[PATCH] ppc64: Fix RTAS races on pSeries

The low level kernel interface to RTAS (the firmware runtime services)
was plagued with races that could cause from bogus results of RTAS
operations to total machine crashes in some circumstances. This patch
fix the ones I could identify, hoping I didn't miss any. I also added
a WARN_ON (well, it's asm equivalent) to enter_rtas to make sure we
never _ever_ try to call that with interrupts enabled.

arch/ppc64/kernel/entry.S

@@ -512,12 +512,25 @@ _GLOBAL(enter_rtas)
 	mfsrr1	r10
 	std	r10,_SRR1(r1)
 
+	/* There is no way it is acceptable to get here with interrupts enabled,
+	 * check it with the asm equivalent of WARN_ON
+	 */
+	mfmsr	r6
+	andi.	r0,r6,MSR_EE
+1:	tdnei	r0,0
+.section __bug_table,"a"
+	.llong	1b,__LINE__ + 0x1000000, 1f, 2f
+.previous
+.section .rodata,"a"
+1:	.asciz	__FILE__
+2:	.asciz "enter_rtas"
+.previous
+	
 	/* Unfortunately, the stack pointer and the MSR are also clobbered,
 	 * so they are saved in the PACA which allows us to restore
 	 * our original state after RTAS returns.
          */
 	std	r1,PACAR1(r13)
-	mfmsr	r6
         std	r6,PACASAVEDMSR(r13)
 
 	/* Setup our real return addr */	

arch/ppc64/kernel/rtas.c

@@ -68,15 +68,20 @@ char rtas_data_buf[RTAS_DATA_BUF_SIZE]__page_aligned;
 void
 call_rtas_display_status(char c)
 {
-	struct rtas_args *rtas = &(get_paca()->xRtas);
+	struct rtas_args *args = &(get_paca()->xRtas);
+	unsigned long s;
+
+	spin_lock_irqsave(&rtas.lock, s);
 
-	rtas->token = 10;
-	rtas->nargs = 1;
-	rtas->nret  = 1;
-	rtas->rets  = (rtas_arg_t *)&(rtas->args[1]);
-	rtas->args[0] = (int)c;
+	args->token = 10;
+	args->nargs = 1;
+	args->nret  = 1;
+	args->rets  = (rtas_arg_t *)&(args->args[1]);
+	args->args[0] = (int)c;
 
-	enter_rtas((void *)__pa((unsigned long)rtas));	
+	enter_rtas((void *)__pa((unsigned long)args));	
+
+	spin_unlock_irqrestore(&rtas.lock, s);
 }
 
 int
@@ -91,8 +96,9 @@ rtas_token(const char *service)
 	return tokp ? *tokp : RTAS_UNKNOWN_SERVICE;
 }
 
-void
-log_rtas_error(struct rtas_args	*rtas_args)
+
+static int
+__log_rtas_error(struct rtas_args *rtas_args)
 {
 	struct rtas_args err_args, temp_args;
 
@@ -113,11 +119,22 @@ log_rtas_error(struct rtas_args	*rtas_args)
 	enter_rtas((void *)__pa((unsigned long)&get_paca()->xRtas));
 	PPCDBG(PPCDBG_RTAS, "\treturned from rtas ...\n");	
 
-
 	err_args = get_paca()->xRtas;
 	get_paca()->xRtas = temp_args;
 
-	if (err_args.rets[0] == 0)
+	return err_args.rets[0];
+}
+
+void
+log_rtas_error(struct rtas_args	*rtas_args)
+{
+	unsigned long s;
+	int rc;
+
+	spin_lock_irqsave(&rtas.lock, s);
+	rc = __log_rtas_error(rtas_args);
+	spin_unlock_irqrestore(&rtas.lock, s);
+	if (rc == 0)
 		log_error(rtas_err_buf, ERR_TYPE_RTAS_LOG, 0);
 }
 
@@ -126,9 +143,10 @@ rtas_call(int token, int nargs, int nret,
 	  unsigned long *outputs, ...)
 {
 	va_list list;
-	int i;
+	int i, logit = 0;
 	unsigned long s;
 	struct rtas_args *rtas_args = &(get_paca()->xRtas);
+	long ret;
 
