[PATCH] s390 (1/13): base s390 fixes.

Martin Schwidefsky (schwidefsky@de.ibm.com)
Mon, 24 Feb 2003 19:07:21 +0100


s390 arch changes/bug fixes:
* add sys_fadvise64 system call
* add initialization of init_sighand
* add support for clone option CLONE_SETTLS
* make use of ptrace_notify
* sig -> sighand rename
* move ptrace_signal_deliver to the correct place
* make eieio a memory barrier
* fix race condition on cpu_vm_mask in __flush_tlb_mm
* add missing KM_SOFTIRQ0, KM_SOFTIRQ1 in asm-s390x/kmap_types.h
* add inline assmelby for _raw_write_trylock

diff -urN linux-2.5.62/arch/s390/kernel/entry.S linux-2.5.62-s390/arch/s390/kernel/entry.S
--- linux-2.5.62/arch/s390/kernel/entry.S Mon Feb 24 18:17:42 2003
+++ linux-2.5.62-s390/arch/s390/kernel/entry.S Mon Feb 24 18:18:24 2003
@@ -616,7 +616,8 @@
.long sys_epoll_ctl /* 250 */
.long sys_epoll_wait
.long sys_set_tid_address
- .rept 255-252
+ .long sys_fadvise64
+ .rept 255-253
.long sys_ni_syscall
.endr

diff -urN linux-2.5.62/arch/s390/kernel/init_task.c linux-2.5.62-s390/arch/s390/kernel/init_task.c
--- linux-2.5.62/arch/s390/kernel/init_task.c Mon Feb 17 23:56:20 2003
+++ linux-2.5.62-s390/arch/s390/kernel/init_task.c Mon Feb 24 18:18:24 2003
@@ -16,6 +16,7 @@
static struct fs_struct init_fs = INIT_FS;
static struct files_struct init_files = INIT_FILES;
static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
+static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
struct mm_struct init_mm = INIT_MM(init_mm);

/*
diff -urN linux-2.5.62/arch/s390/kernel/process.c linux-2.5.62-s390/arch/s390/kernel/process.c
--- linux-2.5.62/arch/s390/kernel/process.c Mon Feb 17 23:56:44 2003
+++ linux-2.5.62-s390/arch/s390/kernel/process.c Mon Feb 24 18:18:24 2003
@@ -196,8 +196,6 @@

/* new return point is ret_from_fork */
frame->gprs[8] = (unsigned long) ret_from_fork;
- /* start disabled because of schedule_tick and rq->lock being held */
- frame->childregs.psw.mask &= ~0x03000000;

/* fake return stack for resume(), don't go back to schedule */
frame->gprs[9] = (unsigned long) frame;
@@ -213,6 +211,10 @@
p->thread.ar4 = get_fs().ar4;
/* Don't copy debug registers */
memset(&p->thread.per_info,0,sizeof(p->thread.per_info));
+
+ /* Set a new TLS ? */
+ if (clone_flags & CLONE_SETTLS)
+ frame->childregs.acrs[0] = regs->gprs[6];
return 0;
}

@@ -335,7 +337,7 @@
int count = 0;
if (!p || p == current || p->state == TASK_RUNNING)
return 0;
- stack_page = (unsigned long) p;
+ stack_page = (unsigned long) p->thread_info;
r15 = p->thread.ksp;
if (!stack_page || r15 < stack_page || r15 >= 8188+stack_page)
return 0;
diff -urN linux-2.5.62/arch/s390/kernel/ptrace.c linux-2.5.62-s390/arch/s390/kernel/ptrace.c
--- linux-2.5.62/arch/s390/kernel/ptrace.c Mon Feb 17 23:56:55 2003
+++ linux-2.5.62-s390/arch/s390/kernel/ptrace.c Mon Feb 24 18:18:24 2003
@@ -359,11 +359,9 @@
return;
if (!(current->ptrace & PT_PTRACED))
return;
- current->exit_code =
- SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD) ? 0x80 : 0);
- current->state = TASK_STOPPED;
- notify_parent(current, SIGCHLD);
- schedule();
+ ptrace_notify(SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD)
+ ? 0x80 : 0));
+
/*
* this isn't the same as continuing with a signal, but it will do
* for normal use. strace only continues with a signal if the
diff -urN linux-2.5.62/arch/s390/kernel/signal.c linux-2.5.62-s390/arch/s390/kernel/signal.c
--- linux-2.5.62/arch/s390/kernel/signal.c Mon Feb 17 23:56:45 2003
+++ linux-2.5.62-s390/arch/s390/kernel/signal.c Mon Feb 24 18:18:24 2003
@@ -391,7 +391,7 @@
handle_signal(unsigned long sig, siginfo_t *info, sigset_t *oldset,
struct pt_regs * regs)
{
- struct k_sigaction *ka = &current->sig->action[sig-1];
+ struct k_sigaction *ka = &current->sighand->action[sig-1];

/* Are we from a system call? */
if (regs->trap == __LC_SVC_OLD_PSW) {
diff -urN linux-2.5.62/arch/s390x/kernel/entry.S linux-2.5.62-s390/arch/s390x/kernel/entry.S
--- linux-2.5.62/arch/s390x/kernel/entry.S Mon Feb 24 18:17:42 2003
+++ linux-2.5.62-s390/arch/s390x/kernel/entry.S Mon Feb 24 18:18:24 2003
@@ -644,7 +644,8 @@
.long SYSCALL(sys_epoll_ctl,sys_ni_syscall)
.long SYSCALL(sys_epoll_wait,sys_ni_syscall)
.long SYSCALL(sys_set_tid_address,sys_ni_syscall)
- .rept 255-252
+ .long SYSCALL(sys_fadvise64,sys_ni_syscall)
+ .rept 255-253
.long SYSCALL(sys_ni_syscall,sys_ni_syscall)
.endr

diff -urN linux-2.5.62/arch/s390x/kernel/init_task.c linux-2.5.62-s390/arch/s390x/kernel/init_task.c
--- linux-2.5.62/arch/s390x/kernel/init_task.c Mon Feb 17 23:56:11 2003
+++ linux-2.5.62-s390/arch/s390x/kernel/init_task.c Mon Feb 24 18:18:24 2003
@@ -16,6 +16,7 @@
static struct fs_struct init_fs = INIT_FS;
static struct files_struct init_files = INIT_FILES;
static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
+static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
struct mm_struct init_mm = INIT_MM(init_mm);

/*
diff -urN linux-2.5.62/arch/s390x/kernel/process.c linux-2.5.62-s390/arch/s390x/kernel/process.c
--- linux-2.5.62/arch/s390x/kernel/process.c Mon Feb 17 23:56:29 2003
+++ linux-2.5.62-s390/arch/s390x/kernel/process.c Mon Feb 24 18:18:24 2003
@@ -203,6 +203,19 @@
p->thread.ar4 = get_fs().ar4;
/* Don't copy debug registers */
memset(&p->thread.per_info,0,sizeof(p->thread.per_info));
+
+ /* Set a new TLS ? */
+ if (clone_flags & CLONE_SETTLS) {
+ if (current->thread.flags & S390_FLAG_31BIT) {
+ frame->childregs.acrs[0] =
+ (unsigned int) regs->gprs[6];
+ } else {
+ frame->childregs.acrs[0] =
+ (unsigned int)(regs->gprs[6] >> 32);
+ frame->childregs.acrs[1] =
+ (unsigned int) regs->gprs[6];
+ }
+ }
return 0;
}

@@ -319,7 +332,7 @@
int count = 0;
if (!p || p == current || p->state == TASK_RUNNING)
return 0;
- stack_page = (unsigned long) p;
+ stack_page = (unsigned long) p->thread_info;
r15 = p->thread.ksp;
if (!stack_page || r15 < stack_page || r15 >= 16380+stack_page)
return 0;
diff -urN linux-2.5.62/arch/s390x/kernel/ptrace.c linux-2.5.62-s390/arch/s390x/kernel/ptrace.c
--- linux-2.5.62/arch/s390x/kernel/ptrace.c Mon Feb 17 23:55:53 2003
+++ linux-2.5.62-s390/arch/s390x/kernel/ptrace.c Mon Feb 24 18:18:24 2003
@@ -598,11 +598,9 @@
return;
if (!(current->ptrace & PT_PTRACED))
return;
- current->exit_code =
- SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD) ? 0x80 : 0);
- current->state = TASK_STOPPED;
- notify_parent(current, SIGCHLD);
- schedule();
+ ptrace_notify(SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD)
+ ? 0x80 : 0));
+
/*
* this isn't the same as continuing with a signal, but it will do
* for normal use. strace only continues with a signal if the
diff -urN linux-2.5.62/arch/s390x/kernel/signal.c linux-2.5.62-s390/arch/s390x/kernel/signal.c
--- linux-2.5.62/arch/s390x/kernel/signal.c Mon Feb 17 23:56:20 2003
+++ linux-2.5.62-s390/arch/s390x/kernel/signal.c Mon Feb 24 18:18:24 2003
@@ -385,7 +385,7 @@
handle_signal(unsigned long sig, siginfo_t *info, sigset_t *oldset,
struct pt_regs * regs)
{
- struct k_sigaction *ka = &current->sig->action[sig-1];
+ struct k_sigaction *ka = &current->sighand->action[sig-1];

