DMA issues with FC4 on a Dell precision M70

[Alfredo Ferrari]

Hi

I have a Dell Precision M70 whose primary hard disk is a SATA disk 
(/dev/sda) running FC4 (kernel 2.6.15-1.1831_FC4).
The secondary hard disk is a PATA one (/dev/hdc, in the laptop modular 
bay) and apparently I cannot get DMA enabled on it, resulting in very poor
performances (2.5 MB/s against 25 MB/s for the same disk in the modular 
bay of a PATA only Dell D800 where it runs happily with DMA enabled).

If I try to enable DMA by hand I get:

[root@pcalfr ~]# hdparm -d1 /dev/hdc

/dev/hdc:
  setting using_dma to 1 (on)
  HDIO_SET_DMA failed: Operation not permitted
  using_dma    =  0 (off)

Every help/suggestion is greatly appreciated

              Alfredo Ferrari

P.S. The only messages of some relevance I can see at boot (from dmesg) 
are:

   DMA zone: 4096 pages, LIFO batch:0
   DMA32 zone: 0 pages, LIFO batch:0
   Normal zone: 225280 pages, LIFO batch:31
   HighMem zone: 294872 pages, LIFO batch:31

...

PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.2

...

Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
ide0: I/O resource 0x1F0-0x1F7 not free.
ide0: ports already in use, skipping probe
Probing IDE interface ide1...
hdc: ST94811A, ATA DISK drive
ide1 at 0x170-0x177,0x376 on irq 15
hdc: max request size: 1024KiB
hdc: 78140160 sectors (40007 MB) w/8192KiB Cache, CHS=16383/255/63
hdc: cache flushes supported
  hdc: hdc1 hdc2 hdc3 < hdc5 hdc6 hdc7 hdc8 >
ide-floppy driver 0.99.newide

...

SCSI subsystem initialized
libata version 1.20 loaded.
ata_piix 0000:00:1f.2: version 1.05
ACPI: PCI Interrupt 0000:00:1f.2[B] -> Link [LNKB] -> GSI 5 (level, low) -> IRQ
5
ata: 0x170 IDE port busy
PCI: Setting latency timer of device 0000:00:1f.2 to 64
ata1: SATA max UDMA/133 cmd 0x1F0 ctl 0x3F6 bmdma 0xBFA0 irq 14
ata1: dev 0 cfg 49:0f00 82:746b 83:7fe9 84:4023 85:7469 86:3e49 87:4023 88:203f
ata1: dev 0 ATA-6, max UDMA/100, 195371568 sectors: LBA48
ata1(0): applying bridge limits
ata1: dev 0 configured for UDMA/100
scsi0 : ata_piix
   Vendor: ATA       Model: Hitachi HTS72101  Rev: MCZO
   Type:   Direct-Access                      ANSI SCSI revision: 05
SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
SCSI device sda: drive cache: write back
SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
SCSI device sda: drive cache: write back
  sda: sda1 sda2 sda3 sda4 < sda5 sda6 sda7<6>input: PS/2 Mouse as
/class/input/input1
  sda8 sda9 >
sd 0:0:0:0: Attached scsi disk sda
input: AlpsPS/2 ALPS GlidePoint as /class/input/input2
kjournald starting.  Commit interval 600 seconds


+----------------------------------------------------------------------------+
|  Alfredo Ferrari                         ||  Tel.: [phone]         |
|  C.E.R.N.                                ||  Fax.: [phone]         |
|  European Laboratory for Particle Physics||                                |
|  AB Division / ATB Group                 ||  e-mail:                       |
|  1211 Geneva 23                          ||     Alfredo.Ferrari(a)cern.ch    |
|  Switzerland                             ||     Alfredo.Ferrari(a)mi.infn.it |
+----------------------------------------------------------------------------+

[Alessandro Brezzi]

2006/2/18, Alfredo Ferrari <list(a)pceet030.cern.ch&gt;:
>  Hi
>
>  I have a Dell Precision M70 whose primary hard disk is a SATA disk
>  (/dev/sda) running FC4 (kernel 2.6.15-1.1831_FC4).
>  The secondary hard disk is a PATA one (/dev/hdc, in the laptop modular
>  bay) and apparently I cannot get DMA enabled on it, resulting in very poor
>  performances (2.5 MB/s against 25 MB/s for the same disk in the modular
>  bay of a PATA only Dell D800 where it runs happily with DMA enabled).
>
>  If I try to enable DMA by hand I get:
>
>  [root@pcalfr ~]# hdparm -d1 /dev/hdc
>
>  /dev/hdc:
>    setting using_dma to 1 (on)
>    HDIO_SET_DMA failed: Operation not permitted
>    using_dma    =  0 (off)
>
>  Every help/suggestion is greatly appreciated
>
>                Alfredo Ferrari
>
>  P.S. The only messages of some relevance I can see at boot (from dmesg)
>  are:
>
>     DMA zone: 4096 pages, LIFO batch:0
>     DMA32 zone: 0 pages, LIFO batch:0
>     Normal zone: 225280 pages, LIFO batch:31
>     HighMem zone: 294872 pages, LIFO batch:31
>
>  ...
>
>  PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.2
>
>  ...
>
>  Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
>  ide: Assuming 33MHz system bus speed for PIO modes; override with
>  idebus=xx
>  ide0: I/O resource 0x1F0-0x1F7 not free.
>  ide0: ports already in use, skipping probe
>  Probing IDE interface ide1...
>  hdc: ST94811A, ATA DISK drive
>  ide1 at 0x170-0x177,0x376 on irq 15
>  hdc: max request size: 1024KiB
>  hdc: 78140160 sectors (40007 MB) w/8192KiB Cache, CHS=16383/255/63
>  hdc: cache flushes supported
>    hdc: hdc1 hdc2 hdc3 < hdc5 hdc6 hdc7 hdc8 >
>  ide-floppy driver 0.99.newide
>
>  ...
>
>  SCSI subsystem initialized
>  libata version 1.20 loaded.
>  ata_piix 0000:00:1f.2: version 1.05
>  ACPI: PCI Interrupt 0000:00:1f.2[B] -> Link [LNKB] -> GSI 5 (level, low)
>  -> IRQ 5
>  ata: 0x170 IDE port busy
>  PCI: Setting latency timer of device 0000:00:1f.2 to 64
>  ata1: SATA max UDMA/133 cmd 0x1F0 ctl 0x3F6 bmdma 0xBFA0 irq 14
>  ata1: dev 0 cfg 49:0f00 82:746b 83:7fe9 84:4023 85:7469 86:3e49 87:4023
>  88:203f
>  ata1: dev 0 ATA-6, max UDMA/100, 195371568 sectors: LBA48
>  ata1(0): applying bridge limits
>  ata1: dev 0 configured for UDMA/100
>  scsi0 : ata_piix
>     Vendor: ATA       Model: Hitachi HTS72101  Rev: MCZO
>     Type:   Direct-Access                      ANSI SCSI revision: 05
>  SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
>  SCSI device sda: drive cache: write back
>  SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
>  SCSI device sda: drive cache: write back
>    sda: sda1 sda2 sda3 sda4 < sda5 sda6 sda7<6>input: PS/2 Mouse as
>  /class/input/input1
>    sda8 sda9 >
>  sd 0:0:0:0: Attached scsi disk sda
>  input: AlpsPS/2 ALPS GlidePoint as /class/input/input2
>  kjournald starting.  Commit interval 600 seconds 


Hi Alfredo,
is your PATA disk  attached to the secondary IDE channel? Not sure about
laptop arch, but in desktop box primary and secondary IDE interface work
different.
HTH

--
Alessandro Brezzi

[Alfredo Ferrari]

Yes

my PATA disk is the master on the second IDE channel (hdc). This is 
actually the laptop modular bay where I can insert the DVD/2nd hard 
drive/2nd battery. The same disk in the same bay on the same channel
of a D800 (where the primary disk is PATA and not SATA as in the M70)
works nicley with DMA enabled at boot time.

Unfortunately it seems to be a known Linux kernel bug/limitation (see 
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=163418 )
a rather frustrating one I must say.

                    Thanks for the help

                       Alfredo

On Sat, 18 Feb 2006, Alessandro Brezzi wrote:

>  2006/2/18, Alfredo Ferrari <list(a)pceet030.cern.ch&gt;:
> >
> > Hi
> >
> > I have a Dell Precision M70 whose primary hard disk is a SATA disk
> > (/dev/sda) running FC4 (kernel 2.6.15-1.1831_FC4).
> > The secondary hard disk is a PATA one (/dev/hdc, in the laptop modular
> > bay) and apparently I cannot get DMA enabled on it, resulting in very poor
> > performances (2.5 MB/s against 25 MB/s for the same disk in the modular
> > bay of a PATA only Dell D800 where it runs happily with DMA enabled).
> >
> > If I try to enable DMA by hand I get:
> >
> > [root@pcalfr ~]# hdparm -d1 /dev/hdc
> >
> > /dev/hdc:
> >   setting using_dma to 1 (on)
> >   HDIO_SET_DMA failed: Operation not permitted
> >   using_dma    =  0 (off)
> >
> > Every help/suggestion is greatly appreciated
> >
> >               Alfredo Ferrari
> >
> > P.S. The only messages of some relevance I can see at boot (from dmesg)
> > are:
> >
> >    DMA zone: 4096 pages, LIFO batch:0
> >    DMA32 zone: 0 pages, LIFO batch:0
> >    Normal zone: 225280 pages, LIFO batch:31
> >    HighMem zone: 294872 pages, LIFO batch:31
> >
> > ...
> >
> > PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.2
> >
> > ...
> >
> > Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
> > ide: Assuming 33MHz system bus speed for PIO modes; override with
> > idebus=xx
> > ide0: I/O resource 0x1F0-0x1F7 not free.
> > ide0: ports already in use, skipping probe
> > Probing IDE interface ide1...
> > hdc: ST94811A, ATA DISK drive
> > ide1 at 0x170-0x177,0x376 on irq 15
> > hdc: max request size: 1024KiB
> > hdc: 78140160 sectors (40007 MB) w/8192KiB Cache, CHS=16383/255/63
> > hdc: cache flushes supported
> >   hdc: hdc1 hdc2 hdc3 < hdc5 hdc6 hdc7 hdc8 >
> > ide-floppy driver 0.99.newide
> >
> > ...
> >
> > SCSI subsystem initialized
> > libata version 1.20 loaded.
> > ata_piix 0000:00:1f.2: version 1.05
> > ACPI: PCI Interrupt 0000:00:1f.2[B] -> Link [LNKB] -> GSI 5 (level, low)
> > -> IRQ 5
> > ata: 0x170 IDE port busy
> > PCI: Setting latency timer of device 0000:00:1f.2 to 64
> > ata1: SATA max UDMA/133 cmd 0x1F0 ctl 0x3F6 bmdma 0xBFA0 irq 14
> > ata1: dev 0 cfg 49:0f00 82:746b 83:7fe9 84:4023 85:7469 86:3e49 87:4023
> > 88:203f
> > ata1: dev 0 ATA-6, max UDMA/100, 195371568 sectors: LBA48
> > ata1(0): applying bridge limits
> > ata1: dev 0 configured for UDMA/100
> > scsi0 : ata_piix
> >    Vendor: ATA       Model: Hitachi HTS72101  Rev: MCZO
> >    Type:   Direct-Access                      ANSI SCSI revision: 05
> > SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
> > SCSI device sda: drive cache: write back
> > SCSI device sda: 195371568 512-byte hdwr sectors (100030 MB)
> > SCSI device sda: drive cache: write back
> >   sda: sda1 sda2 sda3 sda4 < sda5 sda6 sda7<6>input: PS/2 Mouse as
> > /class/input/input1
> >   sda8 sda9 >
> > sd 0:0:0:0: Attached scsi disk sda
> > input: AlpsPS/2 ALPS GlidePoint as /class/input/input2
> > kjournald starting.  Commit interval 600 seconds
>
>
>
>  Hi Alfredo,
>  is your PATA disk  attached to the secondary IDE channel? Not sure about
>  laptop arch, but in desktop box primary and secondary IDE interface work
>  different.
>  HTH
>
>  --
>  Alessandro Brezzi
>  
-- 

+----------------------------------------------------------------------------+
|  Alfredo Ferrari                         ||  Tel.: [phone]         |
|  C.E.R.N.                                ||  Fax.: [phone]         |
|  European Laboratory for Particle Physics||                                |
|  AB Division / ATB Group                 ||  e-mail:                       |
|  1211 Geneva 23                          ||     Alfredo.Ferrari(a)cern.ch    |
|  Switzerland                             ||     Alfredo.Ferrari(a)mi.infn.it |
+----------------------------------------------------------------------------+

[François Patte]

Alfredo Ferrari wrote:
>  Yes
>  
>  my PATA disk is the master on the second IDE channel (hdc). This is
>  actually the laptop modular bay where I can insert the DVD/2nd hard
>  drive/2nd battery. The same disk in the same bay on the same channel
>  of a D800 (where the primary disk is PATA and not SATA as in the M70)
>  works nicley with DMA enabled at boot time.
>  
>  Unfortunately it seems to be a known Linux kernel bug/limitation (see
>  https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=163418 )
>  a rather frustrating one I must say. 
I don't know much about this, but I came accross this problem for
playing dvd with xine.

It seems to me that the problem occurs on laptop with sata hd: there is
a single controller for both ide drive and sata disk and you cannot set
the dma.

This problem does not occur with desktop where you have ide controller.

I have two solutions for this:

1- hdparm -c1 /dev/hdc

You can watch dvd in a better way (not perfect)

2- rip the dvd to the hard drive with dvdbackup and watch them (perfect)

Hope this help.

-- 
François Patte
UFR de mathématiques et informatique
Université Paris 5 - Paris
http://www.math-info.univ-paris5.fr/~patte

[Robin Laing]

François Patte wrote:
>  Alfredo Ferrari wrote:
>  
> >Yes
> >
> >my PATA disk is the master on the second IDE channel (hdc). This is
> >actually the laptop modular bay where I can insert the DVD/2nd hard
> >drive/2nd battery. The same disk in the same bay on the same channel
> >of a D800 (where the primary disk is PATA and not SATA as in the M70)
> >works nicley with DMA enabled at boot time.
> >
> >Unfortunately it seems to be a known Linux kernel bug/limitation (see
> >https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=163418 )
> >a rather frustrating one I must say.
>  
>  
>  I don't know much about this, but I came accross this problem for
>  playing dvd with xine.
>  
>  It seems to me that the problem occurs on laptop with sata hd: there is
>  a single controller for both ide drive and sata disk and you cannot set
>  the dma.
>  
>  This problem does not occur with desktop where you have ide controller.
>  
>  I have two solutions for this:
>  
>  1- hdparm -c1 /dev/hdc
>  
>  You can watch dvd in a better way (not perfect)
>  
>  2- rip the dvd to the hard drive with dvdbackup and watch them (perfect)
>  
>  Hope this help.
>   
Been away for a week so I am late responding.

Having had similar problems on a Toshiba laptop with an Intel ICH6 
controller.  Could not get DVD's to play and could not get DMA to work.

I can now play DVD's perfectly from my DVD player.  I played with the 
  'options libata atapi_enabled=1' settings and the latest kernel 
seems to have all working.  I would have to look at the laptop to see 
the exact changes (at home).  I could not get that libata settings to 
work at first.

I know that the changes did not work in grub but I had to play with 
the module config files.

It was also recommended to look at possible bios settings.
-- 
Robin Laing

[Robin Laing]

Robin Laing wrote:
>  François Patte wrote:
>  
> > Alfredo Ferrari wrote:
> > 
> >
>  
>  Been away for a week so I am late responding.
>  
>  Having had similar problems on a Toshiba laptop with an Intel ICH6 
>  controller.  Could not get DVD's to play and could not get DMA to work.
>  
>  I can now play DVD's perfectly from my DVD player.  I played with the 
>   'options libata atapi_enabled=1' settings and the latest kernel seems 
>  to have all working.  I would have to look at the laptop to see the 
>  exact changes (at home).  I could not get that libata settings to work 
>  at first.
>  
>  I know that the changes did not work in grub but I had to play with the 
>  module config files. 
FWIW, these are the changes I made.

options ide hdc=noprobe  (DVD drive)
options libata atapi_enabled=1

I don't remember making this change but I could have.
alias scsi_hostadapter ata_piix

-- 
Robin Laing