David Wuertele | 6 May 02:53 2003

Compact Flash PIO 4 tuning

Howdy,

I'm working on an embedded system that has a CF card on the IDE bus.
The CF spec only allows PIO mode, and my card can do PIO 4.  However,
when I try to stream data from this card into memory at 2 MBytes per
second, it pegs the CPU at 100%.  This is ruining the performance of
my application, which has to process this data in real time.

I realize that PIO 4 requires CPU involvement, but are there ways I
can tune the kernel for copying large files to memory, so that my
application doesn't get starved of CPU?

Thanks,
Dave

-
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Vojtech Pavlik | 6 May 07:47 2003
Picon

Re: Compact Flash PIO 4 tuning

On Mon, May 05, 2003 at 05:53:35PM -0700, David Wuertele wrote:
> Howdy,
> 
> I'm working on an embedded system that has a CF card on the IDE bus.
> The CF spec only allows PIO mode, and my card can do PIO 4.  However,
> when I try to stream data from this card into memory at 2 MBytes per
> second, it pegs the CPU at 100%.  This is ruining the performance of
> my application, which has to process this data in real time.
> 
> I realize that PIO 4 requires CPU involvement, but are there ways I
> can tune the kernel for copying large files to memory, so that my
> application doesn't get starved of CPU?

If you use the full bandwidth of the PIO device, you use 100% CPU.
That's it and there's no way around. You can check that the IDE/CF
interface is set to PIO4 and that under Linux you have IRQ unmasking
enabled, those two will help, though the first statement is still
valid.

--

-- 
Vojtech Pavlik
SuSE Labs, SuSE CR
-
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Taral | 12 May 06:39 2003
Picon

Modular IDE

I've managed to shoe-horn IDE into compiling itself as a module... Patch
is attached. Any comments?

-- 
Taral <taral <at> taral.net>
This message is digitally signed. Please PGP encrypt mail to me.
"Most parents have better things to do with their time than take care of
their children." -- Me
diff -Nru a/drivers/ide/Makefile b/drivers/ide/Makefile
--- a/drivers/ide/Makefile	Sun May 11 15:26:06 2003
+++ b/drivers/ide/Makefile	Sun May 11 15:26:06 2003
 <at>  <at>  -12,19 +12,23  <at>  <at> 

 # Core IDE code - must come before legacy

-obj-$(CONFIG_BLK_DEV_IDE)		+= ide-io.o ide-probe.o ide-geometry.o ide-iops.o ide-taskfile.o
ide.o ide-lib.o ide-default.o
 obj-$(CONFIG_BLK_DEV_IDEDISK)		+= ide-disk.o
 obj-$(CONFIG_BLK_DEV_IDECD)		+= ide-cd.o
 obj-$(CONFIG_BLK_DEV_IDETAPE)		+= ide-tape.o
 obj-$(CONFIG_BLK_DEV_IDEFLOPPY)		+= ide-floppy.o

-obj-$(CONFIG_BLK_DEV_IDEPCI)		+= setup-pci.o
-obj-$(CONFIG_BLK_DEV_IDEDMA_PCI)	+= ide-dma.o
-obj-$(CONFIG_BLK_DEV_IDE_TCQ)		+= ide-tcq.o
-obj-$(CONFIG_BLK_DEV_IDEPNP)		+= ide-pnp.o
-
+ide-mod-y					:= ide-probe.o ide-io.o ide-geometry.o ide-iops.o ide-taskfile.o ide.o ide-lib.o ide-default.o
(Continue reading)

Helge Hafting | 13 May 11:33 2003
Picon
Picon

[PATCH] Some Kconfig corrections for ide drivers

Ide hardware seems to have moved from drivers/ide
to drivers/ide/pci or drivers/ide/legacy.

This patch updates Kconfig accordingly, so people
reading the help texts are told to look at
<file:drivers/ide/pci/sis5513.c>
instead of the nonexistant
<file:drivers/ide/sis5513.c>
There are similiar updated for the other ide drivers.

Helge Hafting

--- Kconfig.orig        2003-05-05 01:53:14.000000000 +0200
+++ Kconfig     2003-05-13 11:18:19.000000000 +0200
 <at>  <at>  -456,7 +456,7  <at>  <at> 

           If you say Y here, you also need to say Y to "Use DMA by default
           when available", above.  Please read the comments at the top of
-         <file:drivers/ide/alim15x3.c>.
+         <file:drivers/ide/pci/alim15x3.c>.

           If unsure, say N.

 <at>  <at>  -543,7 +543,7  <at>  <at> 
         depends on BLK_DEV_HPT34X && IDEDMA_PCI_WIP
         help
           This is a dangerous thing to attempt currently! Please read the
-         comments at the top of <file:drivers/ide/hpt34x.c>.  If you say Y
+         comments at the top of <file:drivers/ide/pci/hpt34x.c>.  If 
you say Y
(Continue reading)

pacho baratta | 22 May 14:08 2003
Picon

/dev/hda problem with samsung hd

i'm working with a disk (samsung sv1022d) that my system cannot see. i
have a live cd with kernel 2.4.20 and glibc-2.2.5.

now when i boot i find in dmesg:

hda samsung sv1022d my100 ata disk drive
hda: task_no_data_intr 0x51 {DriveRady SeekComplete Error}
hda: task_no_data_intr 0x04 {DriveStatusError}
hda: 1 sectors (0 mb) w/480KiB Cache CHS=1240/255/63

....

hda : hda : bad access : block 0 count 2
end request I/O error dev 03:030 (hda) sector 0

so i tried parted, this is the response:
/dev/hda disk label not recognized
the operating system think the geometry on dev/hda is 0/255/63

and doesn't give me acess to the disk.

same for fdisk:
it simply prints: impossible to read /dev/hda.

hdparm -iv /dev/hda says:
multcount	0 off
i/o_support		0	
unmaskirq	0
dma_supprt		1
keepsettings		0
(Continue reading)

Jens Axboe | 22 May 14:10 2003
Picon

Re: /dev/hda problem with samsung hd

On Thu, May 22 2003, pacho baratta wrote:
> i'm working with a disk (samsung sv1022d) that my system cannot see. i
> have a live cd with kernel 2.4.20 and glibc-2.2.5.
> 
> now when i boot i find in dmesg:
> 
> hda samsung sv1022d my100 ata disk drive
> hda: task_no_data_intr 0x51 {DriveRady SeekComplete Error}
> hda: task_no_data_intr 0x04 {DriveStatusError}
> hda: 1 sectors (0 mb) w/480KiB Cache CHS=1240/255/63

I fixed this some time ago, the patch is in 2.4.21-rc kernels. So please
try 2.4.21-rc2 and see if it works.

--

-- 
Jens Axboe

-
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Bartlomiej Zolnierkiewicz | 22 May 16:11 2003
Picon
Picon

Re: /dev/hda problem with samsung hd


Can you try kernel >= 2.4.21-pre7 (latest is 2.4.21-rc2) ?
It contains fix for some Samsung drives which don't abort unknown
commands.

Regards,
--
Bartlomiej

On 22 May 2003, pacho baratta wrote:

> i'm working with a disk (samsung sv1022d) that my system cannot see. i
> have a live cd with kernel 2.4.20 and glibc-2.2.5.
>
> now when i boot i find in dmesg:
>
> hda samsung sv1022d my100 ata disk drive
> hda: task_no_data_intr 0x51 {DriveRady SeekComplete Error}
> hda: task_no_data_intr 0x04 {DriveStatusError}
> hda: 1 sectors (0 mb) w/480KiB Cache CHS=1240/255/63
>
> ....
>
> hda : hda : bad access : block 0 count 2
> end request I/O error dev 03:030 (hda) sector 0
>
>
> so i tried parted, this is the response:
> /dev/hda disk label not recognized
> the operating system think the geometry on dev/hda is 0/255/63
(Continue reading)

Bartlomiej Zolnierkiewicz | 23 May 02:47 2003
Picon
Picon

Re: /dev/hda problem with samsung hd


Hi Jens,
already solved, it was exactly this issue.
--
Bartlomiej

On Thu, 22 May 2003, Jens Axboe wrote:

> On Thu, May 22 2003, pacho baratta wrote:
> > i'm working with a disk (samsung sv1022d) that my system cannot see. i
> > have a live cd with kernel 2.4.20 and glibc-2.2.5.
> >
> > now when i boot i find in dmesg:
> >
> > hda samsung sv1022d my100 ata disk drive
> > hda: task_no_data_intr 0x51 {DriveRady SeekComplete Error}
> > hda: task_no_data_intr 0x04 {DriveStatusError}
> > hda: 1 sectors (0 mb) w/480KiB Cache CHS=1240/255/63
>
> I fixed this some time ago, the patch is in 2.4.21-rc kernels. So please
> try 2.4.21-rc2 and see if it works.
>
> --
> Jens Axboe

-
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

(Continue reading)

Gutko | 25 May 22:06 2003
Picon

Still no DMA on boot on S-ATA (Asus A7N8X-deluxe)

Hardware:
Asus A7N8X -deluxe v1.04 bios 1004
SI3112A sata chipset
IBM vancouver2 180GXP pata drive connected to sata trough Asus sata<->pata
converter
Kernel 2.4.21-rc3 (and all older 2.4.x with sata support)

Why my hdd is always detected as PIO?
I can enable udma5 using hdparm, later in rc.local but why it is not working
on boot?

----------DMESG-----------
Uniform Multi-Platform E-IDE driver Revision: 7.00beta3-.2.4
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
NFORCE2: IDE controller at PCI slot 00:09.0
NFORCE2: chipset revision 162
NFORCE2: not 100% native mode: will probe irqs later
AMD_IDE: PCI device 10de:0065 (nVidia Corporation) (rev a2) UDMA100
controller on pci00:09.0
    ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:DMA
    ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:DMA
SiI3112 Serial ATA: IDE controller at PCI slot 01:0b.0
SiI3112 Serial ATA: chipset revision 1
SiI3112 Serial ATA: not 100% native mode: will probe irqs later
    ide2: MMIO-DMA at 0xe0f54000-0xe0f54007, BIOS settings: hde:pio, hdf:pio
    ide3: MMIO-DMA at 0xe0f54008-0xe0f5400f, BIOS settings: hdg:pio, hdh:pio
hdc: LITE-ON LTR-48125S, ATAPI CD/DVD-ROM drive
hdd: HL-DT-STDVD-ROM GDR8160B, ATAPI CD/DVD-ROM drive
hde: IC35L180AVV207-1, ATA DISK drive
hdg: no response (status = 0xfe)
(Continue reading)

Gutko | 25 May 22:10 2003
Picon

Still no DMA on boot on S-ATA (Asus A7N8X-deluxe)

Hardware:
Asus A7N8X -deluxe v1.04 bios 1004
SI3112A sata chipset
IBM vancouver2 180GXP pata drive connected to sata trough Asus sata<->pata
converter
Kernel 2.4.21-rc3 (and all older 2.4.x with sata support)

Why my hdd is always detected as PIO?
I can enable udma5 using hdparm, later in rc.local but why it is not working
on boot?

----------DMESG-----------
Uniform Multi-Platform E-IDE driver Revision: 7.00beta3-.2.4
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
NFORCE2: IDE controller at PCI slot 00:09.0
NFORCE2: chipset revision 162
NFORCE2: not 100% native mode: will probe irqs later
AMD_IDE: PCI device 10de:0065 (nVidia Corporation) (rev a2) UDMA100
controller on pci00:09.0
    ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:DMA
    ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:DMA
SiI3112 Serial ATA: IDE controller at PCI slot 01:0b.0
SiI3112 Serial ATA: chipset revision 1
SiI3112 Serial ATA: not 100% native mode: will probe irqs later
    ide2: MMIO-DMA at 0xe0f54000-0xe0f54007, BIOS settings: hde:pio, hdf:pio
    ide3: MMIO-DMA at 0xe0f54008-0xe0f5400f, BIOS settings: hdg:pio, hdh:pio
hdc: LITE-ON LTR-48125S, ATAPI CD/DVD-ROM drive
hdd: HL-DT-STDVD-ROM GDR8160B, ATAPI CD/DVD-ROM drive
hde: IC35L180AVV207-1, ATA DISK drive
hdg: no response (status = 0xfe)
(Continue reading)


Gmane