On Thu, Jan 10, 2013 at 06:00:25PM +0100, Andreas Färber wrote: > Am 10.01.2013 17:02, schrieb Stefan Hajnoczi: > > On Wed, Dec 19, 2012 at 12:09:21PM +0000, Julien Grall wrote: > >> The commit 582299336879504353e60c7937fbc70fea93f3da introduced a 1-shift > >> for > >> some offset in dma emulation. > >> > >> Before the previous commit, which converted ioport_register_* to > >> MemoryRegion, > >> the DMA controller registered 8 ioports with the following formula: > >> base + ((8 + i) << d->shift) where 0 <= i < 8 > >> When an IO occured within a Memory Region, DMA callback receives an offset > >> relative to the started address. Here the started address is: > >> base + (8 << d->shift). > >> The offset should be: (i << d->shift). After the shift is reverted, the > >> offset > >> are 0..7 not 1..8. > >> > >> Cc: 1089...@bugs.launchpad.net > >> Reviewed-by: Andreas Färber <afaer...@suse.de> > >> Reported-by: Andreas Gustafsson <g...@gson.org> > >> Signed-off-by: Julien Grall <julien.gr...@citrix.com> > >> --- > >> > >> Modification between V1 and V2: > >> * Modify the commit message to explain the problem. > >> > >> hw/dma.c | 22 +++++++++++----------- > >> 1 file changed, 11 insertions(+), 11 deletions(-) > > > > This patch resolves "dma: unknown iport 0" warnings for my Windows 8 guest. > > > > Tested-by: Stefan Hajnoczi <stefa...@redhat.com> > > If you don't want to queue it for the trivial tree, I'll queue it > together with Hervé's conversions.
Thanks, please take it through your tree. Stefan