| Message ID | 20111125125020.GG3214@riva.dynamic.greenend.org.uk |
|---|---|
| State | New |
| Headers | show |
applied to Precise
On Fri, Nov 25, 2011 at 12:37:44PM -0700, Tim Gardner wrote:
> applied to Precise
Thanks for applying this.
Actually, though; would you consider simply building in pata_macio on
powerpc? Booting a desktop CD in qemu runs into
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=641210, and it looks as
though the Debian kernel maintainers felt that building the module in
was the simplest way around the lack of autoloading in this case.
Thanks,
On Mon, Nov 28, 2011 at 12:15:12PM +0000, Colin Watson wrote: > On Fri, Nov 25, 2011 at 12:37:44PM -0700, Tim Gardner wrote: > > applied to Precise > > Thanks for applying this. > > Actually, though; would you consider simply building in pata_macio on > powerpc? Booting a desktop CD in qemu runs into > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=641210, and it looks as > though the Debian kernel maintainers felt that building the module in > was the simplest way around the lack of autoloading in this case. > > Thanks, Having read the description in that bug it does seem that these devices can be connected to PCI (discoverable) and some MACIO platform driver (non-discoverable). Pretty much the only way to support those MACIO connected versions is to build this driver in; vile though that is. I have pushed a further update to precise switching this to =y. -apw
On 11/28/2011 06:09 AM, Andy Whitcroft wrote: > On Mon, Nov 28, 2011 at 12:15:12PM +0000, Colin Watson wrote: >> On Fri, Nov 25, 2011 at 12:37:44PM -0700, Tim Gardner wrote: >>> applied to Precise >> >> Thanks for applying this. >> >> Actually, though; would you consider simply building in pata_macio on >> powerpc? Booting a desktop CD in qemu runs into >> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=641210, and it looks as >> though the Debian kernel maintainers felt that building the module in >> was the simplest way around the lack of autoloading in this case. >> >> Thanks, > > Having read the description in that bug it does seem that these devices > can be connected to PCI (discoverable) and some MACIO platform driver > (non-discoverable). Pretty much the only way to support those MACIO > connected versions is to build this driver in; vile though that is. > > I have pushed a further update to precise switching this to =y. > > -apw works for me.
diff --git a/debian.master/d-i/modules/pata-modules b/debian.master/d-i/modules/pata-modules index 623a47d..a58c30a 100644 --- a/debian.master/d-i/modules/pata-modules +++ b/debian.master/d-i/modules/pata-modules @@ -20,6 +20,7 @@ pata_it8213.ko ? pata_it821x.ko ? pata_jmicron.ko ? pata_legacy.ko ? +pata_macio.ko ? pata_marvell.ko ? pata_mpiix.ko ? pata_netcell.ko ?
This is needed to handle Apple MacIO ATA devices. In particular, it's needed to see any disk devices when testing Ubuntu powerpc images with 'qemu-system-ppc -cpu G4 -M mac99'. Signed-off-by: Colin Watson <cjwatson@canonical.com> --- debian.master/d-i/modules/pata-modules | 1 + 1 files changed, 1 insertions(+), 0 deletions(-)