vt8235 fix, hopefully last variant

vt8235 fix, hopefully last variant

[Vojtech Pavlik]

[-- Attachment #1: Type: text/plain, Size: 183 bytes --]

Hi!

Can you guys try out this last take on a fix for your ATAPI device
problems? Applies against clean 2.4.20.

Please report failure/success.

Thanks.

-- 
Vojtech Pavlik
SuSE Labs

[-- Attachment #2: vt8235-atapi --]
[-- Type: text/plain, Size: 1331 bytes --]

ChangeSet@1.884, 2002-12-19 11:23:11+01:00, vojtech@suse.cz
  VIA IDE: Always use slow address setup timings for ATAPI devices.


 via82cxxx.c |   19 ++++++-------------
 1 files changed, 6 insertions(+), 13 deletions(-)


diff -Nru a/drivers/ide/pci/via82cxxx.c b/drivers/ide/pci/via82cxxx.c
--- a/drivers/ide/pci/via82cxxx.c	Thu Dec 19 11:23:42 2002
+++ b/drivers/ide/pci/via82cxxx.c	Thu Dec 19 11:23:42 2002
@@ -1,16 +1,5 @@
 /*
- * $Id: via82cxxx.c,v 3.35-ac2 2002/09/111 Alan Exp $
- *
- *  Copyright (c) 2000-2001 Vojtech Pavlik
- *
- *  Based on the work of:
- *	Michel Aubry
- *	Jeff Garzik
- *	Andre Hedrick
- */
-
-/*
- * Version 3.35
+ * Version 3.36
  *
  * VIA IDE driver for Linux. Supported southbridges:
  *
@@ -152,7 +141,7 @@
 	via_print("----------VIA BusMastering IDE Configuration"
 		"----------------");
 
-	via_print("Driver Version:                     3.35-ac");
+	via_print("Driver Version:                     3.36");
 	via_print("South Bridge:                       VIA %s",
 		via_config->name);
 
@@ -351,6 +340,10 @@
 		ide_timing_compute(peer, peer->current_speed, &p, T, UT);
 		ide_timing_merge(&p, &t, &t, IDE_TIMING_8BIT);
 	}
+
+	/* Always use 4 address setup clocks on ATAPI devices */
+	if (drive->media != ide_disk)
+		t.setup = 4;
 
 	via_set_speed(HWIF(drive)->pci_dev, drive->dn, &t);
 

Re: vt8235 fix, hopefully last variant

[AnonimoVeneziano]

Vojtech Pavlik wrote:

>Hi!
>
>Can you guys try out this last take on a fix for your ATAPI device
>problems? Applies against clean 2.4.20.
>
>Please report failure/success.
>
>Thanks.
>
>  
>
>------------------------------------------------------------------------
>
>ChangeSet@1.884, 2002-12-19 11:23:11+01:00, vojtech@suse.cz
>  VIA IDE: Always use slow address setup timings for ATAPI devices.
>
>
> via82cxxx.c |   19 ++++++-------------
> 1 files changed, 6 insertions(+), 13 deletions(-)
>
>
>diff -Nru a/drivers/ide/pci/via82cxxx.c b/drivers/ide/pci/via82cxxx.c
>--- a/drivers/ide/pci/via82cxxx.c	Thu Dec 19 11:23:42 2002
>+++ b/drivers/ide/pci/via82cxxx.c	Thu Dec 19 11:23:42 2002
>@@ -1,16 +1,5 @@
> /*
>- * $Id: via82cxxx.c,v 3.35-ac2 2002/09/111 Alan Exp $
>- *
>- *  Copyright (c) 2000-2001 Vojtech Pavlik
>- *
>- *  Based on the work of:
>- *	Michel Aubry
>- *	Jeff Garzik
>- *	Andre Hedrick
>- */
>-
>-/*
>- * Version 3.35
>+ * Version 3.36
>  *
>  * VIA IDE driver for Linux. Supported southbridges:
>  *
>@@ -152,7 +141,7 @@
> 	via_print("----------VIA BusMastering IDE Configuration"
> 		"----------------");
> 
>-	via_print("Driver Version:                     3.35-ac");
>+	via_print("Driver Version:                     3.36");
> 	via_print("South Bridge:                       VIA %s",
> 		via_config->name);
> 
>@@ -351,6 +340,10 @@
> 		ide_timing_compute(peer, peer->current_speed, &p, T, UT);
> 		ide_timing_merge(&p, &t, &t, IDE_TIMING_8BIT);
> 	}
>+
>+	/* Always use 4 address setup clocks on ATAPI devices */
>+	if (drive->media != ide_disk)
>+		t.setup = 4;
> 
> 	via_set_speed(HWIF(drive)->pci_dev, drive->dn, &t);
> 
>  
>
It said that 2 of 3 hunks are failed. I don't know why. Error in patch again

Bye

Marcello

Re: vt8235 fix, hopefully last variant

[John Reiser]

[-- Attachment #1: Type: text/plain, Size: 497 bytes --]

Vojtech Pavlik wrote:
> Can you guys try out this last take on a fix for your ATAPI device
> problems? Applies against clean 2.4.20.

