Stuart Henderson:

> from netbsd ata_wdc.c r1.56 and wd.c r1.279 (June 2004), their log:
>
> "Do not force ATA_SINGLE on devices that don't support READ/WRITE MULTI. We
> just have to take an interrupt for each sector.

Makes sense in principle.

> Tested with one laptop disk (which normally runs in DMA mode and was forced
> to single-sector transfers) and 3 CF cards. Increases the performance of
> the CF cards substantially (760KB/s->1240KB/s in one case, 410KB/s->750KB/s
> in the other two cases)."

Interestingly, it doesn't change the maximum read speed for my 8 GB
Transcend. It's still a measly ~1340 kB/s. However, the previously
considerable CPU usage drops to almost nothing.

For comparison, I get some 33 MB/s out of my (4-sector) SanDisk.
That's about 25 times faster.


BTW, I have both cards plugged into an Addonics Quad-CF PCI adapter.
http://addonics.com/products/flash_m...r/ad4cfprj.asp

pciide1 at pci0 dev 9 function 0 "CMD Technology PCI0680" rev 0x02
pciide1: bus-master DMA support present
pciide1: channel 0 configured to native-PCI mode
pciide1: using irq 11 for native-PCI interrupt
wd0 at pciide1 channel 0 drive 0: <SanDisk SDCFX4-8192>
wd0: 4-sector PIO, LBA, 7815MB, 16007040 sectors
wd0(pciide1:0:0): using PIO mode 4, Ultra-DMA mode 4
pciide1: channel 1 configured to native-PCI mode
wd1 at pciide1 channel 1 drive 0: <TRANSCEND>
wd1: 1-sector PIO, LBA, 7775MB, 15924384 sectors
wd1(pciide1:1:0): using PIO mode 4, Ultra-DMA mode 4

--
Christian "naddy" Weisgerber naddy@mips.inka.de