[Kernel-janitors] old patches never die?

[Kernel-janitors] old patches never die?

[Carl Spalletta]

  It seems as if a relatively small percentage of -kj patches ultimately
get merged at kernel.org.  For example the kswapd_init_fail patch from
2.6.3-kj1 and 2.6.4-rc2-kj1 not only didn't get merged but the file (mm/vmscan.c)
has since been patched (for a different purpose) at the same location. So that
patch would have to be redone anyway:

--- linux-263/mm/vmscan.c	2004-02-17 19:57:16.000000000 -0800
+++ linux-263-kj1/mm/vmscan.c	2004-02-18 14:42:03.000000000 -0800
@@ -1087,10 +1087,14 @@ int shrink_all_memory(int nr_pages)
 
 static int __init kswapd_init(void)
 {
+	int ret;
 	pg_data_t *pgdat;
 	swap_setup();
-	for_each_pgdat(pgdat)
-		kernel_thread(kswapd, pgdat, CLONE_KERNEL);
+	for_each_pgdat(pgdat) {
+		ret = kernel_thread(kswapd, pgdat, CLONE_KERNEL);
+		if (ret < 0)
+			panic("%s: unable to initialise kswapd\n", __FUNCTION__);
+	}
 	total_memory = nr_free_pagecache_pages();
 	return 0;
 }

  Perhaps it was deemed unnecessary to panic() if kswapd doesnt start ;).

  Generally speaking, what if anything can a janitor do to get positive or 
negative feedback on a submission?

__________________________________
Do you Yahoo!?
Yahoo! Finance Tax Center - File online. File on time.
http://taxes.yahoo.com/filing.html
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
http://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [Kernel-janitors] old patches never die?

[Zach Brown]

> -	for_each_pgdat(pgdat)
> -		kernel_thread(kswapd, pgdat, CLONE_KERNEL);
> +	for_each_pgdat(pgdat) {
> +		ret = kernel_thread(kswapd, pgdat, CLONE_KERNEL);
> +		if (ret < 0)
> +			panic("%s: unable to initialise kswapd\n", __FUNCTION__);
> +	}

If we're going to the trouble of storing 'ret' instead of just checking
the return of kernel_thread(), it sure would be nice to include it in
the panic() output.

- z

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

Re: [Kernel-janitors] old patches never die?

[Michael Still]

Carl Spalletta wrote:

>   Generally speaking, what if anything can a janitor do to get positive or 
> negative feedback on a submission?

Personally, I've found both this list and janitor@rustcorp.com.au to be 
helpful for getting feedback. You should probably send any given patch 
to only one of those two however, as real humans have to read the mail, 
and it's a shame to duplicate effort.

Cheers,
Mikal

-- 

Michael Still (mikal@stillhq.com) | "All my life I've had one dream,
http://www.stillhq.com            |  to achieve my many goals"
UTC + 11                          |    -- Homer Simpson
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
http://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [Kernel-janitors] old patches never die?

[Randy.Dunlap]

On Sun, 28 Mar 2004 17:49:27 -0800 (PST) Carl Spalletta wrote:

|   It seems as if a relatively small percentage of -kj patches ultimately
| get merged at kernel.org.  For example the kswapd_init_fail patch from
| 2.6.3-kj1 and 2.6.4-rc2-kj1 not only didn't get merged but the file (mm/vmscan.c)
| has since been patched (for a different purpose) at the same location. So that
| patch would have to be redone anyway:

Can you provide any data on the percentage of merges?


| --- linux-263/mm/vmscan.c	2004-02-17 19:57:16.000000000 -0800
| +++ linux-263-kj1/mm/vmscan.c	2004-02-18 14:42:03.000000000 -0800
| @@ -1087,10 +1087,14 @@ int shrink_all_memory(int nr_pages)
|  
|  static int __init kswapd_init(void)
|  {
| +	int ret;
|  	pg_data_t *pgdat;
|  	swap_setup();
| -	for_each_pgdat(pgdat)
| -		kernel_thread(kswapd, pgdat, CLONE_KERNEL);
| +	for_each_pgdat(pgdat) {
| +		ret = kernel_thread(kswapd, pgdat, CLONE_KERNEL);
| +		if (ret < 0)
| +			panic("%s: unable to initialise kswapd\n", __FUNCTION__);
| +	}
|  	total_memory = nr_free_pagecache_pages();
|  	return 0;
|  }
| 
|   Perhaps it was deemed unnecessary to panic() if kswapd doesnt start ;).
| 
|   Generally speaking, what if anything can a janitor do to get positive or 
| negative feedback on a submission?

I'd certainly like to see more involvement by reviewers (as I've
said in the past).  Willy and GregKH are good about reviewing,
and sometimes DaveJ, Acme, jgarzik.  Oh, and Domen, Luiz, Michael, etc.

The patch in question is still in the KJ patchset, so that is some
kind of positive feedback.  I prefer to see some kind of feedback on
all patches, but sometimes there aren't enough hours in a day for that,
so it comes down to:  if I put it into the KJ patchset, that's positive
feedback.  Otherwise it should get some email comments.

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