 	PPCDBG(PPCDBG_RTAS, "Entering rtas_call\n");
 	PPCDBG(PPCDBG_RTAS, "\ttoken    = 0x%x\n", token);
@@ -138,6 +156,9 @@ rtas_call(int token, int nargs, int nret,
 	if (token == RTAS_UNKNOWN_SERVICE)
 		return -1;
 
+	/* Gotta do something different here, use global lock for now... */
+	spin_lock_irqsave(&rtas.lock, s);
+
 	rtas_args->token = token;
 	rtas_args->nargs = nargs;
 	rtas_args->nret  = nret;
@@ -152,24 +173,14 @@ rtas_call(int token, int nargs, int nret,
 	for (i = 0; i < nret; ++i)
 		rtas_args->rets[i] = 0;
 
-#if 0   /* Gotta do something different here, use global lock for now... */
-	spin_lock_irqsave(&rtas_args->lock, s);
-#else
-	spin_lock_irqsave(&rtas.lock, s);
-#endif
 	PPCDBG(PPCDBG_RTAS, "\tentering rtas with 0x%lx\n",
 		(void *)__pa((unsigned long)rtas_args));
 	enter_rtas((void *)__pa((unsigned long)rtas_args));
 	PPCDBG(PPCDBG_RTAS, "\treturned from rtas ...\n");
 
 	if (rtas_args->rets[0] == -1)
-		log_rtas_error(rtas_args);
+		logit = (__log_rtas_error(rtas_args) == 0);
 
-#if 0   /* Gotta do something different here, use global lock for now... */
-	spin_unlock_irqrestore(&rtas_args->lock, s);
-#else
-	spin_unlock_irqrestore(&rtas.lock, s);
-#endif
 	ifppcdebug(PPCDBG_RTAS) {
 		for(i=0; i < nret ;i++)
 			udbg_printf("\tnret[%d] = 0x%lx\n", i, (ulong)rtas_args->rets[i]);
@@ -178,7 +189,15 @@ rtas_call(int token, int nargs, int nret,
 	if (nret > 1 && outputs != NULL)
 		for (i = 0; i < nret-1; ++i)
 			outputs[i] = rtas_args->rets[i+1];
-	return (ulong)((nret > 0) ? rtas_args->rets[0] : 0);
+	ret = (ulong)((nret > 0) ? rtas_args->rets[0] : 0);
+
+	/* Gotta do something different here, use global lock for now... */
+	spin_unlock_irqrestore(&rtas.lock, s);
+
+	if (logit)
+		log_error(rtas_err_buf, ERR_TYPE_RTAS_LOG, 0);
+
+	return ret;
 }
 
 /* Given an RTAS status code of 990n compute the hinted delay of 10^n
@@ -464,12 +483,12 @@ asmlinkage int ppc_rtas(struct rtas_args __user *uargs)
 	enter_rtas((void *)__pa((unsigned long)&get_paca()->xRtas));
 	args = get_paca()->xRtas;
 
+	spin_unlock_irqrestore(&rtas.lock, flags);
+
 	args.rets  = (rtas_arg_t *)&(args.args[nargs]);
 	if (args.rets[0] == -1)
 		log_rtas_error(&args);
 
-	spin_unlock_irqrestore(&rtas.lock, flags);
-
 	/* Copy out args. */
 	if (copy_to_user(uargs->args + nargs,
 			 args.args + nargs,
@@ -485,7 +504,9 @@ asmlinkage int ppc_rtas(struct rtas_args __user *uargs)
 void rtas_stop_self(void)
 {
 	struct rtas_args *rtas_args = &(get_paca()->xRtas);
+	unsigned long s;
 
+	spin_lock_irqsave(&rtas.lock, s);
 	rtas_args->token = rtas_token("stop-self");
 	BUG_ON(rtas_args->token == RTAS_UNKNOWN_SERVICE);
 	rtas_args->nargs = 0;
@@ -495,6 +516,8 @@ void rtas_stop_self(void)
 	printk("%u %u Ready to die...\n",
 	       smp_processor_id(), hard_smp_processor_id());
 	enter_rtas((void *)__pa(rtas_args));
+	spin_unlock_irqrestore(&rtas.lock, s);
+
 	panic("Alas, I survived.\n");
 }
 #endif /* CONFIG_HOTPLUG_CPU */