On Mon, 2019-03-11 at 02:08 +0300, Vitaly Chikunov wrote: > Mimi, > > On Sun, Mar 10, 2019 at 06:26:25PM -0400, Mimi Zohar wrote: > > Hi Vitaly, > > > > On Wed, 2019-02-13 at 01:46 +0300, Vitaly Chikunov wrote: > > > - Make digest name search work just with simple strcmp() and three > > > arrays, dropping strmatch(). > > > - Remove get_digestbyname() which was doing two hash name resolving > > > attempts, reverting to use plain EVP_get_digestbyname(). This will > > > force the user to specify the proper hash name depending on what > > > OpenSSL provides, allowing to specify older hash name in older > > > OpenSSL. > > > - Improve hash_info parser to produce hash_algo_name[] more resembling > > > what is in the kernel, making algocmp() not needed and removed. > > > - Fix indent in get_hash_algo_by_id(). > > > > Thanks, this is a lot better. As there isn't a "Signed-off-by", did > > Forgot this. I can resend with "Signed-off-by". > > > you want to squash this with the original patch - "Extract digest > > algorithms from hash_info.h"? > > Do you mean rebase squash with already upstreamed commit? Did not know > this is allowed. Agreed, normally it shouldn't be done, but it hasn't been included in a release. Defining and then removing algocmp() doesn't make sense. > > If you want to do it, this probably should not be squashed with "Extract > digest algorithms from hash_info.h", but with "Try to load digest by its > alias", because it is simplification of this commit. Unless there is a reason for keeping these patches separate, please squash them. It will be easier for anyone reviewing the code. Thanks! Mimi
