[LTP] [PATCH v2] getcpu01: Reinstate node_id test
Petr Vorel
pvorel@suse.cz
Tue Dec 13 10:53:00 CET 2022
Hi Xu,
> Presently the node_id is only checked on i386 and it is broken. The
> sched_getcpu call was substituted for getcpu when
> available. sched_getcpu does not have the node_id parameter, it's not
> the same thing as getcpu.
> Also we can check the node_id on any platform which has NUMA. Which
> includes more than just x86.
> Signed-off-by: Richard Palethorpe <rpalethorpe@suse.com>
> Cc: Cyril Hrubis <chrubis@suse.cz>
> ---
> V2:
> * Removed all the ifdefs
> * Use libc getcpu when available
> * Remove kernel version check
> configure.ac | 1 +
> include/lapi/sched.h | 7 +++++
> testcases/kernel/syscalls/getcpu/getcpu01.c | 35 ++-------------------
> 3 files changed, 11 insertions(+), 32 deletions(-)
> diff --git a/configure.ac b/configure.ac
> index e9b15c7f7..1ab7cc60d 100644
> --- a/configure.ac
> +++ b/configure.ac
> @@ -101,6 +101,7 @@ AC_CHECK_FUNCS_ONCE([ \
> fstatat \
> getauxval \
> getcontext \
> + getcpu \
> getdents \
> getdents64 \
> io_pgetevents \
> diff --git a/include/lapi/sched.h b/include/lapi/sched.h
> index 1d22a9d7e..1065665d1 100644
> --- a/include/lapi/sched.h
> +++ b/include/lapi/sched.h
> @@ -70,6 +70,13 @@ static inline void clone3_supported_by_kernel(void)
> }
> }
> +#ifndef HAVE_GETCPU
> +static inline int getcpu(unsigned *cpu, unsigned *node)
> +{
> + return tst_syscall(__NR_getcpu, cpu, node, NULL);
> +}
> +#endif
> +
> #ifndef SCHED_DEADLINE
> # define SCHED_DEADLINE 6
> #endif
> diff --git a/testcases/kernel/syscalls/getcpu/getcpu01.c b/testcases/kernel/syscalls/getcpu/getcpu01.c
> index fcc273e29..f6fcc4fc1 100644
> --- a/testcases/kernel/syscalls/getcpu/getcpu01.c
> +++ b/testcases/kernel/syscalls/getcpu/getcpu01.c
> @@ -11,26 +11,12 @@
> #define _GNU_SOURCE
> #include <dirent.h>
> #include <errno.h>
> -#include <sched.h>
> #include <stdio.h>
> #include <stdlib.h>
> #include <sys/types.h>
> -#include "lapi/syscalls.h"
> -#include "lapi/cpuset.h"
> #include "tst_test.h"
> -#include "config.h"
> -
> -static inline int get_cpu(unsigned *cpu_id,
> - unsigned *node_id LTP_ATTRIBUTE_UNUSED,
> - void *cache_struct LTP_ATTRIBUTE_UNUSED)
> -{
> -#ifndef HAVE_SCHED_GETCPU
> - return tst_syscall(__NR_getcpu, cpu_id, node_id, cache_struct);
> -#else
> - *cpu_id = sched_getcpu();
> -#endif
> - return 0;
> -}
> +#include "lapi/cpuset.h"
> +#include "lapi/sched.h"
> static unsigned int max_cpuid(size_t size, cpu_set_t * set)
> {
> @@ -78,7 +64,6 @@ realloc:
> return cpu_max;
> }
> -#ifdef __i386__
> static unsigned int get_nodeid(unsigned int cpu_id)
> {
> DIR *directory_parent, *directory_node;
> @@ -119,33 +104,26 @@ static unsigned int get_nodeid(unsigned int cpu_id)
> }
> return node_id;
> }
> -#endif
> static void run(void)
> {
> unsigned int cpu_id, node_id = 0;
> unsigned int cpu_set;
> -#ifdef __i386__
> unsigned int node_set;
> -#endif
> cpu_set = set_cpu_affinity();
> -#ifdef __i386__
> node_set = get_nodeid(cpu_set);
> -#endif
> - TEST(get_cpu(&cpu_id, &node_id, NULL));
> + TEST(getcpu(&cpu_id, &node_id));
> if (TST_RET == 0) {
> if (cpu_id != cpu_set)
> tst_res(TFAIL, "getcpu() returned wrong value"
> " expected cpuid:%d, returned value cpuid: %d",
> cpu_set, cpu_id);
> -#ifdef __i386__
> else if (node_id != node_set)
> tst_res(TFAIL, "getcpu() returned wrong value"
> " expected node id:%d returned node id:%d",
> node_set, node_id);
> -#endif
> else
> tst_res(TPASS, "getcpu() returned proper"
> " cpuid:%d, node id:%d", cpu_id,
> @@ -156,13 +134,6 @@ static void run(void)
> }
> }
FYI Richie is touching the same code, one of you will need to rebase.
IMHO it's better to remove this in dedicated patchset (i.e. in Xu).
Kind regards,
Petr
> -static void setup(void)
> -{
> - if (tst_kvercmp(2, 6, 20) < 0)
> - tst_brk(TCONF, "kernel >= 2.6.20 required");
> -}
> -
> static struct tst_test test = {
> .test_all = run,
> - .setup = setup,
> };
More information about the ltp
mailing list