Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Hernan G Solari]

Hello
    I am disturbing you with this problem because I think there is
something to be learnt, I have
no urgency or further-problem with my work-around.

Description: trying to boot with kernel 2.6.22  the booting process
stops not finding
the root partition, in the previous line it gives an EMPTY list of
available partitions.
Disabling completely the (Serial) ATA driver and enabling the old SATA
driver is my work-around (a regresion).
Installing linux with debian 4.0 is possible (kernel 2.6.18.4-amd64),
but ubuntu crashes
silently after loading kernel and initrd.image (I guess it is the same
problem).

Description of the machine, dmesg output with kernel 2.6.22, lspci,
modules, cpu, iomem, ioports, acpi-table as well as instalation story at
http://www.df.uba.ar/~solari/acer.html

Hitachi 120Gb drive
scsi 0:0:0:0: Direct-Access ATA Hitachi HTS54161 SBDO PQ: 0 ANSI: 5

ATI controller
00:12.0 IDE interface: ATI Technologies Inc 4379 Serial ATA Controller
(rev 80)
 (prog-if 8f [Master SecP SecO PriP PriO])
Subsystem: Acer Incorporated [ALI] Unknown device 010f
(pci.ids are up-to-day)

The acpi system does not work either, but I have reported the problem to
a more specific list.

I can test patches in case you want.

                               thanks for your attention
                                                                        
      Hernan


-- 
Hernán Gustavo Solari, solari@df.uba.ar, http://www.df.uba.ar/~solari

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Tejun Heo]

Hernan G Solari wrote:
> Hello
>     I am disturbing you with this problem because I think there is
> something to be learnt, I have
> no urgency or further-problem with my work-around.
> 
> Description: trying to boot with kernel 2.6.22  the booting process
> stops not finding
> the root partition, in the previous line it gives an EMPTY list of
> available partitions.
> Disabling completely the (Serial) ATA driver and enabling the old SATA
> driver is my work-around (a regresion).
> Installing linux with debian 4.0 is possible (kernel 2.6.18.4-amd64),
> but ubuntu crashes
> silently after loading kernel and initrd.image (I guess it is the same
> problem).

Can you set up netconsole and collect the kernel log from the failing
boot and post it here?

Thanks.

-- 
tejun

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Tejun Heo]

[restoring LKML and cc'ing linux-ide and linux-acpi]

Hernan Gustavo Solari wrote:
>     I am replying directly to you since, finding that most of the
> material in the kernel list is well beyond my understanding, I cancelled
> my subscription.
> 
>     You can forward as much of my replies to the list as you think it suits

You don't have to be subscribed to the list.  Just keep them cc'd.

>> Can you set up netconsole and collect the kernel log from the failing
>> boot and post it here?
> 
> Let me first update you on the issue. I installed the kernel 2.6.22.8
> when it was the latest stable release (today stable release is 2.6.22.9)
> Trying to reproduce the problem with this kernel I get a crash but
> different from that in the 2.6.22 kernel.
> 
> The kernel is compiled as it comes, with AMD64 generic option set and
> the small systems option set by me.
> 
> With acpi=off it runs.
> With acpi=on crashes after lines
> aer: probe of 0000:00:04.0:pcie01 failed with error 2
> aer: probe of 0000:00:05.0:pcie01 failed with error 2
> aer: probe of 0000:00:06.0:pcie01 failed with error 2
> aer: probe of 0000:00:07.0:pcie01 failed with error 2
> 
> With acpi=ht runs and complaints saying
> aerinit_init: AER service init fails - No ACPI _05C support
> aer: probe of 0000:00:04.0:pcie01 failed with error 1
> and the same message for the other probes.
> 
> The next line is:
> Real Time clock driver v1.12ac
> Linux agpgart interface v0.102 (c) Dave Jones
> Serial: 8250/16550 driver ....
> 
> 
> I have filled a ticket with respect to the ACPI problems, much of them
> can be avoided disabling part of the ACPI system.
> 
> Options: I can compile the 2.6.22 kernel to get the reported problem
>      or try to debug the problem in 2.6.22.8 or something else it
>       fits the kernel-team.
> 
> With respect to netconsole, the kernel parameter would be
> netconsole=[src-port]@[src-ip]/[<dev>],[tgt-port]@<tgt-ip>/[tgt-macaddr]
> According to
> http://www.mjmwired.net/kernel/Documentation/networking/netconsole.txt
> and based on kernel 2.6.22.1
> 
> and I cannot figure out what is [tgt-macaddr] "ethernet MAC address for
> logging agent (broadcast)"
> is it Internet Hardware address? (here
> Ethernet  HWaddr 00:0E:A6:5F:13:B2 ) or something else I cannot imagine?

