LTT for 2.5.44-bk2 7/10: S/390 trace support

LTT for 2.5.44-bk2 7/10: S/390 trace support

Post by Karim Yaghmou » Thu, 31 Oct 2002 00:10:10



D: This patch adds the bare-minimum S/390-specific low-level trace
D: statements.

diffstat:
 arch/s390/kernel/entry.S    |   18 ++++++
 arch/s390/kernel/process.c  |    5 +
 arch/s390/kernel/sys_s390.c |    3 +
 arch/s390/kernel/traps.c    |  117 ++++++++++++++++++++++++++++++++++++++++++++
 arch/s390/mm/fault.c        |   12 ++++
 drivers/s390/cio/cio.c      |    4 +
 drivers/s390/s390mach.c     |   12 ++++
 include/asm-s390/trace.h    |   16 ++++++
 8 files changed, 187 insertions(+)

diff -urpN linux-2.5.44-bk2/arch/s390/kernel/entry.S linux-2.5.44-bk2-ltt/arch/s390/kernel/entry.S
--- linux-2.5.44-bk2/arch/s390/kernel/entry.S   Sat Oct 19 00:01:16 2002
+++ linux-2.5.44-bk2-ltt/arch/s390/kernel/entry.S       Tue Oct 29 15:24:18 2002
@@ -7,6 +7,7 @@
  *    Author(s): Martin Schwidefsky (schwidef...@de.ibm.com),
  *               Hartmut Penner (h...@de.ibm.com),
  *               Denis Joseph Barrow (djbar...@de.ibm.com,barrow...@yahoo.com),
+ *  Portions added by T. Halloran: (C) Copyright 2002 IBM Poughkeepsie, IBM Corporation
  */

 #include <linux/sys.h>
@@ -181,6 +182,14 @@ system_call:
         GET_THREAD_INFO           # load pointer to task_struct to R9
        sll     %r7,2
         stosm   24(%r15),0x03     # reenable interrupts
+/* call to ltt trace done here.  R8 has the syscall (svc) number to trace */
+#if (CONFIG_TRACE) /* tjh - ltt port */
+        /* add call to trace_real_syscall_entry */                          
+        la      %r2,SP_PTREGS(%r15)   # load pt_regs as first parameter    
+        l       %r1,BASED(.Ltracesysent)                                    
+        basr    %r14,%r1                                                    
+        lm      %r0,%r6,SP_R0(%r15) /* restore call clobbered regs tjh */  
+#endif
         l       %r8,sys_call_table-entry_base(%r7,%r13) # get system call addr.
        tm      __TI_flags+3(%r9),_TIF_SYSCALL_TRACE
         bo      BASED(sysc_tracesys)
@@ -189,6 +198,13 @@ system_call:
                                   # ATTENTION: check sys_execve_glue before
                                   # changing anything here !!

+#if (CONFIG_TRACE) /* tjh - ltt port */
+        /* add call to trace_real_syscall_exit */
+        la      %r2,SP_PTREGS(%r15)   # load pt_regs as first parameter  
+        l       %r1,BASED(.Ltracesysext)                                
+        basr    %r14,%r1                                                
+        lm      %r0,%r6,SP_R0(%r15) /* restore call clobbered regs */
+#endif                                                                  
 sysc_return:
        stnsm   24(%r15),0xfc     # disable I/O and ext. interrupts
        tm      __TI_flags+3(%r9),_TIF_WORK_MASK
@@ -928,6 +944,8 @@ restart_go:
 .Lsigaltstack: .long  sys_sigaltstack
 .Ltrace:       .long  syscall_trace
 .Lvfork:       .long  sys_vfork
+.Ltracesysent: .long  trace_real_syscall_entry
+.Ltracesysext: .long  trace_real_syscall_exit
 #if CONFIG_SMP || CONFIG_PREEMPT
 .Lschedtail:   .long  schedule_tail
 #endif
diff -urpN linux-2.5.44-bk2/arch/s390/kernel/process.c linux-2.5.44-bk2-ltt/arch/s390/kernel/process.c
--- linux-2.5.44-bk2/arch/s390/kernel/process.c Sat Oct 19 00:02:27 2002
+++ linux-2.5.44-bk2-ltt/arch/s390/kernel/process.c     Tue Oct 29 15:24:18 2002
@@ -33,6 +33,7 @@
 #include <linux/delay.h>
 #include <linux/reboot.h>
 #include <linux/init.h>
+#include <linux/trace.h>

