I see. Thus the problem is that IPWorksOpenPGP does not create proper OpenPGP private keys. I guess they use OpenSSL with their different CRT parameter style and do not convert them correctly. RFC-4880 says this in 5.5.3:

The secret key is this series of multiprecision integers:
o  MPI of RSA secret exponent d;
o  MPI of RSA secret prime value p;
o  MPI of RSA secret prime value q (p < q);
o  MPI of u, the multiplicative inverse of p, mod q.

At the time the verification is done some output has already been written to the file 'signed'. When checking whether the deprecated abbreviated format

The digest algorithm used is computed based on the preferences in the key if encryption is also used. Thus this should always work and any decent key has sha256 in its preferences. In case sha1 has a higher precedence, as seen on old keys, --personal-digest-preferences can be used to prefer sha256. However, it is way better to fix the key. The easisies way to do that is to change the expiration date - then the new standard preferences will be used.

Thanks so much your helps.
With new version 3.1.6, I can generate key on Kleopatra tool and use key stored in smartcard.

There was indeed a problem. With a test card I could reproduce the issue and fix it.

I contacted Microsoft Security Response Center (MSRC) in regard to this matter. They confirmed the failed PGP key verification, but have not yet any explanation for that.

What are MS doing when they get it right, though? I'd look at the differences between those two to identify what they've messed up here.

Thanks. The mail is a standard, non-crypto mail with one attachment. That attachment is a TNEF file which has according to ytnef(1) just one file. That file has the name gpgolPGP.dat and contains a clearsigned message.

Sure, I zipped the eml which failed and I´ll send it by e-mail to you

Is it possible that you upload or send me a copy of such a mail (wk gnupg.org)? ZIP or tar the eml file and send it in an encrypted mail to me to make sure it won't be modified on the transport.

I checked my mails in detail, and I can confirm that the error occurs only with "Microsoft security update releases". Indeed "Microsoft security advisory notification" and "Microsoft security update summary for..." will be verified correctly.

I agree. It also happens to me. But only with mails coming from "Microsoft security update releases". Mails coming form "Microsoft security advisory notification" and Microsoft security update summary for..." are ok and are signed by the same key. It could be some trouble in MS automated email treatment.

Fair enough. Let's wait and see what others think.

Also consider that it is possible to change the key usage flags. Thus it will never be clear whether one has a fixed or unfixed public key. I'd like to close this bug because it is currently also discussed in the IETF WG.

No info received.

There is another argument for respecting the usage flags: it trims the admissible key space, if key ID in the PKESK packet is zero ('wild card') and thus all private keys have to be considered for decryption.

I disagree, and you don't have to try to convince me, the decision is with werner. I just want to give my opinion:
Bug compatibility is nothing esoteric or bad especially for a general purpose backend tool like gnupg. Being open to accepting broken input is a good thing because it will mean that we can get people out of a "broken tool vendor lock in".

i agree with @Valodim that it would be better to not have a warning at all for an attempt to decrypt from secret key whose public key has never been marked as valid for encryption. A strict failure there (as with a strict failure for lack of mdc) is a better scenario than a warning. If the user controls the secret key and they decide they want to be able to decrypt with it, they should be able to mark it as decryption-capable (if that's really what they want) and retry. But this is an action only for experts.

The same *cannot* be said for a subkey that is marked specifically for certification or signing, and not for decryption.

I understand the real world requirement for decrypting messages that have been encrypted to a revoked or expired key.

I don't see a problem. If you have the private key you can and will use it. I guess your concern is an oracle?

Dear aheinecke,

Hi Adam,

To compute the key validity (trust) more information may be needed and we can only do that after the changes have been saved. Further, no-auto-chec-trustdb will anyway delay that computation until "gpg --check-trustdb" is run (e.g. by a cron job).

success, thank you for the help!

In GnuPG 2.1, secret keys are under control of gpg-agent. Currently, it is not deleted by gpg frontend.
Please run:

$ gpg -K --with-keygrip

IIRC, rfc2440 did not forbid partial length encoding for key-material so gpg could use that. rfc4880 limits partial length encoding to non-key-material which causes this error message.

Well, I'd expect gpg not to alter my digest/compression preferences when changing my cipher preferences and vice versa. So if a user's going to have to lose his previously set preferences for a key in this manner because that's the only reasonably viable way of maintaining backwards compatibility, I think it would be appropriate to let him know beforehand and also suggest that he set it all up at once (as I've so described above) so that nothing is lost in the process.

The way the setpref command works is implementation specific and thus the OpenPGP standard is irrelevant here
.
Are you requesting a change in the behaviour of the setpref command? That would not be easy to implement for backward compatibility.

Well, iff we implement that for gpg we also need to implement it for gpgsm.

A decision must be made what the desired behaviour should be.

• marcus (Marcus Brinkmann) <noreply@dev.gnupg.org> [20170622 16:41]:

So, the default change 7y ago and the world didn't end. Closing this.

So, the default change 7y ago and the world didn't end. Closing this.

Tested this again with 2.1.18 and it works now as expected. Export secret key
just exports a key if it has no passphrase. So I think this issue can be marked
as resolved.

It is a hack in OpenKeychain to allow the use of several devices. Frankly, I am
not sure whether this is really a good idea: The security is limited by the key
for the least secure device.

So i've tested this locally with:

    export GNUPGHOME=$(mktemp -d)
    gpg --quick-gen-key 'test user <test@example.org>'
    gpg --armor --export-secret-key 'test user <test@example.org>'

(choosing no passphrase during the prompts that come up during the quick-gen-key
step). The final export step works fine.

Can you show what steps you're taking that fail for you, Andre?

I'm using latest master and I still can't export a secret key without passphrase.

And Justus also has not closed this bug or wrote that he commited something
more. So I think the 2.1.13 announcement was mistaken and this problem still
exists. (Or am I missing some option / need a different pinentry mode?)

We got it for 2.1: -f or --recipient-file

Hi,
the 2.1.13 announcement has
"""

• gpg: Allow export of non-passphrase protected secret keys.

"""
(from https://lists.gnupg.org/pipermail/gnupg-announce/2016q2/000390.html)
so this defect may be fixed with 2.1.13 I guess, cool!
Probably only need a test to confirm?

Replacing revoked keys made me wonder if we actually need an auto-refresh key.
If we try to return one valid key with --locate-keys wouldn't it make more sense
semantically if we use the auto-key-locate mechanisms with locate-keys when a
key is expired in the local store?

This would also work better for revoked keys where a Parcimonie style auto
refresh would pick up the revocation and locate-keys would then look for a new key.

I can confirm one defect with 2.1.11:
The ability to export a secret key without passphrase available in gnupg2.0
is gone. My use case is to write a testcase that automatically imports the key.

I did not work on this other than what I merged. What I did is to enhance our
fake pinentry program to allow it to supply different passphrases, make it write
a log so that we can quantify the pinentry interaction in test cases, and to add
an export test documenting the status quo.

The question at hand is whether dkg's patch or Justus work is the way to go. I
have not yet reviewed dkg's patch, though.

dgk: You are right that Pinentry may be used even with --batch. In fact gpgme
uses --batch and a Pinentry is used nevertheless.

Right, there are no technical means right now to inhibit the export of private
keys. However, it would be easy to add this by not allowing gpg-agent to tell
the client the key used to encrypt the import/export command of keys.

A user migyt have used no passphrase for a key in the believe that an
unprotected key can't be exported.

I'm not convinced that this policy is effectively implemented in gpg-agent.

The patch series that starts here:

https://lists.gnupg.org/pipermail/gnupg-devel/2016-May/031121.html

resolves the export of secret key material stored as cleartext, and it does so
without modifying gpg-agent at all.

fwiw, I do not agree with T2324 (justus on Apr 18 2016, 05:22 PM / Roundup) that gpg --batch should not use pinentry at
all -- i think it's quite useful to be able to combine --batch with pinentry,
where the key is stored protected, or is otherwise marked by gpg-agent for
limited use.

re: T2324 (justus on Apr 18 2016, 05:22 PM / Roundup)

2. gpg --export-secret-key should export unprotected keys that are stored w/o a passphrase"

That would violate the policy we implement in gpg-agent. The
gpg-agent is responsible for private keys and a client may not use a
private key without the agent's consent. If we would allow that by
default there won't be any protection at all and keys can be easily
exported and used. A required confirmation via the Pinentry would
solve the practical problem. However, there is the question what to
do on unattended systems - the only way it can be done right now is
configuring gpg-agent to use a custom pinentry, or by extending the
loopback mode.

Werner: Yes please.

I have some stashed work to fix this but it is not ready - let me know if you
want to work on it.

*See also T2070

See also issue20170

I understand the reason for re-encrypting -- i'm quite happy that the agent is
sensible about improving the security of the key when it adopts it.

my concern is that users don't know what to expect, and that different workflows
result in different sets of keys stored in the agent.

So i'd recommend that when importing without --batch, if the password fails for
any reason, gpg should fall back to the fast migration "kludge" rather than just
skipping that keyblock. That way the imported secret key material will still be
available and can be cleaned up/hardened on first successful use.

Thank you for your report and the log, but it doesn't have useful information so
that I can debug.

The information of card reader is required, if the problem happens for specific
card reader only. Please include full log which includes card reader information.

There seems to be a problem with your reader. We would need to closer analyze
the log (which I copy below):

DBG: send apdu: c=00 i=A4 p1=00 p2=0C lc=2 le=-1 em=0
DBG: ccid-driver: PC_to_RDR_IccPowerOn:
DBG: ccid-driver: dwLength ..........: 0
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 145
DBG: ccid-driver: bPowerSelect ......: 0x01 (5.0 V)
DBG: ccid-driver: [0008] 00 00
DBG: ccid-driver: RDR_to_PC_DataBlock:
DBG: ccid-driver: dwLength ..........: 21
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 145
DBG: ccid-driver: bStatus ...........: 0
DBG: ccid-driver: [0010] 3B DA 18 FF 81 B1
DBG: ccid-driver: [0016] FE 75 1F 03 00 31 C5 73 C0 01 40 00 90 00 0C
DBG: ccid-driver: PC_to_RDR_XfrBlock:
DBG: ccid-driver: dwLength ..........: 4
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 146
DBG: ccid-driver: bBWI ..............: 0x00
DBG: ccid-driver: wLevelParameter ...: 0x0000
DBG: ccid-driver: [0010] FF 11 18 F6
DBG: ccid-driver: RDR_to_PC_DataBlock:
DBG: ccid-driver: dwLength ..........: 4
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 146
DBG: ccid-driver: bStatus ...........: 0
DBG: ccid-driver: [0010] FF 11 18 F6
DBG: ccid-driver: PC_to_RDR_SetParameters:
DBG: ccid-driver: dwLength ..........: 7
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 147
DBG: ccid-driver: bProtocolNum ......: 0x01
DBG: ccid-driver: [0008] 00 00 18 10 FF 75 00 FE
DBG: ccid-driver: [0016] 10
DBG: ccid-driver: RDR_to_PC_Parameters:
DBG: ccid-driver: dwLength ..........: 7
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 147
DBG: ccid-driver: bStatus ...........: 0
DBG: ccid-driver: protocol ..........: T=1
DBG: ccid-driver: bmFindexDindex ....: 18
DBG: ccid-driver: bmTCCKST1 .........: 10
DBG: ccid-driver: bGuardTimeT1 ......: FF
DBG: ccid-driver: bmWaitingIntegersT1: 75
DBG: ccid-driver: bClockStop ........: 00
DBG: ccid-driver: bIFSC .............: 254
DBG: ccid-driver: bNadValue .........: 16
DBG: ccid-driver: PC_to_RDR_XfrBlock:
DBG: ccid-driver: dwLength ..........: 5
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 148
DBG: ccid-driver: bBWI ..............: 0x00
DBG: ccid-driver: wLevelParameter ...: 0x0000
DBG: ccid-driver: [0010] 10 C1 01 FE 2E
DBG: ccid-driver: RDR_to_PC_DataBlock:
DBG: ccid-driver: dwLength ..........: 4
DBG: ccid-driver: bSlot .............: 0
DBG: ccid-driver: bSeq ..............: 148
DBG: ccid-driver: bStatus ...........: 0
DBG: ccid-driver: [0010] 00 82 00 82
DBG: ccid-driver: invalid response for S-block (Change-IFSD)
apdu_send_simple(0) failed: unknown host status error
DBG: send apdu: c=00 i=A4 p1=04 p2=00 lc=6 le=-1 em=0

Without pcscd running, I get a "Not supported" error. The scd.log is attached.
Using pcscd, it works, except for that special case.

debug 2048
debug 1024

is what I need.

Thanks. We need to know some more detailed information. Please
put

debug 2018
debug 1024
log-file /somewhere/scd.log

into scdaemon.conf, kill scdaemon and try again. It seems you have not yet been
asked for a PIN so the log won't reveal the PIN. Anyway, you may want to send
the log to me by PM (wk@gnupg.org - key 1e42b367).

Fails with 2.0.29 too, compiled from source. With enabled debug-all verbose in
scdaemon.conf, the log ends with:

2016-03-19 10:12:09 scdaemon[1988] DBG: response: sw=6A88 datalen=0
2016-03-19 10:12:09 scdaemon[1988] operation decipher result: Missing item in object
2016-03-19 10:12:09 scdaemon[1988] app_decipher failed: Missing item in object
scdaemon[1988]: chan_7 -> ERR 100663364 Missing item in object <SCD>
scdaemon[1988]: chan_7 <- RESTART
scdaemon[1988]: chan_7 -> OK

We should create a test case for trust signatures before we start to fix it.

The current version is 2.0.29 - please try again using this version.

Marking as resolved since this is available in 2.1 and we are not going to
backport this to 1.4 or 2.0. Thanks.

I guess you are reporting for GnuPG 2.0 or 1.4.
We already implemented your suggestion in 2.1.