Messages in this thread | | | Date | Tue, 15 Dec 1998 16:06:33 +0100 (CET) | From | MOLNAR Ingo <> | Subject | Re: clock too slow (6 times !) in kernel 2.0.36+ (Xeon450/permedia2) |
| |
as a quick workaround, it's more or less safe to fix this up in user-space, just do this in a root process:
#include <asm/io.h> #include <unistd.h>
#define HZ 100 #define LATCH ((CLOCK_TICK_RATE + HZ/2) / HZ) #define CLOCK_TICK_RATE 1193180
main() { iopl(3);
/* set the clock to 100 Hz */ outb(0x34,0x43); outb(LATCH & 0xff , 0x40); /* LSB */ outb(LATCH >> 8 , 0x40); /* MSB */ }
(compile it with -O2. i have not actually tested or compiled this)
-- mingo
On 15 Dec 1998 storner@image.dk wrote:
> In <Pine.LNX.4.05.9812150814410.29586-100000@ace.ulyssis.student.kuleuven.ac.be> Bert de Bruijn <bob@ace.ulyssis.student.kuleuven.ac.be> writes: > > >A strange problem occurs on a Xeon-450 we have here (running RedHat 5.2) : > >the machine has a Permedia 2 AGP video card, that gets set to IRQ10 by > >the BIOS. > >The kernel clock runs fine, until I start X (XFree86-3DLabs-3.3.3). After > >that, the clock runs about six times too slow. > >So it takes about 6 seconds before there have been 100 timer interrupts. > > I've seen this reported on a local (danish) LUG list. It appears to be > a problem with the X server mistakenly reprogramming the timer chip > to interrupt at a different rate. In other words, it is an X server > bug, not a kernel bug. > > -- > Henrik Storner > Praetra censeo, Microsoftem esse delendam! > (Thanks to Erik Corry). > > - > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.rutgers.edu > Please read the FAQ at http://www.tux.org/lkml/ >
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |