* g10/openfile.c (overwrite_filep): Use gnupg_access.
--
As said, this is just an obvious but partial fix. We need to review
things for the output module.
Signed-off-by: Werner Koch <wk@gnupg.org>
* tools/gpgtar-create.c (fillup_entry_w32): Move parentheses.
--
Fixes-commit: 8b8925a2bdbb12dd537dde20a27cdb1416c2f1ae
The bug is so obvious that I wonder why it was not reported more often
on Windows. (Adding 1 to MAXDWORD (0xfffffff) always gives 0 for the
product).
Signed-off-by: Werner Koch <wk@gnupg.org>
* sm/gpgsm.c (opts): Add option --ldapserver and make --keyserver an
alias.
--
We should use "keyserver" for OpenPGP and thus it is better to allow
for "ldapserver" here - it is the same convention as now used in
dirmngr.
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/dirmngr_ldap.c (main): Handle no args case.
--
This is required for example for CRLs. The old code did not require
this because the hos was taken from the URL given has arg.
Signed-off-by: Werner Koch <wk@gnupg.org>
* common/exechelp-w32.c (gnupg_spawn_process): Change to use
CreateProcessW.
(gnupg_spawn_process_fd): Ditto.
(gnupg_spawn_process_detached): Ditto.
* g10/exec.c (w32_system): Ditto.
--
GnuPG-bug-id: 4398
We do not use this weirdo CREATE_UNICODE_ENVIRONMENT flag because it
does not make any sense to have non-ASCII names in the environment. I
can't imagine why this should be used at all and rationale for this
API features is, well, sparse.
(cherry picked from commit cf2f6d8a3f0594c03c383b4989a3041e9c4536d7)
* common/exechelp-w32.c (gnupg_spawn_process_detached): Add
CREATE_BREAKAWAY_FROM_JOB creation flag if required.
--
When the gpg process is assigned to a W32 "Job" the
child processes are killed once the Job is finished.
As we want our detached processes to linger e.g.
gpg-agent the breakaway flag is required in
that case.
GnuPG-Bug-Id: T4333
Thanks to Jan Echternach for reporting this and providing
a patch.
Signed-off-by: Andre Heinecke <aheinecke@gnupg.org>
(cherry picked from commit 03df28b18b92b3fd3d2ba1000903c088dc5b0fcf)
* common/init.c (_init_common_subsystems) [W32]: Set the codepage to
UTF-8 for input and putput. Switch gettext to UTF-8.
* g10/gpg.c (utf8_strings) [W32]: Make sure this is always set.
--
With this patch the former patch to use ReadConsoleW and WriteConsoleW
in ttyio.c are kind of superfluous because the ANSI version of these
functions are also able to read/write UTF-8 directly given the console
code page has been set correctly. However, this seems to work only
with recent versions of Windows-10.
GnuPG-bug-id: 4365
(cherry picked from commit 8c41b8aac3efb78178fe1eaf52d8d1bbc44941a8)
Removed changes for "gpgconf --show-codepages" of the original patch.
* common/ttyio.c (do_get) [W32]: Use ReadConsoleW.
(w32_write_console): New.
(tty_printf, tty_fprintf) [W32]: Use new function.
--
Note that due this change fixed stings (i.e. gettext translations)
printed to the console will not be rendered correctly unless "chcp
65001" has been used. This needs to be fixed by followup patch.
GnuPG-bug-id: 4365
(cherry picked from commit f165c8a737cc968554c9d78932c69869456108ff)
* common/ttyio.c: Remove cruft like EMX and RISCOS support. Translate
a few strings. Re-indent.
--
Backported-from-master: 8622f53994249d8fb49a488cfe480ffbeb8cbfba
* common/t-w32-cmdline.c (test_all): Add new test cases.
* common/w32-misc.c (strip_one_arg): Add arg endquote.
(parse_cmdstring): Take care of backslashes in quotes.
--
I found some new test vectors from Microsoft.
(cherry picked from commit 20c60076866904187a09393de596deef286116f8)
* common/w32-misc.c: New.
* common/t-w32-cmdline.c: New.
* common/init.c: Include w32help.h.
(prepare_w32_commandline): New.
(_init_common_subsystems) [W32]: Call prepare_w32_commandline.
* common/Makefile.am (common_sources) [W32]: Add w32-misc.c
(module_tests): Add t-w32-cmdline
(t_w32_cmdline_LDADD): New.
--
The rules for the command line parser are not cleary specified - if at
all. See the comment in t-w32-cmdline.c.
We can't use the mingw version because that would require to change
all argv handling to be wchar_t and that only for Windows. That would
be too ugly. Parsing the command line into argv by us is much easier
and we can do that only if needed - i.e. if globing is required (we
are prepared for this) or a non-ASCII character has been encountered.
This way we keep things stable and only fix the currently not working
Unicode problem.
GnuPG-bug-id: 4398
(cherry picked from commit deb6c94362c0f179de1cac18707aad2f51a21e10)
* g10/gpg.c (_dowildcard): Remove.
(my_strusage): Enable wildcards using our new system.
--
This patch actually removes the current globing using a mingw
feature. We are not able to use this because it does not handle
Unicode filenames. The plan is to implement this in init.c. This
patch merely configures gpg to request globing.
GnuPG-bug-id: 4398
(cherry picked from commit 8e15506d6680bbee85bc01453da28fc90b4cb673)
* agent/cvt-openpgp.c (convert_from_openpgp_main): Return
GPG_ERR_BAD_SECKEY.
--
Backport master commit of:
21ef425e222ddfa460b37dece63adb67ff2e4dd1
When non-protected case, error at gcry_pk_testkey results
GPG_ERR_BAD_PASSPHRASE. Here, it should be converted to
GPG_ERR_BAD_SECKEY, or else, user will be asked passwd even if it's
not protected.
Signed-off-by: NIIBE Yutaka <gniibe@fsij.org>
* dirmngr/ks-engine-ldap.c (my_ldap_connect): Remove the
password_param thing because we set the password directly without an
intermediate var.
--
Reported-by: Ingo Kloecker
* sm/gpgsm.c (main): Use gpgrt_fcancel on decryption error if gpgrt
supports this.
* sm/decrypt.c (decrypt_gcm_filter): New.
(gpgsm_decrypt): Use this filter if requested. Check authtag.
* common/compliance.c (gnupg_cipher_is_allowed): Allow GCM for gpgsm
in consumer (decrypt) de-vs mode.
--
Backported-from-master: 4980fb3c6dde8c1dda975e8a36d6086c8456a631
We allow GCM in de-vs mode for decryption although this has not been
evaluation. It is decryption and thus no serious harm may happen.
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/server.c (make_keyserver_item): Fix default port for ldaps.
Move a tmpstr out of the blocks.
* dirmngr/ks-engine-ldap.c (my_ldap_connect): Improve diagnostics.
--
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/ks-engine-ldap.c (my_ldap_connect): Use LDAP_OPT_TIMEOUT.
* dirmngr/dirmngr.c (main): Move --ldaptimeout setting to ...
(parse_rereadable_options): here.
--
Note that this has not yet been tested. In fact a test with OpenLDAP
using a modified route got stuck in the connection attempt. Maybe it
works on Windows - will be tested later.
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/dirmngr.c (opts): Add option --ldapserver.
(ldapserver_list_needs_reset): New var.
(parse_rereadable_options): Implement option.
(main): Ignore dirmngr_ldapservers.conf if no --ldapserver is used.
* dirmngr/server.c (cmd_ldapserver): Add option --clear and list
configured servers if none are given.
--
This option allows to specify LDAP keyserver in dirmngr instead of
using gpgsm.conf.
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/server.c (cmd_ldapserver): Strip an optional prefix.
(make_keyserver_item): Handle non-URL ldap specs.
* dirmngr/dirmngr.h (struct ldap_server_s): Add fields starttls,
ldap_over_tls, and ntds.
* dirmngr/ldapserver.c (ldapserver_parse_one): Add for an empty host
string. Improve error messages for the non-file case. Support flags.
* dirmngr/ks-action.c (ks_action_help): Handle non-URL ldap specs.
(ks_action_search, ks_action_get, ks_action_put): Ditto.
* dirmngr/ks-engine-ldap.c: Include ldapserver.h.
(ks_ldap_help): Handle non-URL ldap specs.
(my_ldap_connect): Add args r_host and r_use_tls. Rewrite to support
URLs and non-URL specified keyservers.
(ks_ldap_get): Adjust for changes in my_ldap_connect.
(ks_ldap_search): Ditto.
(ks_ldap_put): Ditto.
--
The idea here is to unify our use of URLS or colon delimited ldap
keyserver specification. The requirement for percent escaping, for
example the bindname in an URLs, is cumbersome and prone to errors.
This we allow our classic colon delimited format as an alternative.
That format makes it also easy to specify flags to tell dirmngr
whether to use starttls or ldap-over-tls. The code is nearly 100%
compatible to existing specification. There is one ambiguity if the
hostname for CRL/X509 searches is just "ldap"; this can be solved by
prefixing it with "ldap:" (already implemented in gpgsm).
GnuPG-bug-id: 5405, 5452
* common/keyserver.h: Remove.
* sm/gpgsm.h (struct keyserver_spec): Remove.
(opt): Change keyserver to a strlist_t.
* sm/gpgsm.c (keyserver_list_free): Remove.
(parse_keyserver_line): Remove.
(main): Store keyserver in an strlist.
* sm/call-dirmngr.c (prepare_dirmngr): Adjust for the strlist. Avoid
an ambiguity in dirmngr by adding a prefix if needed.
* g10/options.h (struct keyserver_spec): Move definition from
keyserver.h to here. Remove most fields.
* g10/keyserver.c (free_keyserver_spec): Adjust.
(cmp_keyserver_spec): Adjust.
(parse_keyserver_uri): Simplify.
(keyidlist): Remove fakev3 arg which does not make any sense because
we don't even support v3 keys.
--
We now rely on the dirmngr to parse the keyserver specs. Thus a bad
specification will not be caught immediately. However, even before
that dirmngr had stricter tests.
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/http.h (HTTP_PARSE_NO_SCHEME_CHECK): New.
* dirmngr/http.c (http_parse_uri): Use this flag. Change all callers
to use the new macro for better readability.
(do_parse_uri): Add pseudo scheme "opaque".
(uri_query_value): New.
--
This scheme can be used to convey arbitrary strings in a parsed_uri_t
object.
Signed-off-by: Werner Koch <wk@gnupg.org>
* scd/app-p15.c (send_certinfo): free labelbuf
(do_sign): goto leave instead of return
* scd/command.c (cmd_genkey): goto leave instead of return
--
Signed-off-by: Jakub Jelen <jjelen@redhat.com>
GnuPG-bug-id: 5393
Modifified for this backport:
* scd/command.c (cmd_genkey): Make it easier to read by replacing
keyno with orig_line.
Signed-off-by: Werner Koch <wk@gnupg.org>
* dirmngr/ks-engine-ldap.c (extract_keys): Return the fingerprint if
available.
(ks_ldap_search): Ditto.
(extract_keys): Make sure to free the ldap values also in corner
cases.
(my_ldap_value_free): New.
(ks_ldap_get): Ditto.
(ks_ldap_search): Ditto.
(my_ldap_connect): Ditto.
--
For background see these comments from gpgme:
/* The output for external keylistings in GnuPG is different from all
the other key listings. We catch this here with a special
preprocessor that reformats the colon handler lines. */
/* The format is:
pub:<keyid>:<algo>:<keylen>:<creationdate>:<expirationdate>:<flags>
as defined in 5.2. Machine Readable Indexes of the OpenPGP
HTTP Keyserver Protocol (draft). Modern versions of the SKS
keyserver return the fingerprint instead of the keyid. We
detect this here and use the v4 fingerprint format to convert
it to a key id.
We want:
pub:o<flags>:<keylen>:<algo>:<keyid>:<creatdate>:<expdate>::::::::
*/
Regarding the freeing of values: I was not able to find a
specification stating it is okay to pass NULL to ldap_value_free, thus
the new wrapper. Also add robustness measures in case ldap_get_value
returns an empty array.
GnuPG-bug-id: 5441
Signed-off-by: Werner Koch <wk@gnupg.org>
