ACPI + Promise IDE = disk corruption :-(((

ACPI + Promise IDE = disk corruption :-(((

[Pavel Roskin]

Hello!

It's just a word of warning for those who are trying ACPI with the latest
kernels.

I enabled ACPI in 2.4.5-ac17 (2.4.5-ac16 works fine with the same config
except ACPI). When I booted I saw a message

ACPI: If experiencing system slowness, try adding "acpi=no-idle" to
cmdline

and after that ...

reiserfs: checking transaction log (device 21:04) ...

Normally it takes few seconds before the system goes further, but that
time it tool much longer (one minute maybe), and the next message was
something like "hde: DMA timeout"

I hit reset hoping to boot the system with "acpi=no-idle", but GRUB
couldn't load stage2, which resides on the root partition (reiserfs).

I had to install Linux on another drive and run reiserfsck. It found many
errors, fixed them all, and now I'm happily running my old system.

I also checked the same kernel with "acpi=no-idle" - it works. The ACPI
messages are now:

ACPI: Core Subsystem version [20010208]
ACPI: Subsystem enabled
ACPI: System firmware supports: C2 C3
ACPI: plvl2lat=10 plvl3lat=20
ACPI: C2 enter=143 C2 exit=35
ACPI: C3 enter=858 C3 exit=71
ACPI: Not using ACPI idle
ACPI: System firmware supports: S0 S1 S5

The config file is here: http://www.red-bean.com/~proski/linux/config
dmesg output is here: http://www.red-bean.com/~proski/linux/dmesg

Possibly important data:

Kernel compiled with gcc-3.0
Motherboard Micron SE440-BX2
BIOS 4S4EB2X0.05A.0016.P15 (the latest)
1 Intel Pentium III 550MHz
Root filesystem is reiserfs
Root filesystem is on Promise PDC20267 ide controller
Local APIC (but not IO-APIC) is used.

I can make more experiments (with other hard drives) if needed.

-- 
Regards,
Pavel Roskin

Re: ACPI + Promise IDE = disk corruption :-(((

[Alan Cox]

> I enabled ACPI in 2.4.5-ac17 (2.4.5-ac16 works fine with the same config
> except ACPI). When I booted I saw a message

> I hit reset hoping to boot the system with "acpi=no-idle", but GRUB
> couldn't load stage2, which resides on the root partition (reiserfs).

I've seen several people report ACPI eats disks. ACPI is incredibly complex
badly designed crud. My advice is never use ACPI. This incidentally appears
to be the advice Microsoft give people too - they tell people to disable
ACPI as one of the first steps to diagnosing strange crashes in machines

ACPI is over complex, new technology. The BIOS stuff is new (and frequently
wrong), the kernel stuff is new (and has plenty of known bugs) and the 
combination is a recipe for disaster.

I've been discussing with a few folk about doing an alternative mini acpi
subset so that Linux can be booted on most 'ACPI only' hardware without 
using all the ACPI junk

Alan

Re: ACPI + Promise IDE = disk corruption :-(((

[Chris Wedgwood]

On Fri, Jun 22, 2001 at 11:16:01PM +0100, Alan Cox wrote:

    ACPI is over complex, new technology. The BIOS stuff is new (and
    frequently wrong)

Show my a BIOS that isn't broken.

I'm serious, EVERY SINGLE MACHINE at my disposal of late has had
buggy ACPI tables or worse.

Ironically, many laptops use a Pheonix BIOS who traditionaly have had
terrible QA too --- their bioses APM would only work under Windows95
initially and then later Windows98.

I'm really started to wish the OpenBios folks had more resources
available; the quality of the average BIOS it seems in the last
couple fo years is really bad.



  --cw

RE: ACPI + Promise IDE = disk corruption :-(((

[Grover, Andrew]

Just a note, in 2.4.6-pre5, the acpi=no-idle option goes away, but you
should no longer experience any corruption issues, either.

Regards -- Andy

PS sorry you experienced problems - glad you could recover.

