From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wi0-x22b.google.com (mail-wi0-x22b.google.com [IPv6:2a00:1450:400c:c05::22b]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by mail.server123.net (Postfix) with ESMTPS for ; Sun, 22 Feb 2015 21:20:14 +0100 (CET) Received: by mail-wi0-f171.google.com with SMTP id hi2so12823577wib.4 for ; Sun, 22 Feb 2015 12:20:13 -0800 (PST) Message-ID: <54EA39FA.7010408@gmail.com> Date: Sun, 22 Feb 2015 21:20:10 +0100 From: Milan Broz MIME-Version: 1.0 References: <2E68FEAA-A69B-4E25-84F9-F30B2E7DFB69@gmail.com> <54E7B46D.8060805@tu-ilmenau.de> <54E82B80.8000607@gmail.com> In-Reply-To: Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit Subject: Re: [dm-crypt] cryptsetup problem on Raspberry Pi 2 w 512bit key-size (works on Raspberry Pi 1, x86_64, 256bit) List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Johannes Ernst Cc: dm-crypt@saout.de, Lars Winterfeld On 02/22/2015 08:40 PM, Johannes Ernst wrote: >> (If us use other hash it works even on Pi? Try sha1 and sha256 at least.) > > It appears independent of the hash involved: I tried sha1, sha256 in addition to the original sha256. The behavior is the same: > 1. cryptsetup open … does not open > 2. cryptsetup open —debug opens on the second attempt to put the password in. ok, this is really strange. One (random) guess from the log: > # Using userspace crypto wrapper to access keyslot area. it means that code is using kernel userspace crypto (and cryptsetup already revealed at least two problems there...) Could you try it without it? (Code should fallback to old dmcrypt temporary devices mode). You can do it either by - disabling/blacklisting kernel modules which provides it: af_alg.ko and algif_skcipher.ko (or disable CRYPTO_USER_API, CRYPTO_USER_API_SKCIPHER when compiling kernel) - try to run older cryptsetup (at least 1.6.4 or older) I am afraid I cannot help more here without reproducing it... (Is it RPI2 only issue or anyone see it on other ARM device?) Milan