RefOf() bug?

RefOf() bug?

[Ville Syrjälä]

My HP OmniBook 6000 has some problems with the battery. The first time the 
battery gets detected (it doesn't matter if I boot with battery 
disconnected or connected) things don't work. Then if I disconnect and 
reconnect the battery things start to work. But the next 
disconnect/reconnect cycle fails again, the next after that succeeds and 
so on.

I've traced the problem to usage of RefOf() operator. Here's a snippet of 
the DSDT:

While (LGreater (Local1, 0x08))
{
  If (LNot (And (UBIF, VTOB (Local1))))
  {
    GBFE (Local2, Local1, RefOf (Local3))
    If (Local3)
    {
      If (LNot (\_SB.PCI0.ISA0.EC0.SMRD (0x0B, 0x16, Local3, RefOf (Local4))))

                                                             ^^^^^^^^^^^^^
      {
        Store (\_SB.PCI0.ISA0.EC0.BCNT, Local5)
        Store (Zero, Local3)
        Store (Zero, ERRC)
        While (LGreater (Local5, Local3))
        {
          GBFE (Local4, Local3, RefOf (Local6))

I've highlighted the offending RefOf(). On the first run of the outer 
while loop SMRD() creates a buffer of size 4 and stores it into Arg3. For 
some reason that size never changes even though SMRD() creates buffers 
with size 5 and 6 on the following runs and when the buffer is supposed 
to be size 6 the following GBFE() tries to access beyond the size 4 limit 
and I get the error:
dsopcode-0526 [528] ds_init_buffer_field  : Field [TIDX] size 40 exceeds Buffer [NULL] size 32 (

The data in the Local4 buffer changes as SMRD copies new stuff to the new 
buffer so it looks like the Store() to Arg3 only copies the buffer 
contents up to the original buffer's size.

Why the battery works on every second reconnect is because the first 
SMRD() call creates a buffer with size 6 on those occasions. I have yet to 
find out why. But that should not matter since it looks like RefOf() is to 
blame here.

I've only glanced at the ACPI spec so I'm not entirely sure my diagnosis 
is correct...

-- 
Ville Syrjälä
syrjala-ORSVBvAovxo@public.gmane.org
http://www.sci.fi/~syrjala/


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

RE: RefOf() bug?

[Yu, Luming]

acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> My HP OmniBook 6000 has some problems with the battery. The
> first time the
> battery gets detected (it doesn't matter if I boot with battery
> disconnected or connected) things don't work. Then if I
> disconnect and
> reconnect the battery things start to work. But the next
> disconnect/reconnect cycle fails again, the next after that
> succeeds and
> so on.
> 
> I've traced the problem to usage of RefOf() operator. Here's
> a snippet of
> the DSDT:
> 
> While (LGreater (Local1, 0x08))
> {
>   If (LNot (And (UBIF, VTOB (Local1))))
>   {
>     GBFE (Local2, Local1, RefOf (Local3))
>     If (Local3)
>     {
>       If (LNot (\_SB.PCI0.ISA0.EC0.SMRD (0x0B, 0x16, Local3,
> RefOf (Local4))))
> 
> 
> ^^^^^^^^^^^^^
>       {
>         Store (\_SB.PCI0.ISA0.EC0.BCNT, Local5)
>         Store (Zero, Local3)
>         Store (Zero, ERRC)
>         While (LGreater (Local5, Local3))
>         {
>           GBFE (Local4, Local3, RefOf (Local6))
> 
> I've highlighted the offending RefOf(). On the first run of the outer
> while loop SMRD() creates a buffer of size 4 and stores it
> into Arg3. For
> some reason that size never changes even though SMRD()
> creates buffers
> with size 5 and 6 on the following runs and when the buffer
> is supposed
> to be size 6 the following GBFE() tries to access beyond the
> size 4 limit
> and I get the error:
> dsopcode-0526 [528] ds_init_buffer_field  : Field [TIDX] size
> 40 exceeds Buffer [NULL] size 32 (
> 
> The data in the Local4 buffer changes as SMRD copies new
> stuff to the new
> buffer so it looks like the Store() to Arg3 only copies the buffer
> contents up to the original buffer's size.
> 
> Why the battery works on every second reconnect is because the first
> SMRD() call creates a buffer with size 6 on those occasions.
> I have yet to
> find out why. But that should not matter since it looks like
> RefOf() is to
> blame here.
> 
> I've only glanced at the ACPI spec so I'm not entirely sure
> my diagnosis
> is correct...

