2.6.11.8 disabled dma

xushi · 04-30-2005, 03:06 PM

For some reason, after upgrading to 2.6.11.8, my system became unbelievably slow, especially when trying to load applications. I just checked hdparm, and it says that my dma is turned off

Code:

/dev/hda1:
 multcount    = 16 (on)
 IO_support   =  0 (default 16-bit)
 unmaskirq    =  0 (off)
 using_dma    =  0 (off)
 keepsettings =  0 (off)
 readonly     =  0 (off)
 readahead    = 256 (on)
 geometry     = 65535/16/63, sectors = 10489618944, start = 135807903

when typing hdparm -d /dev/hda, all i get is

Code:

/dev/hda:
 using_dma    =  0 (off)

hdparm -d1 /dev/hda

Code:

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

searching through my .config file,

Code:

root@xushitop:/usr/src/linux# cat .config | grep DMA
CONFIG_GENERIC_ISA_DMA=y
CONFIG_BLK_DEV_IDEDMA_PCI=y
# CONFIG_BLK_DEV_IDEDMA_FORCED is not set
CONFIG_IDEDMA_PCI_AUTO=y
# CONFIG_IDEDMA_ONLYDISK is not set
CONFIG_BLK_DEV_IDEDMA=y
# CONFIG_IDEDMA_IVB is not set
CONFIG_IDEDMA_AUTO=y

Hmm.. =/

EDIT:
Could it be that i don't have a chipset enabled in ATA/ATAPI/MFM/RLL support?
my chipset afaik is an

Code:

00:1e.0 PCI bridge: Intel Corp. 82801BAM/CAM PCI Bridge (rev 81)
00:1f.0 ISA bridge: Intel Corp. 82801DBM LPC Interface Controller (rev 01)
00:1f.1 IDE interface: Intel Corp. 82801DBM (ICH4) Ultra ATA Storage Controller (rev 01)

The closest i can find in the kernel is the Intel PIIXn chipset support

xushi · 04-30-2005, 03:31 PM

I'm very sorry =/

My dissertation being due at the end of the week makes me panick lately whenever i get any problems with my system.. (remember OO crashing? heh) I should try and solve the problem before posting.

Enabling Intel PIIXn chipset support solved the problem.

On a side note, (not to let the thread go to waste), my partition is 9.5% non contiguous.. Is that an acceptable percentage?

Thanks for the support.

BenneJezzerette · 04-30-2005, 06:33 PM

Hello, I just now, Downloaded this Linux 2.6.11.8 and have not yet decompressed it. From the file Size it appears they have added some new and hopefully improved features. I will take the deep set chance and try this out. So it appears you have successfully made this Kernel work. Maybe then I can too. I have had issues with 2.6.11.5 6 and 7, so maybe the fixed the depreciated intermodule.c in this new Linux 2.6.11.8. Will have to search around it before I do anything with it.

netcrawl · 04-30-2005, 08:13 PM

I've got 2.6.11.8 running on 2 slack-current boxes with absolutely no difficulty, regardless of depreciated(sic) modules. Same as all the previous releases.

BenneJezzerette · 04-30-2005, 08:59 PM

Quote:

Originally posted by netcrawl
I've got 2.6.11.8 running on 2 slack-current boxes with absolutely no difficulty, regardless of depreciated(sic) modules. Same as all the previous releases.

good, I did some modifications to the install.sh and the makefile settings,. chaning the directiories it will write to rather than root. This should make it work right then

http://www.mozillaes.org/components/...les/pengin.gif

carboncopy · 04-30-2005, 10:11 PM

I don't want to hijack the post but, anybody tried compiling kernel with gcc 4.0? It is not beta anymore.

I had issue compiling gnupg 1.4.1 in Mac Os X.4 with gcc4.0

Is there away for me to switch gcc version like the one in Mac Os X?

Code:

gcc_select VERSION_NO

I know this could be done using bash script. But how do I compile it to meet such requirements that all gcc executable would have -VERSION?