Re: [PATCH] dell-wmi: Update code for processing WMI events

["Pali Rohár" <pali.rohar@gmail.com>]

[-- Attachment #1: Type: Text/Plain, Size: 9558 bytes --]

On Wednesday 22 October 2014 12:51:17 Pali Rohár wrote:
> On Tuesday 21 October 2014 23:32:12 Darren Hart wrote:
> > On Tue, Oct 21, 2014 at 12:15:24AM +0200, Pali Rohár wrote:
> > > WMI buffer can contains more events. First value in buffer
> > > is length of event followed by data of specified length.
> > > After that is next length and next data. When length is
> > > zero then there is no more events in bufffer.
> > > 
> > > This patch adds support for processing all events in
> > > buffer (not only first) and parse more event types (not
> > > only hotkey events). Because of variable length of events
> > > sometimes BIOS fills more hotkeys (or other values) into
> > > single WMI event. In this case this patch process also
> > > these multiple hotkeys (and not only first one).
> > > 
> > > Some event types are just ignored because kernel is not
> > > interested for them (e.g. NIC Link status, battery unplug,
> > > ...).
> > > 
> > > This patch is based on DSDT table from Dell Latitude
> > > E6440. Code should be backward compatible so will process
> > > other events of old types same as before this patch.
> > > 
> > > This patch also fixes problem when in kernel log are
> > > written messages about unknown WMI events. Now all know
> > > events are parsed and those which are not interesting for
> > > kernel are dropped without unknown message.
> > 
> > This should probably be done in a separate patch.
> 
> It is not possible, because my patch rewrite code for handling
> events. Kernel does not print "unknown event" messages when it
> parse WMI event and understand specified part.
> 
> > > Signed-off-by: Pali Rohár <pali.rohar@gmail.com>
> > > Tested-by: Pali Rohár <pali.rohar@gmail.com>
> > 
> > Well yes, I should hope so ;-)
> > 
> > > ---
> > > 
> > >  drivers/platform/x86/dell-wmi.c |  157
> > >  +++++++++++++++++++++++++++++++-------- 1 file changed,
> > >  127 insertions(+), 30 deletions(-)
> > > 
> > > diff --git a/drivers/platform/x86/dell-wmi.c
> > > b/drivers/platform/x86/dell-wmi.c index 25721bf..3d15949
> > > 100644
> > > --- a/drivers/platform/x86/dell-wmi.c
> > > +++ b/drivers/platform/x86/dell-wmi.c
> > > @@ -145,11 +145,35 @@ static const u16
> > > bios_to_linux_keycode[256] __initconst = {
> > > 
> > >  static struct input_dev *dell_wmi_input_dev;
> > > 
> > > +static void dell_wmi_process_key(int reported_key)
> > > +{
> > > +	const struct key_entry *key;
> > > +
> > > +	key =
> > > sparse_keymap_entry_from_scancode(dell_wmi_input_dev,
> > > +						reported_key);
> > > +	if (!key) {
> > > +		pr_info("Unknown key %x pressed\n", reported_key);
> > > +		return;
> > > +	}
> > > +
> > > +	pr_debug("Key %x pressed\n", reported_key);
> > > +
> > > +	/* Don't report brightness notifications that will also
> > > come via ACPI */ +	if ((key->keycode == KEY_BRIGHTNESSUP
> > > || +	     key->keycode == KEY_BRIGHTNESSDOWN) &&
> > > acpi_video) +		return;
> > > +
> > > +	sparse_keymap_report_entry(dell_wmi_input_dev, key, 1,
> > > true); +}
> > > +
> > > 
> > >  static void dell_wmi_notify(u32 value, void *context)
> > >  {
> > >  
> > >  	struct acpi_buffer response = { ACPI_ALLOCATE_BUFFER,
> > >  	NULL }; union acpi_object *obj;
> > >  	acpi_status status;
> > > 
> > > +	acpi_size buffer_size;
> > > +	u16 *buffer_entry, *buffer_end;
> > > +	int len, i;
> > > 
> > >  	status = wmi_get_event_data(value, &response);
> > >  	if (status != AE_OK) {
> > > 
> > > @@ -158,44 +182,117 @@ static void dell_wmi_notify(u32
> > > value, void *context)
> > > 
> > >  	}
> > >  	
> > >  	obj = (union acpi_object *)response.pointer;
> > > 
> > > +	if (!obj) {
> > > +		pr_info("no response\n");
> > > +		return;
> > > +	}
> > 
> > If you intend to print this, it should probably be a bit
> > more informative. Is "info" the right level here? I would
> > imagine either WARN if this was a bad thing, or DEBUG if
> > this is more for debugging the driver.
> 
> So what you (or somebody else) prefer? warn or debug?
> 
> > > -	if (obj && obj->type == ACPI_TYPE_BUFFER) {
> > > -		const struct key_entry *key;
> > > -		int reported_key;
> > > -		u16 *buffer_entry = (u16 *)obj->buffer.pointer;
> > > -		int buffer_size = obj->buffer.length/2;
> > > -
> > > -		if (buffer_size >= 2 && dell_new_hk_type &&
> > > buffer_entry[1] != 0x10) { -			pr_info("Received
> 
> unknown
> 
> > > WMI event (0x%x)\n",
> > > -				buffer_entry[1]);
> > > -			kfree(obj);
> > > -			return;
> > > -		}
> > > +	if (obj->type != ACPI_TYPE_BUFFER) {
> > > +		pr_info("bad response type %x\n", obj->type);
> > > +		kfree(obj);
> > > +		return;
> > > +	}
> > > +
> > > +	pr_debug("Received WMI event (%*ph)\n",
> > > +		obj->buffer.length, obj->buffer.pointer);
> > > 
> > > -		if (buffer_size >= 3 && (dell_new_hk_type ||
> > > buffer_entry[1] == 0x0)) -			reported_key =
> > > (int)buffer_entry[2];
> > > +	buffer_entry = (u16 *)obj->buffer.pointer;
> > > +	buffer_size = obj->buffer.length/2;
> > > +
> > > +	if (!dell_new_hk_type) {
> > > +		if (buffer_size >= 3 && buffer_entry[1] == 0x0)
> > > +			dell_wmi_process_key(buffer_entry[2]);
> > > 
> > >  		else if (buffer_size >= 2)
> > > 
> > > -			reported_key = (int)buffer_entry[1] & 0xffff;
> > > -		else {
> > > +			dell_wmi_process_key(buffer_entry[1]);
> > 
> > Why can we drop the 0xffff mask now?
> 
> Because it is useless (or correct me if not!). Variable
> buffer_entry has type u16* so operation "AND 0xFFFF" on 16bit
> integer do nothing.
> 
> > > +		else
> > > 
> > >  			pr_info("Received unknown WMI event\n");
> > > 
> > > -			kfree(obj);
> > > -			return;
> > > +		kfree(obj);
> > > +		return;
> > > +	}
> > > +
> > > +	buffer_end = buffer_entry + buffer_size;
> > > +
> > > +	while (buffer_entry < buffer_end) {
> > > +
> > > +		len = buffer_entry[0];
> > > +		if (len == 0)
> > > +			break;
> > > +
> > > +		len++;
> > > +
> > 
> > Why increment len here? Are you trying to avoid a "len + 1"
> > in the comparisons below? If so, is using "len * 2" in the
> > debug message below correct? Please clarify.
> 
> in buffer_entry[0] (16 bit integer) is stored length of event
> (in 16bit units) without first (length) value. And "%*ph"
> takes size in bytes (u8). So length in bytes (u8) units is 2
> * length in u16 units.
> 
> > > +		if (buffer_entry+len > buffer_end) {
> > 
> > See coding style documentation on operators. Please run
> > patches through checkpatch.
> 
> checkpatch.pl does not show any problem for these lines.
> 
> > > +			pr_info("Invalid length of WMI event\n");
> > 
> > info doesn't see correct here either.
> 
> debug or warn?
> 
> > > +			break;
> > > 
> > >  		}
> > > 
> > > -		key =
> > > sparse_keymap_entry_from_scancode(dell_wmi_input_dev,
> > > -							reported_key);
> > > -		if (!key) {
> > > -			pr_info("Unknown key %x pressed\n", 
reported_key);
> > > -		} else if ((key->keycode == KEY_BRIGHTNESSUP ||
> > > -			    key->keycode == KEY_BRIGHTNESSDOWN) &&
> 
> acpi_video) {
> 
> > > -			/* Don't report brightness notifications that 
will
> 
> also
> 
> > > -			 * come via ACPI */
> > > -			;
> > > -		} else {
> > > -			sparse_keymap_report_entry(dell_wmi_input_dev,
> 
> key,
> 
> > > -						   1, true);
> > > +		pr_debug("Process buffer (%*ph)\n", len*2,
> 
> buffer_entry);
> 
> > > +
> > > +		switch (buffer_entry[1]) {
> > > +		case 0x00:
> > > +			for (i = 2; i < len; ++i) {
> > > +				switch (buffer_entry[i]) {
> > > +				case 0xe043:
> > > +					/* NIC Link is Up */
> > > +					pr_debug("NIC Link is Up\n");
> > > +					break;
> > > +				case 0xe044:
> > > +					/* NIC Link is Down */
> > > +					pr_debug("NIC Link is Down\n");
> > > +					break;
> > > +				case 0xe045:
> > > +					/* Unknown event but defined in DSDT */
> > > +				default:
> > > +					/* Unknown event */
> > > +					pr_info("Unknown WMI event type 0x00: "
> > > +						"0x%x\n", (int)buffer_entry[i]);
> > > +					break;
> > > +				}
> > > +			}
> > > +			break;
> > > +		case 0x10:
> > > +			/* Keys pressed */
> > > +			for (i = 2; i < len; ++i)
> > > +				dell_wmi_process_key(buffer_entry[i]);
> > > +			break;
> > > +		case 0x11:
> > > +			for (i = 2; i < len; ++i) {
> > > +				switch (buffer_entry[i]) {
> > > +				case 0xfff0:
> > > +					/* Battery unplugged */
> > > +					pr_debug("Battery unplugged\n");
> > > +					break;
> > > +				case 0xfff1:
> > > +					/* Battery inserted */
> > > +					pr_debug("Battery inserted\n");
> > > +					break;
> > > +				case 0x01e1:
> > > +				case 0x02ea:
> > > +				case 0x02eb:
> > > +				case 0x02ec:
> > > +				case 0x02f6:
> > > +					/* Keyboard backlight level changed */
> > > +					pr_debug("Keyboard backlight level "
> > > +						 "changed\n");
> > > +					break;
> > > +				default:
> > > +					/* Unknown event */
> > > +					pr_info("Unknown WMI event type 0x11: "
> > > +						"0x%x\n", (int)buffer_entry[i]);
> > > +					break;
> > > +				}
> > > +			}
> > > +			break;
> > > +		default:
> > > +			/* Unknown event */
> > > +			pr_info("Unknown WMI event type 0x%x\n",
> > > +				(int)buffer_entry[1]);
> > > +			break;
> > > 
> > >  		}
> > > 
> > > +
> > > +		buffer_entry += len;
> > > +
> > > 
> > >  	}
> > > 
> > > +
> > > 
> > >  	kfree(obj);
> > >  
> > >  }

Darren: PING. See my comments and questions.

-- 
Pali Rohár
pali.rohar@gmail.com

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]