Bug 1028362

Summary: cryptsetup: malformed hash algorithm specifiers lead to zero-entropy keys
Product: Red Hat Enterprise Linux 7 Reporter: Florian Weimer <fweimer>
Component: cryptsetupAssignee: Ondrej Kozina <okozina>
Status: CLOSED CURRENTRELEASE QA Contact: Release Test Team <release-test-team-automation>
Severity: low Docs Contact:
Priority: medium    
Version: 7.0CC: agk, gmazyland, okozina, pholica, pkotvan, prajnoha
Target Milestone: rcKeywords: Triaged
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: cryptsetup-1.6.3-1.el7 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2014-06-13 09:19:40 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1028357    

Description Florian Weimer 2013-11-08 09:52:15 UTC
If the user specifies --type=plain --hash=sha1:secure (or something similar, the colon is important) during open, cryptsetup will use an all-zero key because it truncates the key-derived hash to length 0.  This is caused by code in crypt_plain_hash() which implements a ":NNN" specifier for reducing key strength:

	/* hash[:hash_length] */
	if ((s = strchr(hash_name_buf, ':'))) {
		*s = '\0';
		hash_size = atoi(++s);
		if (hash_size > key_size) {
			log_dbg("Hash length %zd > key length %zd",
				hash_size, key_size);
			return -EINVAL;
		}
		pad_size = key_size - hash_size;
	} else {
		hash_size = key_size;
		pad_size = 0;
	}
…
	if (r == 0 && pad_size)
		memset(key + hash_size, 0, pad_size);


This code should have more error checking.

Comment 2 Milan Broz 2013-11-08 13:52:22 UTC
The colon will always reduce key, it should be used only in situation when you are dealing with seriously broken cryptosystem (which set some parts of key to zero).

But yes, error checking should be better here. I need to implement it even for --hash=plain (==no hash) because some implementation of cryptoloop/losetup wipes the last byte of key... (and cryptsetup should be able to replace such system).

So this will be fixed upstream and RH should track the patch for RHEL as well, thanks.

Comment 3 Milan Broz 2013-11-10 18:10:18 UTC
Should be fixed upstream by this patch
http://code.google.com/p/cryptsetup/source/detail?r=db561257084fd09f117d589fe9f42980a9ea3ad9#

Thanks for reporting this problem!

Comment 5 Ludek Smid 2014-06-13 09:19:40 UTC
This request was resolved in Red Hat Enterprise Linux 7.0.

Contact your manager or support representative in case you have further questions about the request.