The recent commit
b4a5d26d8835d972995f0a0a2f805a8845bafa0b
"linux: Consolidate sigaction implementation" changed the definition
of struct sigaction for s390 (31bit). Unfortunately the order of the
fields were wrong.
This leads to blocking testcases e.g. nptl/tst-sem11.
A thread which blocks due to sem_wait() is cancelled via pthread_cancel()
and the signal-handler sigcancel_handler (see <glibc-src>/nptl/nptl-init.c
is called.
But it just returns as the siginfo_t argument is not setup by the kernel.
Then the main-thread is blocking due to pthread_join().
The flag SA_SIGINFO is set in sa_flags in struct sigaction and
is copied to the "kernel_sigaction.h" struct by the sigaction() call,
but due to the wrong ordering of the struct fields,
the kernel does not recognize it.
+2018-04-12 Stefan Liebler <stli@linux.vnet.ibm.com>
+
+ * sysdeps/unix/sysv/linux/s390/kernel_sigaction.h
+ (struct kernel_sigaction): Use the same definition on 31bit as is used
+ on 64bit.
+
2018-04-09 Florian Weimer <fweimer@redhat.com>
[BZ #23037]
void (*_sa_sigaction)(int, siginfo_t *, void *);
} _u;
#define k_sa_handler _u._sa_handler
-#ifndef __s390x__
- sigset_t sa_mask;
- unsigned long sa_flags;
- void (*sa_restorer)(void);
-#else
+ /* The 'struct sigaction' definition in s390 kernel header
+ arch/s390/include/uapi/asm/signal.h is used for __NR_rt_sigaction
+ on 64 bits and for __NR_sigaction for 31 bits.
+
+ The expected layout for __NR_rt_sigaction for 31 bits is either
+ 'struct sigaction' from include/linux/signal_types.h or
+ 'struct compat_sigaction' from include/linux/compat.h.
+
+ So for __NR_rt_sigaction we can use the same layout for both s390x
+ and s390. */
unsigned long sa_flags;
void (*sa_restorer)(void);
sigset_t sa_mask;
-#endif
};
#define SET_SA_RESTORER(kact, act) \