Count values in first column where second column occurrences are the same in UNIX

I have to count values in first column of a file for which the second column occurrences are the same , using Unix.

12,abn

23,abn

45,hh

52,hh

89,mm

89,mm

My output should be something like this:

35,abn

97,hh

178,mm

Any help?