IDE DMA/VIA woes on SuSE 2.4.19-167

IDE DMA/VIA woes on SuSE 2.4.19-167

[Matthias Andree]

Hi,

I have some issues here:

Plextor PX-W4824TA 1.03 as hdc (no hdd)
VIA KT133

00:07.1 IDE interface: VIA Technologies, Inc. VT82C586B PIPC Bus Master IDE (rev 10) (prog-if 8a [Master SecP PriP])
        Flags: bus master, medium devsel, latency 32
        I/O ports at ffa0 [size=16]
        Capabilities: [c0] Power Management version 2

When I try to enable DMA (hdparm -d1 or hdparm -d1 -X66), hdparm -tT
chokes, SuSE k_athlon-2.4.19-167. FreeBSD-5 (with atapicam) is fine and
uses UDMA33.

SuSE's kernel seems to be fine on a different hardware (VIA KT333 +
Toshiba SD-M1612).

I've tried 2.4.21-pre5 which crashes on boot, I haven't yet been able to
go fishing for the Ooops and decode it (I have serial console here, so
it's just a matter of finding the time to do that).

ide-scsi makes no difference (not that I had expected that).

Which kernel version should I try next before thinking about this for
longer?

-- 
Matthias Andree

Re: IDE DMA/VIA woes on SuSE 2.4.19-167

[Meino Christian Cramer]

From: Matthias Andree <matthias.andree@gmx.de>
Subject: IDE DMA/VIA woes on SuSE 2.4.19-167
Date: Wed, 5 Mar 2003 03:44:46 +0100

Hi,

 ..."similiar" here:

 CDRW: Plextor 48/24/10
 Linux 2.4.20
 EPoX 8K5A3+ (VIA KT333)

 Copying the contents of CD is VERY SLOW...

 Copying a CD is more worst than a CDRW of comparable contents.

 ....sigh

 Keep Hacking!
 Meino

> Hi,
> 
> I have some issues here:
> 
> Plextor PX-W4824TA 1.03 as hdc (no hdd)
> VIA KT133
> 
> 00:07.1 IDE interface: VIA Technologies, Inc. VT82C586B PIPC Bus Master IDE (rev 10) (prog-if 8a [Master SecP PriP])
>         Flags: bus master, medium devsel, latency 32
>         I/O ports at ffa0 [size=16]
>         Capabilities: [c0] Power Management version 2
> 
> When I try to enable DMA (hdparm -d1 or hdparm -d1 -X66), hdparm -tT
> chokes, SuSE k_athlon-2.4.19-167. FreeBSD-5 (with atapicam) is fine and
> uses UDMA33.
> 
> SuSE's kernel seems to be fine on a different hardware (VIA KT333 +
> Toshiba SD-M1612).
> 
> I've tried 2.4.21-pre5 which crashes on boot, I haven't yet been able to
> go fishing for the Ooops and decode it (I have serial console here, so
> it's just a matter of finding the time to do that).
> 
> ide-scsi makes no difference (not that I had expected that).
> 
> Which kernel version should I try next before thinking about this for
> longer?
> 
> -- 
> Matthias Andree
> -
> 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/
> 

[2.4.20]: Question about myst. interrupt thingy

[Meino Christian Cramer]

Hi,

 System:
 generic Linux 2.4.20 (dRock 1.60 based with heavy changes)
 gcc 2.95.3
 Athlon XP 2400+
 256MB DDR RAM (Samsung)
 EPoX 8K5A3+ (VIA KT333)
 Sound onboard (VIA8233)
 Radeon 7500 AGP (made by Sapphire) 

 Interrupts as follows:
            CPU0       
   0:    3617891          XT-PIC  timer
   1:      27017          XT-PIC  keyboard
   2:          0          XT-PIC  cascade
   5:      55271          XT-PIC  usb-uhci, VIA8233
   9:          0          XT-PIC  usb-uhci
  10:    3045481          XT-PIC  usb-uhci, radeon@PCI:1:0:0
  11:         30          XT-PIC  sym53c8xx
  14:     505263          XT-PIC  ide0
  15:      74476          XT-PIC  ide1
 NMI:          0 
 ERR:          7
 
 
 Now, I got the following message whiel booting the kernel:
 
 PCI: No IRQ known for interrupt pin A of device 00:11.1. Please try using pci=biosirq.
 
 If I am using pci=biosirq I got the same mesage without "Please try
 using pci=biosirq".
 
 What is THIS? What is that mean?
 Anything serious?
 
 Since my box has some "mysterious"  effects (locks for minutes as it
 seems of being under heaviest load, than unlocks again. Sometimes
 programs being "stable for years" (bash ie.) crashes and core dump.) I
 would be very interesting, whether this has something to do with it.
 
 Any help is very appreciated !
 Thank you very much in advance.
 
 Kind regards,
 Meino Cramer

Re: IDE DMA/VIA woes on SuSE 2.4.19-167

[Matthias Andree]

On Wed, 05 Mar 2003, Matthias Andree wrote:

> Plextor PX-W4824TA 1.03 as hdc (no hdd)
> VIA KT133
> 
> 00:07.1 IDE interface: VIA Technologies, Inc. VT82C586B PIPC Bus Master IDE (rev 10) (prog-if 8a [Master SecP PriP])
>         Flags: bus master, medium devsel, latency 32
>         I/O ports at ffa0 [size=16]
>         Capabilities: [c0] Power Management version 2
> 
> When I try to enable DMA (hdparm -d1 or hdparm -d1 -X66), hdparm -tT
> chokes, SuSE k_athlon-2.4.19-167. FreeBSD-5 (with atapicam) is fine and
> uses UDMA33.

Now, seems that PIO manages the default hdparm -tT block size, but DMA
doesn't; but the error messages in the kernel ring buffer aren't
specific.

I let go of hdparm -tT, installed Jörg Schilling's sdd and ran sdd
if=/dev/sr1 -onull -t bs=2048 and lo and behold, it passed and read a
data CD with up to 45x. So the remaining problem is that the ATAPI
drives stick to PIO for data reads.

I tried applying Andrew's ide-akpm on top of SuSE's kernel, to find it
crashes on boot on SuSE's hardware scan.

If that's interesting enough, I can try to dig up the crash messages
(I'll have to use a serial console for that though).