Team yaSSL has recently been considering the addition of the Camellia block cipher to the CyaSSL embedded SSL library, and we would like to request your feedback on this addition.  Do you think it would be a beneficial addition?  If it was added to CyaSSL, would you be interested in using it?  Because we strive to shape our SSL library around what our users want and need, we greatly value your feedback.

If you are unfamiliar with the Camellia block cipher, you can find a Wikipedia article on the cipher here: http://en.wikipedia.org/wiki/Camellia_(cipher), or the Camellia cipher’s English homepage, here: http://info.isl.ntt.co.jp/crypt/eng/camellia/index.html.

Please send any feedback to info@yassl.com or reply to this thread, below.

Thanks!
Team yaSSL

If I understand correctly, there isn't a way to pass a variable to the callback set by CyaSSL_SetIOSend / CyaSSL_SetIOReceive.

If possible, could you please add an extra parameter to the CyaSSL_SetIOSend / CyaSSL_SetIOReceive functions that will be passed to the callbacks?
It'll be easier to pass contexts, classes, etc...

Thanks
Nitay

that way, it appears that the overall memory consumption is much lower, but more importantly, returns back to base level after done transmitting traffic (a large recording, so i can see actual memory shifts...).

this turns my suspicion towards SSL_new and SSL_free - is it possible that XMALLOC and XFREE aren't performing "traditional" malloc/free operations, but some sort of pooling, I was trying to understand from the code but failed to reach certainty.

any help will be greatly appreciated.

thanks,
Dan

I have tcp server which uses multiple threads for reading through an IO Completion Port. I want to add SSL to this server.

What, in your opinion, is the best way to do so?
The writes are less of a problem, but how do I decrypt a SSL packet without calling read?

For now, I've set the write and read callbacks using CyaSSL_SetIORecv / CyaSSL_SetIOSend.
I didn't implement the reads yet.
Also, the accept could be done to a new client before attaching it to the IOCP, so its not really a non-blocking IO...

What do you think?

EDIT: The reads are done through IOCP, so the threads are not blocked on recv() or select(), but on GetQueuedCompletionStatus()

Thanks
Nitay

I'm trying to get to the download page, but it keeps redirecting me back to the information page (and it resets the information fields!)

How do i get a proper download link to cyassl?

Thanks
Nitay

I am cross compiling CyaSSL and getting the following error while creating libcyassl.la

crti.o: No such file: No such file or directory

I am unable to figure out the issue because it does not show me the exact link command. Can someone please tell me how I could do a verbose build so that I can debug such issues?

If someone has any clues about the issue, suggestions are most welcome

Thanks
Rahul

I am testing CyaSSL for embedded devices. Could you please advice the best configuration settings for runtime performance (i.e. speed)? We are running CyaSSL as pure software without any specific hardware support for crypto operations.

Especially for RSA there are a lot of configuration options to use different algorithms. In mpi_class.h I saw various configurations.

Thanks,
Yun

I have used the following switchs

NO_MD4,NO_RABBIT,NO_SESSION_CACHE,NO_CYASSL_SERVER,NO_ERROR_STRINGS,NO_PSK,NO_PWDBASED,NO_HC128,NO_DSA,OPENSSL_EXTRA,NO_64BIT,NO_SHA-384

These are the only imports i made
   
    SSL_CTX_new
    TLSv1_1_client_method
    SSL_CTX_set_verify
    SSL_CTX_set_cipher_list
    SSL_new(ctx);
    SSL_set_fd(ssl, isocket) ;
    SSL_connect
    SSL_get_peer_certificate
    X509_NAME_oneline
    X509_get_issuer_name
    OPENSSL_free
    OPENSSL_free
    X509_free
    SSL_free
    SSL_write
    SSL_read
    SSL_library_init

I am compiling with visual studio (6 or 2008)
The fact is, i have only 50KB to spare due to the nature of the program I am writting.
Reading from the site, i could get footprint as little as 30KB, so what is wrong with my compilation?

Thanks

I had a look at the source code of CyaSSL and I've found out that the compression method used by CyaSSL has number 221. Why 221 instead of 1?

Cheers,
Giancarlo

I am evaluating CyaSSL for our product. Can someone please tell me if CyaSSL is a FIPS compliant library? Same question on  wolfCrypt and YaSSL webserver. Are they FIPS compliant?

Thanks
Rahul

1. is it possible that there is a memory leak in SSL_free / SSL_new?
2. if not, are you using memory pooling with the SSL* objects? clearly I can look in the code, but I want to be sure - I ask because that could account for what I am seeing

thanks in advance,
Dan

The below patch (on cyassl-2.6.0) appears to solve the problem. Looking at the code, building without LARGE_STATIC_BUFFERS (which is what I do) makes things worse.

Index: cyassl/src/internal.c
===================================================================
--- cyassl.orig/src/internal.c    2013-05-08 15:22:15.704663345 +0200
+++ cyassl/src/internal.c    2013-05-08 15:24:14.716658339 +0200
@@ -1791,6 +1791,7 @@
 
 int DtlsPoolSend(CYASSL* ssl)
 {
+    int ret;
     DtlsPool *pool = ssl->dtls_pool;
 
     if (pool != NULL && pool->used > 0) {
@@ -1807,6 +1808,9 @@
             c16toa(ssl->keys.dtls_epoch, dtls->epoch);
             c32to48(ssl->keys.dtls_sequence_number++, dtls->sequence_number);
 
+            if ((ret = CheckAvalaibleSize(ssl, buf->length)) != 0)
+                return ret;
+
             XMEMCPY(ssl->buffers.outputBuffer.buffer, buf->buffer, buf->length);
             ssl->buffers.outputBuffer.idx = 0;
             ssl->buffers.outputBuffer.length = buf->length;