Big problems in acpi/event.c (needs seq_printf?)

From: Pavel Machek
Date: Mon Apr 19 2004 - 11:54:35 EST


acpi/event.c needs some heavy facelifting.

This one is needed so that return value (ERESTARTSYS) is propagated
from bus_receive_event(). If this is not done, suspend/resume causes
cat /proc/acpi/event to fail with -EIO. Please apply.

However, there are more problems with event.c:

* it uses spinlock_irq to protect event_is_open(). That's huge
overkill, atomic_t should be enough here, and it is never accessed
from interrupt anyway.

* its racy w.r.t. concurent readers (and yes, you *can* have concurent
readers, think threads). acpi_system_read_event() uses static
variables, without any locking. Probably seq_printf() converstion is

--- clean/drivers/acpi/event.c 2003-02-15 18:51:16.000000000 +0100
+++ linux/drivers/acpi/event.c 2004-04-19 18:43:58.000000000 +0200
@@ -63,9 +58,8 @@

result = acpi_bus_receive_event(&event);
- if (result) {
- return_VALUE(-EIO);
- }
+ if (result)
+ return_VALUE(result);

chars_remaining = sprintf(str, "%s %s %08x %08x\n",

When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at