31  * When the operating system detects that it is in an invalid state, a panic
34  * a system panic: recording information about the panic in memory (and thus
37  * system to be in one of four states with respect to the panic code:
39  * CALM    - the state of the system prior to any thread initiating a panic
42  *           a system panic records information about the cause of the panic
52  * The panic code itself must be re-entrant because actions taken during any
53  * state may lead to another system panic.  Additionally, any Solaris
54  * thread may initiate a panic at any time, and so we must have synchronization
56  * The panic code makes use of a special locking primitive, a trigger, to
59  * each transition between the four states.  When a thread enters the panic
65  * presumes a panic is underway and simply spins.  Unfortunately, since we
68  * There are two common reasons for re-entering the panic code once a panic
71  * routine will invoke panic(); (2) if the clock routine decides that sync
72  * or dump is not making progress, it will invoke panic() to force a timeout.
74  * unrelated panic while sync or dump is still underway, is extremely unlikely.
79  * The panic code is entered via a call directly to the vpanic() function,
80  * or its varargs wrappers panic() and cmn_err(9F).  The vpanic routine
84  * the common panicsys() routine.  The first thread to initiate a panic
86  * the panic code itself does not overwrite valuable data on that thread's
96  * Every thread which initiates a panic has its T_PANIC flag set so we can
101  * reboots.  We allow the panic_thread to store its register set and panic
104  * state at the time of the panic.  If we do get a dump, the panic buffer
106  * therein (see <sys/panic.h>).
119  * indicate a soft c, and thus the verb and adjective forms of 'panic' are
121  * the ill-conceived "panicing" and "panic'd" is discouraged.
147 #include <sys/panic.h>
153  * Panic variables which are set once during the QUIESCE state by the
154  * first thread to initiate a panic.  These are examined by post-mortem
155  * debugging tools; the inconsistent use of 'panic' versus 'panic_' in
163 label_t panic_pcb;			/* t_pcb at time of panic */
167 clock_t panic_lbolt;			/* lbolt at time of panic */
168 int64_t panic_lbolt64;			/* lbolt64 at time of panic */
169 hrtime_t panic_hrtime;			/* hrtime at time of panic */
170 timespec_t panic_hrestime;		/* hrestime at time of panic */
171 int panic_ipl;				/* ipl on panic_cpu at time of panic */
178  * Panic variables which can be set via /etc/system or patched while
181 char *panic_bootstr = NULL;		/* mdboot string to use after panic */
182 int panic_bootfcn = AD_BOOT;		/* mdboot function to use after panic */
188  * The do_polled_io flag is set by the panic code to inform the SCSI subsystem
195  * panic subsystem that it should not attempt an initial debug_enter.
200  * Triggers for panic state transitions:
258 	 * Record one-time panic information and quiesce the other CPUs.  in panicsys()
259 	 * Then print out the panic message and stack trace.  in panicsys()
332 		 * A hypervisor panic originates outside of Solaris, so we  in panicsys()
333 		 * don't want to prepend the panic message with misleading  in panicsys()
358 				debug_enter("panic: entering debugger "  in panicsys()
361 				debug_enter("panic: entering debugger "  in panicsys()
376 	 * will re-enter panic if we are not making progress with dump.  in panicsys()
389 		debug_enter("panic: entering debugger (continue to reboot)");  in panicsys()
408 panic(const char *format, ...)  in panic()  function