[v2] selftests: riscv: fix v_exec_initval_nolibc.c

[PATCH v2] selftests: riscv: fix v_exec_initval_nolibc.c

Posted by Ignacio Encinas 11 months, 1 week ago

Vector registers are zero initialized by the kernel. Stop accepting
"all ones" as a clean value.

Note that this was not working as expected given that
	value == 0xff
can be assumed to be always false by the compiler as value's range is
[-128, 127]. Both GCC (-Wtype-limits) and clang
(-Wtautological-constant-out-of-range-compare) warn about this.

Reviewed-by: Charlie Jenkins <charlie@rivosinc.com>
Tested-by: Charlie Jenkins <charlie@rivosinc.com>
Signed-off-by: Ignacio Encinas <ignacio@iencinas.com>
---
Changes in v2:

Remove code that becomes useless now that the only "clean" value for
vector registers is 0.

- Link to v1: https://lore.kernel.org/r/20250305-fix-v_exec_initval_nolibc-v1-1-b87b60e43002@iencinas.com
---
 tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c b/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
index 35c0812e32de0c82a54f84bd52c4272507121e35..4dde05e45a04122b566cedc36d20b072413b00e2 100644
--- a/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
+++ b/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
@@ -6,7 +6,7 @@
  * the values. To further ensure consistency, this file is compiled without
  * libc and without auto-vectorization.
  *
- * To be "clean" all values must be either all ones or all zeroes.
+ * To be "clean" all values must be all zeroes.
  */
 
 #define __stringify_1(x...)	#x
@@ -14,9 +14,8 @@
 
 int main(int argc, char **argv)
 {
-	char prev_value = 0, value;
+	char value = 0;
 	unsigned long vl;
-	int first = 1;
 
 	if (argc > 2 && strcmp(argv[2], "x"))
 		asm volatile (
@@ -44,14 +43,11 @@ int main(int argc, char **argv)
 			"vsrl.vi " __stringify(register) ", " __stringify(register) ", 8\n\t" \
 			".option pop\n\t"					\
 			: "=r" (value));					\
-		if (first) {							\
-			first = 0;						\
-		} else if (value != prev_value || !(value == 0x00 || value == 0xff)) { \
+		if (value != 0x00) {						\
 			printf("Register " __stringify(register)		\
 				" values not clean! value: %u\n", value);	\
 			exit(-1);						\
 		}								\
-		prev_value = value;						\
 	}									\
 })
 

---
base-commit: 03d38806a902b36bf364cae8de6f1183c0a35a67
change-id: 20250301-fix-v_exec_initval_nolibc-498d976c372d

Best regards,
-- 
Ignacio Encinas <ignacio@iencinas.com>

Re: [PATCH v2] selftests: riscv: fix v_exec_initval_nolibc.c

Posted by Ignacio Encinas Rubio 10 months, 2 weeks ago

Gentle ping :)

On 6/3/25 20:49, Ignacio Encinas wrote:
> Vector registers are zero initialized by the kernel. Stop accepting
> "all ones" as a clean value.
> 
> Note that this was not working as expected given that
> 	value == 0xff
> can be assumed to be always false by the compiler as value's range is
> [-128, 127]. Both GCC (-Wtype-limits) and clang
> (-Wtautological-constant-out-of-range-compare) warn about this.
> 
> Reviewed-by: Charlie Jenkins <charlie@rivosinc.com>
> Tested-by: Charlie Jenkins <charlie@rivosinc.com>
> Signed-off-by: Ignacio Encinas <ignacio@iencinas.com>
> ---
> Changes in v2:
> 
> Remove code that becomes useless now that the only "clean" value for
> vector registers is 0.
> 
> - Link to v1: https://lore.kernel.org/r/20250305-fix-v_exec_initval_nolibc-v1-1-b87b60e43002@iencinas.com
> ---
>  tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c | 10 +++-------
>  1 file changed, 3 insertions(+), 7 deletions(-)
> 
> diff --git a/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c b/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
> index 35c0812e32de0c82a54f84bd52c4272507121e35..4dde05e45a04122b566cedc36d20b072413b00e2 100644
> --- a/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
> +++ b/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
> @@ -6,7 +6,7 @@
>   * the values. To further ensure consistency, this file is compiled without
>   * libc and without auto-vectorization.
>   *
> - * To be "clean" all values must be either all ones or all zeroes.
> + * To be "clean" all values must be all zeroes.
>   */
>  
>  #define __stringify_1(x...)	#x
> @@ -14,9 +14,8 @@
>  
>  int main(int argc, char **argv)
>  {
> -	char prev_value = 0, value;
> +	char value = 0;
>  	unsigned long vl;
> -	int first = 1;
>  
>  	if (argc > 2 && strcmp(argv[2], "x"))
>  		asm volatile (
> @@ -44,14 +43,11 @@ int main(int argc, char **argv)
>  			"vsrl.vi " __stringify(register) ", " __stringify(register) ", 8\n\t" \
>  			".option pop\n\t"					\
>  			: "=r" (value));					\
> -		if (first) {							\
> -			first = 0;						\
> -		} else if (value != prev_value || !(value == 0x00 || value == 0xff)) { \
> +		if (value != 0x00) {						\
>  			printf("Register " __stringify(register)		\
>  				" values not clean! value: %u\n", value);	\
>  			exit(-1);						\
>  		}								\
> -		prev_value = value;						\
>  	}									\
>  })
>  
> 
> ---
> base-commit: 03d38806a902b36bf364cae8de6f1183c0a35a67
> change-id: 20250301-fix-v_exec_initval_nolibc-498d976c372d
> 
> Best regards,

Re: [PATCH v2] selftests: riscv: fix v_exec_initval_nolibc.c

Posted by Alexandre Ghiti 10 months, 2 weeks ago

Hi ignacio,

On 30/03/2025 13:12, Ignacio Encinas Rubio wrote:
> Gentle ping :)


No worries, it's either being merged in 6.15-rc1 or rc2!

Thanks,

Alex


>
> On 6/3/25 20:49, Ignacio Encinas wrote:
>> Vector registers are zero initialized by the kernel. Stop accepting
>> "all ones" as a clean value.
>>
>> Note that this was not working as expected given that
>> 	value == 0xff
>> can be assumed to be always false by the compiler as value's range is
>> [-128, 127]. Both GCC (-Wtype-limits) and clang
>> (-Wtautological-constant-out-of-range-compare) warn about this.
>>
>> Reviewed-by: Charlie Jenkins <charlie@rivosinc.com>
>> Tested-by: Charlie Jenkins <charlie@rivosinc.com>
>> Signed-off-by: Ignacio Encinas <ignacio@iencinas.com>
>> ---
>> Changes in v2:
>>
>> Remove code that becomes useless now that the only "clean" value for
>> vector registers is 0.
>>
>> - Link to v1: https://lore.kernel.org/r/20250305-fix-v_exec_initval_nolibc-v1-1-b87b60e43002@iencinas.com
>> ---
>>   tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c | 10 +++-------
>>   1 file changed, 3 insertions(+), 7 deletions(-)
>>
>> diff --git a/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c b/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
>> index 35c0812e32de0c82a54f84bd52c4272507121e35..4dde05e45a04122b566cedc36d20b072413b00e2 100644
>> --- a/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
>> +++ b/tools/testing/selftests/riscv/vector/v_exec_initval_nolibc.c
>> @@ -6,7 +6,7 @@
>>    * the values. To further ensure consistency, this file is compiled without
>>    * libc and without auto-vectorization.
>>    *
>> - * To be "clean" all values must be either all ones or all zeroes.
>> + * To be "clean" all values must be all zeroes.
>>    */
>>   
>>   #define __stringify_1(x...)	#x
>> @@ -14,9 +14,8 @@
>>   
>>   int main(int argc, char **argv)
>>   {
>> -	char prev_value = 0, value;
>> +	char value = 0;
>>   	unsigned long vl;
>> -	int first = 1;
>>   
>>   	if (argc > 2 && strcmp(argv[2], "x"))
>>   		asm volatile (
>> @@ -44,14 +43,11 @@ int main(int argc, char **argv)
>>   			"vsrl.vi " __stringify(register) ", " __stringify(register) ", 8\n\t" \
>>   			".option pop\n\t"					\
>>   			: "=r" (value));					\
>> -		if (first) {							\
>> -			first = 0;						\
>> -		} else if (value != prev_value || !(value == 0x00 || value == 0xff)) { \
>> +		if (value != 0x00) {						\
>>   			printf("Register " __stringify(register)		\
>>   				" values not clean! value: %u\n", value);	\
>>   			exit(-1);						\
>>   		}								\
>> -		prev_value = value;						\
>>   	}									\
>>   })
>>   
>>
>> ---
>> base-commit: 03d38806a902b36bf364cae8de6f1183c0a35a67
>> change-id: 20250301-fix-v_exec_initval_nolibc-498d976c372d
>>
>> Best regards,
>
> _______________________________________________
> linux-riscv mailing list
> linux-riscv@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-riscv