6619 2014-08-07 18:28:10 +0000 enable gma500 graphics support for intel-core2-32 BSP 2015-05-04 21:58:37 +0000 1 1 7 Build System, Metadata & Runtime BSPs bsps-meta-intel 1.8 x86 Multiple RESOLVED WORKSFORME http://bugzilla.yoctoproject.org/show_bug.cgi?id=7439 Medium normal 1.8 M4 6362 1 nitin.a.kamble sgw alexandru.c.georgescu ross.burton sgw tom.zanussi georgex.l.musat --- --- No (bug/feature does not impact docs) oldest_to_newest 45022 0 nitin.a.kamble 2014-08-07 18:28:10 +0000 Currently emenlow-noemgd BSP is using the gma500 (actually gma600) graphics driver. The intel-core2-32 BSP is missing this needed gma500 support. The task is to look into emenlow-noemgd BSP, and provide similar gma500 support in the intel-core2-32 BSP. 45023 1 nitin.a.kamble 2014-08-07 18:28:41 +0000 Assigning this to Max. 45313 2 max.eliaser 2014-08-28 21:07:34 +0000 I have an early draft of the patches for this here: http://git.yoctoproject.org/cgit/cgit.cgi/linux-yocto-contrib/log/?h=meliaser/3.14-meta-tunnelcreek-scc Still requires zarusd to be configured so that the gma500_gfx kernel module is disabled on FRI2. 45314 3 nitin.a.kamble 2014-08-28 21:43:01 +0000 I am trying to push ImageConfigurator (in place of zaurusd) to enabling machine configuration for blacklisting of modules. So you can mention the need of blacklisting of the driver in the commit in detail, but it is not a blocker for you. 45341 4 max.eliaser 2014-08-29 17:19:58 +0000 (In reply to comment #3) > I am trying to push ImageConfigurator (in place of zaurusd) to enabling > machine configuration for blacklisting of modules. So you can mention the > need of blacklisting of the driver in the commit in detail, but it is not a > blocker for you. When you say it's not a blocker, should I actually submit a pull request for this, with the understanding that it will break intel-core2-32 on FRI2 until this is resolved? 45379 5 max.eliaser 2014-09-03 21:01:25 +0000 I have sent out the patch series for this. Handing this bug off to Nitin Kamble given my pending departure. 46097 6 nitin.a.kamble 2014-10-03 22:42:04 +0000 This works if proper xorg.conf file is provided. The common BSP will support the emenlow completely with the GMA500 graphics when the machine-setup-tool parts is pull in. 46549 7 nitin.a.kamble 2014-10-23 17:25:05 +0000 the machine setup tool will be upstreamed in the 1.8 m2 timeframe 49018 8 nitin.a.kamble 2015-02-24 19:29:25 +0000 This has been fixed in the machinesetup tools commit. Once those commits are pulled in this bug can be closed. 49111 9 alexandru.c.georgescu 2015-02-27 09:28:33 +0000 (In reply to comment #8) > This has been fixed in the machinesetup tools commit. Once those commits are > pulled in this bug can be closed. Hi Nitin, Any updates on this? 49269 10 ross.burton 2015-03-06 16:14:12 +0000 So it looks like gma500-using hardware only needs a xorg.conf to set "SWCursor on". It's worth testing whether this bug in the kernel driver has since been fixed and the xorg.conf can be removed. 50767 11 sgw 2015-05-04 21:58:37 +0000 This bug is fixed with the lastest kernel and xserver. While running xserver will generate an error in the Xorg.0.log reporting that psb or psb_drv is not found, this is not a problem since the modesetting driver will be used and is correct. This error can be ignored, it occurs since the out-sensing code is first looking for the psb or psb_drv modules. When an xorg.conf file exists and defines a Device this autosensing is skipped.