[PATCH] HPT374 kernel panic - regression in 2.6.8

[PATCH] HPT374 kernel panic - regression in 2.6.8

[Jindrich Makovicka]

Hello,

HighPoint 374 driver in 2.6.8 can cause kernel panic on boot with 
non-33MHz timings because some lines from an older version have been 
included in the source again. After removing the check, HPT374 works 
just fine using internal PLL.

--- old/hpt366.c	2004-08-14 13:06:12.000000000 +0200
+++ new/hpt366.c	2004-08-14 12:59:25.000000000 +0200
@@ -991,11 +991,6 @@
  	if (pci_get_drvdata(dev))
  		goto init_hpt37X_done;
  	
-	if (hpt_minimum_revision(dev,8))
-	{
-		printk(KERN_ERR "HPT374: Only 33MHz PCI timings are supported.\n");
-		return -EOPNOTSUPP;
-	}
  	/*
  	 * adjust PLL based upon PCI clock, enable it, and wait for
  	 * stabilization.

Regards,
-- 
Jindrich Makovicka

Re: [PATCH] HPT374 kernel panic - regression in 2.6.8

[Alan Cox]

On Sad, 2004-08-14 at 12:14, Jindrich Makovicka wrote:
> Hello,
> 
> HighPoint 374 driver in 2.6.8 can cause kernel panic on boot with 
> non-33MHz timings because some lines from an older version have been 
> included in the source again. After removing the check, HPT374 works 
> just fine using internal PLL.

The HPT374 only supports 33Mhz timings. If you are overclocking it then
its probably going to mostly work. The lines added are not "from an
older source file" but form a port forwards of long missing and
important fixes from 2.4 that have been there for a long long time.

So the first question is why does your late series HPT37x think its
running at > 33Mhz ? Are you overclocking it ?

Alan

Re: [PATCH] HPT374 kernel panic - regression in 2.6.8

[Bartlomiej Zolnierkiewicz]

On Saturday 14 August 2004 17:16, Alan Cox wrote:
> On Sad, 2004-08-14 at 12:14, Jindrich Makovicka wrote:
> > Hello,
> >
> > HighPoint 374 driver in 2.6.8 can cause kernel panic on boot with
> > non-33MHz timings because some lines from an older version have been
> > included in the source again. After removing the check, HPT374 works
> > just fine using internal PLL.
>
> The HPT374 only supports 33Mhz timings. If you are overclocking it then
> its probably going to mostly work. The lines added are not "from an
> older source file" but form a port forwards of long missing and
> important fixes from 2.4 that have been there for a long long time.

2.4 driver also misses some fixes...

> So the first question is why does your late series HPT37x think its
> running at > 33Mhz ? Are you overclocking it ?

http://linus.bkbits.net:8080/linux-2.5/cset@40586b50zsHhQgPLGTlje7g_f5wGTw?nav=index.html|
src/|src/drivers|src/drivers/ide|src/drivers/ide/pci|
related/drivers/ide/pci/hpt366.c

please read bugzilla bugs mentioned in the changelong and fix this

Re: [PATCH] HPT374 kernel panic - regression in 2.6.8

[Alan Cox]

On Llu, 2004-08-16 at 23:30, Bartlomiej Zolnierkiewicz wrote:
> http://linus.bkbits.net:8080/linux-2.5/cset@40586b50zsHhQgPLGTlje7g_f5wGTw?nav=index.html|
> src/|src/drivers|src/drivers/ide|src/drivers/ide/pci|
> related/drivers/ide/pci/hpt366.c
> 
> please read bugzilla bugs mentioned in the changelong and fix this

Already did and sent out the patch. 2.4.x is missing that small change
so it might be a good one for Marcelo to merge ?

Re: [PATCH] HPT374 kernel panic - regression in 2.6.8

[Marcelo Tosatti]

On Mon, Aug 16, 2004 at 11:00:34PM +0100, Alan Cox wrote:
> On Llu, 2004-08-16 at 23:30, Bartlomiej Zolnierkiewicz wrote:
> > http://linus.bkbits.net:8080/linux-2.5/cset@40586b50zsHhQgPLGTlje7g_f5wGTw?nav=index.html|
> > src/|src/drivers|src/drivers/ide|src/drivers/ide/pci|
> > related/drivers/ide/pci/hpt366.c
> > 
> > please read bugzilla bugs mentioned in the changelong and fix this
> 
> Already did and sent out the patch. 2.4.x is missing that small change
> so it might be a good one for Marcelo to merge ?

That and the triflex update (which I have around) we miss in 2.4 that I remember. 

I have no changelog for the triflex fixes however - can you write up please Bart?