[PATCH] RISC-V: Clean up the pr106602.c testcase

[PATCH] RISC-V: Clean up the pr106602.c testcase

[Palmer Dabbelt]

The test case that was added is rv64i-specific, as there's better ways
to generate this code on rv32i (where the long/int cast is a NOP) and on
rv64i_zba (where we have word shifts).  This renames the original test
case and adds two more for those targets.

gcc/testsuite/ChangeLog:
	PR target/106602
	* gcc.target/riscv/pr106602.c: Moved to...
	* gcc.target/riscv/pr106602-rv64i.c: ...here.
	* gcc.target/riscv/pr106602-rv32i.c: New test.
	* gcc.target/riscv/pr106602-rv64i_zba.c: New test.
---
The test suite is still running, but it looks like it's made it past
these so I figured I'd send it now as otherwise I might forget.

OK for trunk?  (assuming the tests finish and pass)
---
 gcc/testsuite/gcc.target/riscv/pr106602-rv32i.c   | 14 ++++++++++++++
 .../riscv/{pr106602.c => pr106602-rv64i.c}        |  2 +-
 .../gcc.target/riscv/pr106602-rv64i_zba.c         | 15 +++++++++++++++
 3 files changed, 30 insertions(+), 1 deletion(-)
 create mode 100644 gcc/testsuite/gcc.target/riscv/pr106602-rv32i.c
 rename gcc/testsuite/gcc.target/riscv/{pr106602.c => pr106602-rv64i.c} (88%)
 create mode 100644 gcc/testsuite/gcc.target/riscv/pr106602-rv64i_zba.c

diff --git a/gcc/testsuite/gcc.target/riscv/pr106602-rv32i.c b/gcc/testsuite/gcc.target/riscv/pr106602-rv32i.c
new file mode 100644
index 00000000000..05b54db7486
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/pr106602-rv32i.c
@@ -0,0 +1,14 @@
+/* { dg-do compile { target { riscv64*-*-* } } } */
+/* { dg-options "-O2 -march=rv32i -mabi=ilp32" } */
+
+unsigned long
+foo2 (unsigned long a)
+{
+  return (unsigned long)(unsigned int) a << 6;
+}
+
+/* { dg-final { scan-assembler-times "slli\t" 1 } } */
+/* { dg-final { scan-assembler-not "srli\t" } } */
+/* { dg-final { scan-assembler-not "\tli\t" } } */
+/* { dg-final { scan-assembler-not "addi\t" } } */
+/* { dg-final { scan-assembler-not "and\t" } } */
diff --git a/gcc/testsuite/gcc.target/riscv/pr106602.c b/gcc/testsuite/gcc.target/riscv/pr106602-rv64i.c
similarity index 88%
rename from gcc/testsuite/gcc.target/riscv/pr106602.c
rename to gcc/testsuite/gcc.target/riscv/pr106602-rv64i.c
index 825b1a143b5..ef0719f4a9a 100644
--- a/gcc/testsuite/gcc.target/riscv/pr106602.c
+++ b/gcc/testsuite/gcc.target/riscv/pr106602-rv64i.c
@@ -1,5 +1,5 @@
 /* { dg-do compile { target { riscv64*-*-* } } } */
-/* { dg-options "-O2" } */
+/* { dg-options "-O2 -march=rv64i -mabi=lp64" } */
 
 unsigned long
 foo2 (unsigned long a)
diff --git a/gcc/testsuite/gcc.target/riscv/pr106602-rv64i_zba.c b/gcc/testsuite/gcc.target/riscv/pr106602-rv64i_zba.c
new file mode 100644
index 00000000000..23b9f1e60f6
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/pr106602-rv64i_zba.c
@@ -0,0 +1,15 @@
+/* { dg-do compile { target { riscv64*-*-* } } } */
+/* { dg-options "-O2 -march=rv64i_zba -mabi=lp64" } */
+
+unsigned long
+foo2 (unsigned long a)
+{
+  return (unsigned long)(unsigned int) a << 6;
+}
+
+/* { dg-final { scan-assembler-times "slli.uw\t" 1 } } */
+/* { dg-final { scan-assembler-not "slli\t" } } */
+/* { dg-final { scan-assembler-not "srli\t" } } */
+/* { dg-final { scan-assembler-not "\tli\t" } } */
+/* { dg-final { scan-assembler-not "addi\t" } } */
+/* { dg-final { scan-assembler-not "and\t" } } */
-- 
2.39.2

Re: [PATCH] RISC-V: Clean up the pr106602.c testcase

[Jeff Law]

On 4/11/23 13:03, Palmer Dabbelt wrote:
> The test case that was added is rv64i-specific, as there's better ways
> to generate this code on rv32i (where the long/int cast is a NOP) and on
> rv64i_zba (where we have word shifts).  This renames the original test
> case and adds two more for those targets.
> 
> gcc/testsuite/ChangeLog:
> 	PR target/106602
> 	* gcc.target/riscv/pr106602.c: Moved to...
> 	* gcc.target/riscv/pr106602-rv64i.c: ...here.
> 	* gcc.target/riscv/pr106602-rv32i.c: New test.
> 	* gcc.target/riscv/pr106602-rv64i_zba.c: New test.
> ---
> The test suite is still running, but it looks like it's made it past
> these so I figured I'd send it now as otherwise I might forget.
> 
> OK for trunk?  (assuming the tests finish and pass)
OK
jeff

Re: [PATCH] RISC-V: Clean up the pr106602.c testcase

[Palmer Dabbelt]

On Tue, 11 Apr 2023 14:01:06 PDT (-0700), gcc-patches@gcc.gnu.org wrote:
>
>
> On 4/11/23 13:03, Palmer Dabbelt wrote:
>> The test case that was added is rv64i-specific, as there's better ways
>> to generate this code on rv32i (where the long/int cast is a NOP) and on
>> rv64i_zba (where we have word shifts).  This renames the original test
>> case and adds two more for those targets.
>>
>> gcc/testsuite/ChangeLog:
>> 	PR target/106602
>> 	* gcc.target/riscv/pr106602.c: Moved to...
>> 	* gcc.target/riscv/pr106602-rv64i.c: ...here.
>> 	* gcc.target/riscv/pr106602-rv32i.c: New test.
>> 	* gcc.target/riscv/pr106602-rv64i_zba.c: New test.
>> ---
>> The test suite is still running, but it looks like it's made it past
>> these so I figured I'd send it now as otherwise I might forget.
>>
>> OK for trunk?  (assuming the tests finish and pass)
> OK

Committed.