Re: [PATCH] Replace assert() with WARN_ON in drivers/scsi/libata-core.c

[Aubin LaBrosse <aubin@stormboxes.com>]

[-- Attachment #1: Type: text/plain, Size: 1173 bytes --]

Jeff Garzik wrote:
> Aubin LaBrosse wrote:
>> Replace assert() with WARN_ON in drivers/scsi/libata-core.c
>>
>> This is my first patch to the kernel - this work was suggested by 
>> Jeff Garzik as a janitorial task for libata.  If I could get some 
>> feedback letting me know that this is the correct format and my tools 
>> are setup right, I'll do the rest of these.
>>
>> Signed-off-by: Aubin LaBrosse <aubin@stormboxes.com>
>> ---
>>
>>  drivers/scsi/libata-core.c |   46 
>> ++++++++++++++++++++++----------------------
>>  1 files changed, 23 insertions(+), 23 deletions(-)
>
> Someone already beat you to it :)
>
> I looked over your patch anyway.  It looks acceptable, except we avoid 
> the '~' style in such tests:
>
> +    WARN_ON(qc == NULL);    /* ata_qc_from_tag _might_ return NULL */
> +    WARN_ON(~qc->flags & ATA_QCFLAG_ACTIVE);
>
>     Jeff
>
So what would be the proper way to invert that type of test?  (!(field & 
flag)) will do the same, right?  is that preferred for anything other 
than aesthetic reasons? (which is a fine reason to prefer something, of 
course, I'm just wondering if there's any other reasons, performance or 
otherwise

-aubin

[-- Attachment #2: aubin.vcf --]
[-- Type: text/x-vcard, Size: 160 bytes --]

begin:vcard
fn:Aubin LaBrosse
n:LaBrosse;Aubin
email;internet:aubin@stormboxes.com
tel;work:589-0692
tel;home:429-1520
tel;cell:493-0121
version:2.1
end:vcard