Carl-Daniel Hailfinger wrote:
On 12.02.2009 17:35, Stefan Reinauer wrote:
Carl-Daniel: if you solve the problem wit an algorithm that is easier to understand and/or shorter than the current code with 2 special cases, let's review your _code_ and I'm sure enough people will Ack it. Don't discuss discarding my patches with empty hands, I tend to react allergic to that.
Unfortunately, this patch
- broke every target using CONFIG_GXFUMA (AMD Pistachio and DBM690T)
- is inactive on every other target with UMA graphics and won't solve
the problem there.
Is there any in-tree board where the patch has brought an improvement?
Yes, the i945 (and other chipsets) can improve from this. I have an i945 patch in the queue to implement this, that will come shortly. It was kind of urgent to get the generic, infrastructure parts of the patches public before they bit-rot.
Sorry for the inconvenience and happy valentines day (it's a bit late for this timezone but I didn't wake up until 30min ago)
Stefan