Re: patch: delay_50ms() in ide-probe.c in 2.2.14

From: Khimenko Victor (khim@sch57.msk.ru)
Date: Sun Feb 20 2000 - 11:21:55 EST


In <Pine.LNX.4.10.10002200721010.9714-100000@lap.arijort.home> arijort@valinux.com (arijort@valinux.com) wrote:

> Simple arithmetic tells me this patch

> a) produces same behavior as previous code, and

Are you sure ??? Jiffes is *VOLATILE* variable. And it's REALLY volatile.

> b) does it with fewer calculations.

Yes, this version does DIFFERENT thing thenoriginal one. But faster :-)

> So it has to be good, right?

No.

> But then how do we exit the while loop?

In you version it'll not happen.

> How does jiffies change?

In timer interrupt.

> This patch applies cleanly against 2.2.14
> Against 2.3.45, the relevant file is ide.c
> but otherwise, it's clean.

And it's wrong in 2.2.14 and 2.3.45 :-)

> --- drivers/block/ide-probe.c.orig Sun Feb 20 00:46:54 2000
> +++ drivers/block/ide-probe.c Sun Feb 20 07:28:32 2000
> @@ -154,8 +154,7 @@
> */
> static void delay_50ms (void)
> {
> - unsigned long timeout = jiffies + ((HZ + 19)/20) + 1;
> - while (0 < (signed long)(timeout - jiffies));
> + while (0 < (signed long)((HZ + 39)/20));
> }

> /*

P.S. "Simple arithmetic" can be VERY wrong in multithreaded program. Like linux
kernel.

-
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/



This archive was generated by hypermail 2b29 : Wed Feb 23 2000 - 21:00:25 EST