IDE DMA disabled with via82cxxx on kernels >= 2.5.35

IDE DMA disabled with via82cxxx on kernels >= 2.5.35

[rwhron]

I have two machines with VIA chipsets.  Recent 2.5 
kernels boot with "DMA disabled".  

On Athlon, hdparm -tT /dev/hda is about 8 times higher 
on 2.4 than 2.5 kernels.  hdparm -i says dma is (on) though.

I tried CONFIG_BLK_DEV_IDEDMA_FORCED=y, but that didn't 
eliminated the "DMA disabled" message.  Other IDE settings are:

CONFIG_IDE=y
CONFIG_BLK_DEV_IDE=y
CONFIG_BLK_DEV_IDEDISK=y
CONFIG_IDEDISK_MULTI_MODE=y
CONFIG_BLK_DEV_IDECD=m
CONFIG_BLK_DEV_IDEPCI=y
CONFIG_BLK_DEV_IDEDMA_PCI=y
CONFIG_IDEDMA_PCI_AUTO=y
CONFIG_BLK_DEV_IDEDMA=y
CONFIG_IDEDMA_AUTO=y
CONFIG_BLK_DEV_IDE_MODES=y

boot message:
VP_IDE: IDE controller at PCI slot 00:07.1
VP_IDE: chipset revision 6
VP_IDE: not 100%% native mode: will probe irqs later
VP_IDE: VIA vt82c686b (rev 40) IDE UDMA100 controller on pci00:07.1
    ide0: BM-DMA at 0xd000-0xd007, BIOS settings: hda:DMA, hdb:pio
    ide1: BM-DMA at 0xd008-0xd00f, BIOS settings: hdc:DMA, hdd:pio
hda: IC35L040AVER07-0, ATA DISK drive
hda: IRQ probe failed (0xffffffba)
hda: DMA disabled
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14

lspci
00:00.0 Host bridge: VIA Technologies, Inc. VT8363/8365 [KT133/KM133] (rev 03)
00:01.0 PCI bridge: VIA Technologies, Inc. VT8363/8365 [KT133/KM133 AGP]
00:07.0 ISA bridge: VIA Technologies, Inc. VT82C686 [Apollo Super South] (rev 40)
00:07.1 IDE interface: VIA Technologies, Inc. VT82C586B PIPC Bus Master IDE (rev 06)
00:07.4 Bridge: VIA Technologies, Inc. VT82C686 [Apollo Super ACPI] (rev 40)

I also have a K6/2 with a VIA chipset.  2.5.34 and earlier were fine.
2.5.34-mm4 and subsequent 2.5.x kernels boot with "DMA disabled".

lspci
00:00.0 Host bridge: VIA Technologies, Inc. VT82C598 [Apollo MVP3] (rev 04)
00:01.0 PCI bridge: VIA Technologies, Inc. VT82C598/694x [Apollo MVP3/Pro133x AGP]
00:07.0 ISA bridge: VIA Technologies, Inc. VT82C586/A/B PCI-to-ISA [Apollo VP] (rev 47)
00:07.1 IDE interface: VIA Technologies, Inc. Bus Master IDE (rev 06)

The K6/2 boots with:

Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
VP_IDE: IDE controller at PCI slot 00:07.1
VP_IDE: chipset revision 6
VP_IDE: not 100% native mode: will probe irqs later
VP_IDE: VIA vt82c586b (rev 47) IDE UDMA33 controller on pci00:07.1
ide0: BM-DMA at 0xe000-0xe007, BIOS settings: hda:DMA, hdb:DMA
ide1: BM-DMA at 0xe008-0xe00f, BIOS settings: hdc:DMA, hdd:DMA
hda: Maxtor 51536U3, ATA DISK drive
hdb: ATAPI CDROM, ATAPI CD/DVD-ROM drive
hda: DMA disabled
hdb: DMA disabled
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14

-- 
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html

Re: IDE DMA disabled with via82cxxx on kernels >= 2.5.35

[Alan Cox]

On Wed, 2003-01-08 at 05:16, rwhron@earthlink.net wrote:
> I have two machines with VIA chipsets.  Recent 2.5 
> kernels boot with "DMA disabled".  

It turns it on again but forgets to tell you. Its on the fix list
bu tfor 2.4 first

Re: IDE DMA disabled with via82cxxx on kernels >= 2.5.35

[rwhron]

>> Recent 2.5 kernels boot with "DMA disabled".

> It turns it on again but forgets to tell you. Its on the fix list
> but for 2.4 first

Actual throughput on Athlon is better then hdparm lets on.  A 
dd test on 2.5.53-mm1 shows about 120 MB/sec for a file that fits in 
memory.

# time dd if=/dev/zero of=/tmp/junk bs=4M count=100
100+0 records in
100+0 records out

real    0m3.312s
user    0m0.002s
sys     0m3.267s

# ls -lh /tmp/junk
-rw-r--r--    1 root     root         400M Jan 13 21:43 /tmp/junk

hdparm -tT on the K6/2 shows similar throughput between 2.4 and 2.5.

Maybe my hdparm is just wacky on 2.5.  

2.4.20-pre3-jam1:

/dev/hda:
 Timing buffer-cache reads:   128 MB in  0.70 seconds =182.86 MB/sec
 Timing buffered disk reads:  64 MB in  2.07 seconds = 30.92 MB/sec

2.5.53-mm1

/dev/hda:
 Timing buffer-cache reads:   128 MB in  7.20 seconds = 17.78 MB/sec
 Timing buffered disk reads:  64 MB in 19.66 seconds =  3.26 MB/sec

-- 
Randy Hron
http://home.earthlink.net/~rwhron/kernel/bigbox.html

Re: IDE DMA disabled with via82cxxx on kernels >= 2.5.35

[Andrew Morton]

On Mon January 13 2003 18:49, rwhron@earthlink.net wrote:
>
> 2.4.20-pre3-jam1:
> 
> /dev/hda:
>  Timing buffer-cache reads:   128 MB in  0.70 seconds =182.86 MB/sec
>  Timing buffered disk reads:  64 MB in  2.07 seconds = 30.92 MB/sec
> 
> 2.5.53-mm1
> 
> /dev/hda:
>  Timing buffer-cache reads:   128 MB in  7.20 seconds = 17.78 MB/sec
>  Timing buffered disk reads:  64 MB in 19.66 seconds =  3.26 MB/sec

2.5.53-mm1 had a dud patch in it which made part of the kernel think HZ=100,
other parts think HZ=1000, depending on whether the .c file happened to
include config.h before including param.h.

Hence your nice 10x discrepancy.