freebsd7, radeon, xorg-server -> deadlock or so

[Robert Noland]

On Wed, 2010-02-10 at 19:13 -0800, Norbert Papke wrote:
> On February 10, 2010, Robert Noland wrote:
> > On Wed, 2010-02-10 at 08:43 -0800, Norbert Papke wrote:
> > > On February 10, 2010, Robert Noland wrote:
> > > > Ah, this suggest an issue with interrupts.  If the driver is blocked
> > > > waiting on an event, moving the mouse will interrupt the driver and
> > > > force it to update.
> > >
> > > Based on your suggestion in another thread, I tried to disable MSI.  This
> > > did not make a difference.  I still experience the hang.
> > 
> > Ok, that was kind of a long shot to begin with.  I've never heard of any
> > issues with MSI on radeons, at least as long as MSI works at all on your
> > system.
> 
> There does seem to be a little MSI weirdness, at least it looks that way to 
> me.  From a verbose dmesg:
> 
> vgapci0: <VGA-compatible display> port 0xc000-0xc0ff mem 
> 0xd0000000-0xdfffffff,0xfe9e0000-0xfe9effff irq 16 at device 0.0 on pci1
> drm0: <ATI ATI Radeon HD 3600 XT> on vgapci0
> vgapci0: attempting to allocate 1 MSI vectors (1 supported)
> msi: routing MSI IRQ 256 to local APIC 0 vector 49
> vgapci0: using IRQ 256 for MSI
> info: [drm] MSI enabled 1 message(s)
> vgapci0: Reserved 0x10000 bytes for rid 0x18 type 3 at 0xfe9e0000
> vgapci0: child drm0 requested pci_enable_busmaster
> info: [drm] Initialized radeon 1.31.0 20080613
> hdac0: <ATI RV635 High Definition Audio Controller> mem 0xfe9fc000-0xfe9fffff 
> irq 17 at device 0.1 on pci1
> hdac0: HDA Driver Revision: 20100122_0141
> hdac0: Reserved 0x4000 bytes for rid 0x10 type 3 at 0xfe9fc000
> hdac0: attempting to allocate 1 MSI vectors (1 supported)
> msi: routing MSI IRQ 257 to local APIC 0 vector 50
> hdac0: using IRQ 257 for MSI
> 
> However, looking at the corresponding snippet from "vmstat -ia":
> 
> interrupt                          total       rate
> irq16: uhci0                          16          0
> stray irq16                            0          0
> irq17: dc0                            50          0
> stray irq17                            0          0
> irq256:                                0          0
> stray irq256                           0          0
> irq257: hdac0                          1          0
> stray irq257                           0          0
> 
> Should there not be some evidence of vgapci0 being attached to either IRQ 16 
> or 256?

Interrupts are not enabled on r600 cards right now, so you won't ever
see an interrupt.  On r300 or r500 or whatever interrupts will be
enabled and directed to either MSI or legacy IRQ's.

robert.


> -- Norbert.
> 
> _______________________________________________
> freebsd-x11 at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-x11
> To unsubscribe, send any mail to "freebsd-x11-unsubscribe at freebsd.org"
-- 
Robert Noland <rnoland at FreeBSD.org>
FreeBSD