x-linux, for the latter with both
make check-gcc RUNTESTFLAGS='--target_board=unix/-mno-lra dg.exp=pr107385.c'
where the test is now UNSUPPORTED and
make check-gcc RUNTESTFLAGS='--target_board=unix/-mlra dg.exp=pr107385.c'
where it fails because I don't have libc around.

Ok for trunk?

There is one special case, NVPTX, which is a TARGET_NO_REGISTER_ALLOCATION
target.  I think claiming for it that it is a lra target is strange (even
though it effectively returns true for targetm.lra_p ()), unsure if it
supports asm goto with outputs or not, if it does and we want to test it,
perhaps we should introduce asm_goto_outputs effective target and use
lra || nvptx-*-* for that?

2024-02-16  Jakub Jelinek  <jakub@redhat.com>

	* lib/target-supports.exp (check_effective_target_lra): Rewrite
	to list some heavily used always LRA targets and otherwise check the
	-fdump-rtl-reload-details dump for messages specific to LRA.

--- gcc/testsuite/lib/target-supports.exp.jj	2024-02-15 09:51:34.591064180 +0100
+++ gcc/testsuite/lib/target-supports.exp	2024-02-16 10:50:29.986180603 +0100
@@ -13215,10 +13215,17 @@ proc check_effective_target_powerpc_as_p
 # return 1 if LRA is supported.
 
 proc check_effective_target_lra { } {
-    if { [istarget hppa*-*-*] || [istarget avr-*-*] } {
-	return 0
+    # Start with heavily used targets which are known to always use LRA.
+    if { [istarget i?86-*-*] || [istarget x86_64-*-*]
+	 || [istarget aarch64*-*-*] || [istarget arm*-*-*]
+	 || [istarget powerpc*-*-*] || [istarget riscv*-*-*] } {
+	return 1
     }
-    return 1
+
+    # Otherwise check the reload dump for messages emitted solely by LRA.
+    return [check_no_messages_and_pattern lra "\\\*{9} Local #1: \\\*{9}" rtl-reload {
+        void foo (void) {}
+    } {-O2 -fdump-rtl-reload-details}] ;# LRA notes requires a detailed dump.
 }
 
 # Test whether optimizations are enabled ('__OPTIMIZE__') per the

	Jakub

Re: [PATCH] testsuite: Fix up lra effective target

[Mike Stump]

On Feb 16, 2024, at 2:16 AM, Jakub Jelinek <jakub@redhat.com> wrote:
> 
> Given the recent discussions on IRC started with Andrew P. mentioning that
> an asm goto outputs test should have { target lra } and the lra effective
> target in GCC 11/12 only returning 0 for PA and in 13/14 for PA/AVR, while
> we clearly have 14 other targets which don't support LRA and a couple of
> further ones which have an -mlra/-mno-lra switch (whatever default they
> have), seems to me the effective target is quite broken.
> 
> Ok for trunk?

Ok.

Re: [PATCH] testsuite: Fix up lra effective target

[Mike Stump]

On Feb 16, 2024, at 2:16 AM, Jakub Jelinek <jakub@redhat.com> wrote:
> 
> There is one special case, NVPTX, which is a TARGET_NO_REGISTER_ALLOCATION
> target.  I think claiming for it that it is a lra target is strange (even
> though it effectively returns true for targetm.lra_p ()), unsure if it
> supports asm goto with outputs or not, if it does and we want to test it,
> perhaps we should introduce asm_goto_outputs effective target and use
> lra || nvptx-*-* for that?

Since the port people have to maintain that code in general, I usually leave it to them to try and select a cheap, maintainable way to manage it.

If people want to pave the way, I'd tend to defer to them, having thought about more than I.

Re: [PATCH] testsuite: Fix up lra effective target

[Hans-Peter Nilsson]

> Date: Fri, 16 Feb 2024 11:16:22 +0100
> From: Jakub Jelinek <jakub@redhat.com>

> Given the recent discussions on IRC started with Andrew P. mentioning that
> an asm goto outputs test should have { target lra } and the lra effective
> target in GCC 11/12 only returning 0 for PA and in 13/14 for PA/AVR, while
> we clearly have 14 other targets which don't support LRA and a couple of
> further ones which have an -mlra/-mno-lra switch (whatever default they
> have), seems to me the effective target is quite broken.

Definitely, good riddance to that list.

I suggested a little over a year ago to generalize
check_effective_target_lra to get rid of that flawed target
list but was effectively shut down with a review request
that'd *keep* the faulty non-lra target list. :-(
"https://gcc.gnu.org/pipermail/gcc-patches/2023-February/611531.html"

TL;DR: I based LRA-ness on EBB being scanned in LRA but not
for reload (same empty foo), i.e. matching the string "EBB 2
3".  I don't know which method more stable, but that didn't
require -O2 nor -fdump-rtl-reload-details.

Having said that, I'm glad there's now a generic, working
(non-target-list-dependent) effective_target lra.

brgds, H-P

New effective-target 'asm_goto_with_outputs' (was: [PATCH] testsuite: Fix up lra effective target)

[Thomas Schwinge]

[-- Attachment #1: Type: text/plain, Size: 2652 bytes --]

Hi!

On 2024-02-16T10:48:53-0800, Mike Stump <mikestump@comcast.net> wrote:
> On Feb 16, 2024, at 2:16 AM, Jakub Jelinek <jakub@redhat.com> wrote:
>> 
>> There is one special case, NVPTX, which is a TARGET_NO_REGISTER_ALLOCATION
>> target.  I think claiming for it that it is a lra target is strange (even
>> though it effectively returns true for targetm.lra_p ()), unsure if it
>> supports asm goto with outputs or not, if it does and we want to test it,
>> perhaps we should introduce asm_goto_outputs effective target and use
>> lra || nvptx-*-* for that?
>
> Since the port people have to maintain that code in general, I usually leave it to them to try and select a cheap, maintainable way to manage it.
>
> If people want to pave the way, I'd tend to defer to them, having thought about more than I.

Here I am.  ;-)

After commit e16f90be2dc8af6c371fe79044c3e668fa3dda62
"testsuite: Fix up lra effective target", we get for nvptx target:

    -PASS: gcc.c-torture/compile/asmgoto-2.c   -O0  (test for excess errors)
    +ERROR: gcc.c-torture/compile/asmgoto-2.c   -O0 : no files matched glob pattern "lra1020113.c.[0-9][0-9][0-9]r.reload" for " dg-do 2 compile { target lra } "

Etc.

That is, the current effective-target 'lra' is not suitable for nvptx --
which, I suppose, is OK, given that nvptx neither uses LRA nor doesn't
use LRA.  ;-) (Therefore, effective-target 'lra' shouldn't get used in
test cases that are active for nvptx.)

However, nvptx appears to support 'asm goto' with outputs, including the
new execution test case:

    PASS: gcc.dg/pr107385.c execution test

I'm attaching "[WIP] New effective-target 'asm_goto_with_outputs'", which
does address the effective-target check for nvptx, and otherwise does
's%lra%asm_goto_with_outputs'.  (I have not yet actually merged
'check_effective_target_lra' into
'check_effective_target_asm_goto_with_outputs'.)

I have verified that all current effective-target 'lra' test cases
actually use 'asm goto' with outputs, there is just one exception:
'gcc.dg/pr110079.c' (see
<https://inbox.sourceware.org/Zel5TMMr/3BHgl0g@tucnak>
"bb-reorder: Fix -freorder-blocks-and-partition ICEs on aarch64 with asm goto [PR110079]",
<https://gcc.gnu.org/PR110079>
"ICE with -freorder-blocks-and-partition and inline-asm goto").  That
test case, 'gcc.dg/pr110079.c', currently uses 'target lra', and uses
'asm goto' -- but not with outputs, so is 'asm_goto_with_outputs' not
really applicable?  The test case does PASS for nvptx target (but I've
not verified what it's actually doing/testing).  How to handle that one?


Grüße
 Thomas



[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-WIP-New-effective-target-asm_goto_with_outputs.patch --]
[-- Type: text/x-diff, Size: 8592 bytes --]

From d9f8faaa5026bb970b3246235eb22bf9b5e9fe3a Mon Sep 17 00:00:00 2001
From: Thomas Schwinge <tschwinge@baylibre.com>
Date: Mon, 4 Mar 2024 16:04:11 +0100
Subject: [PATCH] [WIP] New effective-target 'asm_goto_with_outputs'

After commit e16f90be2dc8af6c371fe79044c3e668fa3dda62
"testsuite: Fix up lra effective target", we get for nvptx target:

    -PASS: gcc.c-torture/compile/asmgoto-2.c   -O0  (test for excess errors)
    +ERROR: gcc.c-torture/compile/asmgoto-2.c   -O0 : no files matched glob pattern "lra1020113.c.[0-9][0-9][0-9]r.reload" for " dg-do 2 compile { target lra } "

Etc.

