* [PATCH] toshiba_acpi: Fix machines that don't support HCI_SYSTEM_EVENT
@ 2011-11-16 23:37 Seth Forshee
2011-11-18 21:20 ` Azael Avalos
0 siblings, 1 reply; 3+ messages in thread
From: Seth Forshee @ 2011-11-16 23:37 UTC (permalink / raw)
To: Matthew Garrett; +Cc: platform-driver-x86, linux-kernel
The Satellite C670-10V generates notifications for hotkeys but does
not support HCI_SYSTEM_EVENT. As a result when a hotkey is pressed
it gets stuck in an infinite loop in toshiba_acpi_notify. To fix
this, detect whether or not HCI_SYSTEM_EVENT is supported up-front
and don't try to read system events if it isn't supported. In
addition, limit the number of retries when reading HCI_SYSTEM_EVENT
fails so that this loop cannot run unbounded.
Signed-off-by: Seth Forshee <seth.forshee@canonical.com>
---
drivers/platform/x86/toshiba_acpi.c | 21 ++++++++++++++++-----
1 files changed, 16 insertions(+), 5 deletions(-)
diff --git a/drivers/platform/x86/toshiba_acpi.c b/drivers/platform/x86/toshiba_acpi.c
index 13ef8c3..dcdc1f4 100644
--- a/drivers/platform/x86/toshiba_acpi.c
+++ b/drivers/platform/x86/toshiba_acpi.c
@@ -121,6 +121,7 @@ struct toshiba_acpi_dev {
int illumination_supported:1;
int video_supported:1;
int fan_supported:1;
+ int system_event_supported:1;
struct mutex mutex;
};
@@ -724,7 +725,7 @@ static int keys_proc_show(struct seq_file *m, void *v)
u32 hci_result;
u32 value;
- if (!dev->key_event_valid) {
+ if (!dev->key_event_valid && dev->system_event_supported) {
hci_read1(dev, HCI_SYSTEM_EVENT, &value, &hci_result);
if (hci_result == HCI_SUCCESS) {
dev->key_event_valid = 1;
@@ -964,6 +965,8 @@ static int __devinit toshiba_acpi_add(struct acpi_device *acpi_dev)
/* enable event fifo */
hci_write1(dev, HCI_SYSTEM_EVENT, 1, &hci_result);
+ if (hci_result == HCI_SUCCESS)
+ dev->system_event_supported = 1;
props.type = BACKLIGHT_PLATFORM;
props.max_brightness = HCI_LCD_BRIGHTNESS_LEVELS - 1;
@@ -1032,12 +1035,15 @@ static void toshiba_acpi_notify(struct acpi_device *acpi_dev, u32 event)
{
struct toshiba_acpi_dev *dev = acpi_driver_data(acpi_dev);
u32 hci_result, value;
+ int retries = 3;
- if (event != 0x80)
+ if (!dev->system_event_supported || event != 0x80)
return;
+
do {
hci_read1(dev, HCI_SYSTEM_EVENT, &value, &hci_result);
- if (hci_result == HCI_SUCCESS) {
+ switch (hci_result) {
+ case HCI_SUCCESS:
if (value == 0x100)
continue;
/* act on key press; ignore key release */
@@ -1049,14 +1055,19 @@ static void toshiba_acpi_notify(struct acpi_device *acpi_dev, u32 event)
pr_info("Unknown key %x\n",
value);
}
- } else if (hci_result == HCI_NOT_SUPPORTED) {
+ break;
+ case HCI_NOT_SUPPORTED:
/* This is a workaround for an unresolved issue on
* some machines where system events sporadically
* become disabled. */
hci_write1(dev, HCI_SYSTEM_EVENT, 1, &hci_result);
pr_notice("Re-enabled hotkeys\n");
+ /* fall through */
+ default:
+ retries--;
+ break;
}
- } while (hci_result != HCI_EMPTY);
+ } while (retries && hci_result != HCI_EMPTY);
}
--
1.7.5.4
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] toshiba_acpi: Fix machines that don't support HCI_SYSTEM_EVENT
2011-11-16 23:37 [PATCH] toshiba_acpi: Fix machines that don't support HCI_SYSTEM_EVENT Seth Forshee
@ 2011-11-18 21:20 ` Azael Avalos
2011-11-18 21:28 ` Matthew Garrett
0 siblings, 1 reply; 3+ messages in thread
From: Azael Avalos @ 2011-11-18 21:20 UTC (permalink / raw)
To: Seth Forshee; +Cc: Matthew Garrett, platform-driver-x86, linux-kernel
Qosmio X500-X505 series are also affected by this
"infinite loop", this patch fixes the problem on these
laptops.
Tested-by: Azael Avalos <coproscefalo@gmail.com>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] toshiba_acpi: Fix machines that don't support HCI_SYSTEM_EVENT
2011-11-18 21:20 ` Azael Avalos
@ 2011-11-18 21:28 ` Matthew Garrett
0 siblings, 0 replies; 3+ messages in thread
From: Matthew Garrett @ 2011-11-18 21:28 UTC (permalink / raw)
To: Azael Avalos; +Cc: Seth Forshee, platform-driver-x86, linux-kernel
Cool. I'll merge this for 3.2.
--
Matthew Garrett | mjg59@srcf.ucam.org
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2011-11-18 21:28 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-11-16 23:37 [PATCH] toshiba_acpi: Fix machines that don't support HCI_SYSTEM_EVENT Seth Forshee
2011-11-18 21:20 ` Azael Avalos
2011-11-18 21:28 ` Matthew Garrett
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox