* tools/Makefile.am: gpg-auth is one in libexec_PROGRAMS.
* tools/gpg-auth.c: Support use by root for login user.
--
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
* configure.ac (HAVE_NEWER_LIBGCRYPT): Remove conditional.
* tools/Makefile.am (gpg_pair_tool_SOURCES): We build it always.
--
This is because we require libgcrypt 1.9 anyway.
* tools/symcryptrun.c: Remove.
* tools/Makefile.am: Ditto.
* doc/tools.texi: Remove man page.
* configure.ac: Remove build option and tests used only by this tool.
* Makefile.am (AM_DISTCHECK_CONFIGURE_FLAGS): Do not build
symcryptrun.
--
symcryptrun is too ancient to be of any use and has not been tested in
many years. Thus we should not distribute it anymore.
* tools/Makefile.am (gpgconf-w32): New target. Builds gpgconf with
subsystem windows.
* build-aux/speedo/w32/wixlib.wxs: Package it.
--
This allows us to create scripts that call gpgconf without opening
a console. Using subsystem windows is better then to just close
the console after start as it avoids the console flashing up.
(cherry picked from commit c366e04958481382c3f7b50f169120053186069b)
* tools/card-tool-keys.c: Rename to card-keys.c.
* tools/card-tool-misc.c: Rename to card-misc.c.
* tools/card-tool-yubikey.c: Rename to card-yubikey.c.
* tools/card-tool.h: Rename to gpg-card.h.
* tools/gpg-card-tool-w32info.rc: Rename to gpg-card-w32info.rc
* doc/card-tool.texi: Rename top gpg-card.texi
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/card-tool-yubikey.c: New.
* tools/Makefile.am (gpg_card_tool_SOURCES): Add it.
* tools/card-call-scd.c (scd_apdu): Allow returning data.
* tools/card-tool-misc.c (send_apdu): New. Move from gpg-card-tool.c
and let it return data. Change all callers.
* tools/gpg-card-tool.c (cmd_writecert): Prepend the certref with the
current application type.
(cmd_yubikey): New.
--
This command allows listing of active applications and to enable or
disable selected applications. This is in particular useful to
disable the OpenPGP application so that the PIV support can easily be
tested.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/card-tool-misc.c: New.
* tools/card-tool.h: Rewored data structures for key infos.
* tools/gpg-card-tool.c: Ditto.
* tools/card-call-scd.c: Ditto.
--
Note that this also changes the way the key information is printed.
Formerly we printed it like:
Signature key ....: <openpgp-fingerprint>
created ....: <timestamp>
keygrip ... : <keygrip>
now we do:
Signature key ....: <keygrip>
fingerprint : <openpgp-fingerprint>
created ....: <timestamp>
This is because a keygrip is always available but a fingerprint and
the creation date are properties of an OpenPGP card. A standard way
of listing keys is better than one depending on the type of card.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/card-call-scd.c: New.
* tools/card-tool.h: new.
* tools/gpg-card-tool.c: Largely extended.
--
gpg-card-tool will eventually replace the --card-edit command of gpg
because it makes more sense to have a multi-protocol aware tool and
not just one for OpenPGP cards. Most OpenPGP card things works now
but a few, those which require close interaction with gpg, still need
to be implemented. And of course the whole planned non-interactive
stuff needs to be written.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/gpg-card-tool.c: New.
* tools/gpg-card-tool-w32info.rc: New.
* tools/Makefile.am: Add new tool.
--
To support more cards than the OpenPGP card it is useful to have a
separate tool. It will have have the "gpg --card-edit" style
interactive interface as well as direct command line options for all
commands. In a first step the OpenPGP card will be supported, to
allow its use as an alternative to the gpg command, and the
forthcoming PIV card support.
The tool can be though as a direct interface to scdaemon.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/gpg-zip.in: Remove.
* m4/tar-ustar.m4: Remove.
--
Note that the script was even not anymore installed. See also
GnuPG-bug-id: 4252
Signed-off-by: Werner Koch <wk@gnupg.org>
* configure.ac (GNUPG_CACHE_DIR): New const.
* tools/Makefile.am (libexec_PROGRAMS): Add gpg-pair-tool.
(gpg_pair_tool_SOURCES, gpg_pair_tool_CFLAGS)
(gpg_pair_tool_LDADD): New.
* tools/gpg-pair-tool.c: New.
--
This is a first try on a protocol to pair two devices so that they can
agree on a shared secret to exchange secret keys. The idea is that if
you want to sync your secret keys to another machine (e.g. from
desktop to mobile) you have physical access to both devices and thus a
pairing protocol allows to authenitcate the connection using a short
string. See the source for a protocol description.
How to test:
$ gpg-pair-tool -va --homedir . --initiate >msg.commit
$ gpg-pair-tool -va --homedir 2ndhome --respond \
<msg.commit >msg.dhpart1
$ gpg-pair-tool -va --homedir . --respond \
<msg.dhpart1 >msg.dhpart2
$ gpg-pair-tool -va --homedir 2ndhome --respond \
<msg.dhpart2 >msg.confirm
Now set the SAS as printed by the responder into SAS and run
$ gpg-pair-tool -va --homedir . --respond --sas $SAS <msg.confirm
Storing the secret on disk is obviously not the right thing to do.
With the new PUT_SECRET and GET_SECRET commands of gpg-agent we can
change this to store it all in gpg-agent instead. This will make it
also easier for gpg to access the secret and we won't need an option
to return it from gpg-pair-tool. Thus gpg-pair-tool can be dedicated
to run the protocol and maybe to popup info dialogs.
Adding a second expiration time for running the protocol in addition
to the expiration of the secret is probably a better idea than just
that simple catch-all TTL.
Signed-off-by: Werner Koch <wk@gnupg.org>
We use gpgtar to unpack test data, hence we always build it. If the
user opts out, we simply don't install it.
* configure.ac: Add comment.
* tests/migrations/Makefile.am (required_pgms): Make sure gpgtar is
built.
* tools/Makefile.am: Always build gpgtar, but do not install it if the
user used '--disable-gpgtar'.
Signed-off-by: Justus Winter <justus@g10code.com>
* tools/send-mail.c, tools/send-mail.h: New.
* tools/wks-util.c: New.
* tools/Makefile.am (gpg_wks_server_SOURCES): Add them.
(gpg_wks_client_SOURCES): Ditto.
* tools/gpg-wks.h (opt): Add fields use_sendmail and output.
* tools/gpg-wks-client.c: Add options --send and --output. Rename
command --send to --create.
(command_send, send_confirmation_response): Output via wks_send_mime.
* tools/gpg-wks-server.c: Add options --send and --output.
(send_confirmation_request): Output via wks_send_mime.
(check_and_publish): Add hack for name-value bug.
--
With this code, a dedicated user on the server along with a procmail
script, it was possible to run a basic test.
Signed-off-by: Werner Koch <wk@gnupg.org>
* configure.ac: Add option --enable-wks-tools
* tools/gpg-wks-client.c: New.
* tools/gpg-wks-server.c: New.
* tools/gpg-wks.h: new.
* tools/wks-receive.c: New.
* tools/call-dirmngr.c, tools/call-dirmngr.h: New.
--
Note that this is just a starting point and not a finished
implementation. Here is how to test the system using
foo@test.gnupg.org as example.
Prepare:
mkdir /var/lib/gnupg/wks
chmod o-rwx /var/lib/gnupg/wks
mkdir /var/lib/gnupg/wks/test.gnupg.org
Run the protocol:
./gpg-wks-client -v --send FPR USERID >x
./gpg-wks-server -v --receive <x >y
./gpg-wks-client --receive <y >z
./gpg-wks-server -v --receive <z
You should also setup a cron job to rsync
/var/lib/gnupg/wks/test.gnupg.org/hu/* to the webserver.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tests/openpgp/Makefile.am (CLEANFILES): Add created file
"passphrases".
* tools/Makefile.am (EXTRA_DIST): Add no-libgcrypt.c.
--
dirmngr/Makefile uses a copy no-libgcrypt.c from tools/. tools/ does
not anymore require no-libgcrypt.c but we keep it there in case we
eventually make use of it again.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/Makefile.am (gpgconf_SOURCES): Remove no-libgcrypt.c.
(gpgconf_LDADD): Add LIBGCRYPT_LIBS.
(gpg_connect_agent_LDADD): Ditto.
(gpgtar_LDADD): Ditto.
* dirmngr/Makefile.am (dirmngr_client_LDADD): Ditto.
(t_common_ldadd): Ditto. Remove no-libgcrypt.o.
--
We need this change so that a future code in common/ can use Libgcrypt
functions; in particular hashing.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/Makefile.am: gpgtar now requires neither npth nor libassuan.
* tools/gpgtar-create.c (gpgtar_create): Use the new 'sh-exectool'
helper.
* tools/gpgtar-extract.c (gpgtar_extract): Likewise.
* tools/gpgtar-list.c (gpgtar_list): Likewise.
* tools/gpgtar.c (main): Set default gpg program. Drop the
initialization of npth and libassuan.
Signed-off-by: Justus Winter <justus@g10code.com>
* tools/Makefile.am: Amend CFLAGS and LDADD.
* tools/gpgtar-create.c (gpgtar_create): Add encrypt flag and encrypt
stream if requested.
* tools/gpgtar-extract.c (gpgtar_extract): Likewise for decryption.
* tools/gpgtar-list.c (gpgtar_list): Likewise.
* tools/gpgtar.c (main): Initialize npth and assuan. Parse recipient
and local user, and note which flags are currently ignored. Adapt
calls to gpgtar_list and friends.
(tar_and_encrypt): Drop stub function and prototype.
(decrypt_and_untar): Likewise.
(decrypt_and_list): Likewise.
* tools/gpgtar.h (gpgtar_{create,extract,list}): Add encryption or
decryption argument.
Signed-off-by: Justus Winter <justus@g10code.com>
* tools/gpgsm-gencert.sh: remove deprecated script entirely. It is
fully replaced by gpgsm --gen-key
* doc/tools.texi: remove gpgsm-gencert.sh documentation
* .gitignore: no longer ignore gpgsm-gencert.sh manpage
* doc/Makefile.am: quit making the manpage
* tools/Makefile.am: quit distributing the script
* doc/howto-create-a-server-cert.texi: overhaul documentation to use
gpgsm --gen-key and tweak explanations
--
The commit deprecating gpgsm-gencert.sh
(81972ca7d53ff1996e0086702a09d4405bdc2a7e) dates back exactly 6 years.
https://codesearch.debian.net/results/gpgsm-gencert.sh
suggests that in all of debian it is only referenced in documentation
(for poldi and scute) and example files (libept), and isn't actually
used directly anywhere.
Furthermore, trying to use gpgsm-gencert.sh to make a simple webserver
certificate-signing request failed for me, following the examples in
doc/howto-create-a-server-cert.texi exactly.
It's time we ripped off this band-aid :)
Signed-off-by: Daniel Kahn Gillmor <dkg@fifthhorseman.net>
* gl/: Remove entire tree.
* configure.ac: Remove gnulib tests and the gl/ Makefile.
(setenv): Add to AC_CHECK_FUNCS.
* autogen.rc (extra_aclocal_flags): Set to empty.
* Makefile.am (ACLOCAL_AMFLAGS): Remove -I gl/m4
(SUBDIRS): Remove gl/.
* agent/Makefile.am (common_libs): Remove ../gl/gnulib.a
* common/Makefile.am (t_common_ldadd): Ditto.
* dirmngr/Makefile.am (dirmngr_LDADD): Ditto.
(dirmngr_ldap_LDADD, dirmngr_client_LDADD): Ditto.
* g10/Makefile.am (needed_libs): Ditto.
* g13/Makefile.am (g13_LDADD): Ditto.
* kbx/Makefile.am (kbxutil_LDADD): Ditto.
($(PROGRAMS)): Ditto.
* scd/Makefile.am (scdaemon_LDADD): Ditto.
* sm/Makefile.am (common_libs): Ditto.
* tools/Makefile.am (common_libs, commonpth_libs): Ditto.
* agent/gpg-agent.c: Remove "mkdtemp.h"
* g10/exec.c: Ditto.
* scd/scdaemon.c: Ditto.
* tools/symcryptrun.c: Ditto.
* common/sysutils.c: Remove "setenv.h"
* common/t-timestuff.c: Use putenv if setenv is not available.
--
gnulib has always been a cause of trouble in GnuPG because we used
only a very few functions and the complex include machinery of gnulib
is quite complex and the cause for many build problems for example on
OS X. This is not gnulib's fault but due to our limited use of gnulib
and that we only rarely update the gnulib code to avoid regressions.
In part two we will address the functions
mkdtemp
setenv
unsetenv
strpbrk
which may bot be implemented on all platforms. They are not required
on a libc based system.
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/Makefile.am: Avoid $(DNSLIBS) for dirmngr_ldap
* g10/Makefile.am: $(LIBREADLINE) is only for gpg2; gpgv2 does not
need $(LIBASSUAN_LIBS)
* sm/Makefile.am: gpgsm does not need $(ZLIBS)
* tools/Makefile.am: gpgconf does not need $(NPTH_LIBS)
--
In the course of building GnuPG 2.1.0 beta864 on debian, i found that
several of the installed executables were linked to libraries that
they did not need to be linked to, which would cause unnecessary
package dependencies at runtime.
The changeset here removes these unnecessary libraries from linking.
Something similar could possibly also be done by passing --as-needed
to the linker, but trimming the depenencies seems more parsimonious.
