Re: [PATCH ] scsi-misc-2.6: File System going into read-only mode

From: Stefan Richter <stefanr@s5r6.in-berlin.de>
To: James Bottomley <James.Bottomley@suse.de>
Cc: "Penchala Narasimha Reddy Chilakala,
	TLS-Chennai" <narasimhareddyc@hcl.in>,
	"'linux-scsi@vger.kernel.org'" <linux-scsi@vger.kernel.org>,
	ServeRAID Driver <ServeRAIDDriver@hcl.in>
Subject: Re: [PATCH ]  scsi-misc-2.6:  File System going into read-only mode
Date: Wed, 04 Nov 2009 02:40:04 +0100	[thread overview]
Message-ID: <4AF0DB74.8010906@s5r6.in-berlin.de> (raw)
In-Reply-To: <4AF0D9CE.2050407@s5r6.in-berlin.de>

Stefan Richter wrote:
> Stefan Richter wrote:
>> Before:
>>
>> if (upsg->sg[i].count > (dev->adapter_info.options & AAC_OPT_NEW_COMM)
>>         ? (dev->scsi_host_ptr->max_sectors << 9) : 65536) {
> 
> Doesn't make sense.  dev->adapter_info.options is a __le32 bitfield.
> 
>> After:
>>
>> if (upsg->sg[i].count > (
>>           (dev->adapter_info.options & AAC_OPT_NEW_COMM)
>>                  ? (dev->scsi_host_ptr->max_sectors << 9) : 65536
>>     )) {
> 
> Makes sense.  Don't know if it's correct.

PS:  These things wouldn't happen if
  a) the coding wouldn't be as deeply nested,
  b) sparse with endianess check would be used for QA.
(OK, it's apparently a decade old code.)
-- 
Stefan Richter
-=====-==--= =-== --=--
http://arcgraph.de/sr/