Németh Márton wrote: > From: Márton Németh <[EMAIL PROTECTED]> > > Change the type of Unichrome display adapter with PCI ID=1106:3108 to > get rid of message "irq 16: nobody cared". > > See http://bugzilla.kernel.org/show_bug.cgi?id=8641 for details. > > Signed-off-by: Shankar Giri V <[EMAIL PROTECTED]> > Signed-off-by: Márton Németh <[EMAIL PROTECTED]> > --- > > --- linux-2.6.25-rc5/drivers/char/drm/drm_pciids.h.orig 2008-03-15 > 10:58:49.000000000 +0100 > +++ linux-2.6.25-rc5/drivers/char/drm/drm_pciids.h 2008-03-15 > 10:59:31.000000000 +0100 > @@ -335,7 +335,7 @@ > {0x1106, 0x3118, PCI_ANY_ID, PCI_ANY_ID, 0, 0, VIA_PRO_GROUP_A}, \ > {0x1106, 0x3122, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ > {0x1106, 0x7205, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ > - {0x1106, 0x3108, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ > + {0x1106, 0x3108, PCI_ANY_ID, PCI_ANY_ID, 0, 0, VIA_PRO_GROUP_A}, \ > {0x1106, 0x3344, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ > {0x1106, 0x3343, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ > {0x1106, 0x3230, PCI_ANY_ID, PCI_ANY_ID, 0, 0, VIA_DX9_0}, \ > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > -- > _______________________________________________ > Dri-devel mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/dri-devel > Nope, this is not a correct fix. It enables a set of DMA commands not available on the K8M800, also it makes the via IRQ handler just report "Handled" when in fact the device keeps firing 200 000 or so interrupts per second.
I'm 99.99 % convinced that this is a hw bug, and it also appears in the KM400 chipsets. When VIA was asked about it they replied that IRQ functionality was never verified in the K8M800 since the windows driver never used it. They only use IRQ for video capture transfer, and there is no video-in on K8M800s. Other unichromes with the same IRQ register interface work fine. Having said that, it's not impossible that there may be a hidden register somewhere that makes it all work. The K8M800 irqs have been turned off in newer versions of the openChrome 2D driver to work around this issue. /Thomas ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ -- _______________________________________________ Dri-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dri-devel
