Tomáš Mráz comments

Results 1511 comments of


                                            Tomáš Mráz

rsa: set verify_message to 1 in rsa_sign only when message was digested

Should there be a test?

SIGSEGV in ossl_method_store_do_all

> 2\. The exact line that the process crashed on in ossl_method_store_do_all That would require building with debugging information enabled.

Fix memory leak in x509_pubkey_ex_d2i_ex()

Hmm... I see this is submitted with CLA: trivial but this change is beyond the limits for CLA: trivial submissions. @AITsygunka Would you please consider signing a CLA? https://openssl-library.org/policies/cla/index.html

ECH build artefacts and a bit of code

> And I'm probably doing something wrong with the fixup stuff judging by what's presented above - tell me if that should be done differently/better. It will work as you...

ECH build artefacts and a bit of code

@mattcaswell can you please reconfirm your approval?

SPARC assembly: Don't file aes-cbc on T4 with small sizes.

Hmm... the speed application should error out or at least print a warning when -bytes for the given cipher is not a multiple of its block size. This is basically...

SPARC assembly: Don't file aes-cbc on T4 with small sizes.

> The "openssl speed" can do it and it is part of the test suite. The implication is that openssl speed is wrong - it should report an error if...

SPARC assembly: Don't file aes-cbc on T4 with small sizes.

1. openssl speed should be fixed 2. The question is what is the right behavior for such invalid calls. I would argue that crash might be safer than returning an...

SPARC assembly: Don't file aes-cbc on T4 with small sizes.

> > 1. openssl speed should be fixed > align up to block or a different API? Just failing (and not crashing) would be OK. The testcase should be removed....

SPARC assembly: Don't file aes-cbc on T4 with small sizes.

I've created PR #25792 to fix the crashing speed command.