However, nvptx appears to support 'asm goto' with outputs, including the
new execution test case:

    PASS: gcc.dg/pr107385.c execution test

TODO gcc/testsuite/gcc.dg/pr110079.c
doesn't using 'asm_goto' with outputs, but is PASS for nvptx, and would ERROR for 'target lra'.
---
 gcc/doc/sourcebuild.texi                        | 3 +++
 gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c | 2 +-
 gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c | 2 +-
 gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c | 3 +--
 gcc/testsuite/gcc.c-torture/compile/pr98096.c   | 2 +-
 gcc/testsuite/gcc.dg/pr100590.c                 | 2 +-
 gcc/testsuite/gcc.dg/pr107385.c                 | 2 +-
 gcc/testsuite/gcc.dg/pr108095.c                 | 2 +-
 gcc/testsuite/gcc.dg/pr110079.c                 | 2 +-
 gcc/testsuite/gcc.dg/pr97954.c                  | 2 +-
 gcc/testsuite/gcc.dg/torture/pr100329.c         | 2 +-
 gcc/testsuite/gcc.dg/torture/pr100398.c         | 2 +-
 gcc/testsuite/gcc.dg/torture/pr100519.c         | 2 +-
 gcc/testsuite/gcc.dg/torture/pr110422.c         | 2 +-
 gcc/testsuite/lib/target-supports.exp           | 9 +++++++++
 15 files changed, 25 insertions(+), 14 deletions(-)

diff --git a/gcc/doc/sourcebuild.texi b/gcc/doc/sourcebuild.texi
index b56b9c39733..a176a3c864f 100644
--- a/gcc/doc/sourcebuild.texi
+++ b/gcc/doc/sourcebuild.texi
@@ -2863,6 +2863,9 @@ Target supports weak undefined symbols
 @item R_flag_in_section
 Target supports the 'R' flag in .section directive in assembly inputs.
 
+@item asm_goto_with_outputs
+Target supports 'asm goto' with outputs.
+
 @item automatic_stack_alignment
 Target supports automatic stack alignment.
 
diff --git a/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c b/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c
index 43e597bc59f..234c90e5295 100644
--- a/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c
+++ b/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c
@@ -1,5 +1,5 @@
 /* This test should be switched off for a new target with less than 4 allocatable registers */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 int
 foo (void)
 {
diff --git a/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c b/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c
index e1574a2903a..af1ba5a7001 100644
--- a/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c
+++ b/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c
@@ -1,5 +1,5 @@
 /* Test to generate output reload in asm goto on x86_64.  */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-skip-if "no O0" { { i?86-*-* x86_64-*-* } && { ! ia32 } } { "-O0" } { "" } } */
 
 #if defined __x86_64__
diff --git a/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c b/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c
index 6799b83c20a..cb3c7d711a6 100644
--- a/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c
+++ b/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c
@@ -1,5 +1,4 @@
-
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* PR middle-end/110420 */
 /* PR middle-end/103979 */
 /* PR middle-end/98619 */
diff --git a/gcc/testsuite/gcc.c-torture/compile/pr98096.c b/gcc/testsuite/gcc.c-torture/compile/pr98096.c
index bba3fa5c619..3a1b6159c2e 100644
--- a/gcc/testsuite/gcc.c-torture/compile/pr98096.c
+++ b/gcc/testsuite/gcc.c-torture/compile/pr98096.c
@@ -1,6 +1,6 @@
 /* Test for correct naming of label operands in asm goto in case of presence of
    input/output operands. */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 int i, j;
 int f(void) {
   asm goto ("# %0 %2" : "+r" (i) ::: jmp);
diff --git a/gcc/testsuite/gcc.dg/pr100590.c b/gcc/testsuite/gcc.dg/pr100590.c
index 8d1e1a0d306..64cb717ed0e 100644
--- a/gcc/testsuite/gcc.dg/pr100590.c
+++ b/gcc/testsuite/gcc.dg/pr100590.c
@@ -1,5 +1,5 @@
 /* PR rtl-optimization/100590 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-options "-O1 -fno-dce -w" } */
 
 int
diff --git a/gcc/testsuite/gcc.dg/pr107385.c b/gcc/testsuite/gcc.dg/pr107385.c
index 0cc0655d848..690ad3c1b5e 100644
--- a/gcc/testsuite/gcc.dg/pr107385.c
+++ b/gcc/testsuite/gcc.dg/pr107385.c
@@ -1,5 +1,5 @@
 /* PR middle-end/107385 */
-/* { dg-do run { target lra } } */
+/* { dg-do run { target asm_goto_with_outputs } } */
 /* { dg-options "-O2" } */
 
 __attribute__((noipa)) int
diff --git a/gcc/testsuite/gcc.dg/pr108095.c b/gcc/testsuite/gcc.dg/pr108095.c
index 0a487cf614a..115ea73e39b 100644
--- a/gcc/testsuite/gcc.dg/pr108095.c
+++ b/gcc/testsuite/gcc.dg/pr108095.c
@@ -1,5 +1,5 @@
 /* PR tree-optimization/108095 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-options "-Os -g" } */
 
 int v;
diff --git a/gcc/testsuite/gcc.dg/pr110079.c b/gcc/testsuite/gcc.dg/pr110079.c
index 1682f9c2344..e8ec6666b23 100644
--- a/gcc/testsuite/gcc.dg/pr110079.c
+++ b/gcc/testsuite/gcc.dg/pr110079.c
@@ -1,5 +1,5 @@
 /* PR rtl-optimization/110079 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-options "-O2" } */
 /* { dg-additional-options "-freorder-blocks-and-partition" { target freorder } } */
 
diff --git a/gcc/testsuite/gcc.dg/pr97954.c b/gcc/testsuite/gcc.dg/pr97954.c
index 0be60f500b6..619bed3b40c 100644
--- a/gcc/testsuite/gcc.dg/pr97954.c
+++ b/gcc/testsuite/gcc.dg/pr97954.c
@@ -1,5 +1,5 @@
 /* PR rtl-optimization/97954 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-options "-O2" } */
 
 int
diff --git a/gcc/testsuite/gcc.dg/torture/pr100329.c b/gcc/testsuite/gcc.dg/torture/pr100329.c
index 2a4331ba712..5759131a6ca 100644
--- a/gcc/testsuite/gcc.dg/torture/pr100329.c
+++ b/gcc/testsuite/gcc.dg/torture/pr100329.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-additional-options "--param tree-reassoc-width=2" } */
 
 unsigned int a0;
diff --git a/gcc/testsuite/gcc.dg/torture/pr100398.c b/gcc/testsuite/gcc.dg/torture/pr100398.c
index 4fc1168d22f..eeeaf98e15b 100644
--- a/gcc/testsuite/gcc.dg/torture/pr100398.c
+++ b/gcc/testsuite/gcc.dg/torture/pr100398.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 
 int
 test5_limit (void)
diff --git a/gcc/testsuite/gcc.dg/torture/pr100519.c b/gcc/testsuite/gcc.dg/torture/pr100519.c
index 89dff668a97..b82ce46c529 100644
--- a/gcc/testsuite/gcc.dg/torture/pr100519.c
+++ b/gcc/testsuite/gcc.dg/torture/pr100519.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-additional-options "--param tree-reassoc-width=2" } */
 
 unsigned int foo_a1, foo_a2;
diff --git a/gcc/testsuite/gcc.dg/torture/pr110422.c b/gcc/testsuite/gcc.dg/torture/pr110422.c
index 2a653bdfce3..34de7a2980f 100644
--- a/gcc/testsuite/gcc.dg/torture/pr110422.c
+++ b/gcc/testsuite/gcc.dg/torture/pr110422.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 
 struct T { int x; };
 int foo(void) {
diff --git a/gcc/testsuite/lib/target-supports.exp b/gcc/testsuite/lib/target-supports.exp
index 5a1f0ed5a28..158dc51a71f 100644
--- a/gcc/testsuite/lib/target-supports.exp
+++ b/gcc/testsuite/lib/target-supports.exp
@@ -13211,6 +13211,15 @@ proc check_effective_target_lra { } {
     } {-O2 -fdump-rtl-reload-details}] ;# LRA notes requires a detailed dump.
 }
 
+# Return 1 if 'asm goto' with outputs is supported, 0 otherwise.
+
+proc check_effective_target_asm_goto_with_outputs { } {
+    if { [istarget nvptx-*-*] } {
+	return 1
+    }
+    return [check_effective_target_lra]
+}
+
 # Test whether optimizations are enabled ('__OPTIMIZE__') per the
 # 'current_compiler_flags' (thus don't cache).
 
-- 
2.34.1

Re: New effective-target 'asm_goto_with_outputs'

[Jeff Law]

On 3/21/24 5:20 AM, Thomas Schwinge wrote:
> Hi!
> 
> On 2024-02-16T10:48:53-0800, Mike Stump <mikestump@comcast.net> wrote:
>> On Feb 16, 2024, at 2:16 AM, Jakub Jelinek <jakub@redhat.com> wrote:
>>>
>>> There is one special case, NVPTX, which is a TARGET_NO_REGISTER_ALLOCATION
>>> target.  I think claiming for it that it is a lra target is strange (even
>>> though it effectively returns true for targetm.lra_p ()), unsure if it
>>> supports asm goto with outputs or not, if it does and we want to test it,
>>> perhaps we should introduce asm_goto_outputs effective target and use
>>> lra || nvptx-*-* for that?
>>
>> Since the port people have to maintain that code in general, I usually leave it to them to try and select a cheap, maintainable way to manage it.
>>
>> If people want to pave the way, I'd tend to defer to them, having thought about more than I.
> 
> Here I am.  ;-)
> 
> After commit e16f90be2dc8af6c371fe79044c3e668fa3dda62
> "testsuite: Fix up lra effective target", we get for nvptx target:
> 
>      -PASS: gcc.c-torture/compile/asmgoto-2.c   -O0  (test for excess errors)
>      +ERROR: gcc.c-torture/compile/asmgoto-2.c   -O0 : no files matched glob pattern "lra1020113.c.[0-9][0-9][0-9]r.reload" for " dg-do 2 compile { target lra } "
> 
> Etc.
> 
> That is, the current effective-target 'lra' is not suitable for nvptx --
> which, I suppose, is OK, given that nvptx neither uses LRA nor doesn't
> use LRA.  ;-) (Therefore, effective-target 'lra' shouldn't get used in
> test cases that are active for nvptx.)
> 
> However, nvptx appears to support 'asm goto' with outputs, including the
> new execution test case:
> 
>      PASS: gcc.dg/pr107385.c execution test
> 
> I'm attaching "[WIP] New effective-target 'asm_goto_with_outputs'", which
> does address the effective-target check for nvptx, and otherwise does
> 's%lra%asm_goto_with_outputs'.  (I have not yet actually merged
> 'check_effective_target_lra' into
> 'check_effective_target_asm_goto_with_outputs'.)
> 
> I have verified that all current effective-target 'lra' test cases
> actually use 'asm goto' with outputs, there is just one exception:
> 'gcc.dg/pr110079.c' (see
> <https://inbox.sourceware.org/Zel5TMMr/3BHgl0g@tucnak>
> "bb-reorder: Fix -freorder-blocks-and-partition ICEs on aarch64 with asm goto [PR110079]",
> <https://gcc.gnu.org/PR110079>
> "ICE with -freorder-blocks-and-partition and inline-asm goto").  That
> test case, 'gcc.dg/pr110079.c', currently uses 'target lra', and uses
> 'asm goto' -- but not with outputs, so is 'asm_goto_with_outputs' not
> really applicable?  The test case does PASS for nvptx target (but I've
> not verified what it's actually doing/testing).  How to handle that one?
I'd just make target_lra return false for nvptx rather than creating a 
new selector -- I'm not aware of any features other than asm goto that 
LRA provides that aren't supported reload.

Or perhaps rename the selector entirely to target_asm_goto?

jeff

Re: New effective-target 'asm_goto_with_outputs'

[Jakub Jelinek]

On Fri, Mar 22, 2024 at 12:17:03PM -0600, Jeff Law wrote:
> I'd just make target_lra return false for nvptx rather than creating a new

The lra effective target currently though doesn't check if asm goto can have
outputs, but rather if the target is using lra.

> selector -- I'm not aware of any features other than asm goto that LRA
> provides that aren't supported reload.
> 
> Or perhaps rename the selector entirely to target_asm_goto?

In that case we should just test if asm goto with outputs is allowed
in a cached snippet, rather than testing if there is LRA in the ra dumps.

	Jakub

Re: New effective-target 'asm_goto_with_outputs'

[Jeff Law]

On 3/22/24 12:24 PM, Jakub Jelinek wrote:
> On Fri, Mar 22, 2024 at 12:17:03PM -0600, Jeff Law wrote:
>> I'd just make target_lra return false for nvptx rather than creating a new
> 
> The lra effective target currently though doesn't check if asm goto can have
> outputs, but rather if the target is using lra.
Right.  It's not 100% precise as we lose one testcase for nvptx.  THat's 
a tradeoff I'd be willing to make.

> 
>> selector -- I'm not aware of any features other than asm goto that LRA
>> provides that aren't supported reload.
>>
>> Or perhaps rename the selector entirely to target_asm_goto?
> 
> In that case we should just test if asm goto with outputs is allowed
> in a cached snippet, rather than testing if there is LRA in the ra dumps.
I won't lose any sleep with that approach, I just don't see that it adds 
a lot of value.

jeff

New effective-target 'asm_goto_with_outputs' (was: [PATCH] testsuite: Fix up lra effective target)

[Thomas Schwinge]

[-- Attachment #1: Type: text/plain, Size: 2935 bytes --]

Hi!