> From: Pavel Roskin [mailto:proski@gnu.org]
> Hello!
> 
> It's just a word of warning for those who are trying ACPI 
> with the latest
> kernels.
> 
> I enabled ACPI in 2.4.5-ac17 (2.4.5-ac16 works fine with the 
> same config
> except ACPI). When I booted I saw a message
> 
> ACPI: If experiencing system slowness, try adding "acpi=no-idle" to
> cmdline
> 
> and after that ...
> 
> reiserfs: checking transaction log (device 21:04) ...
> 
> Normally it takes few seconds before the system goes further, but that
> time it tool much longer (one minute maybe), and the next message was
> something like "hde: DMA timeout"
> 
> I hit reset hoping to boot the system with "acpi=no-idle", but GRUB
> couldn't load stage2, which resides on the root partition (reiserfs).
> 
> I had to install Linux on another drive and run reiserfsck. 
> It found many
> errors, fixed them all, and now I'm happily running my old system.
> 
> I also checked the same kernel with "acpi=no-idle" - it 
> works. The ACPI
> messages are now:
> 
> ACPI: Core Subsystem version [20010208]
> ACPI: Subsystem enabled
> ACPI: System firmware supports: C2 C3
> ACPI: plvl2lat=10 plvl3lat=20
> ACPI: C2 enter=143 C2 exit=35
> ACPI: C3 enter=858 C3 exit=71
> ACPI: Not using ACPI idle
> ACPI: System firmware supports: S0 S1 S5
> 
> The config file is here: http://www.red-bean.com/~proski/linux/config
> dmesg output is here: http://www.red-bean.com/~proski/linux/dmesg
> 
> Possibly important data:
> 
> Kernel compiled with gcc-3.0
> Motherboard Micron SE440-BX2
> BIOS 4S4EB2X0.05A.0016.P15 (the latest)
> 1 Intel Pentium III 550MHz
> Root filesystem is reiserfs
> Root filesystem is on Promise PDC20267 ide controller
> Local APIC (but not IO-APIC) is used.
> 
> I can make more experiments (with other hard drives) if needed.
> 
> -- 
> Regards,
> Pavel Roskin
> 
> -
> To unsubscribe from this list: send the line "unsubscribe 
> linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

RE: ACPI + Promise IDE = disk corruption :-(((

[Grover, Andrew]

> From: Alan Cox [mailto:alan@lxorguk.ukuu.org.uk]
> I've seen several people report ACPI eats disks. ACPI is 
> incredibly complex
> badly designed crud. My advice is never use ACPI. This 
> incidentally appears
> to be the advice Microsoft give people too - they tell people 
> to disable
> ACPI as one of the first steps to diagnosing strange crashes 
> in machines

They're just bugs. That's why it's marked experimental, and unlike the rest
of the "experimental" drivers, we aren't kidding.

It's just *one* issue that has generated all the disk corruption reports.
Putting the processor into the C3 power state, in combination with bus
mastering. This is disabled in the most recent release. I'd love to fix this
one, but if it were easy, it'd be fixed by now. Maybe you can shed some
light - if you're willing, let me know and I'll describe the problem in
greater detail.

> I've been discussing with a few folk about doing an 
> alternative mini acpi
> subset so that Linux can be booted on most 'ACPI only' 
> hardware without 
> using all the ACPI junk

Could these discussions be opened up to a wider readership? Perhaps you
could use linux-pm-devel@sourceforge.net, or
acpi@phobos.fachschaften.tu-muenchen.de?

Regards -- Andy

Re: ACPI + Promise IDE = disk corruption :-(((

[Alan Cox]

> It's just *one* issue that has generated all the disk corruption reports.
> Putting the processor into the C3 power state, in combination with bus
> mastering. This is disabled in the most recent release. I'd love to fix this
> one, but if it were easy, it'd be fixed by now. Maybe you can shed some
> light - if you're willing, let me know and I'll describe the problem in
> greater detail.

By all means. I doubt I can help much on that issue but I can at least think
about it

> Could these discussions be opened up to a wider readership? Perhaps you
> could use linux-pm-devel@sourceforge.net, or
> acpi@phobos.fachschaften.tu-muenchen.de?

I have enough mail without reading those lists too. All I am talking about
here is parsing the irq routing and apic tables. 

Re: ACPI + Promise IDE = disk corruption :-(((

[Chris Wedgwood]

On Fri, Jun 22, 2001 at 05:30:19PM -0700, Grover, Andrew wrote:

    It's just *one* issue that has generated all the disk corruption
    reports.

The same code is used for FreeBSD and friends too right? Are they
seeing anywhere near the same number of types of poroblems Linux is?



  --cw

RE: ACPI + Promise IDE = disk corruption :-(((

[Grover, Andrew]

Their processor power state code looks dormant at the moment, so they
haven't hit this particular issue.

They have in the past run into a number of problems, and submitted fixes.
The Linux version is getting much wider testing right now.

-- Andy

PS Just FreeBSD, no Net or OpenBSD just yet.

> From: Chris Wedgwood [mailto:cw@f00f.org]
>>     It's just *one* issue that has generated all the disk corruption
>>     reports.
> 
> The same code is used for FreeBSD and friends too right? Are they
> seeing anywhere near the same number of types of poroblems Linux is?