On 03/01/2012 12:16 PM, John Brown wrote:
>
>
> -----Original Message-----
> From: James Simmons [mailto:[email protected]]
> Sent: Tuesday, 28 February 2012 5:58 AM
> To: John Brown
> Cc: Xavier Bachelot; Openchrome
> Subject: Re: [Openchrome-users] EPIA-M720 (VX900) xorg.conf (EE) No devices
> detected
>
>
>> Which still concerns me, although I accept it doesn't explain why I can't
>> even get a desktop.
>>
>> Apart from the drm and dri errors, the only other errors in the log is:
>>
>> [ 12.006] (EE) open /dev/fb0: No such file or directory
>>
>> Again, any suggestions would be greatly appreciated.
>
> Could you try this code:
>
> http://www.infradead.org/~jsimmons/release_0_3_0-Feb27.tar.bz2
>
> It's the next gen version of the code.
>
>
> THANKS. I installed that release. It no longer produces the error message
> from the monitor, but now the screen just remains completely blank instead.
> This was with no xorg.conf. The Xorg.0.log is attached.
>
> Errors now are:
>
> [ 11.288] (II) Loading /usr/lib/xorg/modules/libfbdevhw.so
> [ 11.297] (II) Module fbdevhw: vendor="X.Org Foundation"
> [ 11.297] compiled for 1.9.0, module version = 0.0.2
> [ 11.297] ABI class: X.Org Video Driver, version 8.0
> [ 11.297] (EE) open /dev/fb0: No such file or directory
>
> [ 11.953] (EE) CHROME(0): [drm] Failed to open DRM device for
> pci:0000:00:01.0: No such file or directory
>
> [ 12.087] (EE) CHROME(0): VIAVidHWDiffInit: Unhandled ChipSet.
>
> Any further ideas or clues?
>
The drm error is because the Chrome9 family of IGPs is not yet fully
supported in regular kernel. This is being worked on, but there's still
significant work to be done.
The VIAVidHWDiffInit error could be fixed by merging some commits from
master to the kms_branch.
James, this are the VX900 Xv support commits
(4406c4826f650e896548b369a73173c504434881 and
aa56f57da6e403060536f4215e3142f38e98d84d). What do you think about
cheery-picking them into kms_branch ?
John, this is harmless and has nothing to do with your problem.
Both above errors will need to be fixed before you can use XvMC though,
but the XvMC code itself needs fweaking too anyway before it can be
useful for CX700 and later IGPs (this does include yours).
Then you're hitting another bug in the kms_branch and X doesn't finish
starting up. James, I think this is the unresolved symbol I posted about
yesterday.
I think the attached patch should workaround that.
Once done, please send the log from the patched driver.
Regards,
Xavier
diff --git a/configure.ac b/configure.ac
index c4e1d73..ce93906 100644
--- a/configure.ac
+++ b/configure.ac
@@ -112,13 +112,12 @@ if test x$XSERVER_LIBPCIACCESS = xyes; then
fi
if test "$DRI" != no; then
- PKG_CHECK_MODULES(DRI, [libdrm >= 2.2 xf86driproto])
- AC_CHECK_HEADER([${sdkdir}/dri.h],
- [have_dri_h="yes"], [have_dri_h="no"],[-])
- AC_CHECK_HEADER([${sdkdir}/sarea.h],
- [have_sarea_h="yes"], [have_sarea_h="no"],[-])
- AC_CHECK_HEADER([${sdkdir}/dristruct.h],
- [have_dristruct_h="yes"], [have_dristruct_h="no"],[-])
+ AC_CHECK_FILE([${sdkdir}/dri.h],
+ [have_dri_h="yes"], [have_dri_h="no"])
+ AC_CHECK_FILE([${sdkdir}/sarea.h],
+ [have_sarea_h="yes"], [have_sarea_h="no"])
+ AC_CHECK_FILE([${sdkdir}/dristruct.h],
+ [have_dristruct_h="yes"], [have_dristruct_h="no"])
fi
AC_MSG_CHECKING([whether to include DRI support])
diff --git a/src/via_driver.h b/src/via_driver.h
index 9a12869..156987c 100644
--- a/src/via_driver.h
+++ b/src/via_driver.h
@@ -51,7 +51,6 @@
#include "fourcc.h"
#include "fb.h"
-#include "xf86dri.h"
#include "xf86Crtc.h"
#include "xf86RandR12.h"
#include "xf86cmap.h"
_______________________________________________
openchrome-users mailing list
[email protected]
http://wiki.openchrome.org/mailman/listinfo/openchrome-users
Main page:
http://www.openchrome.org
Wiki:
http://wiki.openchrome.org
User Forum:
http://wiki.openchrome.org/tikiwiki/tiki-view_forum.php?forumId=1
