Re: IDE cs5530 hda: lost interrupt

Re: IDE cs5530 hda: lost interrupt

[Alan Cox]

Ar Mer, 2006-11-08 am 19:05 -0300, ysgrifennodd Saulo:
> 14:          2    XT-PIC  ide0    >>> just 2 interrupts
> 15:       2964    XT-PIC  ide1
> NMI:         0
> ERR:         0

Thats very odd indeed as the IRQ is hard wired to 14. Are you sure the
system works with other OS's and isn't faulty (I ask this as its the
first 5530 report of this kind I've seen in about ten years, and the
device is in legacy mode which means its hard wired to IRQ 14)

Alan

Re: IDE cs5530 hda: lost interrupt

[Alan Cox]

Ar Mer, 2006-11-08 am 19:48 -0300, ysgrifennodd Saulo:
> I tried others operating systems, this machine work fine with DOS 6.22
> and Windows CE 4.2, but lock on FreeBSD and Linux.

So it works with systems that don't use the interrupt ? Is this some
kind of PDA type device designed for Windows CE where they've not
bothered wiring the interrupt perhaps.

If so you'll need to write a polling CF driver for it, which actually is
one of those things that we could do with for some other embedded cases
too.

Alan

IDE cs5530 hda: lost interrupt

[Saulo]

Hi all,

any help is wellcome...

--------------------
CPU: NSC Geode(TM) Integrated Processor by National Semi stepping 02
...
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
CS5530: ide CONTROLLER AT pci SLOT 0000:00:12.2
CS5530: chipset revision 0
CS5530: not 100% native mode: will probe irqs later
PCI: Enabling bus mastering for device 0000:00:12.2
PCI: Setting latency timer of device 0000:00:12.2 to 64
     ide0: BM-DMA at 0xfc00-0xfc07, BIOS settings: hda:pio, hdb:pio
     ide1: BM-DMA at 0xfc08-0xfc0f, BIOS settings: hdc:pio, hdd:pio
hda: CF 32MB, CFA DISK drive
hda: IRQ probe failed (0xfeba)    >>> I think my problem may start here, 
but when I fix to IRQ 14 in try_to_identify() to hda the problem persist
ide0 at 0x1f0-0x1f7,0x3f6 on irq14
hdc: Hitachi CV 5.1.1, CFA DISK drive
ide1 at 0x170-0x177,0x376 on irq 15 (serialized with ide0)
hda: max request size: 128KiB
hda: 62976 sectors (32MB) w/1KiB Cache, CHS=492/4/32
 hda:<4>hda: lost interrupt
hda: lost interrupt
hda: lost interrupt
...
 hd1
hdc: max request size: 128KiB
hdc: 31488 sectors (16 MB) w/1KiB Cache, CHS=246/4/32
 hdc: hdc1
USB Universal Host Controller Interface driver v3.0
...
INIT: ...
---------------
when I look in my /proc/interrupts I see something strange

0:      929785    XT-PIC  timer
1:         128    XT-PIC  I8042
2:           0    XT-PIC  cascade
6:           3    XT-PIC  floppy
8:           0    XT-PIC  rtc
14:          2    XT-PIC  ide0    >>> just 2 interrupts
15:       2964    XT-PIC  ide1
NMI:         0
ERR:         0
  
pci devices (lspci)

00:00.0 Class 0600: 1078:0001 - Host bridge: Cyrix Corporation:PCI Master
00:12.0 Class 0601: 1078:0100 - ISA bridge: Cyrix Corporation:5530 
Legacy [Kahlua] (rev 30)
00:12.1 Class 0680: 1078:0101 - POWER interface: Cyrix Corporation:5530 
SMI [Kahlua]
00:12.2 Class 0101: 1078:0102 - IDE interface: Cyrix Corporation:5530 
IDE [Kahlua]
00:12.3 Class 0401: 1078:0103 - Multimedia audio controller: Cyrix 
Corporation:5530 Audio [Kahlua]
00:12.4 Class 0300: 1078:0104 - VGA compatible controller: Cyrix 
Corporation:5530 Video [Kahlua]
00:13.0 Class 0c03: 0e11:a0f8 - USB open host controller: Compaq 
Computer Corporation: ZFMicro Chipset USB (rev 06)

this is my machine configuration:
- GEODE GX1 BGA 200MHZ
- 64MB
- CS5530 with 2 controllers IDE master primary CompactFlash and master 
secondary, without slave connectors.

I tried all of present configurations from ide.txt and all that I found 
in the internet about this, and ever I get lots of <hda: lost interrupt>.
The incredible is that hdc work fine, and I undertand why.

Yes I tried all
    - ide=nodma
    - hda=noprobe
    - set default values to ide0
    - ide0=serialize ide1=serialize
    - disabling acpi and apic
    - chage CompactFlash
    - checking

thank´s

-- 
[]´s Saulo Alessandre

Re: IDE cs5530 hda: lost interrupt

[Saulo]

I tried others operating systems, this machine work fine with DOS 6.22 
and Windows CE 4.2, but lock on FreeBSD and Linux.

Other strange think is that after lots and lots of "lost interrupts" 
this work. I can mount and list files but with lots of "lost interrupts" 
and after a long time.

Saulo Alessandre


Alan Cox wrote:

>Ar Mer, 2006-11-08 am 19:05 -0300, ysgrifennodd Saulo:
>  
>
>>14:          2    XT-PIC  ide0    >>> just 2 interrupts
>>15:       2964    XT-PIC  ide1
>>NMI:         0
>>ERR:         0
>>    
>>
>
>Thats very odd indeed as the IRQ is hard wired to 14. Are you sure the
>system works with other OS's and isn't faulty (I ask this as its the
>first 5530 report of this kind I've seen in about ten years, and the
>device is in legacy mode which means its hard wired to IRQ 14)
>
>Alan
>
>  
>

Re: IDE cs5530 hda: lost interrupt

[Saulo]

Other strange think is that after lots and lots of "lost interrupts" 
this work. I can mount and list files but with lots of "lost interrupts" 
and after a long time.

this is the strange.

before ls in primary master

14:          2    XT-PIC  ide0    >>> just 2 interrupts
15:       3893    XT-PIC  ide1

after ls and many "lost interrupts"

14:          2    XT-PIC  ide0    >>> just 2 interrupts
15:       3901    XT-PIC  ide1    >>> more interrupts


[]´s Saulo Alessandre

Re: IDE cs5530 hda: lost interrupt

[Mark Lord]

Saulo wrote:
> Hi all,
> 
> any help is wellcome...
> 
> --------------------
> CPU: NSC Geode(TM) Integrated Processor by National Semi stepping 02
> ...
> ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
> CS5530: ide CONTROLLER AT pci SLOT 0000:00:12.2
> CS5530: chipset revision 0
> CS5530: not 100% native mode: will probe irqs later
> PCI: Enabling bus mastering for device 0000:00:12.2
> PCI: Setting latency timer of device 0000:00:12.2 to 64
>     ide0: BM-DMA at 0xfc00-0xfc07, BIOS settings: hda:pio, hdb:pio
>     ide1: BM-DMA at 0xfc08-0xfc0f, BIOS settings: hdc:pio, hdd:pio
> hda: CF 32MB, CFA DISK drive
> hda: IRQ probe failed (0xfeba)    >>> I think my problem may start here, 
> but when I fix to IRQ 14 in try_to_identify() to hda the problem persist
> ide0 at 0x1f0-0x1f7,0x3f6 on irq14
> hdc: Hitachi CV 5.1.1, CFA DISK drive
> ide1 at 0x170-0x177,0x376 on irq 15 (serialized with ide0)
> hda: max request size: 128KiB
> hda: 62976 sectors (32MB) w/1KiB Cache, CHS=492/4/32
> hda:<4>hda: lost interrupt
> hda: lost interrupt
> hda: lost interrupt
> ...
...

Send me one of those devices and I'll fix it.

Cheers

Re: IDE cs5530 hda: lost interrupt

[Alan Cox]

Ar Iau, 2006-11-09 am 10:30 -0300, ysgrifennodd Saulo:
> No it's not a PDA designed for Windows CE it´s a motherboard designed 
> for us like 386 compatible. According to motherboard´s designers INTR 
> are standard and Windows XP run too without any changes.

Have you tested this.. ? From the docs I don't see any obvious other
explanation.

Re: IDE cs5530 hda: lost interrupt

[Saulo]

Alan Cox wrote:

>Ar Mer, 2006-11-08 am 19:48 -0300, ysgrifennodd Saulo:
>  
>
>>I tried others operating systems, this machine work fine with DOS 6.22
>>and Windows CE 4.2, but lock on FreeBSD and Linux.
>>    
>>
>
>So it works with systems that don't use the interrupt ? Is this some
>kind of PDA type device designed for Windows CE where they've not
>bothered wiring the interrupt perhaps.
>
>If so you'll need to write a polling CF driver for it, which actually is
>one of those things that we could do with for some other embedded cases
>too.
>
>Alan
>
No it's not a PDA designed for Windows CE it´s a motherboard designed 
for us like 386 compatible. According to motherboard´s designers INTR 
are standard and Windows XP run too without any changes.

Saulo

Re: IDE cs5530 hda: lost interrupt

[Saulo]

Mark Lord wrote:

> Saulo wrote:
>
>> Hi all,
>>
>> any help is wellcome...
>>
>> --------------------
>> CPU: NSC Geode(TM) Integrated Processor by National Semi stepping 02
>> ...
>> ide: Assuming 33MHz system bus speed for PIO modes; override with 
>> idebus=xx
>> CS5530: ide CONTROLLER AT pci SLOT 0000:00:12.2
>> CS5530: chipset revision 0
>> CS5530: not 100% native mode: will probe irqs later
>> PCI: Enabling bus mastering for device 0000:00:12.2
>> PCI: Setting latency timer of device 0000:00:12.2 to 64
>>     ide0: BM-DMA at 0xfc00-0xfc07, BIOS settings: hda:pio, hdb:pio
>>     ide1: BM-DMA at 0xfc08-0xfc0f, BIOS settings: hdc:pio, hdd:pio
>> hda: CF 32MB, CFA DISK drive
>> hda: IRQ probe failed (0xfeba)    >>> I think my problem may start 
>> here, but when I fix to IRQ 14 in try_to_identify() to hda the 
>> problem persist
>> ide0 at 0x1f0-0x1f7,0x3f6 on irq14
>> hdc: Hitachi CV 5.1.1, CFA DISK drive
>> ide1 at 0x170-0x177,0x376 on irq 15 (serialized with ide0)
>> hda: max request size: 128KiB
>> hda: 62976 sectors (32MB) w/1KiB Cache, CHS=492/4/32
>> hda:<4>hda: lost interrupt
>> hda: lost interrupt
>> hda: lost interrupt
>> ...
>
> ...
>
> Send me one of those devices and I'll fix it.
>
> Cheers
> -

Sorry... I would like but I can´t send one of this device to you. If you 
have any idea about this problem, any help is wellcome.

Saulo