1683701 - View Security Info shows incorrect information when Require Encryption is selected for S/MIME encryption - until draft message is saved

Status: RESOLVED FIXED

(Thunderbird :: Security, defect)

Description

[Deric Sullivan]

User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:84.0) Gecko/20100101 Firefox/84.0

Steps to reproduce:

In Thunderbird, when writing an email, from the Security menu I select "Require Encryption", then from the same menu I select "View Security Info". The information shown seems incorrect.
If I save the email as draft and then "View Security Info" again, then the information seems correct. Or I can wait a while (presumably until a draft is automatically saved) and the information also becomes correct.

Note that if I send the email while the info seems incorrect, the message is still encrypted, so there does not appear to be a major security concern with this issue.

My setup: Thunderbird version 78.6.0 (64-bit) running on Linux (Ubuntu 20.04) and pulling certificates out of an LDAP server.

Actual results:

View Security Info will say:
"The contents of your message will be sent as follows:
Digitally signed: No
Encrypted: No"

Expected results:

View Security Info should say:
"The contents of your message will be sent as follows:
Digitally signed: Yes
Encrypted: Yes"

Comment 1

[Wayne Mery (:wsmwk)]

Do you also see this?

Comment 2

[Deric Sullivan]

Note that my version of Thunderbird has been upgraded from 78.6.0 to now 78.7.0 and the issue is still present.

Comment 3

[Onno Ekker [:nONoNonO UTC+1]]

I can confirm this for S/MIME encryption: when I select Require Encryption from the Options menu and S/MIME from the Encryption Technology submenu under Options and immediately choose Message Security Info from the View menu, Digitally Signed and Encrypted is still shown as No, although the status of the certificates of the recipients is already shown correctly.

For OpenPGP messages, only the status of the keys of the recipients is shown, to the security info for such messages is shown correctly.

Comment 4

[Magnus Melin [:mkmelin]]

Seems ok on trunk, even if the UI is a bit confusing. Unless all the recipients' keys are known and listed as valid, it will show the no's. If I try to send to myself (for which I of course have the key), it does show Yes, Yes.

Comment 5

[Deric Sullivan]

For me, when sending to a single recipient, the View Security Info shows No, No even if the recipient certificate status shows as Valid.

Comment 6

[Onno Ekker [:nONoNonO UTC+1]]

For me it also shows No No, until the draft is saved...

Comment 7

[Onno Ekker [:nONoNonO UTC+1]]

Attachment: Before saving draft.png

Comment 8

[Deric Sullivan]

Note that the issue can be seen in the reverse situation as well. I can prepare an email for sending with encryption (S/MIME), save as draft so the View Security Info shows Yes, Yes, then de-select Encryption (choose "Do Not Encrypt" from the Security pull down menu), and View Security Info again. This time Yes, Yes will be shown when I would expect it to be No, No. If I again save as draft, the information will be OK if I select View Security Info again.

Comment 9

[Onno Ekker [:nONoNonO UTC+1]]

(In reply to Magnus Melin [:mkmelin] from comment #4)

Seems ok on trunk, even if the UI is a bit confusing. Unless all the recipients' keys are known and listed as valid, it will show the no's. If I try to send to myself (for which I of course have the key), it does show Yes, Yes.

I checked in latest Daily, but I still see this problem. Don't know how you checked? Maybe the draft was already saved.
Furthermore, if you close the draft message after saving and reopen it, again the wrong info is shown, until the next (auto) save of the message.

Comment 10

[Magnus Melin [:mkmelin]]

Yes, I think what happened is the auto-draft got saved. I've found the problem and have a patch I'm testing.
It looks like this is a very old problem.

Comment 11

[Magnus Melin [:mkmelin]]

Attachment: bug1683701_smime_view_secinfo.patch

https://treeherder.mozilla.org/#/jobs?repo=try-comm-central&revision=8dc8190333beac7138f164804e74023ae8a408ba (just started)

Getting rid of the global gSMFields. The changes in enigmail code are not changing anything, just updating the references (dead/uncommented code atm)

Comment 12

[Alessandro Castellani [:aleca]]

Comment on attachment 9200965 [details] [diff] [review]
bug1683701_smime_view_secinfo.patch

Review of attachment 9200965 [details] [diff] [review]:
-----------------------------------------------------------------

Indeed, this fixes the problem.

Comment 13

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/comm-central/rev/8f36cf1cf596
make viewing S/MIME security info during compose show the right values. r=aleca

Comment 14

[Magnus Melin [:mkmelin]]

Comment on attachment 9200965 [details] [diff] [review]
bug1683701_smime_view_secinfo.patch

[Approval Request Comment]
Regression caused by (bug #): not a regression, or then it's a very old regression
User impact if declined: during composition View Message Security shows wrong details until the message is (auto) saved as a draft.
Testing completed (on c-c, etc.): on c-c
Risk to taking this patch (and alternatives if risky): not very risky, it's not a very commonly used UI which is why it was unnoticed for ages

Comment 15

[Wayne Mery (:wsmwk)]

Comment on attachment 9200965 [details] [diff] [review]
bug1683701_smime_view_secinfo.patch

[Triage Comment]
Approved for beta

Comment 16

[Rob Lemley [:rjl]]

Thunderbird 86.0b2:
https://hg.mozilla.org/releases/comm-beta/rev/eb14ce2382c1

Comment 17

[Wayne Mery (:wsmwk)]

Comment on attachment 9200965 [details] [diff] [review]
bug1683701_smime_view_secinfo.patch

[Triage Comment]
Approved for esr78

Comment 18

[Rob Lemley [:rjl]]

Thunderbird 78.8.0:
https://hg.mozilla.org/releases/comm-esr78/rev/59fe07501063

Comment 19

[Deric Sullivan]

My version of Thunderbird is now updated to 78.8.0. I no longer see the original problem that I reported. So as far as I can tell the issue is resolved for me. Thank you for your help.

Comment 20

[Kai Engert [:KaiE:]]

This caused regression bug 1697252.

Comment 21

[Kai Engert [:KaiE:]]

In my testing, the change in function showMessageComposeSecurityStatus() is sufficient to fix this issue.