Re: ACPI key events handling

On Fri, 2005-03-11 at 00:28, Thomas Renninger wrote:

Li, Shaohua wrote:

quoted

How about below patch? It fixes current ec address space handler bugs.

Thanks,
Shaohua


diff -puN drivers/acpi/ec.c~ec_addr_space_handler drivers/acpi/ec.c

--- 2.5/drivers/acpi/ec.c~ec_addr_space_handler	2005-03-08

09:19:05.228720200 +0800

+++ 2.5-root/drivers/acpi/ec.c	2005-03-08 10:35:54.130060696 +0800

@@ -441,7 +441,7 @@ acpi_ec_space_handler (

{
	int			result = 0;
	struct acpi_ec		*ec = NULL;
-	u32			temp = 0;
+	u32			temp = *value;
	acpi_integer		f_v = 0;
	int 			i = 0;

@@ -462,10 +462,9 @@ next_byte:

	switch (function) {
	case ACPI_READ:
		result = acpi_ec_read(ec, (u8) address, &temp);
-		*value = (acpi_integer) temp;
		break;
	case ACPI_WRITE:
-		result = acpi_ec_write(ec, (u8) address, (u8) *value);
+		result = acpi_ec_write(ec, (u8) address, (u8) temp);
		break;
	default:
		result = -EINVAL;

@@ -474,19 +473,18 @@ next_byte:

	}

	bit_width -= 8;
-	if(bit_width){
-
+	if (bit_width){
		if(function == ACPI_READ)
-			f_v |= (acpi_integer) (*value) << 8*i;
+			f_v |= temp << 8 * i;
		if(function == ACPI_WRITE)
-			(*value) >>=8; 
+			temp >>= 8;
		i++;
 

So the actual fix is this?

It also fixes the 'changing *value when write', since we possibly use
the value in some places later. Below line.
-			(*value) >>=8; 

quoted

+		(u8)address ++;
 

I saw it some days ago.
However this only fixes something if you see:
*acpi_ec_space_handler: bit_width should be 8*
messages in your syslog.

Yep.

Could you delete this message as well, please?

I'm ok for it. Maybe the warning should be under 'if (acpi_strict)'.

I have a machine here having this message in syslog
every second...
I didn't try the keys, but I think they work as expected.

I wonder whether  u64 temp should be used, acpi_integer could be u64
and bit_width (QWordAcc) could be 64 bit?

Yeah, looks reasonable to me. I'll update the patch.

quoted

		goto next_byte;
	}

-
	if(function == ACPI_READ){
-		f_v |= (acpi_integer) (*value) << 8*i;
+		f_v |= temp << 8 * i;
		*value = f_v;
	}

@@ -505,8 +503,6 @@ out:

	default:
		return_VALUE(AE_OK);
	}
-	
-
}

 

quoted

-----Original Message-----
From: <redacted> [mailto:acpi-devel-
admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org] On Behalf Of Karol Kozimor
Sent: Thursday, March 10, 2005 4:50 PM
To: Paulo Vitor Magacho da Silva
Cc: <redacted>
Subject: Re: [ACPI] ACPI key events handling

Thus wrote Paulo Vitor Magacho da Silva:
   

quoted

 I have an ACER 163lmi laptop and I am having problem with some of
     

the
 

quoted

Fn
   

quoted

keys. I don't know if this is related to ACPI or not.
 The problem is that when I press some the Fn keys, like the sleep
     

button
   

quoted

(Fn+F4), and I look at the /proc/acpi/event, sometimes the key works
sometimes it doesn't. I did also the check of the Fn keys with the
     

acerhk
 

quoted

quoted

driver. My laptop has the mail, internet, p1, p2, wireless, buttons.
     

And
 

quoted

quoted

when I press those buttons, sometimes I get the correct reading from
     

the
 

quoted

quoted

acerhk driver, sometimes when I press the mail buttons it reads back
     

the
 

quoted

quoted

wireless, or any other key, and sometimes it just reads garbage.
     

Sounds like another instance of
http://bugzilla.kernel.org/show_bug.cgi?id=4124

Best regards,

   

-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click