Linux - GeneralThis Linux forum is for general Linux questions and discussion.
If it is Linux Related and doesn't seem to fit in any other forum then this is the place.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
/dev/hda:
Timing buffer-cache reads: 128 MB in 0.82 seconds =156.10 MB/sec
Timing buffered disk reads: 64 MB in 1.28 seconds = 50.00 MB/sec
I don't know if any other variables effect this... I've got a lot of RAM on this box, so it's not like it had to swap to perform the test. It's got a decently fast processor as well. ::shrug::
Its all ATA controller for the most part, RAM can come into it on the very low end machines. You have an ATA 100 card running a drive at UDMA5, ATA100, that has to be a 7200... good rig. The drive can do 133 according to its parm though.
Everything I had in that huge list was running at a pio mode or mdma, except for that last one which was UDMA4, or ATA66. Still, 50 is a great speed for that rig.
Location: SW Coast of Florida, USA-- in fact, ground zero for Charley is where my town is
Distribution: Mandrake 10 Community, SuSE 9+
Posts: 167
Rep:
Half-elf, try this:
As su, in console:
hdparm -d1 /dev/hda
hdparm -d1 /dev/hdb
See if things radically speed up when you repeat the becnchmarking. This turns on DMA mode, and when I turned mine off with:
hdparm -d0 /dev/hdd
the second rate of 46 MB/sec ended up being 3.4 MB/sec instead. Lesson, try turning just the DMA on, to start with. Some distros default it ON, some default it OFF when they install.
Linux does not know my Intel (partly a Brookdale, but an EARLY ICH4 South Bridge) chipset either, but it does like DMA much better than non-DMA mode.
In your case, Xiphias, check and see if in device manager|hard drives the properties for the HDs show that they are NOT in DMA mode, then if you can change to DMA mnode try that-- this is in Windows. It sounds like you have a slightly older HD or HDs, or the CD-ROM or CD-RW if you have one is in DMA mode and the rest of the IDE needs to be. I would leave Linux alone.
On the other hand, if the HDs are in DMA mode when you first check, get them out of it with those results-- Windows can handle HDs not in DMA mode usually, and sometimes is faster that way. Linux does NOT do this, so I would not alter Linux's settings.
WOOOW, now my benchmark is around 37mb/sec!!!
that's good enought for me If I was not using DMA, can you tell me what I was using? PIO?
now if I reboot, will I lost this setting?
Also, how can I know what's the speed of the DMA I'm using (there are DMA33,66,100,133 true?)
If I remember my hardware course right, 100 and 133 are UDMA mode, can I set this up? and can I know witout dangerous trying if my computer can handle this?
Location: SW Coast of Florida, USA-- in fact, ground zero for Charley is where my town is
Distribution: Mandrake 10 Community, SuSE 9+
Posts: 167
Rep:
Um,
DMA tells the IDE controller to talk to the HD in a special way. it does not, with Linux, interfere with the UDMA\ATA settings, whihc can be looked like like this:
ATA0 or ATA1 (actually not UDMA, this is 8MB old IDE or less)
UDMA2 = 16.66 MB/sec
UDMA3 = 33 MB/sec
UDMA4 = 66MB/sec
UDMA5 = 100MB/sec
UDMA6 = 133MB/sec and hdparm does not grok this one yet.
Many mobos limit effective throughput to the speed of PCI-- which is usually about 33 MB/sec. And it looks like it is so in your case. My chipset actually allows 100 MB/sec throughput, but my stock kernel does not know how to handle a P4 chipset yet-- it thinks it is a P II (also used for early P III's) chipset. SO, in my case the max upper thorughput I could get in theory out of my drive is 66 MB/sec. But the HDPARM thing is calcing time from start of asking for request to dump buffer to tiem data is written to /tmp. So, since X is running while I do this I am losing a lot of efficiency adsn getting about 50% to 67% depending on drive-- and in fact I benched each partition and thus found out it even allows for seek time variation this way.
Location: SW Coast of Florida, USA-- in fact, ground zero for Charley is where my town is
Distribution: Mandrake 10 Community, SuSE 9+
Posts: 167
Rep:
PIO is only used usually for drives that are not 33 MB/sec or faster-- most drives that are that fast or faster set themselves to PIO4 when in what is called LBA mode and UDMA mode is 3 or better. PIO becomes something you do not need to play with given what happened.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.