commit 0d5ee5304f9141098e839f67cb0ce19bf127f485
parent 5d8be559bb9a31f8077253fc54f5062f0bb011f0
Author: Kyle Milz <krwmilz@gmail.com>
Date: Sat, 30 Apr 2022 19:10:25 +0000
rewrite README to be better
Diffstat:
| M | README | | | 53 | ++++++++++++++++++++++++++++++++++++----------------- |
1 file changed, 36 insertions(+), 17 deletions(-)
diff --git a/README b/README
@@ -1,17 +1,36 @@
-The ATI Radeon HD4850M in my 2009 iMac 11,1 does not work with the
-Linux radeondrm stack. Although I use OpenBSD, it has the radeondrm(4)
-stack from Linux, and suffers the same problem. There are reports
-(https://bugs.freedesktop.org/show_bug.cgi?id=98046) of other people
-having the same problem too.
-
-Latest Release: 70 on July 20 2021
-Black screen: imac_radeondrm_atombios_encoders.diff
-Bad video: imac_radeondrm_display.diff
-Install: cd /sys/dev/pci/drm && patch < imac_radeondrm_*.diff</code>
-
-There are two distinct problems:
-
- 1) The first problem is the display goes black when the radeondrm
- code initializes the GPU.
- 2) Once the first has been figured out, the video output is
- corrupted resulting in strange rhythmic lines on the screen.
+ iMac 11,1 radeondrm(4)
+========================
+
+The ATI Radeon HD4850M in my 2009 iMac 11,1 does not work with
+radeondrm(4) because of two distinct problems.
+Both Linux (https://bugs.freedesktop.org/show_bug.cgi?id=98046) and OpenBSD
+have this issue.
+
+Latest Release: 71 on April 30 2022
+Install: # cd /sys/dev/pci/drm && patch < imac_radeondrm_*.diff
+
+There are two distinct problems, each fixed with their own diff.
+
+
+ 1 - Black screen
+-----------------
+
+The first problem is the display goes black, with backlight off, when
+radeondrm(4) initializes the GPU. The fix for this issue is similar to
+the workaround added for iMac 10,1 machines.
+
+The following patches this issue in the OpenBSD kernel:
+
+ # cd /sys/dev/pci/drm && patch < imac_radeondrm_atombios_encoders.diff
+
+
+ 2 - PLL computation
+--------------------
+
+There is something wrong with the new PLL code that was introduced around
+2012 that causes the PLL computation to be incorrect for this board.
+Reverting that change fixes this issue.
+
+The following patches this issue in the OpenBSD kernel:
+
+ # cd /sys/dev/pci/drm && patch < imac_radeondrm_display.diff
