Did you intend to include the changes in 1.6.1? It doesn't appear that they made
it into the release.
- Queries
- All Stories
- Search
- Advanced Search
- Transactions
- Transaction Logs
Advanced Search
Jan 29 2014
Done and 1.6.1 released
Jan 28 2014
Thanks. I backport it tomorrow.
I just tested with the changes applied to the LIBGCRYPT-1-6-BRANCH head, minus
the sha1.c changes (since the NEON parts weren't there), and it is building and
running fine for me. I assume it should work fine in master given this, and can
test that later if you would like, I'm just currently missing the libgpg-error >=
1.13 dependency for doing that. Thanks for your help getting this fixed up.
Fixed in master with commit 52f7c48.
Can you test it with master or should I do a backport to 1.6 first?
Fixed with commit fb8d8517 for 1.6.1
Backported for 1.6.1
Fixed for master with commit cbdc355.
Jan 27 2014
Thanks.
Changing category to dismngr to remind about doing a release.
Jan 24 2014
Looks good, compiling with libgcrypt-1.6.0 and creating a signed S/MIME e-mail
with gpgsm works.
Oh well, backporting is easy enough. Commit id f6bd8ed, will go into 1.6.1.
Will you be so kind and test the attached patch?
Jan 20 2014
This has meanwhile been fixed in master with commit 9edcf109.
I don't see a reason tobackport it to 1.6.
I don't think it makes sense to backport it to 1.5 - it has been this way for so
long. Users of 1.5 should upgrade to 1.6.0.
And in master with commit 5f2af6c2
Jan 13 2014
Fixed for 1.6.1 with commit 36214bfa.
Thanks.
Jan 8 2014
Fixed in master. Thanks.
Dec 31 2013
Dec 30 2013
Dec 24 2013
Dec 19 2013
Dec 16 2013
References:
OpenSSL
http://git.openssl.org/gitweb/?
p=openssl.git;a=blob;f=crypto/ecdsa/ecs_ossl.c;h=adab1f74b41daf6e719ca1fdae1ba81
7085c7802;hb=HEAD#l309
Nettle:
http://git.lysator.liu.se/nettle/nettle/blobs/master/ecc-ecdsa-sign.c#line86
http://git.lysator.liu.se/nettle/nettle/blobs/master/ecc-hash.c
NSS:
https://hg.mozilla.org/projects/nss/file/49360b638350/lib/freebl/ec.c#l746
Done something similar for master; needs to be backported to 1.5.
Dec 10 2013
Support is already in master and in the soon to be released 1.6.0
Libgcrypt 1.6 will be released this year.
Oct 7 2013
Duplicate of T1547
See T1547
Sep 24 2013
Aug 26 2013
Apr 18 2013
Fixed in master.
Fixed. see T1384.
Fixed for 1.5 and master. Only yet tested for 1.5 on an ultra64 (gcc63).
Duplicate of T1384
Apr 5 2013
Marking as resolved, as this seems to work fine now as far as I can tell (I'm
certainly happy).
scrypt has now been implemented.
Mar 22 2013
Mar 20 2013
Done in master.
Ah yes, that Serpent case again. We have a fix for that but a brief inspection
made me believe that it will only happen if the serpent.c code is used outside
of the Libgcrypt. It is an alignment problem. Please give me a few days to
prepare an updated patch.
I just verified that the issue only appears on 64 bit sparc, all others like 32
bit i386, 32 bit sparc and 64 bit amd64 work fine, so it is most likely a byte-
order issue.
Mar 18 2013
Argh, I did quite some fixes for the 1.5.1 release and assumed that I applied
this one as well. Now pushed but too late for 1.5.1 :-(. However, there is no
need to use automake for a released version.
You are using version 1.4.5 and not 1.5.0.
Two hints: Your are using the Solaris as and ld - better try with a native GNU
toolchain. You have some global CLFAGS set which add a lot of extra and useless
include paths. Please remove them.
I close the bug because it is pretty old. Feel free to re-open it if you have
newer info and tried with 1.5.0.