Please check your ACPI version.
--Luming


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: RefOf() bug?

[Ville Syrjälä]

On Tue, Mar 30, 2004 at 06:53:01PM +0800, Yu, Luming wrote:
> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> > My HP OmniBook 6000 has some problems with the battery. The
> > first time the
> > battery gets detected (it doesn't matter if I boot with battery
> > disconnected or connected) things don't work. Then if I
> > disconnect and
> > reconnect the battery things start to work. But the next
> > disconnect/reconnect cycle fails again, the next after that
> > succeeds and
> > so on.
> 
> Please check your ACPI version.

ACPI 20040220 + kernel 2.6.4-rc1-mm2

-- 
Ville Syrjälä
syrjala-ORSVBvAovxo@public.gmane.org
http://www.sci.fi/~syrjala/


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

RE: RefOf() bug?

[Yu, Luming]

acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> On Tue, Mar 30, 2004 at 06:53:01PM +0800, Yu, Luming wrote:
>> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
>>> My HP OmniBook 6000 has some problems with the battery. The first
>>> time the battery gets detected (it doesn't matter if I boot with
>>> battery disconnected or connected) things don't work. Then if I
>>> disconnect and reconnect the battery things start to work. But the
>>> next disconnect/reconnect cycle fails again, the next after that
>>> succeeds and so on.
>> 
>> Please check your ACPI version.
> 
> ACPI 20040220 + kernel 2.6.4-rc1-mm2

Hmm, could you try new one? I suppose the most recent ACPI seldom have 
battery status issue. 

--Luming




-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: RefOf() bug?

[Ville Syrjälä]

On Tue, Mar 30, 2004 at 08:14:22PM +0800, Yu, Luming wrote:
> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> > On Tue, Mar 30, 2004 at 06:53:01PM +0800, Yu, Luming wrote:
> >> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> >>> My HP OmniBook 6000 has some problems with the battery. The first
> >>> time the battery gets detected (it doesn't matter if I boot with
> >>> battery disconnected or connected) things don't work. Then if I
> >>> disconnect and reconnect the battery things start to work. But the
> >>> next disconnect/reconnect cycle fails again, the next after that
> >>> succeeds and so on.
> >> 
> >> Please check your ACPI version.
> > 
> > ACPI 20040220 + kernel 2.6.4-rc1-mm2
> 
> Hmm, could you try new one? I suppose the most recent ACPI seldom have 
> battery status issue. 

ACPI 20040326 + kernel 2.6.5-rc2-mm5

Problem remains :(

-- 
Ville Syrjälä
syrjala-ORSVBvAovxo@public.gmane.org
http://www.sci.fi/~syrjala/


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

RE: RefOf() bug?

[Yu, Luming]

acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> On Tue, Mar 30, 2004 at 08:14:22PM +0800, Yu, Luming wrote:
>> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
>>> On Tue, Mar 30, 2004 at 06:53:01PM +0800, Yu, Luming wrote:
>>>> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
>>>>> My HP OmniBook 6000 has some problems with the battery. The first
>>>>> time the battery gets detected (it doesn't matter if I boot with
>>>>> battery disconnected or connected) things don't work. Then if I
>>>>> disconnect and reconnect the battery things start to work. But the
>>>>> next disconnect/reconnect cycle fails again, the next after that
>>>>> succeeds and so on.
>>>> 
>>>> Please check your ACPI version.
>>> 
>>> ACPI 20040220 + kernel 2.6.4-rc1-mm2
>> 
>> Hmm, could you try new one? I suppose the most recent ACPI seldom
>> have battery status issue.
> 
> ACPI 20040326 + kernel 2.6.5-rc2-mm5
> 
> Problem remains :(

:-(,  	Send me dmesg , as well as /proc/acpi/dsdt privately. 

--Luming


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

RE: RefOf() bug?

[Moore, Robert]

Please send me the dmesg and dsdt also, I would like to look at this
problem.

Bob


-----Original Message-----
From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
[mailto:acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org] On Behalf Of Yu, Luming
Sent: Tuesday, March 30, 2004 6:38 AM
To: Ville Syrj?l?; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: RE: [ACPI] RefOf() bug?

acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> On Tue, Mar 30, 2004 at 08:14:22PM +0800, Yu, Luming wrote:
>> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
>>> On Tue, Mar 30, 2004 at 06:53:01PM +0800, Yu, Luming wrote:
>>>> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
>>>>> My HP OmniBook 6000 has some problems with the battery. The first
>>>>> time the battery gets detected (it doesn't matter if I boot with
>>>>> battery disconnected or connected) things don't work. Then if I
>>>>> disconnect and reconnect the battery things start to work. But the
>>>>> next disconnect/reconnect cycle fails again, the next after that
>>>>> succeeds and so on.
>>>> 
>>>> Please check your ACPI version.
>>> 
>>> ACPI 20040220 + kernel 2.6.4-rc1-mm2
>> 
>> Hmm, could you try new one? I suppose the most recent ACPI seldom
>> have battery status issue.
> 
> ACPI 20040326 + kernel 2.6.5-rc2-mm5
> 
> Problem remains :(

:-(,  	Send me dmesg , as well as /proc/acpi/dsdt privately. 

--Luming


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=ick
_______________________________________________
Acpi-devel mailing list
Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/acpi-devel


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

RE: RefOf() bug?

[Moore, Robert]

The problem is related to doing the indirect store through the Arg when
there is an already existing buffer stored at that location.

This code reproduces the problem.  At exit, Local0 contains a buffer of
length 4, not 6.

    Method (BT2, 1)
    {
        Store (Buffer (4){}, Arg0)
        Store (Buffer (6){}, Arg0)
        Return ()
    }
    Method (BTST)
    {
        BT2 (RefOf (Local0))
        Return ()
    }

I will continue to debug the problem and fix it for the next release.

Bob


-----Original Message-----
From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
[mailto:acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org] On Behalf Of Moore,
Robert
Sent: Tuesday, March 30, 2004 8:49 AM
To: Yu, Luming; Ville Syrj?l?; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: RE: [ACPI] RefOf() bug?

Please send me the dmesg and dsdt also, I would like to look at this
problem.

Bob


-----Original Message-----
From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
[mailto:acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org] On Behalf Of Yu, Luming
Sent: Tuesday, March 30, 2004 6:38 AM
To: Ville Syrj?l?; acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: RE: [ACPI] RefOf() bug?

acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
> On Tue, Mar 30, 2004 at 08:14:22PM +0800, Yu, Luming wrote:
>> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
>>> On Tue, Mar 30, 2004 at 06:53:01PM +0800, Yu, Luming wrote:
>>>> acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org wrote:
>>>>> My HP OmniBook 6000 has some problems with the battery. The first
>>>>> time the battery gets detected (it doesn't matter if I boot with
>>>>> battery disconnected or connected) things don't work. Then if I
>>>>> disconnect and reconnect the battery things start to work. But the
>>>>> next disconnect/reconnect cycle fails again, the next after that
>>>>> succeeds and so on.
>>>> 
>>>> Please check your ACPI version.
>>> 
>>> ACPI 20040220 + kernel 2.6.4-rc1-mm2
>> 
>> Hmm, could you try new one? I suppose the most recent ACPI seldom
>> have battery status issue.
> 
> ACPI 20040326 + kernel 2.6.5-rc2-mm5
> 
> Problem remains :(

:-(,  	Send me dmesg , as well as /proc/acpi/dsdt privately. 

--Luming


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=ick
_______________________________________________
Acpi-devel mailing list
Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/acpi-devel


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=ick
_______________________________________________
Acpi-devel mailing list
Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/acpi-devel


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: RefOf() bug?

[Ville Syrjälä]

On Wed, Mar 31, 2004 at 05:13:22PM -0800, Moore, Robert wrote:
> The problem is related to doing the indirect store through the Arg when
> there is an already existing buffer stored at that location.
> 
> This code reproduces the problem.  At exit, Local0 contains a buffer of
> length 4, not 6.
> 
>     Method (BT2, 1)
>     {
>         Store (Buffer (4){}, Arg0)
>         Store (Buffer (6){}, Arg0)
>         Return ()
>     }
>     Method (BTST)
>     {
>         BT2 (RefOf (Local0))
>         Return ()
>     }
> 
> I will continue to debug the problem and fix it for the next release.

Great. I'll look forward to testing the next release then.

-- 
Ville Syrjälä
syrjala-ORSVBvAovxo@public.gmane.org
http://www.sci.fi/~syrjala/


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click