Yes, the vt8235-atapi patch works for me; so did vt8235-dvd, but not vt8235-min.
Mitsumi FX4830T ATAPI CD-ROM, MSI KT3 Ultra2 mainboard (KT333), vt8235.

I modified the patch so that it would apply cleanly to RedHatLinux8.0
kernel-2.4.18-18.8.0 as part of "rpm -bb --target i686 kernel-2.4.18.spec".
The result is attached.

-- 
John Reiser, jreiser@BitWagon.com

[-- Attachment #2: vt8235-atapi-jreiser.patch --]
[-- Type: text/plain, Size: 870 bytes --]

--- linux/drivers/ide/via82cxxx.c.orig	Thu Dec 19 08:35:20 2002
+++ linux/drivers/ide/via82cxxx.c	Thu Dec 19 08:43:10 2002
@@ -1,5 +1,5 @@
 /*
- * Version 3.35
+ * Version 3.36jreiser
  *
  * VIA IDE driver for Linux. Supported southbridges:
  *
@@ -129,7 +129,7 @@
 
 	via_print("----------VIA BusMastering IDE Configuration----------------");
 
-	via_print("Driver Version:                     3.35");
+	via_print("Driver Version:                     3.36jreiser");
 	via_print("South Bridge:                       VIA %s", via_config->name);
 
 	pci_read_config_byte(isa_dev, PCI_REVISION_ID, &t);
@@ -295,6 +295,10 @@
 		ide_timing_merge(&p, &t, &t, IDE_TIMING_8BIT);
 	}
 
+	/* Always use 4 address setup clocks on ATAPI devices */
+	if (drive->media != ide_disk)
+		t.setup = 4;
+ 
 	via_set_speed(HWIF(drive)->pci_dev, drive->dn, &t);
 
 	if (!drive->init_speed)

Re: vt8235 fix, hopefully last variant

[Patrick Petermair]

Hi!

Sorry for not replying this week, but I was not at home...now back to 
business:

I tried the last patch, but it didn't work - I got Hunk messages:

starbase:/usr/src/linux-2.4.20# patch -p1 < vt8235-atapi 
patching file drivers/ide/via82cxxx.c
Hunk #1 FAILED at 1.
Hunk #2 FAILED at 141.
Hunk #3 succeeded at 283 (offset -57 lines).
2 out of 3 hunks FAILED -- saving rejects to file 
drivers/ide/via82cxxx.c.rej
starbase:/usr/src/linux-2.4.20# cat drivers/ide/via82cxxx.c.rej 
***************
*** 1,16 ****
  /*
-  * $Id: via82cxxx.c,v 3.35-ac2 2002/09/111 Alan Exp $
-  *
-  *  Copyright (c) 2000-2001 Vojtech Pavlik
-  *
-  *  Based on the work of:
-  *    Michel Aubry
-  *    Jeff Garzik
-  *    Andre Hedrick
-  */
- 
- /*
-  * Version 3.35
   *
   * VIA IDE driver for Linux. Supported southbridges:
   *
--- 1,5 ----
  /*
+  * Version 3.36
   *
   * VIA IDE driver for Linux. Supported southbridges:
   *
***************
*** 152,158 ****
        via_print("----------VIA BusMastering IDE Configuration"
                "----------------");
  
-       via_print("Driver Version:                     3.35-ac");
        via_print("South Bridge:                       VIA %s",
                via_config->name);
  
--- 141,147 ----
        via_print("----------VIA BusMastering IDE Configuration"
                "----------------");
  
+       via_print("Driver Version:                     3.36");
        via_print("South Bridge:                       VIA %s",
                via_config->name);
  
starbase:/usr/src/linux-2.4.20# 

MSI KT3Ultra2, TOSHIBA DVD-ROM SD-M1302
So far my kernel is running your first patch and it works fine. Would 
love to get the final patch running too, but I don't know what that 
messages in via82cxxx.c.rej mean and how to modify the patch so that it 
works - ah, btw: I tried to apply it to a clean 2.4.20 kernel source.

Thanks so far for your great work.

