[2.6.15-git6,-git7] hard lockup on FC4 exiting X (Intel I915)

[Alessandro Suardi <alessandro.suardi@gmail.com>]

Dell Latitude D610, Pentium M @ 1.86Ghz, 2GB RAM
 running uptodate FC4. 100% reproducable since
 2.6.15-git6, no problem up to -git5. Log in to VT1, run
 startx, fire up a gnome-terminal, exit it, Desktop->Logout...
 at this point the mouse arrow stills and the box locks up,
 keyboard dead, no response to pings.

Card is, according to lspci -v,

00:02.0 VGA compatible controller: Intel Corporation Mobile
915GM/GMS/910GML Express Graphics Controller (rev 03) (prog-if 00
[VGA])
        Subsystem: Dell: Unknown device 0182
        Flags: bus master, fast devsel, latency 0, IRQ 11
        Memory at dff00000 (32-bit, non-prefetchable) [size=512K]
        I/O ports at ec38 [size=8]
        Memory at c0000000 (32-bit, prefetchable) [size=256M]
        Memory at dfec0000 (32-bit, non-prefetchable) [size=256K]
        Capabilities: [d0] Power Management version 2

Doesn't happen on -git6 on my older laptop (Latitude C640,
 Pentium IV @ 1.8Ghz, 1GB RAM, Radeon Mobility 7500), so
 it definitely is hardware-related.

Diff'ing the dmesg boot log and the Xorg log from the latest
 working version to the latest available version, the relevant
 lines would seem to be these:

[root@sandman ~]# diff /tmp/dmesg-2615git* | grep -v audit
1c1
< Linux version 2.6.15-git5 (asuardi@sandman) (gcc version 4.0.2
20051125 (Red Hat 4.0.2-8)) #1 Thu Jan 12 01:01:05 CET 2006
---
> Linux version 2.6.15-git7 (asuardi@sandman) (gcc version 4.0.2 20051125 (Red Hat 4.0.2-8)) #1 Thu Jan 12 00:48:04 CET 2006
[...]
78a79
> PCI: Bus #04 (-#07) may be hidden behind transparent bridge #03 (-#04) (try 'pci=assign-busses')
145c146
---
157,159c158,160
< Allocate Port Service[pcie00]
< Allocate Port Service[pcie02]
< Allocate Port Service[pcie03]
---
> Allocate Port Service[0000:00:1c.0:pcie00]
> Allocate Port Service[0000:00:1c.0:pcie02]
> Allocate Port Service[0000:00:1c.0:pcie03]
[...]

[root@sandman ~]# diff /tmp/Xorg-2615git*
6c6
< Current Operating System: Linux sandman 2.6.15-git5 #1 Thu Jan 12
01:01:05 CET 2006 i686
---
> Current Operating System: Linux sandman 2.6.15-git7 #1 Thu Jan 12 00:48:04 CET 2006 i686
[...]
2101,2103c2101,2103
< (II) I810(0): Allocated 4 kB for HW cursor at 0xffff000 (0x37c27000)
< (II) I810(0): Allocated 16 kB for HW (ARGB) cursor at 0xfffb000 (0x35b28000)
< (II) I810(0): Allocated 4 kB for Overlay registers at 0xfffa000 (0x35ae5000).
---
> (II) I810(0): Allocated 4 kB for HW cursor at 0xffff000 (0x36b84000)
> (II) I810(0): Allocated 16 kB for HW (ARGB) cursor at 0xfffb000 (0x36bf0000)
> (II) I810(0): Allocated 4 kB for Overlay registers at 0xfffa000 (0x36b87000).
2121,2122c2121,2122
< (II) I810(0): [drm] added 8192 byte SAREA at 0xf8a5a000
< (II) I810(0): [drm] mapped SAREA 0xf8a5a000 to 0xb7f55000
---
> (II) I810(0): [drm] added 8192 byte SAREA at 0xf8af0000
> (II) I810(0): [drm] mapped SAREA 0xf8af0000 to 0xb7f77000

Something to do with the PCI Express stuff ?


Available for more detailed info and/or testing to interested parties :)

Thanks in advance, ciao,

--alessandro

 "Somehow all you ever need is, never really quite enough, you know"

   (Bruce Springsteen - "Reno")
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/