[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Date: Sun, 6 Sep 2015 13:51:44 -0500
From: JimF <jfoug@....net>
To: john-dev@...ts.openwall.com
Subject: Re: auditing our use of FMT_* flags
Without looking at code, it may only need to init if the FMT_DYNAMIC is set.
Dyna is the only think out there which late builds the format data. All
other's
use the statically defined structure data. Within dyna, there is no format
structures at all, until init, and within 'thin' formats, the format
structure
ONLY contains data that has to be valid before init() is called.
Now, in hind sight, is this flag something required prior to init() If
so, then
it should be there, even if it will be 'reset' to whatever dynamic wants it
to be after init(). I do not think split is available before init() is
it, other
than possibly in the ST code.
On 9/6/2015 1:47 PM, Solar Designer wrote:
> On Sun, Sep 06, 2015 at 01:40:50PM -0500, JimF wrote:
>> These formats 100% handle this flag properly. They just do not set it
>> into their format structure in a static manner.
> Should Kai's tests init() the formats before checking flags, then?
>
> Or, for speed, should they only init() the formats when the test
> appears to fail, and re-test?
>
> Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
