[PATCH] Some systems have a dynamic PAGE_SIZE value and do not add a definition for PAGE_SIZE. This value will have to be retrieved using getpagesize() or sysconf().

[PATCH] Some systems have a dynamic PAGE_SIZE value and do not add a definition for PAGE_SIZE. This value will have to be retrieved using getpagesize() or sysconf().

[Thierry FAUCK - IBM LTC]

From: Thierry Fauck <thierry@linux.vnet.ibm.com>

Signed-off-by: Thierry Fauck <thierry@linux.vnet.ibm.com>
---
 tools/testing/selftests/ptrace/peeksiginfo.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/tools/testing/selftests/ptrace/peeksiginfo.c b/tools/testing/selftests/ptrace/peeksiginfo.c
index d46558b..c34cd8a 100644
--- a/tools/testing/selftests/ptrace/peeksiginfo.c
+++ b/tools/testing/selftests/ptrace/peeksiginfo.c
@@ -31,6 +31,10 @@ static int sys_ptrace(int request, pid_t pid, void *addr, void *data)
 #define TEST_SICODE_PRIV	-1
 #define TEST_SICODE_SHARE	-2
 
+#ifndef PAGE_SIZE
+#define PAGE_SIZE sysconf(_SC_PAGESIZE)
+#endif
+
 #define err(fmt, ...)						\
 		fprintf(stderr,					\
 			"Error (%s:%d): " fmt,			\
-- 
2.0.0

Re: [PATCH] Some systems have a dynamic PAGE_SIZE value and do not add a definition for PAGE_SIZE. This value will have to be retrieved using getpagesize() or sysconf().

[Andrew Morton]

On Wed,  6 Aug 2014 13:34:26 +0200 Thierry FAUCK - IBM LTC <thierry@linux.vnet.ibm.com> wrote:

> --- a/tools/testing/selftests/ptrace/peeksiginfo.c
> +++ b/tools/testing/selftests/ptrace/peeksiginfo.c
> @@ -31,6 +31,10 @@ static int sys_ptrace(int request, pid_t pid, void *addr, void *data)
>  #define TEST_SICODE_PRIV	-1
>  #define TEST_SICODE_SHARE	-2
>  
> +#ifndef PAGE_SIZE
> +#define PAGE_SIZE sysconf(_SC_PAGESIZE)
> +#endif
> +
>  #define err(fmt, ...)						\
>  		fprintf(stderr,					\
>  			"Error (%s:%d): " fmt,			\

More details, please.  What are "some systems"?

Re: [PATCH] Some systems have a dynamic PAGE_SIZE value and do not add a definition for PAGE_SIZE. This value will have to be retrieved using getpagesize() or sysconf().

[Thierry Fauck@linux.vnet.ibm.com]

On 08/06/2014 10:06 PM, Andrew Morton wrote:
> On Wed,  6 Aug 2014 13:34:26 +0200 Thierry FAUCK - IBM LTC <thierry@linux.vnet.ibm.com> wrote:
>
>> --- a/tools/testing/selftests/ptrace/peeksiginfo.c
>> +++ b/tools/testing/selftests/ptrace/peeksiginfo.c
>> @@ -31,6 +31,10 @@ static int sys_ptrace(int request, pid_t pid, void *addr, void *data)
>>  #define TEST_SICODE_PRIV	-1
>>  #define TEST_SICODE_SHARE	-2
>>  
>> +#ifndef PAGE_SIZE
>> +#define PAGE_SIZE sysconf(_SC_PAGESIZE)
>> +#endif
>> +
>>  #define err(fmt, ...)						\
>>  		fprintf(stderr,					\
>>  			"Error (%s:%d): " fmt,			\
> More details, please.  What are "some systems"?
>
Sorry, about that "common description".
A more appropriate one could be : On IBM powerpc where multiple page
size value are supported, current ppc64 and ppc64el distro don't define
the PAGE_SIZE variable in /usr/include as this is a dynamic value
retrieved by the getpagesize() or sysconf() defined in unistd.h. The
PAGE_SIZE variable sounds defined when only one value is supported by
the kernel.
As such, when the PAGE_SIZE definition doesn't exist system should
retrieve the dynamic value.

Is that a better explanation ?

-- 
Thierry Fauck @ linux.vnet.ibm