[PATCH] fix ppc32 sys_swapcontext

[Paul Mackerras]

The patch below fixes a bug in the swapcontext system call on ppc32.
On ppc32, the system call entry only saves the volatile registers,
except in the case of a few system calls (e.g. fork) which need all
the registers saved.  Swapcontext needs all the registers but we
weren't saving them all.  This patch fixes that.

Please apply.

Thanks,
Paul.

diff -urN pmac-2.5/arch/ppc/kernel/entry.S g5-ppc64/arch/ppc/kernel/entry.S
--- pmac-2.5/arch/ppc/kernel/entry.S	2004-02-16 08:19:48.000000000 +1100
+++ g5-ppc64/arch/ppc/kernel/entry.S	2004-03-24 09:47:12.000000000 +1100
@@ -414,6 +414,14 @@
 	stw	r0,TRAP(r1)		/* register set saved */
 	b	sys_clone
 
+	.globl	ppc_swapcontext
+ppc_swapcontext:
+	SAVE_NVGPRS(r1)
+	lwz	r0,TRAP(r1)
+	rlwinm	r0,r0,0,0,30		/* clear LSB to indicate full */
+	stw	r0,TRAP(r1)		/* register set saved */
+	b	sys_swapcontext
+
 /*
  * This routine switches between two different tasks.  The process
  * state of one is saved on its kernel stack.  Then the state
diff -urN pmac-2.5/arch/ppc/kernel/misc.S g5-ppc64/arch/ppc/kernel/misc.S
--- pmac-2.5/arch/ppc/kernel/misc.S	2004-03-16 20:11:56.000000000 +1100
+++ g5-ppc64/arch/ppc/kernel/misc.S	2004-03-24 09:46:40.000000000 +1100
@@ -1363,7 +1363,7 @@
 	.long sys_clock_gettime
 	.long sys_clock_getres
 	.long sys_clock_nanosleep
-	.long sys_swapcontext
+	.long ppc_swapcontext
 	.long sys_tgkill	/* 250 */
 	.long sys_utimes
 	.long sys_statfs64
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/