/* Are we from a system call? */
if (regs->trap == __LC_SVC_OLD_PSW) {
diff -urN linux-2.5.62/include/asm-s390/signal.h linux-2.5.62-s390/include/asm-s390/signal.h
--- linux-2.5.62/include/asm-s390/signal.h Mon Feb 17 23:56:40 2003
+++ linux-2.5.62-s390/include/asm-s390/signal.h Mon Feb 24 18:18:24 2003
@@ -159,6 +159,9 @@
struct k_sigaction {
struct sigaction sa;
};
+
+#define ptrace_signal_deliver(regs, cookie) do { } while (0)
+
#else
/* Here we must cater to libcs that poke about in kernel headers. */

@@ -175,8 +178,6 @@
#define sa_handler _u._sa_handler
#define sa_sigaction _u._sa_sigaction

-#define ptrace_signal_deliver(regs, cookie) do { } while (0)
-
#endif /* __KERNEL__ */

typedef struct sigaltstack {
diff -urN linux-2.5.62/include/asm-s390/spinlock.h linux-2.5.62-s390/include/asm-s390/spinlock.h
--- linux-2.5.62/include/asm-s390/spinlock.h Mon Feb 17 23:57:22 2003
+++ linux-2.5.62-s390/include/asm-s390/spinlock.h Mon Feb 24 18:24:47 2003
@@ -122,4 +122,17 @@
: "+m" ((rw)->lock) : "a" (&(rw)->lock) \
: "2", "3", "cc" )

+extern inline int _raw_write_trylock(rwlock_t *rw)
+{
+ unsigned int result, reg;
+
+ __asm__ __volatile__(" lhi %0,1\n"
+ " sll %0,31\n"
+ " basr %1,0\n"
+ "0: cs %0,%1,0(%3)\n"
+ : "=&d" (result), "=&d" (reg), "+m" (rw->lock)
+ : "a" (&rw->lock) : "cc" );
+ return !result;
+}
+
#endif /* __ASM_SPINLOCK_H */
diff -urN linux-2.5.62/include/asm-s390/system.h linux-2.5.62-s390/include/asm-s390/system.h
--- linux-2.5.62/include/asm-s390/system.h Mon Feb 17 23:57:19 2003
+++ linux-2.5.62-s390/include/asm-s390/system.h Mon Feb 24 18:18:24 2003
@@ -222,7 +222,7 @@
* all memory ops have completed wrt other CPU's ( see 7-15 POP DJB ).
*/

-#define eieio() __asm__ __volatile__ ("BCR 15,0")
+#define eieio() __asm__ __volatile__ ( "bcr 15,0" : : : "memory" )
# define SYNC_OTHER_CORES(x) eieio()
#define mb() eieio()
#define rmb() eieio()
diff -urN linux-2.5.62/include/asm-s390/tlbflush.h linux-2.5.62-s390/include/asm-s390/tlbflush.h
--- linux-2.5.62/include/asm-s390/tlbflush.h Mon Feb 17 23:57:18 2003
+++ linux-2.5.62-s390/include/asm-s390/tlbflush.h Mon Feb 24 18:18:24 2003
@@ -91,15 +91,17 @@

static inline void __flush_tlb_mm(struct mm_struct * mm)
{
+ preempt_disable();
if (mm->cpu_vm_mask != (1UL << smp_processor_id())) {
/* mm was active on more than one cpu. */
if (mm == current->active_mm &&
- atomic_read(&mm->mm_count) == 1)
+ atomic_read(&mm->mm_users) == 1)
/* this cpu is the only one using the mm. */
mm->cpu_vm_mask = 1UL << smp_processor_id();
global_flush_tlb();
} else
local_flush_tlb();
+ preempt_enable();
}

static inline void flush_tlb(void)
diff -urN linux-2.5.62/include/asm-s390/unistd.h linux-2.5.62-s390/include/asm-s390/unistd.h
--- linux-2.5.62/include/asm-s390/unistd.h Mon Feb 17 23:57:22 2003
+++ linux-2.5.62-s390/include/asm-s390/unistd.h Mon Feb 24 18:18:24 2003
@@ -248,6 +248,7 @@
#define __NR_epoll_ctl 250
#define __NR_epoll_wait 251
#define __NR_set_tid_address 252
+#define __NR_fadvise64 253