On 2024-03-21T12:20:38+0100, I wrote:
> On 2024-02-16T10:48:53-0800, Mike Stump <mikestump@comcast.net> wrote:
>> On Feb 16, 2024, at 2:16 AM, Jakub Jelinek <jakub@redhat.com> wrote:
>>> 
>>> There is one special case, NVPTX, which is a TARGET_NO_REGISTER_ALLOCATION
>>> target.  I think claiming for it that it is a lra target is strange (even
>>> though it effectively returns true for targetm.lra_p ()), unsure if it
>>> supports asm goto with outputs or not, if it does and we want to test it,
>>> perhaps we should introduce asm_goto_outputs effective target and use
>>> lra || nvptx-*-* for that?
>>
>> Since the port people have to maintain that code in general, I usually leave it to them to try and select a cheap, maintainable way to manage it.
>>
>> If people want to pave the way, I'd tend to defer to them, having thought about more than I.
>
> Here I am.  ;-)
>
> After commit e16f90be2dc8af6c371fe79044c3e668fa3dda62
> "testsuite: Fix up lra effective target", we get for nvptx target:
>
>     -PASS: gcc.c-torture/compile/asmgoto-2.c   -O0  (test for excess errors)
>     +ERROR: gcc.c-torture/compile/asmgoto-2.c   -O0 : no files matched glob pattern "lra1020113.c.[0-9][0-9][0-9]r.reload" for " dg-do 2 compile { target lra } "
>
> Etc.
>
> That is, the current effective-target 'lra' is not suitable for nvptx --
> which, I suppose, is OK, given that nvptx neither uses LRA nor doesn't
> use LRA.  ;-) (Therefore, effective-target 'lra' shouldn't get used in
> test cases that are active for nvptx.)
>
> However, nvptx appears to support 'asm goto' with outputs, including the
> new execution test case:
>
>     PASS: gcc.dg/pr107385.c execution test
>
> I'm attaching "[WIP] New effective-target 'asm_goto_with_outputs'", which
> does address the effective-target check for nvptx, and otherwise does
> 's%lra%asm_goto_with_outputs'.  (I have not yet actually merged
> 'check_effective_target_lra' into
> 'check_effective_target_asm_goto_with_outputs'.)
>
> I have verified that all current effective-target 'lra' test cases
> actually use 'asm goto' with outputs, there is just one exception:
> 'gcc.dg/pr110079.c' (see
> <https://inbox.sourceware.org/Zel5TMMr/3BHgl0g@tucnak>
> "bb-reorder: Fix -freorder-blocks-and-partition ICEs on aarch64 with asm goto [PR110079]",
> <https://gcc.gnu.org/PR110079>
> "ICE with -freorder-blocks-and-partition and inline-asm goto").  That
> test case, 'gcc.dg/pr110079.c', currently uses 'target lra', and uses
> 'asm goto' -- but not with outputs, so is 'asm_goto_with_outputs' not
> really applicable?  The test case does PASS for nvptx target (but I've
> not verified what it's actually doing/testing).  How to handle that one?

I've now pushed a v2 version to trunk branch in
commit 3fa8bff30ab58bd8b8018764d390ec2fcc8153bb
"New effective-target 'asm_goto_with_outputs'", see attached.


Grüße
 Thomas



[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-New-effective-target-asm_goto_with_outputs.patch --]
[-- Type: text/x-diff, Size: 10332 bytes --]

From 3fa8bff30ab58bd8b8018764d390ec2fcc8153bb Mon Sep 17 00:00:00 2001
From: Thomas Schwinge <tschwinge@baylibre.com>
Date: Mon, 4 Mar 2024 16:04:11 +0100
Subject: [PATCH] New effective-target 'asm_goto_with_outputs'

After commit e16f90be2dc8af6c371fe79044c3e668fa3dda62
"testsuite: Fix up lra effective target", we get for nvptx target:

    -PASS: gcc.c-torture/compile/asmgoto-2.c   -O0  (test for excess errors)
    +ERROR: gcc.c-torture/compile/asmgoto-2.c   -O0 : no files matched glob pattern "lra1020113.c.[0-9][0-9][0-9]r.reload" for " dg-do 2 compile { target lra } "

Etc.

However, nvptx appears to support 'asm goto' with outputs, including the
new execution test case:

    PASS: gcc.dg/pr107385.c execution test

Therefore, generally use new effective-target 'asm_goto_with_outputs' instead
of 'lra'.  One exceptions is 'gcc.dg/pr110079.c', which doesn't use 'asm goto'
with outputs, and continues using effective-target 'lra', with special-casing
nvptx target, to avoid ERROR for 'lra'.

	gcc/
	* doc/sourcebuild.texi (Effective-Target Keywords): Document
	'asm_goto_with_outputs'.  Add comment to 'lra'.
	gcc/testsuite/
	* lib/target-supports.exp (check_effective_target_lra): Add
	comment.
	(check_effective_target_asm_goto_with_outputs): New.
	* gcc.c-torture/compile/asmgoto-2.c: Use it.
	* gcc.c-torture/compile/asmgoto-5.c: Likewise.
	* gcc.c-torture/compile/asmgoto-6.c: Likewise.
	* gcc.c-torture/compile/pr98096.c: Likewise.
	* gcc.dg/pr100590.c: Likewise.
	* gcc.dg/pr107385.c: Likewise.
	* gcc.dg/pr108095.c: Likewise.
	* gcc.dg/pr97954.c: Likewise.
	* gcc.dg/torture/pr100329.c: Likewise.
	* gcc.dg/torture/pr100398.c: Likewise.
	* gcc.dg/torture/pr100519.c: Likewise.
	* gcc.dg/torture/pr110422.c: Likewise.
	* gcc.dg/pr110079.c: Special-case nvptx target.
---
 gcc/doc/sourcebuild.texi                        |  6 ++++++
 gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c |  2 +-
 gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c |  2 +-
 gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c |  3 +--
 gcc/testsuite/gcc.c-torture/compile/pr98096.c   |  2 +-
 gcc/testsuite/gcc.dg/pr100590.c                 |  2 +-
 gcc/testsuite/gcc.dg/pr107385.c                 |  2 +-
 gcc/testsuite/gcc.dg/pr108095.c                 |  2 +-
 gcc/testsuite/gcc.dg/pr110079.c                 |  2 +-
 gcc/testsuite/gcc.dg/pr97954.c                  |  2 +-
 gcc/testsuite/gcc.dg/torture/pr100329.c         |  2 +-
 gcc/testsuite/gcc.dg/torture/pr100398.c         |  2 +-
 gcc/testsuite/gcc.dg/torture/pr100519.c         |  2 +-
 gcc/testsuite/gcc.dg/torture/pr110422.c         |  2 +-
 gcc/testsuite/lib/target-supports.exp           | 13 ++++++++++++-
 15 files changed, 31 insertions(+), 15 deletions(-)

