Skip to contentSkip to conversation

amiserial: Prepare for locking relaxation in caller.

PATCH In review Alan Cox 2 participants 3 messages Feb 20, 2008 – Feb 23, 2008 Char and misc drivers

lore
Share & tools
lore

Alan Cox

author
#0 lore raw

PATCH 1/1 amiserial: Prepare for locking relaxation in caller.

Just wrap this one in a lock_kernel. As I understand it there is no M68K SMP anyway.

drivers/char/amiserial.c +11 -2
diff -u --new-file --recursive --exclude-from /usr/src/exclude linux.vanilla-2.6.25-rc2-mm1/drivers/char/amiserial.c linux-2.6.25-rc2-mm1/drivers/char/amiserial.c
@@ -1074,6 +1074,7 @@
 	if (!retinfo)
 		return -EFAULT;
 	memset(&tmp, 0, sizeof(tmp));
+	lock_kernel();
 	tmp.type = state->type;
 	tmp.line = state->line;
 	tmp.port = state->port;
@@ -1084,6 +1085,7 @@
 	tmp.close_delay = state->close_delay;
 	tmp.closing_wait = state->closing_wait;
 	tmp.custom_divisor = state->custom_divisor;
+	unlock_kernel();
 	if (copy_to_user(retinfo,&tmp,sizeof(*retinfo)))
 		return -EFAULT;
 	return 0;
@@ -1099,13 +1101,17 @@
 
 	if (copy_from_user(&new_serial,new_info,sizeof(new_serial)))
 		return -EFAULT;
+		
+	lock_kernel();
 	state = info->state;
 	old_state = *state;
   
 	change_irq = new_serial.irq != state->irq;
 	change_port = (new_serial.port != state->port);
-	if(change_irq || change_port || (new_serial.xmit_fifo_size != state->xmit_fifo_size))
+	if(change_irq || change_port || (new_serial.xmit_fifo_size != state->xmit_fifo_size)) {
+	  unlock_kernel();
 	  return -EINVAL;
+	}
   
 	if (!serial_isroot()) {
 		if ((new_serial.baud_base != state->baud_base) ||
@@ -1122,8 +1128,10 @@
 		goto check_and_exit;
 	}
 
-	if (new_serial.baud_base < 9600)
+	if (new_serial.baud_base < 9600) {
+		unlock_kernel();
 		return -EINVAL;
+	}
 
 	/*
 	 * OK, past this point, all the error checking has been done.
@@ -1157,6 +1165,7 @@
 		}
 	} else
 		retval = startup(info);
+	unlock_kernel();
 	return retval;
 }
 
#1 lore raw
Quoted 4 lines · On Wed, 20 Feb 2008 20:11:01 +0000 Alan Cox <[email protected]> wrote:

Just wrap this one in a lock_kernel. As I understand it there is no M68K SMP anyway.

I assume that all the bkl/locking patches are prerequisites for tty-bkl-pushdown.patch, and that they don't depend upon tty-bkl-pushdown.patch?

author
#2 lore raw
Quoted 10 lines · On Sat, 23 Feb 2008 00:05:25 -0800 Andrew Morton <[email protected]>…
Quoted 4 lines · On Wed, 20 Feb 2008 20:11:01 +0000 Alan Cox <[email protected]> wrote:

Just wrap this one in a lock_kernel. As I understand it there is no M68K SMP anyway.

I assume that all the bkl/locking patches are prerequisites for tty-bkl-pushdown.patch, and that they don't depend upon tty-bkl-pushdown.patch?

They are all prequisites for some changes I've yet to send you and will do once I'm sure I've not upset anything in the meantime. There should be no interdependancies although I've obviously not tested every possible combination

Alan