summaryrefslogtreecommitdiff
path: root/README
diff options
context:
space:
mode:
Diffstat (limited to 'README')
-rw-r--r--README17
1 files changed, 17 insertions, 0 deletions
diff --git a/README b/README
index a57ce7a..cf6d120 100644
--- a/README
+++ b/README
@@ -310,6 +310,21 @@ TODO: looks optional for a simple running system
- CONFIG_FB_GEODE_LX=y
=> for the Alix 1.E
+TODO: needed? and for which cards?
+- CONFIG_FB_SIMPLE=y
+
+- CONFIG_FB_DEVICE=y
+ => to get a /dev/fb*
+ => TODO: check if XF86fbdev need that
+
+TODO: which one of the hardware acceleration options are needed for
+which cards and have performace benefits (Cirrus for instance should
+have BtBlt functions)
+- CONFIG_FB_TILEBLITTING=y
+
+TODO: what happens when we have multiple framebuffer drivers, which one
+takes precedence (for instance SVGA over VGA16 over CIRRUS?).
+
- CONFIG_MOUSE_PS2=m
- CONFIG_INPUT_MOUSEDEV=m
CONFIG_INPUT_MOUSEDEV_PSAUX=y
@@ -540,6 +555,8 @@ Call Trace:
<SOFTIRQ>
- also real hardware (Unisys): no graphic VESA, only text-VESA modes, missing
=> cirrus framebuffer drivers at boot time (cannot be a module)?
+ => same behaviour with cirrus in qemu and 86box, enabling a PCI bus suddently
+ gives us cirrusfb messages and a working /dev/fb* (not more modes though)
- ${BASE} and @@BASE@@ with sed are both used to pass the build pathes to
builds in patches. ${BASE} sounds like it is working more by accident..
- 'cd DIR' used everywhere without 'cd DIR || exit 1'