Re: Segault Nettle.Hash()->crypt_hash()
Thomas Gusenleitner <[email protected]> Thu, 13 Jul 2017 08:48:17 +0200
| Newsgroups | gmane.comp.lang.pike.user |
|---|---|
| Message-ID | <CABAiGoAxh1Os6Q0OocXm02kQK4xOU8jDycQVp3keJTvgZarSTg@mail.gmail.com> |
Hi Grubba, thanks a lot - thats what i'm looking for. is there a way to define a salt in Crypto.Password.hash() ? greetings Thomas 2017-07-12 19:18 GMT+02:00 Henrik Grubbström <[email protected]>: > On Wed, 12 Jul 2017, Thomas Gusenleitner wrote: > > Hi, >> > > Hi. > > i try to figure out how to use Nettle.Hash()->crypt_hash(). >> >> here a simple test. >> Pike v8.0 release 438 running Hilfe v3.5 (Incremental Pike Frontend) >> >>> Nettle.Hash()->crypt_hash("password","$5$123456",5000); >>> >> Segmentation fault (core dumped) >> > > Oops, seems a NULL check was missing in crypt_hash(). > > Thanks for the report, fixed in Pike 8.0 and 8.1. > > The Nettle.Hash class is typically not used directly, but via some > module that inherits it and implements some basic hash algorithm. > > Pike v8.0 release 438 running Hilfe v3.5 (Incremental Pike Frontend) > >> Crypto.SHA256.crypt_hash("password", "123456", 5000); >> > (1) Result: "GZFZL7s1iPJu9f5mqAsHYBa/62DQI058psTiu9XgwgC" > > In your specific case it looks like you may be interested in the > Crypto.Password module: > > Pike v8.0 release 438 running Hilfe v3.5 (Incremental Pike Frontend) > >> Crypto.Password.hash("password", "$5$", 5000); >> > (1) Result: "$5$KIKLrEluYuV+c8+V$W4rf2hyb5c/EBDHrPTVJvUagnzYhOaGGVXEy0KO > 6173" > >> Crypto.Password.verify("password", "$5$KIKLrEluYuV+c8+V$W4rf2hyb5 >> c/EBDHrPTVJvUagnzYhOaGGVXEy0KO6173"); >> > (2) Result: 1 > >> Crypto.Password.verify("badpass", "$5$KIKLrEluYuV+c8+V$W4rf2hyb5 >> c/EBDHrPTVJvUagnzYhOaGGVXEy0KO6173"); >> > (3) Result: 0 > > Did this help? > > /grubba > > -- > Henrik Grubbström [email protected] > Roxen Internet Software AB