Yes, correct, ethernet hardware address.  Things surrounded by brackets
are optional, BTW.

> PS: using netconsole may take me a few days since I do not have a
> network at home, I have to find the network at work available.

I see.  If you have access to a second machine, you can connect the two
directly.

Thanks.

-- 
tejun

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Hernan Gustavo Solari]

Tejun

    netconsole, pritty nice debunging system... but (yes, there is 
always a but) it does not get to run.
    the method was well implemented, adding the acpi=off it sends the 
information to the receiving machine (I can even see passing a 
netconsole probing message in the machine under testing), but without 
turning off acpi it does not reach the point of loading and, 
consequently, it does not send a byte to the receiving machine.

Hence, result: empty output.

Hernan


-- 
Hernan G. Solari (solari@df.uba.ar http://www.df.uba.ar/~solari)

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Tejun Heo]

Hernan Gustavo Solari wrote:
> Tejun
> 
>    netconsole, pritty nice debunging system... but (yes, there is always
> a but) it does not get to run.
>    the method was well implemented, adding the acpi=off it sends the
> information to the receiving machine (I can even see passing a
> netconsole probing message in the machine under testing), but without
> turning off acpi it does not reach the point of loading and,
> consequently, it does not send a byte to the receiving machine.
> 
> Hence, result: empty output.

Hmmm.. Are netconsole support and network driver built into the kernel?

-- 
tejun

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Hernan G Solari]

>>   netconsole, pritty nice debunging system... but (yes, there is always
>>a but) it does not get to run.
>>   the method was well implemented, adding the acpi=off it sends the
>>information to the receiving machine (I can even see passing a
>>netconsole probing message in the machine under testing), but without
>>turning off acpi it does not reach the point of loading and,
>>consequently, it does not send a byte to the receiving machine.
>>
>>Hence, result: empty output.
>>    
>>
>
>Hmmm.. Are netconsole support and network driver built into the kernel?
>
>  
>
yes, they are
                      Hernan

-- 
Hernán Gustavo Solari, solari@df.uba.ar, http://www.df.uba.ar/~solari

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Tejun Heo]

Hernan G Solari wrote:
>>>   netconsole, pritty nice debunging system... but (yes, there is always
>>> a but) it does not get to run.
>>>   the method was well implemented, adding the acpi=off it sends the
>>> information to the receiving machine (I can even see passing a
>>> netconsole probing message in the machine under testing), but without
>>> turning off acpi it does not reach the point of loading and,
>>> consequently, it does not send a byte to the receiving machine.
>>>
>>> Hence, result: empty output.
>>>    
>>>
>> Hmmm.. Are netconsole support and network driver built into the kernel?
>>
> yes, they are

If you can set up a serial console, it would be better.  If not, can you
please take a photo of the crash and post it?

-- 
tejun

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Hernan G Solari]

> If you can set up a serial console, it would be better.  If not, can you
> please take a photo of the crash and post it?
>   
I am running short of elements, no digital camera o movil phone for the 
case or serial link.
However, what it is on the screen when it crashes follows:

KERNEL 2.6.23-rc8
-----------------------------------------------------begin screen dump
PCI: Bridge: 0000:00:14.4
  IO window: a000-afff
  MEM window: d0300000-d03fffff
  PREFETCH window: 50000000-53ffffff
PCI: Enabling device 0000:00:04.0 (0000 -> 0002)
PCI: Enabling device 0000:08:01.0 (0000 -> 0003)
ACPI: PCI Interrupt 0000:08:01.0[A] -> GSI 23 (level, low) -> IRQ 23
NET: Registered protocol family 2
IP route cache hash table entries: 32768 (order: 6, 262144 bytes)
TCP established hash table entries: 131072 (order: 9, 3145728 bytes)
TCP bind hash table entries: 65536 (order: 8, 1048576 bytes)
TCP: Hash tables configured (established 131072 bind 65536)
TCP reno registered
Simple Boot Flag at 0x36 set to 0x1
Total HugeTLB memory allocated, 0
Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
io scheduler noop registered
io scheduler deadline registered
io scheduler cfq registered (default)
PCI: MSI quirk detected. MSI deactivated.
assign_interrupt_mode Found MSI capability
assign_interrupt_mode Found MSI capability
assign_interrupt_mode Found MSI capability
assign_interrupt_mode Found MSI capability
-------------------------------end screen dump

Not so difficult since most of these lines are also in a succesful boot 
with some of the ACPI turned off.

