[openssl-dev] [openssl.org #3964] Fix OPENSSL_NO_STDIO build

[David Woodhouse]

On Thu, 2016-02-04 at 03:04 +0000, Rich Salz via RT wrote:
> So guys, sorry for dropping the ball. Where are we on this now?

Dropping rt@ since the OPENSSL_NO_STDIO build is actually solved now so
RT#3964 looks like it can be closed. I'm choosing to interpret your
question in the wider sense of the UEFI build.

I have submitted https://github.com/openssl/openssl/pull/631 which
matches the EDK2 tree at http://git.infradead.org/users/dwmw2/edk2.git

The OpenSSL bits look OK to me; it's mostly just a few fixes for some
of the recent cleanups, along with the patch that's been outstanding
for a while for RT#3628.

I am moderately unhappy with the last commit on the EDK2 side, though.

The problem is that the API that the EDK2 CryptoPkg offers to its
clients includes a function to get the size of the context. Perhaps we
should change that... Qin, is that feasible?

In the mean time, I have a horrid "#include <../hmac/hmac_lcl.h>' to
make it actually work.

Having already done that, I was desensitised enough that elsewhere in
the patch I include <../evp/evp_locl.h> just so that we can continue to
have an EVP_MD_CTX locally on the stack instead of having to use
dynamic allocation for it. Which I concede probably wants cleaning up.
But I don't *like* being forced to switch to dynamic allocation. That
just introduces failure modes that didn't exist before.


-- 
David Woodhouse                            Open Source Technology Centre
David.Woodhouse at intel.com                              Intel Corporation

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 5691 bytes
Desc: not available
URL: <http://mta.openssl.org/pipermail/openssl-dev/attachments/20160205/2fd733d4/attachment.bin>