Patrick

Re: vt8235 fix, hopefully last variant

[Vojtech Pavlik]

On Fri, Dec 20, 2002 at 09:12:58PM +0100, Patrick Petermair wrote:
> Hi!
> 
> Sorry for not replying this week, but I was not at home...now back to 
> business:
> 
> I tried the last patch, but it didn't work - I got Hunk messages:
> 
> starbase:/usr/src/linux-2.4.20# patch -p1 < vt8235-atapi 
> patching file drivers/ide/via82cxxx.c
> Hunk #1 FAILED at 1.
> Hunk #2 FAILED at 141.
> Hunk #3 succeeded at 283 (offset -57 lines).
> 2 out of 3 hunks FAILED -- saving rejects to file 
> starbase:/usr/src/linux-2.4.20# 

Since they're in comments only, you can ignore them.

> MSI KT3Ultra2, TOSHIBA DVD-ROM SD-M1302
> So far my kernel is running your first patch and it works fine. Would 
> love to get the final patch running too, but I don't know what that 
> messages in via82cxxx.c.rej mean and how to modify the patch so that it 
> works - ah, btw: I tried to apply it to a clean 2.4.20 kernel source.

Sorry, then I probably made a mistake in generating it.

> Thanks so far for your great work.

Can you please try even when the first two hunks failed? Only the third
one really matters.

-- 
Vojtech Pavlik
SuSE Labs

Re: vt8235 fix, hopefully last variant

[Patrick Petermair]

Vojtech Pavlik:

> > starbase:/usr/src/linux-2.4.20# patch -p1 < vt8235-atapi
> > patching file drivers/ide/via82cxxx.c
> > Hunk #1 FAILED at 1.
> > Hunk #2 FAILED at 141.
> > Hunk #3 succeeded at 283 (offset -57 lines).
> > 2 out of 3 hunks FAILED -- saving rejects to file
> > starbase:/usr/src/linux-2.4.20#
>
> Since they're in comments only, you can ignore them.
>
> Can you please try even when the first two hunks failed? Only the
> third one really matters.

I patched the clean 2.4.20 source and it boots just fine after 
compiling. DMA works fine too.
It's now my default kernel so I'll test it a bit more....

MSI KT3Ultra2, TOSHIBA DVD-ROM SD-M1302

Thanks,
Patrick

Re: vt8235 fix, hopefully last variant

[Vojtech Pavlik]

On Sat, Dec 21, 2002 at 11:51:50AM +0100, Patrick Petermair wrote:
> Vojtech Pavlik:
> 
> > > starbase:/usr/src/linux-2.4.20# patch -p1 < vt8235-atapi
> > > patching file drivers/ide/via82cxxx.c
> > > Hunk #1 FAILED at 1.
> > > Hunk #2 FAILED at 141.
> > > Hunk #3 succeeded at 283 (offset -57 lines).
> > > 2 out of 3 hunks FAILED -- saving rejects to file
> > > starbase:/usr/src/linux-2.4.20#
> >
> > Since they're in comments only, you can ignore them.
> >
> > Can you please try even when the first two hunks failed? Only the
> > third one really matters.
> 
> I patched the clean 2.4.20 source and it boots just fine after 
> compiling. DMA works fine too.
> It's now my default kernel so I'll test it a bit more....

Thanks a lot for the report. If it keeps working, I'll send the patch to
Marcello.

-- 
Vojtech Pavlik
SuSE Labs

Re: vt8235 fix, hopefully last variant

[Christian Seidemann]

Vojtech Pavlik wrote:

> Hi!
>
> Can you guys try out this last take on a fix for your ATAPI device
> problems? Applies against clean 2.4.20.
>
> Please report failure/success.
>
> Thanks.

I also tried the vt8235-atapi patch and it works for me.
MSI KT3 Ultra2 mainboard (KT333), vt8235.
LG DVD-ROM DRD-8160B

Thanks
Christian

Re: vt8235 fix, hopefully last variant

[AnonimoVeneziano]

Christian Seidemann wrote:

>Vojtech Pavlik wrote:
>
>  
>
>>Hi!
>>
>>Can you guys try out this last take on a fix for your ATAPI device
>>problems? Applies against clean 2.4.20.
>>
>>Please report failure/success.
>>
>>Thanks.
>>    
>>
>
>I also tried the vt8235-atapi patch and it works for me.
>MSI KT3 Ultra2 mainboard (KT333), vt8235.
>LG DVD-ROM DRD-8160B
>
>Thanks
>Christian
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/
>
>  
>
For me this doesn't  seems to work :(


Bye

Marcello

Re: vt8235 fix, hopefully last variant

[Patrick Petermair]

AnonimoVeneziano:

> For me this doesn't  seems to work :(

What is it exactly that doesn't work? The patching, compiling, booting, 
dma,...?

