|-----Original Message----- |From: Sevatio [mailto:[EMAIL PROTECTED]] |Sent: Friday, October 26, 2001 4:47 PM |To: [EMAIL PROTECTED] |Cc: [EMAIL PROTECTED] |Subject: Gigabyte GA-7ZX-1 & Bad System Clock (was Re: |[newbie] gigabyte GA-7ZXH) | | |Jose, | |I rebooted my GA-7ZX-1 to Windoze98 and have had it there for |more than a |day. The time remained accurate to the second. As far as I |know, Windoze98 |uses only the hardware clock for its timekeeping. | Does this board use the Winbond chip? According to the kernel stuff (which is where I got all of this, actually READING the timer.* source...) this problem only applies to Winbond timer chips. If you don't have one, you don't have this particular problem. |OTOH, LM8.0 & LM8.1 uses a combination of the PC's hardware |clock & its |system clock. The system clock's time can drift away from the |hardware |clock's time by over an hour per day. The problem is further |exacerbated |when the shutdown process instructs the hardware clock to set |its time to the |screwed-up systemclock. | In my two machines affected by the bug, just the opposite occurs. Setting the hardware clock to the system clock fixes things. The client machines adjust their time to a master, which in turn adjusts it's time via NTP to the standard. So when a shutdown occurs the hardware clock gets set to the correct time. |Normally "adjtimex" can be used to adjust the drift rate of |the system clock, |LM8.0 & LM8.1's drift rate fluctuates wildly rendering |adjtimex useless. | This is hard for me to judge since my systems set themselves so often... |Since Wind98 and LM7.2 (pre 2.4 era) did not have this |problem, I must say |that this timer issue is not OS independent and is related to |LM8.0 & LM8.1 |and perhaps any distro out there that uses the 2.4 kernel. | It could be, but I don't know. I guess I could turn off NTP updates and find out...
Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com