 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
@@ -147,6 +148,10 @@ int kernel_thread(int (*fn)(void *), voi

        /* Ok, create the new process.. */
        p = do_fork(flags | CLONE_VM, 0, &regs, 0, NULL);
+#if (CONFIG_TRACE)
+       if(!IS_ERR(p))
+               TRACE_PROCESS(TRACE_EV_PROCESS_KTHREAD, p->pid, (int) fn);
+#endif
        return IS_ERR(p) ? PTR_ERR(p) : p->pid;
 }

diff -urpN linux-2.5.44-bk2/arch/s390/kernel/sys_s390.c linux-2.5.44-bk2-ltt/arch/s390/kernel/sys_s390.c
--- linux-2.5.44-bk2/arch/s390/kernel/sys_s390.c        Sat Oct 19 00:01:49 2002
+++ linux-2.5.44-bk2-ltt/arch/s390/kernel/sys_s390.c    Tue Oct 29 15:24:18 2002
@@ -24,6 +24,7 @@
 #include <linux/mman.h>
 #include <linux/file.h>
 #include <linux/utsname.h>
+#include <linux/trace.h>

 #include <asm/uaccess.h>
 #include <asm/ipc.h>
@@ -145,6 +146,8 @@ asmlinkage int sys_ipc (uint call, int f
         struct ipc_kludge tmp;
        int ret;

+        TRACE_IPC(TRACE_EV_IPC_CALL, call, first);
+
         switch (call) {
         case SEMOP:
                 return sys_semop (first, (struct sembuf *)ptr, second);
diff -urpN linux-2.5.44-bk2/arch/s390/kernel/traps.c linux-2.5.44-bk2-ltt/arch/s390/kernel/traps.c
--- linux-2.5.44-bk2/arch/s390/kernel/traps.c   Sat Oct 19 00:01:22 2002
+++ linux-2.5.44-bk2-ltt/arch/s390/kernel/traps.c       Tue Oct 29 15:24:18 2002
@@ -5,6 +5,7 @@
  *    Copyright (C) 1999,2000 IBM Deutschland Entwicklung GmbH, IBM Corporation
  *    Author(s): Martin Schwidefsky (schwidef...@de.ibm.com),
  *               Denis Joseph Barrow (djbar...@de.ibm.com,barrow...@yahoo.com),
+ *  Portions added by T. Halloran: (C) Copyright 2002 IBM Poughkeepsie, IBM Corporation
  *
  *  Derived from "arch/i386/kernel/traps.c"
  *    Copyright (C) 1991, 1992 Linus Torvalds
@@ -28,6 +29,8 @@
 #include <linux/delay.h>
 #include <linux/module.h>

+#include <linux/trace.h>
+
 #include <asm/system.h>
 #include <asm/uaccess.h>
 #include <asm/io.h>
@@ -283,6 +286,9 @@ void die(const char * str, struct pt_reg
 static void inline do_trap(long interruption_code, int signr, char *str,
                            struct pt_regs *regs, siginfo_t *info)
 {
+         trapid_t ltt_interruption_code;
+         char * ic_ptr = (char *) &ltt_interruption_code;
+
        /*
         * We got all needed information from the lowcore and can
         * now safely switch on interrupts.
@@ -290,6 +296,10 @@ static void inline do_trap(long interrup
         if (regs->psw.mask & PSW_MASK_PSTATE)
                local_irq_enable();

+       memset(&ltt_interruption_code,0,sizeof(ltt_interruption_code));
+       memcpy(ic_ptr+4,&interruption_code,sizeof(interruption_code));
+       TRACE_TRAP_ENTRY(ltt_interruption_code, (regs->psw.addr & PSW_ADDR_MASK));
+
         if (regs->psw.mask & PSW_MASK_PSTATE) {
                 struct task_struct *tsk = current;

@@ -318,6 +328,7 @@ static void inline do_trap(long interrup
                 else
                         die(str, regs, interruption_code);
         }
+       TRACE_TRAP_EXIT();
 }

 static inline void *get_check_address(struct pt_regs *regs)
@@ -415,6 +426,8 @@ asmlinkage void illegal_op(struct pt_reg
 {
         __u8 opcode[6];
        __u16 *location;
+        trapid_t ltt_interruption_code;
+        char * ic_ptr = (char *) &ltt_interruption_code;
        int signal = 0;

        location = (__u16 *)(regs->psw.addr-S390_lowcore.pgm_ilc);
@@ -426,6 +439,10 @@ asmlinkage void illegal_op(struct pt_reg
        if (regs->psw.mask & PSW_MASK_PSTATE)
                local_irq_enable();

+       memset(&ltt_interruption_code,0,sizeof(ltt_interruption_code));
+       memcpy(ic_ptr+4,&interruption_code,sizeof(interruption_code));
+       TRACE_TRAP_ENTRY(ltt_interruption_code, (regs->psw.addr & PSW_ADDR_MASK));
+
        if (regs->psw.mask & PSW_MASK_PSTATE)
                get_user(*((__u16 *) opcode), location);
        else
@@ -466,6 +483,7 @@ asmlinkage void illegal_op(struct pt_reg
         else if (signal)
                do_trap(interruption_code, signal,
                        "illegal operation", regs, NULL);
+        TRACE_TRAP_EXIT();
 }

@@ -476,6 +494,8 @@ specification_exception(struct pt_regs *
 {
         __u8 opcode[6];
        __u16 *location = NULL;
+        trapid_t ltt_interruption_code;
+        char * ic_ptr = (char *) &ltt_interruption_code;
        int signal = 0;

        location = (__u16 *) get_check_address(regs);
@@ -486,6 +506,10 @@ specification_exception(struct pt_regs *
         */
        if (regs->psw.mask & PSW_MASK_PSTATE)
                local_irq_enable();
+
+        memset(&ltt_interruption_code,0,sizeof(ltt_interruption_code));
+        memcpy(ic_ptr+4,&interruption_code,sizeof(interruption_code));
+        TRACE_TRAP_ENTRY(ltt_interruption_code, (regs->psw.addr & PSW_ADDR_MASK));

         if (regs->psw.mask & PSW_MASK_PSTATE) {
                get_user(*((__u16 *) opcode), location);
@@ -530,6 +554,7 @@ specification_exception(struct pt_regs *
                do_trap(interruption_code, signal,
                        "specification exception", regs, &info);
        }
+        TRACE_TRAP_EXIT();
 }
 #else
 DO_ERROR_INFO(SIGILL, "specification exception", specification_exception,
@@ -539,6 +564,8 @@ DO_ERROR_INFO(SIGILL, "specification exc
 asmlinkage void data_exception(struct pt_regs * regs, long interruption_code)
 {
        __u16 *location;
+        trapid_t ltt_interruption_code;
+        char * ic_ptr = (char *) &ltt_interruption_code;
        int signal = 0;

        location = (__u16 *) get_check_address(regs);
@@ -550,6 +577,10 @@ asmlinkage void data_exception(struct pt
        if (regs->psw.mask & PSW_MASK_PSTATE)
                local_irq_enable();

+       memset(&ltt_interruption_code,0,sizeof(ltt_interruption_code));
+       memcpy(ic_ptr+4,&interruption_code,sizeof(interruption_code));
+       TRACE_TRAP_ENTRY(ltt_interruption_code, (regs->psw.addr & PSW_ADDR_MASK));
+
        if (MACHINE_HAS_IEEE)
                __asm__ volatile ("stfpc %0\n\t"
                                  : "=m" (current->thread.fp_regs.fpc));
@@ -625,6 +656,7 @@ asmlinkage void data_exception(struct pt
                do_trap(interruption_code, signal,
                        "data exception", regs, &info);
        }
+        TRACE_TRAP_EXIT();
 }