Patrick

Re: vt8235 fix, hopefully last variant

[AnonimoVeneziano]

Patrick Petermair wrote:

>AnonimoVeneziano:
>
>  
>
>>For me this doesn't  seems to work :(
>>    
>>
>
>What is it exactly that doesn't work? The patching, compiling, booting, 
>dma,...?
>
>Patrick
>
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/
>
>  
>
The patching .... it said something about HUNKS failed....

Bye

Marcello

Re: vt8235 fix, hopefully last variant

[Patrick Petermair]

AnonimoVeneziano:

> >What is it exactly that doesn't work? The patching, compiling,
> > booting, dma,...?
> >
> The patching .... it said something about HUNKS failed....

It's normal that Hunk 1 and 2 fail... Hunk 3 is the important one. So 
you can compile the kernel even with Hunks 1&2 failing.

Quote from Vojtech:
-----

> I tried the last patch, but it didn't work - I got Hunk messages:
> 
> starbase:/usr/src/linux-2.4.20# patch -p1 < vt8235-atapi 
> patching file drivers/ide/via82cxxx.c
> Hunk #1 FAILED at 1.
> Hunk #2 FAILED at 141.
> Hunk #3 succeeded at 283 (offset -57 lines).
> 2 out of 3 hunks FAILED -- saving rejects to file 
> starbase:/usr/src/linux-2.4.20# 

Since they're in comments only, you can ignore them.

-----

Patrick

Re: vt8235 fix, hopefully last variant

[AnonimoVeneziano]

Patrick Petermair wrote:

>AnonimoVeneziano:
>
>  
>
>>>What is it exactly that doesn't work? The patching, compiling,
>>>booting, dma,...?
>>>
>>>      
>>>
>>The patching .... it said something about HUNKS failed....
>>    
>>
>
>It's normal that Hunk 1 and 2 fail... Hunk 3 is the important one. So 
>you can compile the kernel even with Hunks 1&2 failing.
>
>Quote from Vojtech:
>-----
>
>  
>
>>I tried the last patch, but it didn't work - I got Hunk messages:
>>
>>starbase:/usr/src/linux-2.4.20# patch -p1 < vt8235-atapi 
>>patching file drivers/ide/via82cxxx.c
>>Hunk #1 FAILED at 1.
>>Hunk #2 FAILED at 141.
>>Hunk #3 succeeded at 283 (offset -57 lines).
>>2 out of 3 hunks FAILED -- saving rejects to file 
>>starbase:/usr/src/linux-2.4.20# 
>>    
>>
>
>Since they're in comments only, you can ignore them.
>
>-----
>
>Patrick
>
>
>  
>
Ah, ok, thank you very much, tomorrow I'll try to compile.I've tried to 
compile the patch vt8235-min (with 2 hunks failed) but It wasn't work. 
With vt8235-atapi I haven't tried at all when I've saw the failed hunks. 
I haven't compile.

Thank you

Bye

Marcello

Re: vt8235 fix, hopefully last variant

[John Reiser]

Looking at the important piece:
-----
> +	/* Always use 4 address setup clocks on ATAPI devices */
> +	if (drive->media != ide_disk)
> +		t.setup = 4;
-----
it seems to me that using 4 for t.setup also should be done unless _all_ devices
on that cable are ide_disk.  In particular, if one device is ide_disk but the
other is not, then "t.setup = 4;" should be used even when talking to the ide_disk.
Otherwise the non-ide_disk device might get confused, respond when it should not,
and garble the transaction.  Or, is such a bad thing prevented in some other way?

-- 
John Reiser, jreiser@BitWagon.com

Re: vt8235 fix, hopefully last variant

[Vojtech Pavlik]

On Sat, Dec 21, 2002 at 09:57:42AM -0800, John Reiser wrote:

> Looking at the important piece:
> -----
> > +	/* Always use 4 address setup clocks on ATAPI devices */
> > +	if (drive->media != ide_disk)
> > +		t.setup = 4;
> -----
> it seems to me that using 4 for t.setup also should be done unless _all_ devices
> on that cable are ide_disk.  In particular, if one device is ide_disk but the
> other is not, then "t.setup = 4;" should be used even when talking to the ide_disk.
> Otherwise the non-ide_disk device might get confused, respond when it should not,
> and garble the transaction.  Or, is such a bad thing prevented in some other way?

Good idea.

-- 
Vojtech Pavlik
SuSE Labs