From mboxrd@z Thu Jan  1 00:00:00 1970
From: Irwan Djajadi <irwan.djajadi@iname.com>
Date: Mon, 24 Oct 2005 23:46:01 +0000
Subject: Re: [KJ] [PATCH 2.6.14-rc5 1/1] drivers/block/acsi_slm.c: replace
Message-Id: <435D7239.50600@iname.com>
List-Id: <kernel-janitors.vger.kernel.org>
References: <20051023025351.GC2438@poopie.dyndns.org>
In-Reply-To: <20051023025351.GC2438@poopie.dyndns.org>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: kernel-janitors@vger.kernel.org

Alexey Dobriyan wrote:

>On Sat, Oct 22, 2005 at 09:53:51PM -0500, irwan.djajadi@iname.com wrote:
>  
>
>>drivers/block/acsi_slm.c: replace interruptible_sleep_on() with
>>wait_event_interruptible()
>>    
>>
>
>  
>
>>--- 2.6.14-rc5/drivers/block/acsi_slm.c
>>+++ mod/drivers/block/acsi_slm.c
>>@@ -627,7 +627,7 @@ static ssize_t slm_write( struct file *f
>> 
>> 	while( SLMState = PRINTING ||
>> 		   (SLMState = FILLING && SLMBufOwner != device) ) {
>>-		interruptible_sleep_on( &slm_wait );
>>+		wait_event_interruptible( slm_wait, SLMState=IDLE );
>> 		if (signal_pending(current))
>> 			return( -ERESTARTSYS );
>> 	}
>>    
>>
>
>acsi_slm_replace_interruptible_sleep_on_with_wait_event_interruptible.patch
>from -kj looks more correct.
>
>From: Nishanth Aravamudan
>
>Use wait_event_interruptible() instead of the deprecated
>interruptible_sleep_on(). The sleep_on() call later in the same
>function is replaced with inline wait-queue code which achieves the
>same. This required adding a local wait-queue, though.
>
>--- a/drivers/block/acsi_slm.c
>+++ b/drivers/block/acsi_slm.c
>@@ -625,12 +626,10 @@ static ssize_t slm_write( struct file *f
> 	int		device = iminor(node);
> 	int		n, filled, w, h;
> 
>-	while( SLMState = PRINTING ||
>-		   (SLMState = FILLING && SLMBufOwner != device) ) {
>-		interruptible_sleep_on( &slm_wait );
>-		if (signal_pending(current))
>-			return( -ERESTARTSYS );
>-	}
>+	wait_event_interruptible(slm_wait, (SLMState != PRINTING &&
>+				(SLMState != FILLING || SLMBufOwner = device)));
>+	if (signal_pending(current))
>+		return -ERESTARTSYS;
> 	if (SLMState = IDLE) {
> 		/* first data of page: get current page size  */
> 		if (slm_get_pagesize( device, &w, &h ))
>
>  
>
That does look more correct. Please ignore my patch.
Thanks!
--
Irwan
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors