On Tue, Aug 18, 2009 at 6:14 PM, Cyril Plisko<[email protected]> wrote:
> On Thu, Nov 27, 2008 at 9:33 AM, Cyril Plisko<[email protected]>
> wrote:
>> On Thu, Nov 27, 2008 at 12:30 AM, Peter Buckingham
>> <[email protected]> wrote:
>>> Cyril Plisko wrote:
>>>>
>>>> 6777065 COMSTAR sources should be cstyle clean
>>>> 6777077 COMSTAR drivers unnecessary NUL-terminate strings
>>>> 6777085 COMSTAR Makefiles can take advantage of being ON citizens
>>>
>>> Cool! Can you request a sponsor too? We'll work on getting someone internal
>>> to take the ball from there.
>>
>
> After almost a year I finally found some time to get back this CRs.
>
> I updated the fix to the current ON code and want people to review them.
> The webrev is here - http://cr.opensolaris.org/~imp/comstar-fixes3/
I was asked privately about the inconsistency in the use of
INFO_BUF_SIZE in discovery.c and other places.
Here is the story behind it.
usr/src/uts/common/io/comstar/port/fct/discovery.c does not have
INFO_BUF_SIZE and INFO_BUF_SIZE_DOUBLE as its brethren, but rather
only have INFO_BUF_SIZE.
Original file has one place where 80 bytes long buffer and some other
places with 160 bytes long buffers. Unfortunately this single
occurrence of 80 bytes buffer was not enough to hold the string that
snprintf() is trying to put in. So I decided to have only 160 bytes
buffers across all the discovery.c
While original code doesn't constitute a serious issue (the message
would be cut, but no harm will be done) I think it is better to fix
it.
I also felt that having INFO_BUF_SIZE_DOUBLE w/o INFO_BUF_SIZE would
be a little bit strange, so decided to drop _DOUBLE in this file.
Admittedly it creates dissimilarity with respect to the other files
[while saving us about a hundred bytes of code ;)]
--
Regards,
Cyril
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss
