Hello Robert, I run again the openssl command with version 1.1.1 and now the 
output is:

Private-Key: (2048 bit, 2 primes)
modulus:
publicExponent: 65537 (0x10001)
privateExponent:
prime1:
prime2:
exponent1:
exponent2:
coefficient:

The previous output was with openssl 1.0.2e
Kind regards,
[Unione della Romagna Faentina]
[cid:0707f32f-186c-48bf-a2f3-ecd8c973753a]      dr. Alessandro Baldoni
[cid:670760e4-95b0-4cc8-aeb9-e9226ada149e]      Servizio Informatica
Via Severoli 7
48018 Faenza RA
[cid:23f236d9-3050-45e9-9e56-17a3afcecee3]      0546 691224
[cid:7d875fa5-071d-4e39-9265-1023abd18e9a]      
[email protected]
[cid:3d9b4127-2678-4d02-9faa-11baf517420a]      [email protected]
________________________________
From: Robert Segall <[email protected]>
Sent: Monday, October 19, 2020 14:48
To: Alessandro Baldoni <[email protected]>; [email protected] 
<[email protected]>
Subject: Re: [pound] Pound-3.0e: Error when reading PEM file

Hallo Alessandro

By "wrong values" I meant primes that do not result in the advertised
number of bits (for example). This is very unlikely, but not outright
impossible.

What worries me more is that in your printout I see "Private-Key: (2048
bit)" rather than "RSA Private-Key: (2048 bit, 2 primes)". I believe
mbedssl (like other TLS1.3 implementations) is rather picky about the
tags used. Could you possibly check with your certificate provider for
the reasons? Perhaps they could generate a new certificate with fully
compliant tags just for testing purposes? Alternately, I know the
latest versions of openssl generate these tags, so perhaps you could
create a self-signed certificate just for testing?

BTW: this could also explain the issues people had with Pound 2.8:
using a newer openssl version may have a similar effect.

On Mon, 2020-10-19 at 10:40 +0000, Alessandro Baldoni wrote:
> Hello Robert, this is the output of the SSL command (values removed):
>
> Private-Key: (2048 bit)
> modulus:
> publicExponent: 65537 (0x10001)
> privateExponent:
> prime1:
> prime2:
> exponent1:
> exponent2:
> coefficient:
>
> What do you mean with "a problem of wrong values"?
>
> Kind regards,
>
> [Unione della Romagna Faentina]
> [cid:7d8f8d83-a9e4-4bf0-84b3-9e1aeeb31a71]      dr. Alessandro
> Baldoni
> [cid:19f2ff9f-e848-4fe4-ac3c-65bd6301f0a7]      Servizio Informatica
> Via Severoli 7
> 48018 Faenza RA
> [cid:03df2d07-aef8-437c-8826-30d9d43e5250]      0546 691224
> [cid:61e85ff2-c4dd-4fb5-a25b-25e1039aa233]
> [email protected]
> [cid:447cce9f-3bab-4731-81a1-c49b0721e761]
> [email protected]
> ________________________________
> From: Robert Segall via pound <[email protected]>
> Sent: Monday, October 19, 2020 11:12
> To: [email protected] <[email protected]>
> Cc: Robert Segall <[email protected]>
> Subject: Re: [pound] Pound-3.0e: Error when reading PEM file
>
> Hallo Alessandro
>
> Please have a look at your private key and check what it contains. To
> see it use the command "openssl rsa -noout -text -in cert.pem". The
> expected output:
>
> RSA Private-Key: (... bit, 2 primes)
> modulus:
> ...
> publicExponent: ... (...)
> privateExponent:
> ...
> prime1:
> ...
> prime2:
> ...
> exponent1:
> ...
> exponent2:
> ...
> coefficient:
> ...
>
> If your key looks different it may cause issues, otherwise it may be
> a
> problem of wrong values.
--
Robert Segall
Apsis GmbH
Postfach, Uetikon am See, CH-8707
Tel: +41-32-512 30 19

-- 
pound mailing list
[email protected]
https://admin.hostpoint.ch/mailman/listinfo/pound_apsis.ch

Reply via email to