Discussion:
[CRYPTO] add alg. type to /proc/crypto output
Erlend Aasland
2003-09-05 14:38:59 UTC
Permalink
Hi

Here is a patch that add alg. type output to /proc/crypto. Booted and
tested.


Erlend Aasland

diff -urN linux-2.6.0-test4-bk7/crypto/proc.c linux-2.6.0-test4-bk7-dirty/crypto/proc.c
--- linux-2.6.0-test4-bk7/crypto/proc.c 2003-08-23 01:53:43.000000000 +0200
+++ linux-2.6.0-test4-bk7-dirty/crypto/proc.c 2003-09-05 18:17:35.000000000 +0200
@@ -57,6 +57,7 @@

switch (alg->cra_flags & CRYPTO_ALG_TYPE_MASK) {
case CRYPTO_ALG_TYPE_CIPHER:
+ seq_printf(m, "type : cipher\n");
seq_printf(m, "blocksize : %u\n", alg->cra_blocksize);
seq_printf(m, "min keysize : %u\n",
alg->cra_cipher.cia_min_keysize);
@@ -65,10 +66,17 @@
break;

case CRYPTO_ALG_TYPE_DIGEST:
+ seq_printf(m, "type : digest\n");
seq_printf(m, "blocksize : %u\n", alg->cra_blocksize);
seq_printf(m, "digestsize : %u\n",
alg->cra_digest.dia_digestsize);
break;
+ case CRYPTO_ALG_TYPE_COMPRESS:
+ seq_printf(m, "type : compression\n");
+ break;
+ default:
+ seq_printf(m, "type : unknown\n");
+ break;
}

seq_putc(m, '\n');
David S. Miller
2003-09-05 14:34:03 UTC
Permalink
On Fri, 5 Sep 2003 16:38:59 +0200
Post by Erlend Aasland
Here is a patch that add alg. type output to /proc/crypto. Booted and
tested.
Is it even useful?

When you see names like "md5" and parameters such as "digestsize"
listed, do you really have no clue that it is a "digest"? :-)

The information seems completely redundant to me.
Erlend Aasland
2003-09-06 10:08:18 UTC
Permalink
Post by David S. Miller
Is it even useful?
Check http://samba.org/~jamesm/crypto/TODO
It's listed under medium priority, so I thought I was doing a good
thing implementing it.
Post by David S. Miller
When you see names like "md5" and parameters such as "digestsize"
listed, do you really have no clue that it is a "digest"? :-)
Good point.
James Morris
2003-09-06 16:43:31 UTC
Permalink
Post by Erlend Aasland
Post by David S. Miller
Is it even useful?
Check http://samba.org/~jamesm/crypto/TODO
It's listed under medium priority, so I thought I was doing a good
thing implementing it.
I think someone asked for it once, not sure exactly why, but I guess it
makes parsing simpler and theoretically less likely to break if e.g. a
field is added or removed for an algorithm type.


- James
--
James Morris
<***@intercode.com.au>
V***@vt.edu
2003-09-07 00:21:36 UTC
Permalink
Post by Erlend Aasland
Post by David S. Miller
When you see names like "md5" and parameters such as "digestsize"
listed, do you really have no clue that it is a "digest"? :-)
Good point.
On the flip side, it makes it possible for an *AUTOMATED* process to search for
appropriate types of entries. So for instance, a GUI can say "The following
3 digest functions are available, please choose one" or "The following 4 symmetric
encryptions are available".....
David S. Miller
2003-09-11 23:35:25 UTC
Permalink
On Fri, 5 Sep 2003 16:38:59 +0200
Post by Erlend Aasland
Here is a patch that add alg. type output to /proc/crypto. Booted and
tested.
I've applied this, thank you.

Loading...