@@ -679,6 +711,11 @@ void __init trap_init(void)

 void handle_per_exception(struct pt_regs *regs)
 {
+       trapid_t ltt_interruption_code;
+       char * ic_ptr = (char *) &ltt_interruption_code;
+       memset(&ltt_interruption_code,0,sizeof(ltt_interruption_code));
+       memcpy(ic_ptr+6,&S390_lowcore.pgm_code,2); /* copy the interrupt code */
+       TRACE_TRAP_ENTRY(ltt_interruption_code,(regs->psw.addr & PSW_ADDR_MASK));
        if (regs->psw.mask & PSW_MASK_PSTATE) {
                per_struct *per_info=&current->thread.per_info;
                per_info->lowcore.words.perc_atmid=S390_lowcore.per_perc_atmid;
@@ -693,5 +730,85 @@ void handle_per_exception(struct pt_regs
                /* Hopefully switching off per tracing will help us survive */
                regs->psw.mask &= ~PSW_MASK_PER;
        }
+        TRACE_TRAP_EXIT();
+}
+
+#if (CONFIG_TRACE)
+asmlinkage void trace_real_syscall_entry(struct pt_regs *regs)
+{
+       int use_depth;
+       int use_bounds;
+       int depth = 0;
+       int seek_depth;
+       unsigned long lower_bound;
+       unsigned long upper_bound;
+       unsigned long addr;
+       unsigned long *stack;
+       unsigned long temp_stack;
+       trace_syscall_entry trace_syscall_event;
+       /* Set the syscall ID                               */
+       /* Register 8 is setup just prior to the call       */
+       /* This instruction is just following linkage       */
+       /* so it's ok.  If moved and chance of R8 being     */
+       /* clobbered, would need to dig it out of the stack */
+       __asm__ volatile ("  stc  8,%0\n\t"
+                         :"=m" (trace_syscall_event.syscall_id));
+       /* get the psw address */
+       trace_syscall_event.address = regs->psw.addr;
+       /* and off the hi-order bit */
+       trace_syscall_event.address &= PSW_ADDR_MASK;
+       if (!(user_mode(regs))) /* if kernel mode, return */
+               goto trace_syscall_end;
+       /* Get the trace configuration - if none, return */
+       if (trace_get_config(&use_depth,
+                            &use_bounds,
+                            &seek_depth,
+                            (void *) &lower_bound,
+                            (void *) &upper_bound) < 0)
+               goto trace_syscall_end;
+       /* Do we have to search for an instruction pointer address range */
+       if ((use_depth == 1) || (use_bounds == 1)) {
+               /* Start at the top of the stack */
+               /* stack pointer is register 15 */
+               stack = (unsigned long *) regs->gprs[15];    /* stack pointer */
+               /* Keep on going until we reach the end of the process' stack limit */
+               do {
+                       get_user(addr, stack + 14);     /* get the program address +0x38 */
+                       /* and off the hi-order bit */
+                       addr &= PSW_ADDR_MASK;
+                       /* Does this LOOK LIKE an address in the program */
+                       if ((addr > current->mm->start_code)
+                           && (addr < current->mm->end_code)) {
+                               /* Does this address fit the description */
+                               if (((use_depth == 1) && (depth == seek_depth))
+                                   || ((use_bounds == 1) && (addr > lower_bound)
+                                       && (addr < upper_bound))) {
+                                       /* Set the address */
+                                       trace_syscall_event.address = addr;
+                                       /* We're done */
+                                       goto trace_syscall_end;
+                               } else
+                                       /* We're one depth more */
+                                       depth++;
+                       }
+                       /* Go on to the next address */
+                       get_user(temp_stack, stack);    /* get contents of stack */
+                       temp_stack &= PSW_ADDR_MASK;        /* and off hi order bit */
+                       stack = (unsigned long *) temp_stack;   /* move into stack */
+                       /* stack may or may not go to zero when end hit               */
+                       /* using 0x7fffffff-_STK_LIM to validate that the address is  */
+                       /* within the range of a valid stack address                  */
+                       /* If outside that range, exit the loop, stack end must have  */
+                       /* been hit.                                                  */
+               } while (stack >= (unsigned long *) (0x7fffffff - _STK_LIM));
+       }
+trace_syscall_end:
+       /* Trace the event */
+       trace_event(TRACE_EV_SYSCALL_ENTRY, &trace_syscall_event);
+}
+asmlinkage void trace_real_syscall_exit(void)
+{
+       trace_event(TRACE_EV_SYSCALL_EXIT, NULL);
 }

+#endif                         /* (CONFIG_TRACE) */
diff -urpN linux-2.5.44-bk2/arch/s390/mm/fault.c linux-2.5.44-bk2-ltt/arch/s390/mm/fault.c
--- linux-2.5.44-bk2/arch/s390/mm/fault.c       Sat Oct 19 00:02:28 2002
+++ linux-2.5.44-bk2-ltt/arch/s390/mm/fault.c   Tue Oct 29 15:24:18 2002
@@ -5,6 +5,7 @@
  *    Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
  *    Author(s): Hartmut Penner (h...@de.ibm.com)
  *               Ulrich Weigand (uweig...@de.ibm.com)
+ *  Portions added by T. Halloran: (C) Copyright 2002 IBM Poughkeepsie, IBM Corporation
  *
  *  Derived from "arch/i386/mm/fault.c"
  *    Copyright (C) 1995  Linus Torvalds
@@ -25,6 +26,7 @@
 #include <linux/compatmac.h>
 #include <linux/init.h>
 #include <linux/console.h>
+#include <linux/trace.h>

 #include <asm/system.h>
 #include <asm/uaccess.h>
@@ -153,6 +155,8 @@ extern inline void do_exception(struct p
        int user_address;
         unsigned long fixup;
        int si_code = SEGV_MAPERR;
+        trapid_t ltt_interruption_code;                
+        char * ic_ptr = (char *) &ltt_interruption_code;

         tsk = current;
         mm = tsk->mm;
@@ -200,6 +204,9 @@ extern inline void do_exception(struct p
         */
        local_irq_enable();

+        memset(&ltt_interruption_code,0,sizeof(ltt_interruption_code));
+        memcpy(ic_ptr+4,&error_code,sizeof(error_code));
+        TRACE_TRAP_ENTRY(ltt_interruption_code,(regs->psw.addr & PSW_ADDR_MASK));
         down_read(&mm->mmap_sem);

         vma = find_vma(mm, address);
@@ -248,6 +255,7 @@ survive:
        }

         up_read(&mm->mmap_sem);
+        TRACE_TRAP_EXIT();
         return;

 /*
@@ -262,6 +270,7 @@ bad_area:
                 tsk->thread.prot_addr = address;
                 tsk->thread.trap_no = error_code;
                force_sigsegv(regs, error_code, si_code, address);
+                TRACE_TRAP_EXIT();
                 return;
        }

@@ -269,6 +278,7 @@ no_context:
         /* Are we prepared to handle this kernel fault?  */
         if ((fixup = search_exception_table(regs->psw.addr)) != 0) {
                 regs->psw.addr = fixup;
+                TRACE_TRAP_EXIT();
                 return;
         }

