lkml.org 
[lkml]   [2002]   [Feb]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [patch][looking for maintainers] jiffies compare fixups
Date
From
Tim Schmielau <tim@physik3.uni-rostock.de> said:
> At the end of December, I made a patch to fix comparisons of the jiffies
> counter that would break at jiffies wraparound.

I think you should just forward the remaining bits to Marcelo directly.

[...]

> sent to: bjornw@axis.com Cc: dev-etrax@axis.com
>
> CRIS PORT
> P: Bjorn Wesen
> M: bjornw@axis.com
> L: dev-etrax@axis.com
> W: http://developer.axis.com
> S: Maintained
>
> --- linux-2.4.18-pre9/arch/cris/drivers/ethernet.c Mon Oct 8 20:43:54 200
> 1
> +++ linux-2.4.18-pre9-jiffies64/arch/cris/drivers/ethernet.c Fri Dec 28 15:1
> 6:22 2001
[...]

> @@ -1322,7 +1322,7 @@
> static void
> e100_clear_network_leds(unsigned long dummy)
> {
> - if (led_active && jiffies > led_next_time) {
> + if (led_active && jiffies > time_after(jiffies, led_next_time)) {
> e100_set_network_leds(NO_NETWORK_ACTIVITY);
>
> /* Set the earliest time we may set the LED */

This hunk is surely wrong.


[...]

> sent to: fibrechannel@compaq.com Cc: compaqandlinux@cpqlin.van-dijk.net
>
> COMPAQ FIBRE CHANNEL 64-bit/66MHz PCI non-intelligent HBA
> P: Amy Vanzant-Hodge
> M: Amy Vanzant-Hodge (fibrechannel@compaq.com)
> L: compaqandlinux@cpqlin.van-dijk.net
> W: ftp.compaq.com/pub/products/drivers/linux
> S: Supported
>
> --- linux-2.4.18-pre1/drivers/scsi/cpqfcTSstructs.h Thu Oct 25 22:53:50 200
> 1
> +++ linux-2.4.18-pre1-jiffies64/drivers/scsi/cpqfcTSstructs.h Fri Dec 28 12:4
> 8:20 2001
> @@ -27,7 +27,7 @@
>
> #define DbgDelay(secs) { int wait_time; printk( " DbgDelay %ds ", secs); \
> for( wait_time=jiffies + (secs*HZ); \
> - wait_time > jiffies ;) ; }
> + time_before(jiffies, wait_time) ;) ; }
>
> #define CPQFCTS_DRIVER_VER(maj,min,submin) ((maj<<16)|(min<<8)|(submin))
> // don't forget to also change MODULE_DESCRIPTION in cpqfcTSinit.c

Could use a bit of clarification. Or should be axed: Wait several _seconds_
in a busy loop?!

[...]

> sent to: grif@cs.ucr.edu
>
> /*
> * QLogic ISP2x00 SCSI-FCP
> * Written by Erik H. Moe, ehm@cris.com
> * Copyright 1995, Erik H. Moe
> *
> [...]
> */
>
> /* Renamed and updated to 1.3.x by Michael Griffith <grif@cs.ucr.edu> */
>
> --- linux-2.4.18-pre1/drivers/scsi/qlogicfc.c Thu Oct 25 22:53:51 2001
> +++ linux-2.4.18-pre1-jiffies64/drivers/scsi/qlogicfc.c Fri Dec 28 13:41:14 200
> 1

[...]

> @@ -820,7 +820,7 @@
> some time before recognizing it is attached to a fabric */
>
> #if ISP2x00_FABRIC
> - for (wait_time = jiffies + 5 * HZ; wait_time > jiffies;) {
> + for (wait_time = jiffies + 5 * HZ; time_before(jiffies, wait_time);) {
> barrier();
> cpu_relax();
> }

Again: Loop for a few seconds?

> sent to: jes@trained-monkey.org Cc: linux-hippi@sunsite.dk
>
> HIPPI
> P: Jes Sorensen
> M: jes@trained-monkey.org
> L: linux-hippi@sunsite.dk
> S: Maintained
>
> --- linux-2.4.18-pre1/drivers/net/rrunner.c Thu Nov 29 23:29:57 2001
> +++ linux-2.4.18-pre1-jiffies64/drivers/net/rrunner.c Fri Dec 28 12:15:34 200
> 1
> @@ -770,7 +770,7 @@
> * Give the FirmWare time to chew on the `get running' command.
> */
> myjif = jiffies + 5 * HZ;
> - while ((jiffies < myjif) && !rrpriv->fw_running);
> + while (time_before(jiffies, myjif) && !rrpriv->fw_running);
>
> netif_start_queue(dev);

Again a long loop?

> sent to: lizzi@cnam.fr Cc: davem@redhat.com
>
> /*
> $Id: fore200e.c,v 1.5 2000/04/14 10:10:34 davem Exp $
>
> A FORE Systems 200E-series driver for ATM on Linux.
> Christophe Lizzi (lizzi@cnam.fr), October 1999-March 2000.
>
>
> --- linux-2.4.18-pre1/drivers/atm/fore200e.c Fri Sep 14 00:21:32 2001
> +++ linux-2.4.18-pre1-jiffies64/drivers/atm/fore200e.c Fri Dec 28 14:36:32 200
> 1
> @@ -252,7 +252,7 @@
> fore200e_spin(int msecs)
> {
> unsigned long timeout = jiffies + MSECS(msecs);
> - while (jiffies < timeout);
> + while (time_before(jiffies, timeout));
> }

Better use <linux/delay.h> or thereabouts?

[...]

> sent to: simon@tk.uni-linz.ac.at Cc: frodol@dds.nl,linux-i2c@pelican.tk.uni-lin
> z.ac.at
>
> I2C DRIVERS
> P: Simon Vogl
> M: simon@tk.uni-linz.ac.at
> P: Frodo Looijaard
> M: frodol@dds.nl
> L: linux-i2c@pelican.tk.uni-linz.ac.at
> W: http://www.tk.uni-linz.ac.at/~simon/private/i2c
> S: Maintained
>
> --- linux-2.4.18-pre1/drivers/i2c/i2c-adap-ite.c Thu Oct 11 17:05:47 200
> 1
> +++ linux-2.4.18-pre1-jiffies64/drivers/i2c/i2c-adap-ite.c Fri Dec 28 11:3
> 8:01 2001
> @@ -82,7 +82,7 @@
> unsigned long j = jiffies + 10;
>
> DEB3(printk(" Write 0x%02x to 0x%x\n",(unsigned short)val, ctl&0xff));
> - DEB3({while (jiffies < j) schedule();})
> + DEB3({while (time_before(jiffies, j)) schedule();})
> outw(val,ctl);
> }

schedule_timeout()?

> --- linux-2.4.18-pre1/drivers/i2c/i2c-algo-bit.c Thu Oct 11 17:05:47 200
> 1
> +++ linux-2.4.18-pre1-jiffies64/drivers/i2c/i2c-algo-bit.c Fri Dec 28 11:3
> 9:08 2001
> @@ -49,7 +49,7 @@
> /* respectively. This makes sure that the algorithm works. Some chips */
> /* might not like this, as they have an internal timeout of some mils */
> /*
> -#define SLO_IO jif=jiffies;while(jiffies<=jif+i2c_table[minor].veryslow)\
> +#define SLO_IO jif=jiffies;while(time_before_eq(jiffies, jif+i2c_table[mi
> nor].veryslow))\
> if (need_resched) schedule();
> */

shedule_timeout() perhaps? In any case, this #define is dangerous, it
should be wrapped in "do {} while(0)" at least. And the behind-the-back
setting of jif is unsocial behaviour. Or just axe it, it is inside a
comment.
--
Horst von Brand http://counter.li.org # 22616
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2005-03-22 13:24    [from the cache]
©2003-2011 Jasper Spaans