Why are we testing an intermediate result in ahash?

[Gary R Hook]

Commit 466d7b9f6 (cryptodev-2.6) added code to testmgr to populate, for 
async hash operations, the result buffer with a known value and to test 
the buffer against that value at intermediate steps. If the result 
buffer changes the operation is failed.

My question is: why?
What problem does this solve? Has this requirement existed all along, or 
is it new?

I'm now seeing complaints for AES/CMAC and SHA in my driver. I have no 
problem updating the driver, of course, but I'd like to better 
understand the precipitating issue for the commit.

Mar  2 12:30:56 sosxen2 kernel: [   60.919198] alg: No test for cfb(aes) 
(cfb-aes-ccp)
Mar  2 12:30:56 sosxen2 kernel: [   60.924787] 367: alg: hash: update 
failed on test 3 for cmac-aes-ccp: used req->result
Mar  2 12:30:56 sosxen2 kernel: [   60.946571] 367: alg: hash: update 
failed on test 4 for sha1-ccp: used req->result
Mar  2 12:30:56 sosxen2 kernel: [   60.956461] 367: alg: hash: update 
failed on test 1 for hmac-sha1-ccp: used req->result
Mar  2 12:30:56 sosxen2 kernel: [   60.966117] 367: alg: hash: update 
failed on test 4 for sha224-ccp: used req->result
...

Thanks,
Gary