diff --git a/gcc/doc/sourcebuild.texi b/gcc/doc/sourcebuild.texi
index 7ef82fc9b00..7c0df90e822 100644
--- a/gcc/doc/sourcebuild.texi
+++ b/gcc/doc/sourcebuild.texi
@@ -2871,6 +2871,9 @@ Target supports weak undefined symbols
 @item R_flag_in_section
 Target supports the 'R' flag in .section directive in assembly inputs.
 
+@item asm_goto_with_outputs
+Target supports 'asm goto' with outputs.
+
 @item automatic_stack_alignment
 Target supports automatic stack alignment.
 
@@ -2945,6 +2948,9 @@ Target is using an LLVM assembler and/or linker, instead of GNU Binutils.
 
 @item lra
 Target supports local register allocator (LRA).
+This must not be called (results in @code{ERROR}) for targets that
+don't do register allocation, and therefore neither use nor don't use
+LRA.
 
 @item lto
 Compiler has been configured to support link-time optimization (LTO).
diff --git a/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c b/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c
index 43e597bc59f..234c90e5295 100644
--- a/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c
+++ b/gcc/testsuite/gcc.c-torture/compile/asmgoto-2.c
@@ -1,5 +1,5 @@
 /* This test should be switched off for a new target with less than 4 allocatable registers */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 int
 foo (void)
 {
diff --git a/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c b/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c
index e1574a2903a..af1ba5a7001 100644
--- a/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c
+++ b/gcc/testsuite/gcc.c-torture/compile/asmgoto-5.c
@@ -1,5 +1,5 @@
 /* Test to generate output reload in asm goto on x86_64.  */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-skip-if "no O0" { { i?86-*-* x86_64-*-* } && { ! ia32 } } { "-O0" } { "" } } */
 
 #if defined __x86_64__
diff --git a/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c b/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c
index 6799b83c20a..cb3c7d711a6 100644
--- a/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c
+++ b/gcc/testsuite/gcc.c-torture/compile/asmgoto-6.c
@@ -1,5 +1,4 @@
-
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* PR middle-end/110420 */
 /* PR middle-end/103979 */
 /* PR middle-end/98619 */
diff --git a/gcc/testsuite/gcc.c-torture/compile/pr98096.c b/gcc/testsuite/gcc.c-torture/compile/pr98096.c
index bba3fa5c619..3a1b6159c2e 100644
--- a/gcc/testsuite/gcc.c-torture/compile/pr98096.c
+++ b/gcc/testsuite/gcc.c-torture/compile/pr98096.c
@@ -1,6 +1,6 @@
 /* Test for correct naming of label operands in asm goto in case of presence of
    input/output operands. */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 int i, j;
 int f(void) {
   asm goto ("# %0 %2" : "+r" (i) ::: jmp);
diff --git a/gcc/testsuite/gcc.dg/pr100590.c b/gcc/testsuite/gcc.dg/pr100590.c
index 8d1e1a0d306..64cb717ed0e 100644
--- a/gcc/testsuite/gcc.dg/pr100590.c
+++ b/gcc/testsuite/gcc.dg/pr100590.c
@@ -1,5 +1,5 @@
 /* PR rtl-optimization/100590 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-options "-O1 -fno-dce -w" } */
 
 int
diff --git a/gcc/testsuite/gcc.dg/pr107385.c b/gcc/testsuite/gcc.dg/pr107385.c
index 0cc0655d848..690ad3c1b5e 100644
--- a/gcc/testsuite/gcc.dg/pr107385.c
+++ b/gcc/testsuite/gcc.dg/pr107385.c
@@ -1,5 +1,5 @@
 /* PR middle-end/107385 */
-/* { dg-do run { target lra } } */
+/* { dg-do run { target asm_goto_with_outputs } } */
 /* { dg-options "-O2" } */
 
