On 19 Jan, Ronald G Minnich wrote:
> On Fri, 19 Jan 2001 [EMAIL PROTECTED] wrote:
>
>> >From intel_set_mtrr() in mtrr.c:
>>
>> while (ramsizeK != 0 && reg <= 6) {
>> intel_post(0x60 + reg);
>>
>> range_wb = 1 << (fms(ramsizeK) /* + 1 */);
>> range_uc = range_wb - ramsizeK;
>>
>> if ((range_uc==0) || ((ramsizeK % range_uc) == 0)) {
>>
>> By commenting out the +1 it calculated the correct size. However,
>> doing (ramsizeK % 0) for the case when range_uc is 0 caused the system
>> to crash. Probably a /0 exception that isn't getting handled. For
>> this I added the 0 test for range_uc.
>>
>> Do these changes screw up anything?
>>
>
>
> I would rather not commit them until Ollie gets back and has a chance to
> comment. He is off for the next week; is it OK to wait that long?
>
> ron
Yup. I don't think that is my problem and I can keep that patch
seperate until then.
Ty
--
Tyson D Sawyer iRobot Corporation
Senior Systems Engineer Real World Interface Div.
[EMAIL PROTECTED] Robots for the Real World
603-532-6900 ext 206 http://www.irobot.com