@@ -316,6 +326,8 @@ do_sigbus:
        /* Kernel mode? Handle exceptions or die */
        if (!(regs->psw.mask & PSW_MASK_PSTATE))
                goto no_context;
+
+       TRACE_TRAP_EXIT();
 }

 void do_protection_exception(struct pt_regs *regs, unsigned long error_code)
diff -urpN linux-2.5.44-bk2/drivers/s390/cio/cio.c linux-2.5.44-bk2-ltt/drivers/s390/cio/cio.c
--- linux-2.5.44-bk2/drivers/s390/cio/cio.c     Sat Oct 19 00:02:34 2002
+++ linux-2.5.44-bk2-ltt/drivers/s390/cio/cio.c Tue Oct 29 15:24:18 2002
@@ -18,6 +18,8 @@
 #include <linux/config.h>
 #include <linux/slab.h>

+#include <linux/trace.h>
+
 #include <asm/io.h>
 #include <asm/irq.h>
 #include <asm/delay.h>
@@ -945,9 +947,11 @@ do_IRQ (struct pt_regs regs)
                        }

                        irq_enter ();
+                       TRACE_IRQ_ENTRY(irq, !(((regs).psw.mask&PSW_PROBLEM_STATE) != 0));
                        s390irq_spin_lock (irq);
                        s390_process_IRQ (irq);
                        s390irq_spin_unlock (irq);
+                       TRACE_IRQ_EXIT();
                        irq_exit ();
                }

diff -urpN linux-2.5.44-bk2/drivers/s390/s390mach.c linux-2.5.44-bk2-ltt/drivers/s390/s390mach.c
--- linux-2.5.44-bk2/drivers/s390/s390mach.c    Sat Oct 19 00:02:34 2002
+++ linux-2.5.44-bk2-ltt/drivers/s390/s390mach.c        Tue Oct 29 15:24:18 2002
@@ -5,6 +5,7 @@
  *  S390 version
  *    Copyright (C) 2000 IBM Deutschland Entwicklung GmbH, IBM Corporation
  *    Author(s): Ingo Adlung (adl...@de.ibm.com)
+ *  Portions added by T. Halloran: (C) Copyright 2002 IBM Poughkeepsie, IBM Corporation
  */

 #include <linux/config.h>
@@ -12,6 +13,7 @@
 #include <linux/init.h>
 #include <linux/slab.h>
 #include <linux/completion.h>
+#include <linux/trace.h>
 #ifdef CONFIG_SMP
 #include <linux/smp.h>
 #endif
@@ -179,10 +181,20 @@ s390_do_machine_check(void)
 {
        int crw_count;
        mcic_t mcic;
+        trapid_t ltt_interruption_code;
+        uint32_t ltt_old_psw;

        DBG(KERN_INFO "s390_do_machine_check : starting ...\n");

        memcpy(&mcic, &S390_lowcore.mcck_interruption_code, sizeof (__u64));
+       memcpy( &ltt_interruption_code,
+               &S390_lowcore.mcck_interruption_code,
+               sizeof(__u64));
+       memcpy( &ltt_old_psw,
+               &S390_lowcore.mcck_old_psw,
+               sizeof(uint32_t));
+        ltt_old_psw &=  PSW_ADDR_MASK;
+        TRACE_TRAP_ENTRY(ltt_interruption_code,ltt_old_psw);

        if (mcic.mcc.mcd.sd)    /* system damage */
                s390_handle_damage("received system damage machine check\n");
diff -urpN linux-2.5.44-bk2/include/asm-s390/trace.h linux-2.5.44-bk2-ltt/include/asm-s390/trace.h
--- linux-2.5.44-bk2/include/asm-s390/trace.h   Wed Dec 31 19:00:00 1969
+++ linux-2.5.44-bk2-ltt/include/asm-s390/trace.h       Tue Oct 29 15:24:18 2002
@@ -0,0 +1,16 @@
+/*
+ * linux/include/asm-s390/trace.h
+ *
+ * Copyright (C) 2002, Karim Yaghmour
+ *
+ * S/390 definitions for tracing system
+ */
+
+#include <linux/trace.h>
+#include <asm-generic/trace.h>
+
+/* Current arch type */
+#define TRACE_ARCH_TYPE TRACE_ARCH_TYPE_S390
+
+/* Current variant type */
+#define TRACE_ARCH_VARIANT TRACE_ARCH_VARIANT_NONE
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/