2 #+TITLE: Hacking notes for Libgcrypt
7 The following stuff explains some basic procedures you need to
8 follow if you want to contribute code or documentation.
10 ** No more ChangeLog files
12 Do not modify any of the ChangeLog files in Libgcrypt. Starting on
13 December 1st, 2011 we put change information only in the GIT commit
14 log, and generate a top-level ChangeLog file from logs at "make
15 dist" time. As such, there are strict requirements on the form of
16 the commit log messages. The old ChangeLog files have all be
17 renamed to ChangeLog-2011
19 ** Commit log requirements
21 Your commit log should always start with a one-line summary, the
22 second line should be blank, and the remaining lines are usually
23 ChangeLog-style entries for all affected files. However, it's fine
24 -- even recommended -- to write a few lines of prose describing the
25 change, when the summary and ChangeLog entries don't give enough of
26 the big picture. Omit the leading TABs that you're used to seeing
27 in a "real" ChangeLog file, but keep the maximum line length at 72
28 or smaller, so that the generated ChangeLog lines, each with its
29 leading TAB, will not exceed 80 columns.
33 Libgcrypt is currently licensed under the LGPLv2+ with tools and the
34 manual being under the GPLv2+. We may eventually update to a newer
35 version of the licenses or a combination of them. It is thus
36 important, that all contributed code allows for an update of the
37 license; for example we can't accept code under the LGPLv2(only).
39 Libgcrypt used to have a strict policy of requiring copyright
40 assignments to the FSF. To avoid this major organizational overhead
41 and to allow inclusion of code, not copyrighted by the FSF, this
42 policy has been relaxed. It is now also possible to contribute code
43 by asserting that the contribution is in accordance to the
44 "Libgcrypt Developer's Certificate of Origin" as found in the file
45 "DCO". (Except for a slight wording change, this DCO is identical
46 to the one used by the Linux kernel.)
48 If your want to contribute code or documentation to Libgcrypt and
49 you didn't signed a copyright assignment with the FSF in the past,
50 you need to take these simple steps:
52 - Decide which mail address you want to use. Please have your real
53 name in the address and not a pseudonym. Anonymous contributions
54 can only be done if you find a proxy who certifies for you.
56 - If your employer or school might claim ownership of code written
57 by you; you need to talk to them to make sure that you have the
58 right to contribute under the DCO.
60 - Send an OpenPGP signed mail to the gcrypt-devel@gnupg.org mailing
61 list from your mail address. Include a copy of the DCO as found
62 in the official master branch. Insert your name and email address
63 into the DCO in the same way you want to use it later. Example:
65 Signed-off-by: Joe R. Hacker <joe@example.org>
67 (If you really need it, you may perform simple transformations of
68 the mail address: Replacing "@" by " at " or "." by " dot ".)
70 - That's it. From now on you only need to add a "Signed-off-by:"
71 line with your name and mail address to the commit message. It is
72 recommended to send the patches using a PGP/MIME signed mail.
76 Please follow the GNU coding standards. If you are in doubt consult
77 the existing code as an example. Do no re-indent code without a
78 need. If you really need to do it, use a separate commit for such a
83 ** Taking optimized MPI code out of GMP:
85 I generated the pentium4/* files by glueing the existing assembler
86 prologues to the GMP 4.2.1 assembler files generated with the m4
87 tool in GMP's build process, for example:
89 $ m4 -DHAVE_CONFIG_H -D__GMP_WITHIN_GMP -DOPERATION_rshift -DPIC \
90 rshift.asm >tmp-rshift.s
92 Then tmp-rshift will contain the assembler instructions for the
93 configured platform. Unfortunately, this way the comments are lost.
94 For most files I re-inserted some of the comments, but this is
100 ** Debugging math stuff:
102 While debugging the ECC code in libgcrypt, I was in need for some
103 computer algebra system which would allow me to verify the numbers
104 in the debugging easily. I found that PARI (pari-gp package in
105 Debian) has support for elliptic curves. The below commands shows
106 how they are set up and used with an example.
109 hextodec(s)=local(v=Vec(s),a=10,b=11,c=12,d=13,e=14,f=15,A=10,B=11,C=12,D=13,E=14,F=15,h);if(#setunion(Set(v),Vec("0123456789ABCDEFabcdef"))>22,error);for(i=1,#v,h=shift(h,4)+eval(v[i]));h
111 p = hextodec("01FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF")
112 a = hextodec("01FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFC")
113 b = hextodec("51953EB9618E1C9A1F929A21A0B68540EEA2DA725B99B315F3B8B489918EF109E156193951EC7E937B1652C0BD3BB1BF073573DF883D2C34F1EF451FD46B503F00")
115 /* Set up y^2 = x^3 + ax + b mod (p). */
116 e = ellinit(Mod(1,p)*[0,0,0,a,b]);
118 gx = hextodec ("00C6858E06B70404E9CD9E3ECB662395B4429C648139053FB521F828AF606B4D3DBAA14B5E77EFE75928FE1DC127A2FFA8DE3348B3C1856A429BF97E7E31C2E5BD66")
119 gy = hextodec ("011839296A789A3BC0045C8A5FB42C7D1BD998F54449579B446817AFBD17273E662C97EE72995EF42640C550B9013FAD0761353C7086A272C24088BE94769FD16650")
122 n = hextodec ("01FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFA51868783BF2F966B7FCC0148F709A5D03BB5C9B8899C47AEBB6FB71E91386409")
124 /* Verify that G is on the curve, and that n is the order. */
129 d = hextodec ("018F9573F25059571BDF614529953DE2540497CEDABD04F3AF78813BED7BB163A2FD919EECF822848FCA39EF55E500F8CE861C7D53D371857F7774B79428E887F81B")
131 qx = hextodec ("00316AAAD3E905875938F588BD9E8A4785EF9BDB76D62A83A5340F82CB8E800B25619F5C3EA02B7A4FA43D7497C7702F7DFBEAC8E8F92C3CAABD9F84182FDA391B3B")
132 /* Note: WRONG! (It is apparent that this is the same as X shifted by
134 qy = hextodec ("0000316AAAD3E905875938F588BD9E8A4785EF9BDB76D62A83A5340F82CB8E800B25619F5C3EA02B7A4FA43D7497C7702F7DFBEAC8E8F92C3CAABD9F84182FDA391B")
137 /* Calculate what Q should be given d. */
140 /* This is not 0 and thus shows that libgcrypt gave Q and d that do
143 ====8<=====================