Re[2]: DMA !NOT ONLY! for triton again...

Re[2]: DMA !NOT ONLY! for triton again...

[Guennadi Liakhovetski]

-----Original Message-----An interesting addition:
I've just got a reply from WD - they say my disk only supports PIO4 and not DMA...
> I'm taking the case off the machine right now, i can guarantee you its not UDMA compatible, simply because this thing was made in early1997. :)
> 
> Here we go:
> 
> MDL WDAC21600-00H
> P/N 99-004199-000
> CCC F3 20 FEB 97
> DCM: BHBBKLP
> 
> I've got various of these hard drives in service, for the last 4 years.  Many run in windows pc's, and DMA mode in osr2 and newer, works, and is noticeablely faster.
> 
> Guennadi Liakhovetski wrote:
> 
> > Glad all this discussion helped at least one of us:-))
> >
> > As for me, as I already mentioned in my last posting - I don't know why BIOS makes the difference (as in your case) if ide.txt says it shouldn't?! Ok, chipset, perhaps, is fine. But what about the hard drive? You told you had WDC AC21600H. Can you PLEASE check waht CCC is marked on its label? PLEASE! I am trying to get an answer from WD on this, but not yet alas...
> >
> > And - COME ON, GUYS! - somebody MUST know the answer - how to spot the guilty one - kernel configuration / BIOS / chipset / disk???
> >
> > Guennadi
> >
> > > back in, started playing in the bios.  Finally fixed it.  I was getting > the same operation not permitted, that you
> > > were,until i got that bios setting. But it's making me
> > > wonder if it's something similar in your bios!
> > > I know it wasn't the actual UDMA setting in the bios, i'm
> > > wondering what it was though.  I'll put a keyboard on it,
> > > and poke around tonight or this weekend.
> 
> 

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: DMA !NOT ONLY! for triton again...

[Mike Dresser]

Now, the question is, can we trust a hard drive manufacturer support tech to know what they're talking about, with evidence to the contrary? :)

Somewhat like the 6 days i spent with a Cisco 802 and an 804, trying to get a link up.  Set it to National ISDN-1, dial 9 before dialing outbound number.  Connection comes up, but receive light is on, almost solid.  I figure I got the wrong switch type, so i give AT&T a call.  Bell tech comes out, finds out the line wasn't physically punched down into the block properly.

AT&T tells me it's a Northern DMS-100, and don't dial 9.  I set it up, can't get the connection to come up.  Invalid bear-caps.  I put the 9 back in, AT&T gives me a call, tells me to stop dialing 9.

Bell gets a call.  Turns out, it IS a National ISDN-1, and you're supposed to dial 9.  Seems that if you don't dial 9, it goes through AT&T's VOICE system.  For an ISDN call.  As you can tell, this doens't work.  :P  So after 6 days of learning more about cisco than I really wanted to, I get the link up.

Moral of the story?

Never trust a manufacturer. :)  That, and don't listen to your boss, when he keeps saying my Cisco Config must be bad, or the router is DOA.  Trust your instincts.

Guennadi Liakhovetski wrote:

> -----Original Message-----An interesting addition:
> I've just got a reply from WD - they say my disk only supports PIO4 and not DMA...
> > I'm taking the case off the machine right now, i can guarantee you its not UDMA compatible, simply because this thing was made in early1997. :)
> >
> > Here we go:
> >
> > MDL WDAC21600-00H
> > P/N 99-004199-000
> > CCC F3 20 FEB 97
> > DCM: BHBBKLP
> >
> > I've got various of these hard drives in service, for the last 4 years.  Many run in windows pc's, and DMA mode in osr2 and newer, works, and is noticeablely faster.
> >
> > Guennadi Liakhovetski wrote:
> >
> > > Glad all this discussion helped at least one of us:-))
> > >
> > > As for me, as I already mentioned in my last posting - I don't know why BIOS makes the difference (as in your case) if ide.txt says it shouldn't?! Ok, chipset, perhaps, is fine. But what about the hard drive? You told you had WDC AC21600H. Can you PLEASE check waht CCC is marked on its label? PLEASE! I am trying to get an answer from WD on this, but not yet alas...
> > >
> > > And - COME ON, GUYS! - somebody MUST know the answer - how to spot the guilty one - kernel configuration / BIOS / chipset / disk???
> > >
> > > Guennadi
> > >
> > > > back in, started playing in the bios.  Finally fixed it.  I was getting > the same operation not permitted, that you
> > > > were,until i got that bios setting. But it's making me
> > > > wonder if it's something similar in your bios!
> > > > I know it wasn't the actual UDMA setting in the bios, i'm
> > > > wondering what it was though.  I'll put a keyboard on it,
> > > > and poke around tonight or this weekend.
> >
> >

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

2.40.t11: Unresolved symbols while insmodding ip_tables.

