Via VT8237A kernel support

[Jay Cliburn]

For reasons I cannot fathom -- and because the cognizant gatekeepers at 
LKML won't respond to pleas for feedback -- these two simple patches, 
which add PATA and SATA support for the VIA VT8237A Southbridge chip now 
found on certain new motherboards, have not been applied (or acked or 
nacked or anything else).

PATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115513979609546&w=2

SATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115552504713314&w=2

I know for a fact the the SATA patch works; I wrote it and I've been 
using it for weeks.  I've read the PATA patch works, too, but I haven't 
personally tried it (although I'd be willing to try if it helps).

I don't want to waste anyone's time by filing a Fedora bugzilla that 
will never be implemented, so if I do, what are the prospects for 
getting these patches into the Fedora development tree until such time 
as they're applied upstream (if ever)?

Maybe my expectations are too high.  For the kerneldev-savvy readers, 
what is the typical gestation period for upstream patches of this type?

Thanks,
Jay

[Hans de Goede]

Jay Cliburn wrote:
> For reasons I cannot fathom -- and because the cognizant gatekeepers at 
> LKML won't respond to pleas for feedback -- these two simple patches, 
> which add PATA and SATA support for the VIA VT8237A Southbridge chip now 
> found on certain new motherboards, have not been applied (or acked or 
> nacked or anything else).
>
> PATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115513979609546&w=2
>
> SATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115552504713314&w=2
>
> I know for a fact the the SATA patch works; I wrote it and I've been 
> using it for weeks.  I've read the PATA patch works, too, but I haven't 
> personally tried it (although I'd be willing to try if it helps).
>
> I don't want to waste anyone's time by filing a Fedora bugzilla that 
> will never be implemented, so if I do, what are the prospects for 
> getting these patches into the Fedora development tree until such time 
> as they're applied upstream (if ever)?
>
> Maybe my expectations are too high.  For the kerneldev-savvy readers, 
> what is the typical gestation period for upstream patches of this type?

For a patch as trival as this it should be picked up real soon. But it 
looks wrong that PCI ID seems completely unrelated form the other via 
sata ID's whereas the others are somewhat consecutive numbers. Are you 
sure you've got the right ID (iow the ID for the sata controller 
function of the southbridge?)

Regards,

Hans

[Jay Cliburn]

Hans de Goede wrote:
> For a patch as trival as this it should be picked up real soon. But it 
> looks wrong that PCI ID seems completely unrelated form the other via 
> sata ID's whereas the others are somewhat consecutive numbers. Are you 
> sure you've got the right ID (iow the ID for the sata controller 
> function of the southbridge?)

I'm positive the device ID is correct.

[root@hawk ~]# lspci -v -s 00:0f.0
00:0f.0 IDE interface: VIA Technologies, Inc. VT8237A SATA 2-Port 
Controller (rev 80) (prog-if 8f [Master SecP SecO PriP PriO])
         Subsystem: VIA Technologies, Inc. VT8237A SATA 2-Port Controller
         Flags: bus master, medium devsel, latency 64, IRQ 82
         I/O ports at dc00 [size=8]
         I/O ports at d880 [size=4]
         I/O ports at d800 [size=8]
         I/O ports at d480 [size=4]
         I/O ports at d400 [size=16]
         I/O ports at d000 [size=256]
         Capabilities: [c0] Power Management version 2

[root@hawk ~]# lspci -n -v -s 00:0f.0
00:0f.0 0101: 1106:0591 (rev 80) (prog-if 8f)
         Subsystem: 1106:0591
         Flags: bus master, medium devsel, latency 64, IRQ 82
         I/O ports at dc00 [size=8]
         I/O ports at d880 [size=4]
         I/O ports at d800 [size=8]
         I/O ports at d480 [size=4]
         I/O ports at d400 [size=16]
         I/O ports at d000 [size=256]
         Capabilities: [c0] Power Management version 2

[Arjan van de Ven]

On Wed, 2006-08-30 at 09:26 -0500, Jay Cliburn wrote:
> For reasons I cannot fathom -- and because the cognizant gatekeepers at 
> LKML won't respond to pleas for feedback -- these two simple patches, 
> which add PATA and SATA support for the VIA VT8237A Southbridge chip now 
> found on certain new motherboards, have not been applied (or acked or 
> nacked or anything else).
>
> PATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115513979609546&w=2
>
> SATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115552504713314&w=2
>
> I know for a fact the the SATA patch works; I wrote it and I've been 
> using it for weeks.  I've read the PATA patch works, too, but I haven't 
> personally tried it (although I'd be willing to try if it helps).
>
> I don't want to waste anyone's time by filing a Fedora bugzilla that 
> will never be implemented, so if I do, what are the prospects for 
> getting these patches into the Fedora development tree until such time 
> as they're applied upstream (if ever)?
>
> Maybe my expectations are too high.  For the kerneldev-savvy readers, 
> what is the typical gestation period for upstream patches of this type?

maybe you should post them on the lkml mailing list instead; that is
much more active than the linux-ide list. Maybe CC linux-scsi as well
since the sata stuff lives in drivers/scsi ...

[Jay Cliburn]

Arjan van de Ven wrote:
> maybe you should post them on the lkml mailing list instead; that is
> much more active than the linux-ide list. Maybe CC linux-scsi as well
> since the sata stuff lives in drivers/scsi ...

I'd do that, but the driver source provides explicit instructions for 
patch submittal.  I've heard nothing whatsoever from Mr. Garzik or 
anyone else on the linux-ide list.  Furthermore, the first time I 
submitted the patch, I indeed (and apparently erroneously) submitted it 
to the linux-scsi list.  Again, no response.

[jcliburn@osprey ~]$ head -n 5 kernel-work/linux-2.6/Makefile
VERSION = 2
PATCHLEVEL = 6
SUBLEVEL = 18
EXTRAVERSION = -rc4
NAME=Crazed Snow-Weasel

[jcliburn@osprey ~]$ head -n 8 \ 
kernel-work/linux-2.6/drivers/scsi/sata_via.c
/*
  *  sata_via.c - VIA Serial ATA controllers
  *
  *  Maintained by:  Jeff Garzik 
  *                 Please ALWAYS copy [e-mail]
                    on emails.
  *
  *  Copyright 2003-2004 Red Hat, Inc.  All rights reserved.

[Alan Cox]

On Wed, Aug 30, 2006 at 04:40:34PM +0200, Arjan van de Ven wrote:
> maybe you should post them on the lkml mailing list instead; that is
> much more active than the linux-ide list. Maybe CC linux-scsi as well
> since the sata stuff lives in drivers/scsi ...

Not any more 8) Its now in drivers/ata

[Alan Cox]

On Wed, Aug 30, 2006 at 09:26:19AM -0500, Jay Cliburn wrote:
> found on certain new motherboards, have not been applied (or acked or 
> nacked or anything else).
>
> PATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115513979609546&w=2

This one is in -mm (and has been supported by the new libata PATA code for
a couple of months). It ought to be in 2.6.18 final, I'll make sure it gets
pushed if it hasn't already. If it didn't its an oversight.
> SATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115552504713314&w=2
>
> I know for a fact the the SATA patch works; I wrote it and I've been 
> using it for weeks.  I've read the PATA patch works, too, but I haven't

The 0x0591 identifer is in 2.6.18-rc4-mm2 so again presumably should get
pushed.
> I don't want to waste anyone's time by filing a Fedora bugzilla that 
> will never be implemented, so if I do, what are the prospects for

Actually I have a tracking bug in bugzilla for a couple of ide bits already

Alan

[Jay Cliburn]

Alan Cox wrote:
> On Wed, Aug 30, 2006 at 09:26:19AM -0500, Jay Cliburn wrote:
> > found on certain new motherboards, have not been applied (or acked or 
> > nacked or anything else).
> >
> > PATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115513979609546&w=2
>
> This one is in -mm (and has been supported by the new libata PATA code for
> a couple of months). It ought to be in 2.6.18 final, I'll make sure it gets
> pushed if it hasn't already. If it didn't its an oversight.
> > SATA:  http://marc.theaimsgroup.com/?l=linux-ide&m=115552504713314&w=2
> >
> > I know for a fact the the SATA patch works; I wrote it and I've been 
> > using it for weeks.  I've read the PATA patch works, too, but I haven't
>
> The 0x0591 identifer is in 2.6.18-rc4-mm2 so again presumably should get
> pushed.
> > I don't want to waste anyone's time by filing a Fedora bugzilla that 
> > will never be implemented, so if I do, what are the prospects for
>
> Actually I have a tracking bug in bugzilla for a couple of ide bits already
>
> Alan

Thank you VERY much for this information.  Great news indeed for those
of us who own the 8237a.

I've not heard a peep as a result of my patch submittal.  Is there 
another kernel mailing list I should be watching on which I'd have seen 
the patch applied?  (I watch -ide, -scsi, and netdev.)

[Alan Cox]

On Wed, Aug 30, 2006 at 01:44:21PM -0500, Jay Cliburn wrote:
> I've not heard a peep as a result of my patch submittal.  Is there 
> another kernel mailing list I should be watching on which I'd have seen 
> the patch applied?  (I watch -ide, -scsi, and netdev.)

I'd guess it got ignored as it was already in the upstream trees.

Thanks for noticing the PATA ident was missing I've poked Andrew and Linus
and it should be resolved for .18.

[Jay Cliburn]

Alan Cox wrote:
> On Wed, Aug 30, 2006 at 01:44:21PM -0500, Jay Cliburn wrote:
> > I've not heard a peep as a result of my patch submittal.  Is there 
> > another kernel mailing list I should be watching on which I'd have seen 
> > the patch applied?  (I watch -ide, -scsi, and netdev.)
>
> I'd guess it got ignored as it was already in the upstream trees.
>
> Thanks for noticing the PATA ident was missing I've poked Andrew and Linus
> and it should be resolved for .18.

Alan,

Maybe I should've just kept my mouth shut about the whole thing, because 
it now looks like there are *two* 1106:0591 lines in the 2.6.18-rc4-mm3 
version of drivers/ata/sata_via.c.  It doesn't seem to cause any harm, 
though, since I'm running rc4-mm3 on the system right now.

This patch removes one of the dupes.

[jcliburn@osprey kernel-2.6.18-rc4-mm3]$ diff -u
linux-2.6.18-rc4-mm3/drivers/ata/sata_via.c
linux-2.6.18-rc4-mm3-work/drivers/ata/sata_via.c
--- linux-2.6.18-rc4-mm3/drivers/ata/sata_via.c 2006-08-31
16:21:40.000000000 -0500
+++ linux-2.6.18-rc4-mm3-work/drivers/ata/sata_via.c    2006-08-31
18:13:24.000000000 -0500
@@ -80,7 +80,6 @@
         { 0x1106, 0x0591, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6420 },
         { 0x1106, 0x3149, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6420 },
         { 0x1106, 0x3249, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6421 },
-       { 0x1106, 0x0591, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6420 },
         { }     /* terminate list */
  };

To whom should I submit the patch that fixes this, or, more generally,
who is the correct recipient for patches in the new drivers/ata tree?

Thanks,
Jay

[Alan Cox]

On Fri, Sep 01, 2006 at 12:49:30PM -0500, Jay Cliburn wrote:
> Maybe I should've just kept my mouth shut about the whole thing, because 
> it now looks like there are *two* 1106:0591 lines in the 2.6.18-rc4-mm3 
> version of drivers/ata/sata_via.c.  It doesn't seem to cause any harm, 
> though, since I'm running rc4-mm3 on the system right now.

Patch is good at that
> This patch removes one of the dupes.
>
> [jcliburn@osprey kernel-2.6.18-rc4-mm3]$ diff -u
> linux-2.6.18-rc4-mm3/drivers/ata/sata_via.c
> linux-2.6.18-rc4-mm3-work/drivers/ata/sata_via.c
> --- linux-2.6.18-rc4-mm3/drivers/ata/sata_via.c 2006-08-31
> 16:21:40.000000000 -0500
> +++ linux-2.6.18-rc4-mm3-work/drivers/ata/sata_via.c    2006-08-31
> 18:13:24.000000000 -0500
> @@ -80,7 +80,6 @@
>         { 0x1106, 0x0591, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6420 },
>         { 0x1106, 0x3149, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6420 },
>         { 0x1106, 0x3249, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6421 },
>   { 0x1106, 0x0591, PCI_ANY_ID, PCI_ANY_ID, 0, 0, vt6420 },
>   { }     /* terminate list */
>
> };
>
> To whom should I submit the patch that fixes this, or, more generally,
> who is the correct recipient for patches in the new drivers/ata tree?

[e-mail] and cc [e-mail] for -mm bugs (and linux-kernel)