Hernan

-- 
Hernán Gustavo Solari, solari@df.uba.ar, http://www.df.uba.ar/~solari

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Tejun Heo]

Hello, Hernan.

Hernan G Solari wrote:
> 
>> If you can set up a serial console, it would be better.  If not, can you
>> please take a photo of the crash and post it?
>>   
> I am running short of elements, no digital camera o movil phone for the
> case or serial link.
> However, what it is on the screen when it crashes follows:
> 
> KERNEL 2.6.23-rc8
> -----------------------------------------------------begin screen dump
> PCI: Bridge: 0000:00:14.4
>  IO window: a000-afff
>  MEM window: d0300000-d03fffff
>  PREFETCH window: 50000000-53ffffff
> PCI: Enabling device 0000:00:04.0 (0000 -> 0002)
> PCI: Enabling device 0000:08:01.0 (0000 -> 0003)
> ACPI: PCI Interrupt 0000:08:01.0[A] -> GSI 23 (level, low) -> IRQ 23
> NET: Registered protocol family 2
> IP route cache hash table entries: 32768 (order: 6, 262144 bytes)
> TCP established hash table entries: 131072 (order: 9, 3145728 bytes)
> TCP bind hash table entries: 65536 (order: 8, 1048576 bytes)
> TCP: Hash tables configured (established 131072 bind 65536)
> TCP reno registered
> Simple Boot Flag at 0x36 set to 0x1
> Total HugeTLB memory allocated, 0
> Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
> io scheduler noop registered
> io scheduler deadline registered
> io scheduler cfq registered (default)
> PCI: MSI quirk detected. MSI deactivated.
> assign_interrupt_mode Found MSI capability
> assign_interrupt_mode Found MSI capability
> assign_interrupt_mode Found MSI capability
> assign_interrupt_mode Found MSI capability

Hmmm... and the kernel freezes after this?  libata driver isn't
initialized yet.  I don't think libata has anything to do with here.

Thanks.

-- 
tejun

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Hernan G Solari]

> Hmmm... and the kernel freezes after this?  libata driver isn't
> initialized yet.  I don't think libata has anything to do with here.
>
>   
as I told you, the ATA problem was with the 2.6.22 version of the kernel
and is not longer related to the freeze with kernel 2.6.23-rc8  I asked 
whether you wanted to see the ATA problem in that kernel or the present 
crash.

> Options: I can compile the 2.6.22 kernel to get the reported problem
>       or try to debug the problem in 2.6.22.8 or something else it
>        fits the kernel-team.

was my offer.

As you did not answer that point (you answered other points) I thought 
it was useless to debug a problem that is not longer there. I still 
thing that way.
The booting is apparently allocating port services at the  point of 
freezing but not too close of the disk driver, that much I  knew.

regards
                Hernan




-- 
Hernán Gustavo Solari, solari@df.uba.ar, http://www.df.uba.ar/~solari

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Tejun Heo]

Hello,

Hernan G Solari wrote:
>> Hmmm... and the kernel freezes after this?  libata driver isn't
>> initialized yet.  I don't think libata has anything to do with here.
>>
>>   
> as I told you, the ATA problem was with the 2.6.22 version of the kernel
> and is not longer related to the freeze with kernel 2.6.23-rc8  I asked
> whether you wanted to see the ATA problem in that kernel or the present
> crash.
> 
>> Options: I can compile the 2.6.22 kernel to get the reported problem
>>       or try to debug the problem in 2.6.22.8 or something else it
>>        fits the kernel-team.
> 
> was my offer.
> 
> As you did not answer that point (you answered other points) I thought
> it was useless to debug a problem that is not longer there. I still
> thing that way.
> The booting is apparently allocating port services at the  point of
> freezing but not too close of the disk driver, that much I  knew.

Right, sorry about the confusion.  I think the screen capture would be
useful to ACPI developers.  Thanks.

-- 
tejun

Re: Serial ATA does not find partitions (Hitachi HD, new? ATI controller) where old SATA works

[Hernan G Solari]

> Right, sorry about the confusion.  I think the screen capture would be
> useful to ACPI developers.  Thanks.
>   
In any case, I appreciate your williness to help, plus I learned things 
I would have not learned otherwise, many thanks to you

Hernan

-- 
Hernán Gustavo Solari, solari@df.uba.ar, http://www.df.uba.ar/~solari