[Meino Christian Cramer]

Hi,

 I got problems using ip_tables.
 Insmodding ip_tables gives me:

 /lib/modules/2.4.0-test11/kernel/net/ipv4/netfilter/ip_tables.o:
 unresolved symbol nf_unregister_sockopt
 /lib/modules/2.4.0-test11/kernel/net/ipv4/netfilter/ip_tables.o:
 unresolved symbol nf_register_sockopt
 /lib/modules/2.4.0-test11/kernel/net/ipv4/netfilter/ip_tables.o:
 insmod
 /lib/modules/2.4.0-test11/kernel/net/ipv4/netfilter/ip_tables.o fail

 I tried it with modutils 2.3.18 and the newest version 2.3.21, but
 both fail.

 The System.map does have these two symbols. netsyms.o does also
 have these symbols. Why haven't _I_ these symbols ;o)

 Any help very appreciated!

 Thank you very much in advance.

  keep hacking the right site of life ! :-)
  Meino  
   

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re[2]: DMA !NOT ONLY! for triton again...

[Guennadi Liakhovetski]

Well, yes, I thought they could not have known:-)) I'm absolutely stuck. If disk is fine, chipset is fine and supported by the kernel, then BIOS doesn't (or shouldn't) make a difference... Then WHAT ON THE EARTH??? Mike, have you been able to recall what BIOS option turned DMA on? Shall I write to Andre Hedrick directly? Or is there a mailing-list smth. like linux-ide?

> Now, the question is, can we trust a hard drive manufacturer
> support tech to know what they're talking about, with evidence to
> the contrary? :)

Thanks
Guennadi
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: Re[2]: DMA !NOT ONLY! for triton again...

[Andre Hedrick]

Guennadi,

I have watched this and even if UDMA is not supported cleanly by the
drive, the classic ATA-2 Multi-wrod DMA should be.  There was a time in
the past where WDC had some problems, but they have fixed most if not all
with "modern" drives.  I will be at WDC in two weeks, and I can raise the
issues with them.  Please spell them out completely.

Regards,

On Mon, 4 Dec 2000, Guennadi Liakhovetski wrote:

> Well, yes, I thought they could not have known:-)) I'm absolutely stuck. If disk is fine, chipset is fine and supported by the kernel, then BIOS doesn't (or shouldn't) make a difference... Then WHAT ON THE EARTH??? Mike, have you been able to recall what BIOS option turned DMA on? Shall I write to Andre Hedrick directly? Or is there a mailing-list smth. like linux-ide?
> 
> > Now, the question is, can we trust a hard drive manufacturer
> > support tech to know what they're talking about, with evidence to
> > the contrary? :)
> 
> Thanks
> Guennadi
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> Please read the FAQ at http://www.tux.org/lkml/
> 

Andre Hedrick
CTO Timpanogas Research Group
EVP Linux Development, TRG
Linux ATA Development

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re[4]: DMA !NOT ONLY! for triton again...

[Guennadi Liakhovetski]

Thanks, Andre! Briefly, the problem is that I can't turn on DMA for a WDC AC21600H CCC F6 drive, all the details are on this mailing list, but I'll put them all together in a single message and email it to you personally, ok? There is still a SMALL chance, that the problem is solvable locally by myself, or maybe it simply can't be solved in principle for this hardware. I very much doubt that my BIOS is quite capable of handling DMA on IDE, but that shouldn't prevent the kernel from using it, right? BTW, Mike, you might want to have a look at http://www.linux.com/tuneup/database.phtml/Hardware/000838.html - I just found it. Somebody there suggests, that you actually have to turn OFF PIO support in the kernel. Will million-check this later... But I wouldn't like to corrupt the disk by 'disabling all BIOS settings' as the author there suggests... At least not LBA...

Thanks again to all
Guennadi

-----Original Message-----

> 
> Guennadi,
> 
> I have watched this and even if UDMA is not supported cleanly by the
> drive, the classic ATA-2 Multi-wrod DMA should be.  There was a time in
> the past where WDC had some problems, but they have fixed most if not all
> with "modern" drives.  I will be at WDC in two weeks, and I can raise the
> issues with them.  Please spell them out completely.
> 
> Regards,
> 
> On Mon, 4 Dec 2000, Guennadi Liakhovetski wrote:
> 
> > Well, yes, I thought they could not have known:-)) I'm absolutely stuck. If disk is fine, chipset is fine and supported by the kernel, then BIOS doesn't (or shouldn't) make a difference... Then WHAT ON THE EARTH??? Mike, have you been able to recall what BIOS option turned DMA on? Shall I write to Andre Hedrick directly? Or is there a mailing-list smth. like linux-ide?
> > 
> > > Now, the question is, can we trust a hard drive manufacturer
> > > support tech to know what they're talking about, with evidence to
> > > the contrary? :)
> > 
> > Thanks
> > Guennadi
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > Please read the FAQ at http://www.tux.org/lkml/
> > 
> 
> Andre Hedrick
> CTO Timpanogas Research Group
> EVP Linux Development, TRG
> Linux ATA Development
> 
> 
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: Re[4]: DMA !NOT ONLY! for triton again...

[Alan Cox]

Certain older WDC drives are explicitly blacklisted due to firmware bugs.
WDC put out firmware upgrades but given no answer from them on how to be sure
a drive was upgraded we play safe

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: DMA !NOT ONLY! for triton again...

[Mike Dresser]

Well, i just checked ide-dma.c, and the WDC 21600 isn't listed.  31600 is, but no
21600
I've been using the 21600 for awhile now with DMA enabled, under reasonable load,
and it seems to hold up.
Guennadi: I don't suppose you can get your hands on a different size/brand drive
long enough to plug it in, and see if it allows DMA?

Alan Cox wrote:

> Certain older WDC drives are explicitly blacklisted due to firmware bugs.
> WDC put out firmware upgrades but given no answer from them on how to be sure
> a drive was upgraded we play safe

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: DMA !NOT ONLY! for triton again...

[Guennadi Liakhovetski]

Hi everybody

Tried turning PIO and/or PCI burst in BIOS off - no use:-( The former one
only slows the disk a lot.

Andre, I've sent you an email with the entire story, forgot to mention
there - a few people did draw my attention to 'PIIX: neither IDE port
enabled (BIOS)' and neighbouring messages in dmesg, but nobody was able to
explain what particularly is going on (BIOS does recognise HD and CD-ROM
fine) and how to fix it.

Re: PCI clock... Something somewhere (can't find now) made me think that
my MB is setting the PCI clock synchronously with the CPU clock, i.e. it
is 25MHz in my case... Any ideas where I could see it?:-) In BIOS 'Latency
Timer (PCI Clocks)' is shown as 66MHz, does it mean that my PCI bus clock
IS @ 66/2=33MHz?

> Guennadi: I don't suppose you can get your hands on a different size/brand drive
> long enough to plug it in, and see if it allows DMA?

I'll try... Not sure though...

Thanks
Guennadi
___

Dr. Guennadi V. Liakhovetski
Department of Applied Mathematics
University of Sheffield, U.K.
email: G.Liakhovetski@sheffield.ac.uk





-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: DMA !NOT ONLY! for triton again...

[Guennadi Liakhovetski]

> Re: PCI clock... Something somewhere (can't find now) made me think that
> my MB is setting the PCI clock synchronously with the CPU clock, i.e. it
> is 25MHz in my case... Any ideas where I could see it?:-)

I found it - it's in ide.txt too. Isn't Mark Lord around, btw?:-) Maybe he
could spread some light on the problem...

Thanks
Guennadi
___

Dr. Guennadi V. Liakhovetski
Department of Applied Mathematics
University of Sheffield, U.K.
email: G.Liakhovetski@sheffield.ac.uk


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: DMA !NOT ONLY! for triton again...

[Andre Hedrick]

On Wed, 6 Dec 2000, Guennadi Liakhovetski wrote:

> > Re: PCI clock... Something somewhere (can't find now) made me think that
> > my MB is setting the PCI clock synchronously with the CPU clock, i.e. it
> > is 25MHz in my case... Any ideas where I could see it?:-)
> 
> I found it - it's in ide.txt too. Isn't Mark Lord around, btw?:-) Maybe he
> could spread some light on the problem...
> 
> Thanks
> Guennadi
> ___
> 
> Dr. Guennadi V. Liakhovetski
> Department of Applied Mathematics
> University of Sheffield, U.K.
> email: G.Liakhovetski@sheffield.ac.uk

True Mark Lord is gone, but I do not go by my middle name Mark.
I thought I gave you answer.

Cheers,

Andre (Mark) Hedrick
CTO Timpanogas Research Group
EVP Linux Development, TRG
Linux ATA Development

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/