/* user-visible error numbers are in the range -1 - -122: see <asm-s390/errno.h> */
diff -urN linux-2.5.62/include/asm-s390x/kmap_types.h linux-2.5.62-s390/include/asm-s390x/kmap_types.h
--- linux-2.5.62/include/asm-s390x/kmap_types.h Mon Feb 17 23:56:13 2003
+++ linux-2.5.62-s390/include/asm-s390x/kmap_types.h Mon Feb 24 18:18:24 2003
@@ -14,6 +14,8 @@
KM_PTE1,
KM_IRQ0,
KM_IRQ1,
+ KM_SOFTIRQ0,
+ KM_SOFTIRQ1,
KM_TYPE_NR
};

diff -urN linux-2.5.62/include/asm-s390x/signal.h linux-2.5.62-s390/include/asm-s390x/signal.h
--- linux-2.5.62/include/asm-s390x/signal.h Mon Feb 17 23:56:19 2003
+++ linux-2.5.62-s390/include/asm-s390x/signal.h Mon Feb 24 18:18:24 2003
@@ -159,6 +159,9 @@
struct k_sigaction {
struct sigaction sa;
};
+
+#define ptrace_signal_deliver(regs, cookie) do { } while (0)
+
#else
/* Here we must cater to libcs that poke about in kernel headers. */

@@ -175,8 +178,6 @@
#define sa_handler _u._sa_handler
#define sa_sigaction _u._sa_sigaction

-#define ptrace_signal_deliver(regs, cookie) do { } while (0)
-
#endif /* __KERNEL__ */

typedef struct sigaltstack {
diff -urN linux-2.5.62/include/asm-s390x/spinlock.h linux-2.5.62-s390/include/asm-s390x/spinlock.h
--- linux-2.5.62/include/asm-s390x/spinlock.h Mon Feb 17 23:55:57 2003
+++ linux-2.5.62-s390/include/asm-s390x/spinlock.h Mon Feb 24 18:24:47 2003
@@ -139,5 +139,17 @@
: "a" (&(rw)->lock), "i" (__DIAG44_OPERAND) \
: "2", "3", "cc" )

+extern inline int _raw_write_trylock(rwlock_t *rw)
+{
+ unsigned int result, reg;
+
+ __asm__ __volatile__(" llihh %0,0x8000\n"
+ " basr %1,0\n"
+ "0: csg %0,%1,0(%3)\n"
+ : "=&d" (result), "=&d" (reg), "+m" (rw->lock)
+ : "a" (&rw->lock) : "cc" );
+ return !result;
+}
+
#endif /* __ASM_SPINLOCK_H */

diff -urN linux-2.5.62/include/asm-s390x/system.h linux-2.5.62-s390/include/asm-s390x/system.h
--- linux-2.5.62/include/asm-s390x/system.h Mon Feb 17 23:56:59 2003
+++ linux-2.5.62-s390/include/asm-s390x/system.h Mon Feb 24 18:18:24 2003
@@ -233,7 +233,7 @@
* all memory ops have completed wrt other CPU's ( see 7-15 POP DJB ).
*/

-#define eieio() __asm__ __volatile__ ("BCR 15,0")
+#define eieio() __asm__ __volatile__ ( "bcr 15,0" : : : "memory" )
# define SYNC_OTHER_CORES(x) eieio()
#define mb() eieio()
#define rmb() eieio()
diff -urN linux-2.5.62/include/asm-s390x/tlbflush.h linux-2.5.62-s390/include/asm-s390x/tlbflush.h
--- linux-2.5.62/include/asm-s390x/tlbflush.h Mon Feb 17 23:56:25 2003
+++ linux-2.5.62-s390/include/asm-s390x/tlbflush.h Mon Feb 24 18:18:24 2003
@@ -88,15 +88,17 @@

static inline void __flush_tlb_mm(struct mm_struct * mm)
{
+ preempt_disable();
if (mm->cpu_vm_mask != (1UL << smp_processor_id())) {
/* mm was active on more than one cpu. */
if (mm == current->active_mm &&
- atomic_read(&mm->mm_count) == 1)
+ atomic_read(&mm->mm_users) == 1)
/* this cpu is the only one using the mm. */
mm->cpu_vm_mask = 1UL << smp_processor_id();
global_flush_tlb();
} else
local_flush_tlb();
+ preempt_enable();
}

static inline void flush_tlb(void)
diff -urN linux-2.5.62/include/asm-s390x/unistd.h linux-2.5.62-s390/include/asm-s390x/unistd.h
--- linux-2.5.62/include/asm-s390x/unistd.h Mon Feb 17 23:56:25 2003
+++ linux-2.5.62-s390/include/asm-s390x/unistd.h Mon Feb 24 18:18:24 2003
@@ -215,6 +215,7 @@
#define __NR_epoll_ctl 250
#define __NR_epoll_wait 251
#define __NR_set_tid_address 252
+#define __NR_fadvise64 253


/* user-visible error numbers are in the range -1 - -122: see <asm-s390/errno.h> */

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/