 __attribute__((noipa)) int
diff --git a/gcc/testsuite/gcc.dg/pr108095.c b/gcc/testsuite/gcc.dg/pr108095.c
index 0a487cf614a..115ea73e39b 100644
--- a/gcc/testsuite/gcc.dg/pr108095.c
+++ b/gcc/testsuite/gcc.dg/pr108095.c
@@ -1,5 +1,5 @@
 /* PR tree-optimization/108095 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-options "-Os -g" } */
 
 int v;
diff --git a/gcc/testsuite/gcc.dg/pr110079.c b/gcc/testsuite/gcc.dg/pr110079.c
index 1682f9c2344..f87064d5f2c 100644
--- a/gcc/testsuite/gcc.dg/pr110079.c
+++ b/gcc/testsuite/gcc.dg/pr110079.c
@@ -1,5 +1,5 @@
 /* PR rtl-optimization/110079 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target { nvptx-*-* || lra } } } */
 /* { dg-options "-O2" } */
 /* { dg-additional-options "-freorder-blocks-and-partition" { target freorder } } */
 
diff --git a/gcc/testsuite/gcc.dg/pr97954.c b/gcc/testsuite/gcc.dg/pr97954.c
index 0be60f500b6..619bed3b40c 100644
--- a/gcc/testsuite/gcc.dg/pr97954.c
+++ b/gcc/testsuite/gcc.dg/pr97954.c
@@ -1,5 +1,5 @@
 /* PR rtl-optimization/97954 */
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-options "-O2" } */
 
 int
diff --git a/gcc/testsuite/gcc.dg/torture/pr100329.c b/gcc/testsuite/gcc.dg/torture/pr100329.c
index 2a4331ba712..5759131a6ca 100644
--- a/gcc/testsuite/gcc.dg/torture/pr100329.c
+++ b/gcc/testsuite/gcc.dg/torture/pr100329.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-additional-options "--param tree-reassoc-width=2" } */
 
 unsigned int a0;
diff --git a/gcc/testsuite/gcc.dg/torture/pr100398.c b/gcc/testsuite/gcc.dg/torture/pr100398.c
index 4fc1168d22f..eeeaf98e15b 100644
--- a/gcc/testsuite/gcc.dg/torture/pr100398.c
+++ b/gcc/testsuite/gcc.dg/torture/pr100398.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 
 int
 test5_limit (void)
diff --git a/gcc/testsuite/gcc.dg/torture/pr100519.c b/gcc/testsuite/gcc.dg/torture/pr100519.c
index 89dff668a97..b82ce46c529 100644
--- a/gcc/testsuite/gcc.dg/torture/pr100519.c
+++ b/gcc/testsuite/gcc.dg/torture/pr100519.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 /* { dg-additional-options "--param tree-reassoc-width=2" } */
 
 unsigned int foo_a1, foo_a2;
diff --git a/gcc/testsuite/gcc.dg/torture/pr110422.c b/gcc/testsuite/gcc.dg/torture/pr110422.c
index 2a653bdfce3..34de7a2980f 100644
--- a/gcc/testsuite/gcc.dg/torture/pr110422.c
+++ b/gcc/testsuite/gcc.dg/torture/pr110422.c
@@ -1,4 +1,4 @@
-/* { dg-do compile { target lra } } */
+/* { dg-do compile { target asm_goto_with_outputs } } */
 
 struct T { int x; };
 int foo(void) {
diff --git a/gcc/testsuite/lib/target-supports.exp b/gcc/testsuite/lib/target-supports.exp
index 45435586de2..49f2482686a 100644
--- a/gcc/testsuite/lib/target-supports.exp
+++ b/gcc/testsuite/lib/target-supports.exp
@@ -13204,7 +13204,9 @@ proc check_effective_target_powerpc_as_p10_htm { } {
     }]
 }
 
-# return 1 if LRA is supported.
+# Return 1 if LRA is supported.  This must not be called (results in ERROR) for
+# targets that don't do register allocation, and therefore neither use nor
+# don't use LRA.
 
 proc check_effective_target_lra { } {
     # Start with heavily used targets which are known to always use LRA.
@@ -13220,6 +13222,15 @@ proc check_effective_target_lra { } {
     } {-O2 -fdump-rtl-reload-details}] ;# LRA notes requires a detailed dump.
 }
 
+# Return 1 if 'asm goto' with outputs is supported, 0 otherwise.
+
+proc check_effective_target_asm_goto_with_outputs { } {
+    if { [istarget nvptx-*-*] } {
+	return 1
+    }
+    return [check_effective_target_lra]
+}
+
 # Test whether optimizations are enabled ('__OPTIMIZE__') per the
 # 'current_compiler_flags' (thus don't cache).
 
-- 
2.34.1