[KJ] [PATCH 1/21] polling loops: change exit condition to

[KJ] [PATCH 1/21] polling loops: change exit condition to

[Marcin Slusarz]

hello everyone
here is the first part of patch which changes many polling loops to use
jiffies and time_after / time_before instead of counters
most of the changes are in drivers directory (except first patch)
as it is my first patch for linux kernel, feel free to point out any
mistake you'll find

(this task is the last entry of kernel-janitors todo: http://janitor.kernelnewbies.org/TODO)

LINUX FOR POWER MACINTOSH
P:	Benjamin Herrenschmidt
M:	benh@kernel.crashing.org

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
diff -upr -X linux-2.6.15-rc4/Documentation/dontdiff linux-2.6.15-rc4-orig/arch/ppc/platforms/pmac_smp.c linux-2.6.15-rc4/arch/ppc/platforms/pmac_smp.c
--- linux-2.6.15-rc4-orig/arch/ppc/platforms/pmac_smp.c	2005-12-03 15:22:31.000000000 +0100
+++ linux-2.6.15-rc4/arch/ppc/platforms/pmac_smp.c	2005-12-03 16:51:58.000000000 +0100
@@ -675,11 +675,10 @@ void cpu_die(void)
 
 void __cpu_die(unsigned int cpu)
 {
-	int timeout;
+	unsigned long end_time = jiffies + msecs_to_jiffies(1000);
 
-	timeout = 1000;
 	while (!cpu_dead[cpu]) {
-		if (--timeout = 0) {
+		if (time_after(jiffies, end_time)) {
 			printk("CPU %u refused to die!\n", cpu);
 			break;
 		}
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [KJ] [PATCH 1/21] polling loops: change exit condition to

[Arnd Bergmann]

Am Sonntag 04 Dezember 2005 01:12 schrieb Marcin Slusarz:
> hello everyone
> here is the first part of patch which changes many polling loops to use
> jiffies and time_after / time_before instead of counters
> most of the changes are in drivers directory (except first patch)
> as it is my first patch for linux kernel, feel free to point out any
> mistake you'll find

Hmm, what happened to the poll_event API that Nish had suggested
some time ago for this problem? Was there an actual reason against
merging it or did the patch just get lost?

	Arnd <><
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [KJ] [PATCH 1/21] polling loops: change exit condition to

[Nishanth Aravamudan]

[-- Attachment #1: Type: text/plain, Size: 764 bytes --]

On 04.12.2005 [20:57:52 +0100], Arnd Bergmann wrote:
> Am Sonntag 04 Dezember 2005 01:12 schrieb Marcin Slusarz:
> > hello everyone
> > here is the first part of patch which changes many polling loops to use
> > jiffies and time_after / time_before instead of counters
> > most of the changes are in drivers directory (except first patch)
> > as it is my first patch for linux kernel, feel free to point out any
> > mistake you'll find
> 
> Hmm, what happened to the poll_event API that Nish had suggested
> some time ago for this problem? Was there an actual reason against
> merging it or did the patch just get lost?

A good question. I'll try to dig up my patch from my old machine and
take a look if it can be used instead of all these changes.

Thanks,
Nish

[-- Attachment #2: Type: text/plain, Size: 168 bytes --]

_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [KJ] [PATCH 1/21] polling loops: change exit condition

[Marcin Slusarz]

Nishanth Aravamudan wrote:
>>>here is the first part of patch which changes many polling loops to use
>>>jiffies and time_after / time_before instead of counters
>>>most of the changes are in drivers directory (except first patch)
>>>as it is my first patch for linux kernel, feel free to point out any
>>>mistake you'll find
>>
>>Hmm, what happened to the poll_event API that Nish had suggested
>>some time ago for this problem? Was there an actual reason against
>>merging it or did the patch just get lost?
> 
> 
> A good question. I'll try to dig up my patch from my old machine and
> take a look if it can be used instead of all these changes.
i think you ment "with", not "instead" ;)
in my ("kernel newbie" ;) humble opinion these macros make code harder
too read

(last version of your patch i found is here: 
http://lists.osdl.org/pipermail/kernel-janitors/2005-March/003699.html)

regards,
Marcin
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors